diff --git a/.gitignore b/.gitignore
index e5f932a..50723d9 100644
--- a/.gitignore
+++ b/.gitignore
@@ -21,3 +21,6 @@ __pycache__/
 .envrc
 devbox.json
 devbox.lock
+
+# Git worktrees
+worktrees/
diff --git a/GotoTest/.gitignore b/GotoTest_garage/.gitignore
similarity index 100%
rename from GotoTest/.gitignore
rename to GotoTest_garage/.gitignore
diff --git a/GotoTest/GotoTest b/GotoTest_garage/GotoTest
similarity index 100%
rename from GotoTest/GotoTest
rename to GotoTest_garage/GotoTest
diff --git a/GotoTest/GotoTest.json b/GotoTest_garage/GotoTest.json
similarity index 100%
rename from GotoTest/GotoTest.json
rename to GotoTest_garage/GotoTest.json
diff --git a/GotoTest/GotoTest.nim b/GotoTest_garage/GotoTest.nim
similarity index 100%
rename from GotoTest/GotoTest.nim
rename to GotoTest_garage/GotoTest.nim
diff --git a/GotoTest/GotoTest.nimble b/GotoTest_garage/GotoTest.nimble
similarity index 100%
rename from GotoTest/GotoTest.nimble
rename to GotoTest_garage/GotoTest.nimble
diff --git a/GotoTest/config.nims b/GotoTest_garage/config.nims
similarity index 100%
rename from GotoTest/config.nims
rename to GotoTest_garage/config.nims
diff --git a/OscillatorBot/config.nims b/OscillatorBot/config.nims
deleted file mode 100644
index 5457dba..0000000
--- a/OscillatorBot/config.nims
+++ /dev/null
@@ -1 +0,0 @@
---path:"../libs"
diff --git a/OscillatorBot_garage/OscillatorBot b/OscillatorBot_garage/OscillatorBot
new file mode 100755
index 0000000..6beaae2
Binary files /dev/null and b/OscillatorBot_garage/OscillatorBot differ
diff --git a/OscillatorBot/OscillatorBot.json b/OscillatorBot_garage/OscillatorBot.json
similarity index 100%
rename from OscillatorBot/OscillatorBot.json
rename to OscillatorBot_garage/OscillatorBot.json
diff --git a/OscillatorBot/OscillatorBot.nim b/OscillatorBot_garage/OscillatorBot.nim
similarity index 100%
rename from OscillatorBot/OscillatorBot.nim
rename to OscillatorBot_garage/OscillatorBot.nim
diff --git a/OscillatorBot/OscillatorBot.sh b/OscillatorBot_garage/OscillatorBot.sh
similarity index 100%
rename from OscillatorBot/OscillatorBot.sh
rename to OscillatorBot_garage/OscillatorBot.sh
diff --git a/OscillatorBot_garage/config.nims b/OscillatorBot_garage/config.nims
new file mode 100644
index 0000000..be6555f
--- /dev/null
+++ b/OscillatorBot_garage/config.nims
@@ -0,0 +1 @@
+--path:"../common_libs"
diff --git a/PPO_Bot/.gitignore b/PPO_Bot_garage/.gitignore
similarity index 100%
rename from PPO_Bot/.gitignore
rename to PPO_Bot_garage/.gitignore
diff --git a/PPO_Bot/PPO_Bot b/PPO_Bot_garage/PPO_Bot
similarity index 100%
rename from PPO_Bot/PPO_Bot
rename to PPO_Bot_garage/PPO_Bot
diff --git a/PPO_Bot/PPO_Bot.json b/PPO_Bot_garage/PPO_Bot.json
similarity index 100%
rename from PPO_Bot/PPO_Bot.json
rename to PPO_Bot_garage/PPO_Bot.json
diff --git a/PPO_Bot/PPO_Bot.nim b/PPO_Bot_garage/PPO_Bot.nim
similarity index 100%
rename from PPO_Bot/PPO_Bot.nim
rename to PPO_Bot_garage/PPO_Bot.nim
diff --git a/PPO_Bot/PPO_Bot.nimble b/PPO_Bot_garage/PPO_Bot.nimble
similarity index 78%
rename from PPO_Bot/PPO_Bot.nimble
rename to PPO_Bot_garage/PPO_Bot.nimble
index 49f065e..e35363c 100644
--- a/PPO_Bot/PPO_Bot.nimble
+++ b/PPO_Bot_garage/PPO_Bot.nimble
@@ -7,6 +7,6 @@ bin           = @["PPO_Bot"]
 
 # Dependencies
 requires "nim >= 2.0.0"
-# tankroyale_botapi is vendored in-tree (libs/tankroyale_botapi) and wired via
+# tankroyale_botapi is vendored in-tree (common_libs/tankroyale_botapi) and wired via
 # config.nims --path; no nimble dependency so builds never touch ~/.nimble/pkgs2.
 requires "arraymancer >= 0.7.0"
diff --git a/PPO_Bot/PPO_Bot.sh b/PPO_Bot_garage/PPO_Bot.sh
similarity index 100%
rename from PPO_Bot/PPO_Bot.sh
rename to PPO_Bot_garage/PPO_Bot.sh
diff --git a/PPO_Bot/actions.nim b/PPO_Bot_garage/actions.nim
similarity index 100%
rename from PPO_Bot/actions.nim
rename to PPO_Bot_garage/actions.nim
diff --git a/PPO_Bot/config.nims b/PPO_Bot_garage/config.nims
similarity index 79%
rename from PPO_Bot/config.nims
rename to PPO_Bot_garage/config.nims
index c95fd68..99bb126 100644
--- a/PPO_Bot/config.nims
+++ b/PPO_Bot_garage/config.nims
@@ -6,8 +6,8 @@ switch("threads", "on")
 when withDir(thisDir(), system.fileExists("nimble.paths")):
   include "nimble.paths"
 # end Nimble config
-# Use the repo-vendored Tank Royale bot API (libs/) instead of the nimble pkg.
+# Use the repo-vendored Tank Royale bot API (common_libs/) instead of the nimble pkg.
 # The pkg copy lived in ~/.nimble/pkgs2 and was patched ad-hoc; vendoring makes
 # the build self-contained and keeps the cross-thread Channel fix in-tree.
 # Must come AFTER the nimble.paths include: later --path wins the import search.
-switch("path", thisDir() & "/../libs/tankroyale_botapi")
+switch("path", thisDir() & "/../common_libs/tankroyale_botapi")
diff --git a/PPO_Bot/controllers.nim b/PPO_Bot_garage/controllers.nim
similarity index 100%
rename from PPO_Bot/controllers.nim
rename to PPO_Bot_garage/controllers.nim
diff --git a/PPO_Bot/enemy_tracker.nim b/PPO_Bot_garage/enemy_tracker.nim
similarity index 100%
rename from PPO_Bot/enemy_tracker.nim
rename to PPO_Bot_garage/enemy_tracker.nim
diff --git a/PPO_Bot/network.nim b/PPO_Bot_garage/network.nim
similarity index 100%
rename from PPO_Bot/network.nim
rename to PPO_Bot_garage/network.nim
diff --git a/PPO_Bot/shell.nix b/PPO_Bot_garage/shell.nix
similarity index 100%
rename from PPO_Bot/shell.nix
rename to PPO_Bot_garage/shell.nix
diff --git a/PPO_Bot/state_vector.nim b/PPO_Bot_garage/state_vector.nim
similarity index 100%
rename from PPO_Bot/state_vector.nim
rename to PPO_Bot_garage/state_vector.nim
diff --git a/PPO_Bot_garage/tests/diag_ppo_fullbuffer b/PPO_Bot_garage/tests/diag_ppo_fullbuffer
new file mode 100755
index 0000000..53ae8e1
Binary files /dev/null and b/PPO_Bot_garage/tests/diag_ppo_fullbuffer differ
diff --git a/PPO_Bot_garage/tests/diag_ppo_fullbuffer.nim b/PPO_Bot_garage/tests/diag_ppo_fullbuffer.nim
new file mode 100644
index 0000000..dcddb1d
--- /dev/null
+++ b/PPO_Bot_garage/tests/diag_ppo_fullbuffer.nim
@@ -0,0 +1,36 @@
+## diag_ppo_fullbuffer.nim — run ppoUpdate on a FULL 8192-transition buffer.
+## Verifies whether the round-10 death (first ppoUpdate at buffer capacity)
+## is a real crash in ppoUpdate or purely the saveAdamStates empty-shape bug.
+
+import std/[math, random]
+import arraymancer
+import "../network"
+import "../training"
+
+var ac = initActorCritic()
+var adam: ACAdamStates  # uninitialised → ppoUpdate must reinit (fresh-process path)
+var buf = initTrajectoryBuffer()
+
+randomize(1)
+while buf.len < MAX_TRANSITIONS:
+  var t: Transition
+  for i in 0..<STATE_DIM: t.state[i] = rand(1.0'f32) - 0.5'f32
+  for i in 0..<ACTION_DIM: t.action[i] = rand(1.0'f32) - 0.5'f32
+  t.logProb = rand(1.0'f32) - 1.0'f32
+  t.reward = rand(0.02'f32) - 0.01'f32
+  t.value = rand(0.1'f32)
+  t.done = buf.len mod 300 == 299
+  buf.add(t)
+
+echo "buffer.len = ", buf.len, " (MAX=", MAX_TRANSITIONS, ")"
+var ac2 = ac
+var adam2: ACAdamStates
+try:
+  let m = ppoUpdate(ac2, buf, lastValue = 0.0'f32, adamStates = adam2)
+  echo "ppoUpdate OK: aLoss=", m.actorLoss, " vLoss=", m.valueLoss, " gNorm=", m.gradNorm
+except CatchableError as e:
+  echo "CAUGHT CatchableError: ", e.msg
+  echo getStackTrace(e)
+except Defect as e:
+  echo "CAUGHT Defect: ", e.msg
+  echo getStackTrace(e)
\ No newline at end of file
diff --git a/PPO_Bot_garage/tests/diag_savecheckpoint b/PPO_Bot_garage/tests/diag_savecheckpoint
new file mode 100755
index 0000000..7ac2859
Binary files /dev/null and b/PPO_Bot_garage/tests/diag_savecheckpoint differ
diff --git a/PPO_Bot_garage/tests/diag_savecheckpoint.nim b/PPO_Bot_garage/tests/diag_savecheckpoint.nim
new file mode 100644
index 0000000..9fcce98
--- /dev/null
+++ b/PPO_Bot_garage/tests/diag_savecheckpoint.nim
@@ -0,0 +1,31 @@
+## diag_savecheckpoint.nim — reproduce the startup+saveCheckpoint path that
+## crashes with `io_npy.nim(143, 3) `0 < t.shape.len`` under the boot server.
+## Run: nim c -d:release tests/diag_savecheckpoint.nim && ./tests/diag_savecheckpoint
+
+import std/[os]
+import arraymancer
+import "../network"
+import "../training"
+import "../weights"
+
+const weightsRoot = currentSourcePath().parentDir.parentDir / "weights"
+
+var ac = initActorCritic()
+var adam: ACAdamStates
+
+let loadResult = loadBestAvailable(ac, adam, weightsRoot)
+echo "loaded=", loadResult.loaded, " roundNum=", loadResult.roundNum
+echo "adam.initialized=", adam.initialized
+echo "adam.aw1.m.shape=", adam.aw1.m.shape, " aw1.v.shape=", adam.aw1.v.shape
+echo "adam.logStd.m.shape=", adam.logStd.m.shape
+echo "adam.aw1.t=", adam.aw1.t
+
+try:
+  saveCheckpoint(ac, adam, weightsRoot, 1)
+  echo "saveCheckpoint OK"
+except CatchableError as e:
+  echo "CAUGHT CatchableError: ", e.msg
+  echo getStackTrace(e)
+except Defect as e:
+  echo "CAUGHT Defect: ", e.msg
+  echo getStackTrace(e)
\ No newline at end of file
diff --git a/PPO_Bot/tests/test_actions b/PPO_Bot_garage/tests/test_actions
similarity index 100%
rename from PPO_Bot/tests/test_actions
rename to PPO_Bot_garage/tests/test_actions
diff --git a/PPO_Bot/tests/test_actions.nim b/PPO_Bot_garage/tests/test_actions.nim
similarity index 100%
rename from PPO_Bot/tests/test_actions.nim
rename to PPO_Bot_garage/tests/test_actions.nim
diff --git a/PPO_Bot/tests/test_controllers b/PPO_Bot_garage/tests/test_controllers
similarity index 100%
rename from PPO_Bot/tests/test_controllers
rename to PPO_Bot_garage/tests/test_controllers
diff --git a/PPO_Bot/tests/test_controllers.nim b/PPO_Bot_garage/tests/test_controllers.nim
similarity index 100%
rename from PPO_Bot/tests/test_controllers.nim
rename to PPO_Bot_garage/tests/test_controllers.nim
diff --git a/PPO_Bot/tests/test_network b/PPO_Bot_garage/tests/test_network
similarity index 100%
rename from PPO_Bot/tests/test_network
rename to PPO_Bot_garage/tests/test_network
diff --git a/PPO_Bot/tests/test_network.nim b/PPO_Bot_garage/tests/test_network.nim
similarity index 100%
rename from PPO_Bot/tests/test_network.nim
rename to PPO_Bot_garage/tests/test_network.nim
diff --git a/PPO_Bot/tests/test_radar_lock.nim b/PPO_Bot_garage/tests/test_radar_lock.nim
similarity index 100%
rename from PPO_Bot/tests/test_radar_lock.nim
rename to PPO_Bot_garage/tests/test_radar_lock.nim
diff --git a/PPO_Bot/tests/test_state b/PPO_Bot_garage/tests/test_state
similarity index 100%
rename from PPO_Bot/tests/test_state
rename to PPO_Bot_garage/tests/test_state
diff --git a/PPO_Bot/tests/test_state.nim b/PPO_Bot_garage/tests/test_state.nim
similarity index 100%
rename from PPO_Bot/tests/test_state.nim
rename to PPO_Bot_garage/tests/test_state.nim
diff --git a/PPO_Bot/tests/test_training b/PPO_Bot_garage/tests/test_training
similarity index 100%
rename from PPO_Bot/tests/test_training
rename to PPO_Bot_garage/tests/test_training
diff --git a/PPO_Bot/tests/test_training.nim b/PPO_Bot_garage/tests/test_training.nim
similarity index 100%
rename from PPO_Bot/tests/test_training.nim
rename to PPO_Bot_garage/tests/test_training.nim
diff --git a/PPO_Bot/tests/test_weights b/PPO_Bot_garage/tests/test_weights
similarity index 100%
rename from PPO_Bot/tests/test_weights
rename to PPO_Bot_garage/tests/test_weights
diff --git a/PPO_Bot/tests/test_weights.nim b/PPO_Bot_garage/tests/test_weights.nim
similarity index 100%
rename from PPO_Bot/tests/test_weights.nim
rename to PPO_Bot_garage/tests/test_weights.nim
diff --git a/PPO_Bot/training b/PPO_Bot_garage/training
similarity index 100%
rename from PPO_Bot/training
rename to PPO_Bot_garage/training
diff --git a/PPO_Bot/training.nim b/PPO_Bot_garage/training.nim
similarity index 100%
rename from PPO_Bot/training.nim
rename to PPO_Bot_garage/training.nim
diff --git a/PPO_Bot/weights.nim b/PPO_Bot_garage/weights.nim
similarity index 100%
rename from PPO_Bot/weights.nim
rename to PPO_Bot_garage/weights.nim
diff --git a/PPO_Bot/weights/latest/actor_b1.npy b/PPO_Bot_garage/weights/latest/actor_b1.npy
similarity index 100%
rename from PPO_Bot/weights/latest/actor_b1.npy
rename to PPO_Bot_garage/weights/latest/actor_b1.npy
diff --git a/PPO_Bot/weights/latest/actor_b2.npy b/PPO_Bot_garage/weights/latest/actor_b2.npy
similarity index 100%
rename from PPO_Bot/weights/latest/actor_b2.npy
rename to PPO_Bot_garage/weights/latest/actor_b2.npy
diff --git a/PPO_Bot/weights/latest/actor_b3.npy b/PPO_Bot_garage/weights/latest/actor_b3.npy
similarity index 100%
rename from PPO_Bot/weights/latest/actor_b3.npy
rename to PPO_Bot_garage/weights/latest/actor_b3.npy
diff --git a/PPO_Bot/weights/latest/actor_w1.npy b/PPO_Bot_garage/weights/latest/actor_w1.npy
similarity index 100%
rename from PPO_Bot/weights/latest/actor_w1.npy
rename to PPO_Bot_garage/weights/latest/actor_w1.npy
diff --git a/PPO_Bot/weights/latest/actor_w2.npy b/PPO_Bot_garage/weights/latest/actor_w2.npy
similarity index 100%
rename from PPO_Bot/weights/latest/actor_w2.npy
rename to PPO_Bot_garage/weights/latest/actor_w2.npy
diff --git a/PPO_Bot/weights/latest/actor_w3.npy b/PPO_Bot_garage/weights/latest/actor_w3.npy
similarity index 100%
rename from PPO_Bot/weights/latest/actor_w3.npy
rename to PPO_Bot_garage/weights/latest/actor_w3.npy
diff --git a/PPO_Bot/weights/latest/adam_ab1_m.npy b/PPO_Bot_garage/weights/latest/adam_ab1_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_ab1_m.npy
rename to PPO_Bot_garage/weights/latest/adam_ab1_m.npy
diff --git a/PPO_Bot/weights/latest/adam_ab1_v.npy b/PPO_Bot_garage/weights/latest/adam_ab1_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_ab1_v.npy
rename to PPO_Bot_garage/weights/latest/adam_ab1_v.npy
diff --git a/PPO_Bot/weights/latest/adam_ab2_m.npy b/PPO_Bot_garage/weights/latest/adam_ab2_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_ab2_m.npy
rename to PPO_Bot_garage/weights/latest/adam_ab2_m.npy
diff --git a/PPO_Bot/weights/latest/adam_ab2_v.npy b/PPO_Bot_garage/weights/latest/adam_ab2_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_ab2_v.npy
rename to PPO_Bot_garage/weights/latest/adam_ab2_v.npy
diff --git a/PPO_Bot/weights/latest/adam_ab3_m.npy b/PPO_Bot_garage/weights/latest/adam_ab3_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_ab3_m.npy
rename to PPO_Bot_garage/weights/latest/adam_ab3_m.npy
diff --git a/PPO_Bot/weights/latest/adam_ab3_v.npy b/PPO_Bot_garage/weights/latest/adam_ab3_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_ab3_v.npy
rename to PPO_Bot_garage/weights/latest/adam_ab3_v.npy
diff --git a/PPO_Bot/weights/latest/adam_aw1_m.npy b/PPO_Bot_garage/weights/latest/adam_aw1_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_aw1_m.npy
rename to PPO_Bot_garage/weights/latest/adam_aw1_m.npy
diff --git a/PPO_Bot/weights/latest/adam_aw1_v.npy b/PPO_Bot_garage/weights/latest/adam_aw1_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_aw1_v.npy
rename to PPO_Bot_garage/weights/latest/adam_aw1_v.npy
diff --git a/PPO_Bot/weights/latest/adam_aw2_m.npy b/PPO_Bot_garage/weights/latest/adam_aw2_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_aw2_m.npy
rename to PPO_Bot_garage/weights/latest/adam_aw2_m.npy
diff --git a/PPO_Bot/weights/latest/adam_aw2_v.npy b/PPO_Bot_garage/weights/latest/adam_aw2_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_aw2_v.npy
rename to PPO_Bot_garage/weights/latest/adam_aw2_v.npy
diff --git a/PPO_Bot/weights/latest/adam_aw3_m.npy b/PPO_Bot_garage/weights/latest/adam_aw3_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_aw3_m.npy
rename to PPO_Bot_garage/weights/latest/adam_aw3_m.npy
diff --git a/PPO_Bot/weights/latest/adam_aw3_v.npy b/PPO_Bot_garage/weights/latest/adam_aw3_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_aw3_v.npy
rename to PPO_Bot_garage/weights/latest/adam_aw3_v.npy
diff --git a/PPO_Bot/weights/latest/adam_cb1_m.npy b/PPO_Bot_garage/weights/latest/adam_cb1_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cb1_m.npy
rename to PPO_Bot_garage/weights/latest/adam_cb1_m.npy
diff --git a/PPO_Bot/weights/latest/adam_cb1_v.npy b/PPO_Bot_garage/weights/latest/adam_cb1_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cb1_v.npy
rename to PPO_Bot_garage/weights/latest/adam_cb1_v.npy
diff --git a/PPO_Bot/weights/latest/adam_cb2_m.npy b/PPO_Bot_garage/weights/latest/adam_cb2_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cb2_m.npy
rename to PPO_Bot_garage/weights/latest/adam_cb2_m.npy
diff --git a/PPO_Bot/weights/latest/adam_cb2_v.npy b/PPO_Bot_garage/weights/latest/adam_cb2_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cb2_v.npy
rename to PPO_Bot_garage/weights/latest/adam_cb2_v.npy
diff --git a/PPO_Bot/weights/latest/adam_cb3_m.npy b/PPO_Bot_garage/weights/latest/adam_cb3_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cb3_m.npy
rename to PPO_Bot_garage/weights/latest/adam_cb3_m.npy
diff --git a/PPO_Bot/weights/latest/adam_cb3_v.npy b/PPO_Bot_garage/weights/latest/adam_cb3_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cb3_v.npy
rename to PPO_Bot_garage/weights/latest/adam_cb3_v.npy
diff --git a/PPO_Bot/weights/latest/adam_cw1_m.npy b/PPO_Bot_garage/weights/latest/adam_cw1_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cw1_m.npy
rename to PPO_Bot_garage/weights/latest/adam_cw1_m.npy
diff --git a/PPO_Bot/weights/latest/adam_cw1_v.npy b/PPO_Bot_garage/weights/latest/adam_cw1_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cw1_v.npy
rename to PPO_Bot_garage/weights/latest/adam_cw1_v.npy
diff --git a/PPO_Bot/weights/latest/adam_cw2_m.npy b/PPO_Bot_garage/weights/latest/adam_cw2_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cw2_m.npy
rename to PPO_Bot_garage/weights/latest/adam_cw2_m.npy
diff --git a/PPO_Bot/weights/latest/adam_cw2_v.npy b/PPO_Bot_garage/weights/latest/adam_cw2_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cw2_v.npy
rename to PPO_Bot_garage/weights/latest/adam_cw2_v.npy
diff --git a/PPO_Bot/weights/latest/adam_cw3_m.npy b/PPO_Bot_garage/weights/latest/adam_cw3_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cw3_m.npy
rename to PPO_Bot_garage/weights/latest/adam_cw3_m.npy
diff --git a/PPO_Bot/weights/latest/adam_cw3_v.npy b/PPO_Bot_garage/weights/latest/adam_cw3_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_cw3_v.npy
rename to PPO_Bot_garage/weights/latest/adam_cw3_v.npy
diff --git a/PPO_Bot/weights/latest/adam_logstd_m.npy b/PPO_Bot_garage/weights/latest/adam_logstd_m.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_logstd_m.npy
rename to PPO_Bot_garage/weights/latest/adam_logstd_m.npy
diff --git a/PPO_Bot/weights/latest/adam_logstd_v.npy b/PPO_Bot_garage/weights/latest/adam_logstd_v.npy
similarity index 100%
rename from PPO_Bot/weights/latest/adam_logstd_v.npy
rename to PPO_Bot_garage/weights/latest/adam_logstd_v.npy
diff --git a/PPO_Bot/weights/latest/adam_t.txt b/PPO_Bot_garage/weights/latest/adam_t.txt
similarity index 100%
rename from PPO_Bot/weights/latest/adam_t.txt
rename to PPO_Bot_garage/weights/latest/adam_t.txt
diff --git a/PPO_Bot/weights/latest/critic_b1.npy b/PPO_Bot_garage/weights/latest/critic_b1.npy
similarity index 100%
rename from PPO_Bot/weights/latest/critic_b1.npy
rename to PPO_Bot_garage/weights/latest/critic_b1.npy
diff --git a/PPO_Bot/weights/latest/critic_b2.npy b/PPO_Bot_garage/weights/latest/critic_b2.npy
similarity index 100%
rename from PPO_Bot/weights/latest/critic_b2.npy
rename to PPO_Bot_garage/weights/latest/critic_b2.npy
diff --git a/PPO_Bot/weights/latest/critic_b3.npy b/PPO_Bot_garage/weights/latest/critic_b3.npy
similarity index 100%
rename from PPO_Bot/weights/latest/critic_b3.npy
rename to PPO_Bot_garage/weights/latest/critic_b3.npy
diff --git a/PPO_Bot/weights/latest/critic_w1.npy b/PPO_Bot_garage/weights/latest/critic_w1.npy
similarity index 100%
rename from PPO_Bot/weights/latest/critic_w1.npy
rename to PPO_Bot_garage/weights/latest/critic_w1.npy
diff --git a/PPO_Bot/weights/latest/critic_w2.npy b/PPO_Bot_garage/weights/latest/critic_w2.npy
similarity index 100%
rename from PPO_Bot/weights/latest/critic_w2.npy
rename to PPO_Bot_garage/weights/latest/critic_w2.npy
diff --git a/PPO_Bot/weights/latest/critic_w3.npy b/PPO_Bot_garage/weights/latest/critic_w3.npy
similarity index 100%
rename from PPO_Bot/weights/latest/critic_w3.npy
rename to PPO_Bot_garage/weights/latest/critic_w3.npy
diff --git a/PPO_Bot/weights/latest/log_std.npy b/PPO_Bot_garage/weights/latest/log_std.npy
similarity index 100%
rename from PPO_Bot/weights/latest/log_std.npy
rename to PPO_Bot_garage/weights/latest/log_std.npy
diff --git a/PPO_Bot/weights/round_counter.txt b/PPO_Bot_garage/weights/round_counter.txt
similarity index 100%
rename from PPO_Bot/weights/round_counter.txt
rename to PPO_Bot_garage/weights/round_counter.txt
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b1.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b1.npy
new file mode 100644
index 0000000..5798989
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b2.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b2.npy
new file mode 100644
index 0000000..e04078a
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b3.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b3.npy
new file mode 100644
index 0000000..05ae486
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w1.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w1.npy
new file mode 100644
index 0000000..32cd9bc
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w2.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w2.npy
new file mode 100644
index 0000000..12a6354
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w3.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w3.npy
new file mode 100644
index 0000000..ca68c7f
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/actor_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab1_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab1_m.npy
new file mode 100644
index 0000000..568fbd1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab1_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab1_v.npy
new file mode 100644
index 0000000..9b208da
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab2_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab2_m.npy
new file mode 100644
index 0000000..68d932d
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab2_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab2_v.npy
new file mode 100644
index 0000000..e5e46d2
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab3_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab3_m.npy
new file mode 100644
index 0000000..3386dce
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab3_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab3_v.npy
new file mode 100644
index 0000000..a3e44bf
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_ab3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw1_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw1_m.npy
new file mode 100644
index 0000000..b5e6652
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw1_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw1_v.npy
new file mode 100644
index 0000000..0ae1f1f
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw2_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw2_m.npy
new file mode 100644
index 0000000..d3739b6
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw2_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw2_v.npy
new file mode 100644
index 0000000..4a9c271
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw3_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw3_m.npy
new file mode 100644
index 0000000..a788692
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw3_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw3_v.npy
new file mode 100644
index 0000000..96b94ca
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_aw3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb1_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb1_m.npy
new file mode 100644
index 0000000..cc7f046
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb1_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb1_v.npy
new file mode 100644
index 0000000..0be07a5
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb2_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb2_m.npy
new file mode 100644
index 0000000..37c9348
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb2_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb2_v.npy
new file mode 100644
index 0000000..d126bb4
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb3_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb3_m.npy
new file mode 100644
index 0000000..560fb0b
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb3_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb3_v.npy
new file mode 100644
index 0000000..af4900d
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cb3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw1_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw1_m.npy
new file mode 100644
index 0000000..c6dc6bf
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw1_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw1_v.npy
new file mode 100644
index 0000000..9de6cf0
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw2_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw2_m.npy
new file mode 100644
index 0000000..4e781b4
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw2_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw2_v.npy
new file mode 100644
index 0000000..73180e4
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw3_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw3_m.npy
new file mode 100644
index 0000000..e25442c
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw3_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw3_v.npy
new file mode 100644
index 0000000..73bed0c
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_cw3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_logstd_m.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_logstd_m.npy
new file mode 100644
index 0000000..f259e29
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_logstd_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_logstd_v.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_logstd_v.npy
new file mode 100644
index 0000000..9bc916a
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_logstd_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_t.txt b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_t.txt
new file mode 100644
index 0000000..1054059
--- /dev/null
+++ b/PPO_Bot_garage/weights_pre_fix_20260818/latest/adam_t.txt
@@ -0,0 +1,13 @@
+1056
+1056
+1056
+1056
+1056
+1056
+1056
+1056
+1056
+1056
+1056
+1056
+1056
\ No newline at end of file
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b1.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b1.npy
new file mode 100644
index 0000000..4bdcdfa
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b2.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b2.npy
new file mode 100644
index 0000000..60e32ea
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b3.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b3.npy
new file mode 100644
index 0000000..00e5f5f
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w1.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w1.npy
new file mode 100644
index 0000000..0b3e2cf
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w2.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w2.npy
new file mode 100644
index 0000000..0469371
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w3.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w3.npy
new file mode 100644
index 0000000..648a5bc
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/critic_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/latest/log_std.npy b/PPO_Bot_garage/weights_pre_fix_20260818/latest/log_std.npy
new file mode 100644
index 0000000..726585b
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fix_20260818/latest/log_std.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fix_20260818/round_counter.txt b/PPO_Bot_garage/weights_pre_fix_20260818/round_counter.txt
new file mode 100644
index 0000000..8e2afd3
--- /dev/null
+++ b/PPO_Bot_garage/weights_pre_fix_20260818/round_counter.txt
@@ -0,0 +1 @@
+17
\ No newline at end of file
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b1.npy
new file mode 100644
index 0000000..741d564
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b2.npy
new file mode 100644
index 0000000..e29c3b7
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b3.npy
new file mode 100644
index 0000000..e7f33f9
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w1.npy
new file mode 100644
index 0000000..8951e35
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w2.npy
new file mode 100644
index 0000000..c13b317
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w3.npy
new file mode 100644
index 0000000..ba9a020
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/actor_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b1.npy
new file mode 100644
index 0000000..3a71af4
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b2.npy
new file mode 100644
index 0000000..82c7e3f
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b3.npy
new file mode 100644
index 0000000..be4f787
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w1.npy
new file mode 100644
index 0000000..083c6b6
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w2.npy
new file mode 100644
index 0000000..e77a2be
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w3.npy
new file mode 100644
index 0000000..b0c7929
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/critic_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/log_std.npy b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/log_std.npy
new file mode 100644
index 0000000..bf1ea76
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/checkpoint_1/log_std.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b1.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b2.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b3.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w1.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w2.npy
new file mode 100644
index 0000000..68e8784
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w3.npy
new file mode 100644
index 0000000..5575a6c
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/actor_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab1_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab1_m.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab1_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab1_v.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab2_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab2_m.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab2_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab2_v.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab3_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab3_m.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab3_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab3_v.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_ab3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw1_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw1_m.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw1_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw1_v.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw2_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw2_m.npy
new file mode 100644
index 0000000..68e8784
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw2_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw2_v.npy
new file mode 100644
index 0000000..68e8784
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw3_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw3_m.npy
new file mode 100644
index 0000000..5575a6c
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw3_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw3_v.npy
new file mode 100644
index 0000000..5575a6c
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_aw3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb1_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb1_m.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb1_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb1_v.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb2_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb2_m.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb2_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb2_v.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb3_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb3_m.npy
new file mode 100644
index 0000000..787c1c7
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb3_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb3_v.npy
new file mode 100644
index 0000000..787c1c7
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cb3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw1_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw1_m.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw1_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw1_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw1_v.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw1_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw2_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw2_m.npy
new file mode 100644
index 0000000..68e8784
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw2_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw2_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw2_v.npy
new file mode 100644
index 0000000..68e8784
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw2_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw3_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw3_m.npy
new file mode 100644
index 0000000..e69fa63
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw3_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw3_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw3_v.npy
new file mode 100644
index 0000000..e69fa63
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_cw3_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_logstd_m.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_logstd_m.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_logstd_m.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_logstd_v.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_logstd_v.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_logstd_v.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_t.txt b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_t.txt
new file mode 100644
index 0000000..01a2257
--- /dev/null
+++ b/PPO_Bot_garage/weights_pre_fresh_backup/latest/adam_t.txt
@@ -0,0 +1,13 @@
+3172
+3172
+3172
+3172
+3172
+3172
+3172
+3172
+3172
+3172
+3172
+3172
+3172
\ No newline at end of file
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b1.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b2.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b3.npy
new file mode 100644
index 0000000..787c1c7
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w1.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w2.npy
new file mode 100644
index 0000000..dd53485
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w3.npy
new file mode 100644
index 0000000..e69fa63
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/critic_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest/log_std.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest/log_std.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest/log_std.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b1.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b2.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b3.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w1.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w2.npy
new file mode 100644
index 0000000..68e8784
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w3.npy
new file mode 100644
index 0000000..5575a6c
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/actor_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b1.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b2.npy
new file mode 100644
index 0000000..5c8aaae
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b3.npy
new file mode 100644
index 0000000..787c1c7
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w1.npy
new file mode 100644
index 0000000..bc250f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w2.npy
new file mode 100644
index 0000000..dd53485
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w3.npy
new file mode 100644
index 0000000..e69fa63
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/critic_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/log_std.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/log_std.npy
new file mode 100644
index 0000000..4bd2c62
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_bk/log_std.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b1.npy
new file mode 100644
index 0000000..417fdf5
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b2.npy
new file mode 100644
index 0000000..a2cfa74
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b3.npy
new file mode 100644
index 0000000..c7227d1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w1.npy
new file mode 100644
index 0000000..88bcaba
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w2.npy
new file mode 100644
index 0000000..b9aca1f
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w3.npy
new file mode 100644
index 0000000..cdab5f1
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/actor_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b1.npy
new file mode 100644
index 0000000..c2e4b97
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b2.npy
new file mode 100644
index 0000000..e539163
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b3.npy
new file mode 100644
index 0000000..bbe2e48
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_b3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w1.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w1.npy
new file mode 100644
index 0000000..8692b51
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w1.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w2.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w2.npy
new file mode 100644
index 0000000..d7292f6
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w2.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w3.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w3.npy
new file mode 100644
index 0000000..5f48393
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/critic_w3.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/log_std.npy b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/log_std.npy
new file mode 100644
index 0000000..ffc8e2b
Binary files /dev/null and b/PPO_Bot_garage/weights_pre_fresh_backup/latest_pre_shaping_backup/log_std.npy differ
diff --git a/PPO_Bot_garage/weights_pre_fresh_backup/round_counter.txt b/PPO_Bot_garage/weights_pre_fresh_backup/round_counter.txt
new file mode 100644
index 0000000..86ee83a
--- /dev/null
+++ b/PPO_Bot_garage/weights_pre_fresh_backup/round_counter.txt
@@ -0,0 +1 @@
+40
\ No newline at end of file
diff --git a/QBot_garage/QBot b/QBot_garage/QBot
new file mode 100755
index 0000000..b533462
Binary files /dev/null and b/QBot_garage/QBot differ
diff --git a/QBot_garage/QBot.json b/QBot_garage/QBot.json
new file mode 100644
index 0000000..f49293a
--- /dev/null
+++ b/QBot_garage/QBot.json
@@ -0,0 +1,11 @@
+{
+  "name": "QBot",
+  "version": "0.1.0",
+  "authors": ["Davide Cappellini"],
+  "description": "Tabular Q-learning bot with chained action decomposition",
+  "homepage": "",
+  "countryCodes": ["IT"],
+  "gameTypes": ["classic", "1v1"],
+  "platform": "Nim",
+  "programmingLang": "Nim"
+}
diff --git a/QBot_garage/QBot.nim b/QBot_garage/QBot.nim
new file mode 100644
index 0000000..adb65cb
--- /dev/null
+++ b/QBot_garage/QBot.nim
@@ -0,0 +1,369 @@
+# QBot — gun-only Q-learning prototype, v2.
+# Chained action decomposition: T1 (aim offset, 7 actions) + T2 (fire, 3 actions).
+# Bullet tracking ring buffer: Q-updates happen ONLY on bullet resolution (hit/wall).
+# 120 states × (7 + 3) actions = 1,200 Q-table entries total.
+# All debug output goes to stderr. printToStdOut is never called.
+#
+# Bullet tracking protocol:
+#   1. run(): setFire succeeds → push BulletRecord with valid=false, id=-1 ("pending")
+#   2. onBulletFired(): stamp the oldest pending record with the real bulletId, set valid=true
+#   3. onBulletHit / onBulletHitWall(): resolveBullet() finds by id, does Q-update, clears slot
+
+import std/[random, math, os, streams, strutils]
+import tankroyale_botapi
+import radar_lock/radar_lock as radar_lock
+
+# ── Constants ──────────────────────────────────────────────────────────────────
+
+const botJsonPath = currentSourcePath().parentDir / "QBot.json"
+const qtablesDir  = currentSourcePath().parentDir / "qtables"
+
+const
+  # State features
+  N_AIM_ERR  = 5   # gun aim error: far-left / left / on-target / right / far-right
+  N_DIST     = 3   # close / mid / far
+  N_LAT_SPD  = 4   # enemy lateral speed: fast-left / slow-left / slow-right / fast-right
+  N_HEAT     = 2   # gun ready / cooling
+  N_STATES   = N_AIM_ERR * N_DIST * N_LAT_SPD * N_HEAT  # 120
+
+  # Action tables
+  N_AIM_ACT  = 7   # aim offset actions
+  N_FIRE_ACT = 3   # fire decision actions
+
+  ALPHA     = 0.3
+  # No GAMMA: contextual bandit (γ=0). Each shot is independent; no future state.
+  # Q[s,a] converges to expected reward of action a in state s.
+  EPS_INIT  = 0.5
+  EPS_DECAY = 0.995
+  EPS_FLOOR = 0.03
+  Q_INIT    = 0.0   # neutral init; signal from +1 hit / -0.2 miss drives learning
+
+  # Bullet tracking ring buffer — max bullets in flight simultaneously
+  BULLET_BUF_CAP = 5
+
+# Aim offset per action (degrees relative to direct bearing to enemy)
+const AIM_OFFSETS: array[N_AIM_ACT, float] = [-12.0, -6.0, -3.0, 0.0, 3.0, 6.0, 12.0]
+
+# Fire power per action (0.0 = don't fire)
+const FIRE_POWERS: array[N_FIRE_ACT, float] = [0.0, 1.0, 3.0]
+
+# ── Bullet tracking ───────────────────────────────────────────────────────────
+
+type BulletRecord = object
+  bulletId: int    # -1 = pending (waiting for onBulletFired to stamp real ID)
+  state:    int
+  aimAct:   int
+  fireAct:  int
+  valid:    bool   # true = slot occupied (either pending or id-stamped)
+
+# ── Bot type ───────────────────────────────────────────────────────────────────
+
+type QBot = ref object of Bot
+  # Enemy tracking
+  enemyX, enemyY: float
+  enemyDir:       float
+  enemySpeed:     float
+  hasContact:     bool
+  ticksSinceScan: int
+
+  # Q-tables (flat arrays)
+  # T1: aim offset — index [state * N_AIM_ACT + aimAction]
+  qt1: array[N_STATES * N_AIM_ACT, float64]
+  # T2: fire decision — index [state * N_FIRE_ACT + fireAction]
+  qt2: array[N_STATES * N_FIRE_ACT, float64]
+
+  # Bullet tracking ring buffer
+  bullets:    array[BULLET_BUF_CAP, BulletRecord]
+  bulletHead: int  # next write position (ring)
+
+  # Round stats
+  totalRounds: int
+  epsilon:     float
+  roundHits:   int
+  roundMisses: int
+  roundShots:  int
+
+# ── State discretization ──────────────────────────────────────────────────────
+
+proc discretizeState(gunAimError, dist, lateralSpeed, gunHeat: float): int {.inline.} =
+  # gunAimError: signed angle from gun to direct enemy bearing (negative = gun left of enemy)
+  let aimBin =
+    if gunAimError < -20.0: 0
+    elif gunAimError < -5.0: 1
+    elif gunAimError <= 5.0: 2
+    elif gunAimError <= 20.0: 3
+    else: 4
+
+  let distBin =
+    if dist < 200.0: 0 elif dist <= 500.0: 1 else: 2
+
+  # lateralSpeed: enemy velocity component perpendicular to our line of sight
+  let latBin =
+    if lateralSpeed < -3.0: 0
+    elif lateralSpeed < 0.0: 1
+    elif lateralSpeed <= 3.0: 2
+    else: 3
+
+  # heat ready: threshold slightly above 0 since fire 3.0 generates heat 1.6
+  let heatBin = if gunHeat <= 0.2: 0 else: 1
+
+  aimBin * (N_DIST * N_LAT_SPD * N_HEAT) +
+    distBin * (N_LAT_SPD * N_HEAT) +
+    latBin * N_HEAT +
+    heatBin
+
+# ── Q-learning helpers ────────────────────────────────────────────────────────
+
+proc selectAimAction(bot: QBot, state: int): int =
+  if rand(1.0) < bot.epsilon:
+    return rand(N_AIM_ACT - 1)
+  var best = -1e30
+  var bestA = 0
+  for a in 0 ..< N_AIM_ACT:
+    let v = bot.qt1[state * N_AIM_ACT + a]
+    if v > best:
+      best  = v
+      bestA = a
+  bestA
+
+proc selectFireAction(bot: QBot, state: int): int =
+  if rand(1.0) < bot.epsilon:
+    return rand(N_FIRE_ACT - 1)
+  var best = -1e30
+  var bestA = 0
+  for a in 0 ..< N_FIRE_ACT:
+    let v = bot.qt2[state * N_FIRE_ACT + a]
+    if v > best:
+      best  = v
+      bestA = a
+  bestA
+
+proc updateQT1(bot: QBot, state, action: int, reward: float) {.inline.} =
+  # γ=0 bandit update: Q ← Q + α*(r - Q)
+  let idx = state * N_AIM_ACT + action
+  bot.qt1[idx] += ALPHA * (reward - bot.qt1[idx])
+
+proc updateQT2(bot: QBot, state, action: int, reward: float) {.inline.} =
+  let idx = state * N_FIRE_ACT + action
+  bot.qt2[idx] += ALPHA * (reward - bot.qt2[idx])
+
+# ── Bullet ring buffer ────────────────────────────────────────────────────────
+
+proc pushPendingBullet(bot: QBot, state, aimAct, fireAct: int) =
+  # Push a pending record; onBulletFired will stamp the real ID.
+  bot.bullets[bot.bulletHead] = BulletRecord(
+    bulletId: -1, state: state, aimAct: aimAct, fireAct: fireAct, valid: true)
+  bot.bulletHead = (bot.bulletHead + 1) mod BULLET_BUF_CAP
+
+proc stampBulletId(bot: QBot, realId: int) =
+  # Find the oldest pending (id=-1) record and stamp it with the real bullet ID.
+  # Search backwards from bulletHead (most recent push is just before head).
+  for i in 1 .. BULLET_BUF_CAP:
+    let idx = (bot.bulletHead - i + BULLET_BUF_CAP) mod BULLET_BUF_CAP
+    if bot.bullets[idx].valid and bot.bullets[idx].bulletId == -1:
+      bot.bullets[idx].bulletId = realId
+      return
+
+proc resolveBullet(bot: QBot, bulletId: int, reward: float) =
+  for i in 0 ..< BULLET_BUF_CAP:
+    if bot.bullets[i].valid and bot.bullets[i].bulletId == bulletId:
+      let rec = bot.bullets[i]
+      bot.bullets[i].valid = false
+      updateQT1(bot, rec.state, rec.aimAct, reward)
+      updateQT2(bot, rec.state, rec.fireAct, reward)
+      return
+  # Not found — bullet from previous round or missed stamp; silently ignore.
+
+# ── Persistence ───────────────────────────────────────────────────────────────
+
+proc saveQTable(bot: QBot) =
+  createDir(qtablesDir)
+  let s = newFileStream(qtablesDir / "gun_q.bin", fmWrite)
+  if s == nil: return
+  for v in bot.qt1: s.write(v)
+  for v in bot.qt2: s.write(v)
+  s.close()
+  writeFile(qtablesDir / "gun_meta.txt",
+    "rounds=" & $bot.totalRounds & "\n" &
+    "epsilon=" & $bot.epsilon & "\n")
+
+proc loadQTable(bot: QBot) =
+  let path = qtablesDir / "gun_q.bin"
+  if not fileExists(path): return
+  let s = newFileStream(path, fmRead)
+  if s == nil: return
+  var i = 0
+  while not s.atEnd and i < bot.qt1.len:
+    bot.qt1[i] = s.readFloat64()
+    inc i
+  i = 0
+  while not s.atEnd and i < bot.qt2.len:
+    bot.qt2[i] = s.readFloat64()
+    inc i
+  s.close()
+  let mp = qtablesDir / "gun_meta.txt"
+  if not fileExists(mp): return
+  for line in lines(mp):
+    if line.startsWith("rounds="):
+      bot.totalRounds = parseInt(line[7..^1])
+    elif line.startsWith("epsilon="):
+      bot.epsilon = parseFloat(line[8..^1])
+
+# ── Logging ───────────────────────────────────────────────────────────────────
+
+proc logRoundEnd(bot: QBot) =
+  let hitRate = if bot.roundShots > 0:
+    float(bot.roundHits) / float(bot.roundShots)
+  else: 0.0
+  stderr.writeLine("[QBot] round=" & $bot.totalRounds &
+    " shots=" & $bot.roundShots &
+    " hits=" & $bot.roundHits &
+    " hitRate=" & formatFloat(hitRate, ffDecimal, 3) &
+    " eps=" & formatFloat(bot.epsilon, ffDecimal, 3))
+
+  # Top 5 T1 Q-values
+  type QEntry = tuple[v: float; s, a: int]
+  var top: array[5, QEntry]
+  var topN = 0
+  for s in 0 ..< N_STATES:
+    for a in 0 ..< N_AIM_ACT:
+      let v = bot.qt1[s * N_AIM_ACT + a]
+      if topN < 5:
+        top[topN] = (v, s, a)
+        inc topN
+      else:
+        var minIdx = 0
+        for j in 1 ..< 5:
+          if top[j].v < top[minIdx].v: minIdx = j
+        if v > top[minIdx].v:
+          top[minIdx] = (v, s, a)
+  # Sort descending (simple insertion sort over 5 elements)
+  for i in 0 ..< topN - 1:
+    for j in i + 1 ..< topN:
+      if top[j].v > top[i].v:
+        let tmp = top[i]; top[i] = top[j]; top[j] = tmp
+  stderr.writeLine("[QBot] top T1 Q-values:")
+  for i in 0 ..< topN:
+    stderr.writeLine("  s=" & $top[i].s & " aimAct=" & $top[i].a &
+      " offset=" & formatFloat(AIM_OFFSETS[top[i].a], ffDecimal, 1) & "deg" &
+      " q=" & formatFloat(top[i].v, ffDecimal, 4))
+
+# ── Event handlers ────────────────────────────────────────────────────────────
+
+method onScannedBot*(bot: QBot, e: ScannedBotEvent) =
+  bot.enemyX     = e.x
+  bot.enemyY     = e.y
+  bot.enemyDir   = e.direction
+  bot.enemySpeed = e.speed
+  bot.hasContact = true
+  bot.ticksSinceScan = 0
+
+method onBulletFired*(bot: QBot, e: BulletFiredEvent) =
+  # Stamp the pending bullet record with the real ID assigned by the server.
+  bot.stampBulletId(e.bullet.bulletId)
+
+method onBulletHit*(bot: QBot, e: BulletHitBotEvent) =
+  if e.victimId != getMyId():
+    inc bot.roundHits
+    bot.resolveBullet(e.bullet.bulletId, +1.0)
+
+method onBulletHitWall*(bot: QBot, e: BulletHitWallEvent) =
+  if e.bullet.ownerId == getMyId():
+    inc bot.roundMisses
+    bot.resolveBullet(e.bullet.bulletId, -0.2)
+
+method onRoundStarted*(bot: QBot, e: RoundStartedEvent) =
+  setAdjustGunForBodyTurn(true)
+  setAdjustRadarForBodyTurn(true)
+  setAdjustRadarForGunTurn(true)
+  radar_lock.init()
+  bot.hasContact    = false
+  bot.roundHits     = 0
+  bot.roundMisses   = 0
+  bot.roundShots    = 0
+  bot.ticksSinceScan = 0
+  # Invalidate bullet buffer between rounds
+  for i in 0 ..< BULLET_BUF_CAP: bot.bullets[i].valid = false
+  bot.bulletHead = 0
+  setTargetSpeed(0.0)
+  setTurnRate(0.0)
+  bot.epsilon = max(EPS_FLOOR, bot.epsilon * EPS_DECAY)
+
+method onRoundEnded*(bot: QBot, e: RoundEndedEventForBot) =
+  inc bot.totalRounds
+  # Bullets still in flight at round end → treat as misses
+  for i in 0 ..< BULLET_BUF_CAP:
+    if bot.bullets[i].valid:
+      updateQT1(bot, bot.bullets[i].state, bot.bullets[i].aimAct, -0.2)
+      updateQT2(bot, bot.bullets[i].state, bot.bullets[i].fireAct, -0.2)
+      bot.bullets[i].valid = false
+  logRoundEnd(bot)
+
+method onGameStarted*(bot: QBot, e: GameStartedEventForBot) =
+  loadQTable(bot)
+  randomize()
+
+method onGameEnded*(bot: QBot, e: GameEndedEventForBot) =
+  saveQTable(bot)
+  stderr.writeLine("[QBot] game ended rounds=" & $bot.totalRounds &
+    " eps=" & formatFloat(bot.epsilon, ffDecimal, 3))
+
+# ── Main loop ─────────────────────────────────────────────────────────────────
+
+method run*(bot: QBot) =
+  while isRunning():
+    setTargetSpeed(0.0)
+    setTurnRate(0.0)
+
+    inc bot.ticksSinceScan
+    if bot.ticksSinceScan > 5:
+      bot.hasContact = false
+
+    if not bot.hasContact:
+      setRadarTurnRate(45.0)
+      go()
+      continue
+
+    let myX    = getX()
+    let myY    = getY()
+    let gunDir = getGunDirection()
+    let dist   = hypot(myX - bot.enemyX, myY - bot.enemyY)
+    # Absolute bearing from us to enemy
+    let bearingToEnemy = directionTo(myX, myY, bot.enemyX, bot.enemyY)
+    # Gun aim error: how far off our gun is from direct bearing to enemy
+    let gunAimError  = normalizeRelativeAngle(bearingToEnemy - gunDir)
+    # Lateral speed: enemy velocity perpendicular to our line of sight
+    let lateralSpeed = bot.enemySpeed * sin(degToRad(bot.enemyDir - bearingToEnemy))
+
+    let state = discretizeState(gunAimError, dist, lateralSpeed, getGunHeat())
+
+    # Select actions (no per-tick Q-update — updates happen only on bullet resolution)
+    let aimAct  = bot.selectAimAction(state)
+    let fireAct = bot.selectFireAction(state)
+
+    # Compute gun turn: aim at enemy bearing + selected offset, clamped to ±20 deg/tick
+    let aimOffset  = AIM_OFFSETS[aimAct]
+    let desiredDelta = normalizeRelativeAngle(bearingToEnemy + aimOffset - gunDir)
+    setGunTurnRate(desiredDelta.clamp(-20.0, 20.0))
+
+    # Fire if the action calls for it and the gun is ready
+    let firePower = FIRE_POWERS[fireAct]
+    if firePower > 0.0 and getGunHeat() <= 0.0:
+      if setFire(firePower):
+        inc bot.roundShots
+        # Push a pending record; onBulletFired (fired next tick) stamps the real ID
+        bot.pushPendingBullet(state, aimAct, fireAct)
+
+    # Radar lock
+    setRadarTurnRate(radar_lock.doRadar(getRadarDirection(), bearingToEnemy))
+
+    go()
+
+# ── Entry point ───────────────────────────────────────────────────────────────
+
+when isMainModule:
+  createDir(qtablesDir)
+  var bot = QBot(epsilon: EPS_INIT)
+  for i in 0 ..< bot.qt1.len: bot.qt1[i] = Q_INIT
+  for i in 0 ..< bot.qt2.len: bot.qt2[i] = Q_INIT
+  start(bot, botJsonPath)
diff --git a/QBot_garage/QBot.nimble b/QBot_garage/QBot.nimble
new file mode 100644
index 0000000..e8bf9a9
--- /dev/null
+++ b/QBot_garage/QBot.nimble
@@ -0,0 +1,11 @@
+# Package
+version       = "0.1.0"
+author        = "Davide Cappellini"
+description   = "Tabular Q-learning Tank Royale bot — chained action decomposition prototype"
+license       = "MIT"
+bin           = @["QBot"]
+
+# Dependencies
+requires "nim >= 2.0.0"
+# tankroyale_botapi and radar_lock are vendored in-tree (common_libs/) and wired via
+# config.nims --path; no nimble dependency so builds never touch ~/.nimble/pkgs2.
diff --git a/QBot_garage/QBot.sh b/QBot_garage/QBot.sh
new file mode 100755
index 0000000..4a0fe28
--- /dev/null
+++ b/QBot_garage/QBot.sh
@@ -0,0 +1,3 @@
+#!/bin/sh
+cd "$(dirname "$0")"
+exec ./QBot 2>> /tmp/qbot_stderr.log
diff --git a/QBot_garage/config.nims b/QBot_garage/config.nims
new file mode 100644
index 0000000..be6555f
--- /dev/null
+++ b/QBot_garage/config.nims
@@ -0,0 +1 @@
+--path:"../common_libs"
diff --git a/SAC_LSTM_Bot/docs/campaign_dashboard.svg b/SAC_LSTM_Bot/docs/campaign_dashboard.svg
deleted file mode 100644
index 66993f4..0000000
--- a/SAC_LSTM_Bot/docs/campaign_dashboard.svg
+++ /dev/null
@@ -1,355 +0,0 @@
-<svg xmlns="http://www.w3.org/2000/svg" width="1400" height="1850" viewBox="0 0 1400 1850" font-family="sans-serif">
-<rect width="1400" height="1850" fill="white"/>
-<text x="700" y="32" text-anchor="middle" font-size="21" font-weight="bold">SAC-LSTM campaign dashboard - live run (current only)</text>
-<text x="700" y="56" text-anchor="middle" font-size="12" fill="#555">generated 2026-08-24 11:10:31 - auto-reloads every 60 s (open this file in Chrome)</text>
-<text x="70" y="100" font-size="15" font-weight="bold">Test matches - win % vs opponents</text>
-<text x="70" y="115" font-size="11" fill="#555">raw dots = single test matches, thick = rolling-mean-10</text>
-<text x="70" y="133" font-size="12" font-style="italic" fill="#555">What: How often the bot wins against each opponent in test battles.</text>
-<text x="70" y="149" font-size="12" font-style="italic" fill="#555">Axes: X = training progress (match number); Y = win rate, 0-100%.</text>
-<text x="70" y="165" font-size="12" font-style="italic" fill="#555">Better: higher is better.</text>
-<line x1="70" y1="784.0" x2="697" y2="784.0" stroke="#dddddd"/>
-<line x1="70" y1="664.0" x2="697" y2="664.0" stroke="#dddddd"/>
-<line x1="70" y1="544.0" x2="697" y2="544.0" stroke="#dddddd"/>
-<line x1="70" y1="424.0" x2="697" y2="424.0" stroke="#dddddd"/>
-<line x1="70" y1="304.0" x2="697" y2="304.0" stroke="#dddddd"/>
-<line x1="70" y1="184.0" x2="697" y2="184.0" stroke="#dddddd"/>
-<line x1="70" y1="784" x2="697" y2="784" stroke="black"/>
-<line x1="70" y1="784" x2="70" y2="184" stroke="black"/>
-<line x1="70.0" y1="784" x2="70.0" y2="788" stroke="black"/>
-<text x="70.0" y="801" text-anchor="middle" font-size="11">1</text>
-<line x1="226.8" y1="784" x2="226.8" y2="788" stroke="black"/>
-<text x="226.8" y="801" text-anchor="middle" font-size="11">2</text>
-<line x1="383.5" y1="784" x2="383.5" y2="788" stroke="black"/>
-<text x="383.5" y="801" text-anchor="middle" font-size="11">3</text>
-<line x1="540.2" y1="784" x2="540.2" y2="788" stroke="black"/>
-<text x="540.2" y="801" text-anchor="middle" font-size="11">4</text>
-<line x1="697.0" y1="784" x2="697.0" y2="788" stroke="black"/>
-<text x="697.0" y="801" text-anchor="middle" font-size="11">5</text>
-<line x1="66" y1="784.0" x2="70" y2="784.0" stroke="black"/>
-<text x="63" y="788.0" text-anchor="end" font-size="11">0</text>
-<line x1="66" y1="664.0" x2="70" y2="664.0" stroke="black"/>
-<text x="63" y="668.0" text-anchor="end" font-size="11">20</text>
-<line x1="66" y1="544.0" x2="70" y2="544.0" stroke="black"/>
-<text x="63" y="548.0" text-anchor="end" font-size="11">40</text>
-<line x1="66" y1="424.0" x2="70" y2="424.0" stroke="black"/>
-<text x="63" y="428.0" text-anchor="end" font-size="11">60</text>
-<line x1="66" y1="304.0" x2="70" y2="304.0" stroke="black"/>
-<text x="63" y="308.0" text-anchor="end" font-size="11">80</text>
-<line x1="66" y1="184.0" x2="70" y2="184.0" stroke="black"/>
-<text x="63" y="188.0" text-anchor="end" font-size="11">100</text>
-<text x="383" y="817" text-anchor="middle" font-size="12">test match number (each opponent)</text>
-<text x="16" y="484" text-anchor="middle" font-size="12" transform="rotate(-90 16 484)">win rate (%)</text>
-<circle cx="70.0" cy="784.0" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="226.8" cy="784.0" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="383.5" cy="784.0" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="540.2" cy="724.0" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="697.0" cy="784.0" r="2" fill="#d62728" opacity="0.25"/>
-<polyline points="70.0,784.0 226.8,784.0 383.5,784.0 540.2,769.0 697.0,772.0" fill="none" stroke="#d62728" stroke-width="3.5" opacity="1.0"/>
-<circle cx="70.0" cy="664.0" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="226.8" cy="784.0" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="383.5" cy="724.0" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="540.2" cy="184.0" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="697.0" cy="784.0" r="2" fill="#1f77b4" opacity="0.25"/>
-<polyline points="70.0,664.0 226.8,724.0 383.5,724.0 540.2,589.0 697.0,628.0" fill="none" stroke="#1f77b4" stroke-width="3.5" opacity="1.0"/>
-<circle cx="70.0" cy="784.0" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="226.8" cy="784.0" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="383.5" cy="784.0" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="540.2" cy="784.0" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="697.0" cy="784.0" r="2" fill="#2ca02c" opacity="0.25"/>
-<polyline points="70.0,784.0 226.8,784.0 383.5,784.0 540.2,784.0 697.0,784.0" fill="none" stroke="#2ca02c" stroke-width="3.5" opacity="1.0"/>
-<line x1="82" y1="836" x2="110" y2="836" stroke="#d62728" stroke-width="3"/>
-<text x="116" y="840" font-size="12">Corners - 5 evals</text>
-<line x1="82" y1="854" x2="110" y2="854" stroke="#1f77b4" stroke-width="3"/>
-<text x="116" y="858" font-size="12">Crazy - 5 evals</text>
-<line x1="82" y1="872" x2="110" y2="872" stroke="#2ca02c" stroke-width="3"/>
-<text x="116" y="876" font-size="12">Target - 5 evals</text>
-<text x="747" y="100" font-size="15" font-weight="bold">Training losses (log) &amp; alpha (linear)</text>
-<text x="747" y="115" font-size="11" fill="#555">training_metrics.jsonl - big early spikes are normal</text>
-<text x="747" y="133" font-size="12" font-style="italic" fill="#555">What: How well the brain is learning: losses should fall; alpha sets explore/exploit.</text>
-<text x="747" y="149" font-size="12" font-style="italic" fill="#555">Axes: X = training progress; left Y (log) = losses; right Y (linear) = alpha.</text>
-<text x="747" y="165" font-size="12" font-style="italic" fill="#555">Better: lower losses = better; alpha falls over time as the bot gets confident.</text>
-<line x1="747" y1="784.0" x2="1329" y2="784.0" stroke="#dddddd"/>
-<line x1="747" y1="634.0" x2="1329" y2="634.0" stroke="#dddddd"/>
-<line x1="747" y1="484.0" x2="1329" y2="484.0" stroke="#dddddd"/>
-<line x1="747" y1="334.0" x2="1329" y2="334.0" stroke="#dddddd"/>
-<line x1="747" y1="184.0" x2="1329" y2="184.0" stroke="#dddddd"/>
-<line x1="747" y1="784" x2="1329" y2="784" stroke="black"/>
-<line x1="747" y1="784" x2="747" y2="184" stroke="black"/>
-<line x1="747.0" y1="784" x2="747.0" y2="788" stroke="black"/>
-<text x="747.0" y="801" text-anchor="middle" font-size="11">1</text>
-<line x1="892.5" y1="784" x2="892.5" y2="788" stroke="black"/>
-<text x="892.5" y="801" text-anchor="middle" font-size="11">24</text>
-<line x1="1038.0" y1="784" x2="1038.0" y2="788" stroke="black"/>
-<text x="1038.0" y="801" text-anchor="middle" font-size="11">46</text>
-<line x1="1183.5" y1="784" x2="1183.5" y2="788" stroke="black"/>
-<text x="1183.5" y="801" text-anchor="middle" font-size="11">69</text>
-<line x1="1329.0" y1="784" x2="1329.0" y2="788" stroke="black"/>
-<text x="1329.0" y="801" text-anchor="middle" font-size="11">92</text>
-<line x1="743" y1="784.0" x2="747" y2="784.0" stroke="black"/>
-<text x="740" y="788.0" text-anchor="end" font-size="11">0.01</text>
-<line x1="743" y1="634.0" x2="747" y2="634.0" stroke="black"/>
-<text x="740" y="638.0" text-anchor="end" font-size="11">0.1</text>
-<line x1="743" y1="484.0" x2="747" y2="484.0" stroke="black"/>
-<text x="740" y="488.0" text-anchor="end" font-size="11">1</text>
-<line x1="743" y1="334.0" x2="747" y2="334.0" stroke="black"/>
-<text x="740" y="338.0" text-anchor="end" font-size="11">10</text>
-<line x1="743" y1="184.0" x2="747" y2="184.0" stroke="black"/>
-<text x="740" y="188.0" text-anchor="end" font-size="11">100</text>
-<text x="1038" y="817" text-anchor="middle" font-size="12">metric line number</text>
-<text x="16" y="484" text-anchor="middle" font-size="12" transform="rotate(-90 16 484)">loss (log)</text>
-<polyline points="747.0,345.9 753.4,317.6 759.8,307.9 766.2,277.5 772.6,271.6 779.0,450.8 785.4,488.4 791.8,371.1 798.2,357.6 804.6,346.6 811.0,306.7 817.4,235.4 823.7,222.6 830.1,233.7 836.5,255.2 842.9,317.6 849.3,391.8 855.7,689.2 862.1,570.4 868.5,438.2 874.9,489.9 881.3,415.1 887.7,421.9 894.1,394.0 900.5,315.4 906.9,403.4 913.3,395.4 919.7,347.3 926.1,384.4 932.5,403.4 938.9,359.6 945.3,281.5 951.7,479.1 958.1,530.4 964.5,598.4 970.8,352.1 977.2,326.5 983.6,473.0 990.0,506.0 996.4,331.4 1002.8,746.1 1009.2,663.4 1015.6,420.9 1022.0,617.1 1028.4,361.8 1034.8,514.9 1041.2,377.6 1047.6,387.0 1054.0,361.7 1060.4,357.1 1066.8,327.5 1073.2,392.5 1079.6,519.5 1086.0,436.0 1092.4,399.4 1098.8,385.9 1105.2,493.5 1111.5,483.4 1117.9,456.3 1124.3,575.1 1130.7,672.1 1137.1,312.5 1143.5,463.3 1149.9,745.8 1156.3,554.2 1162.7,381.5 1169.1,301.5 1175.5,292.7 1181.9,241.6 1188.3,212.1 1194.7,234.2 1201.1,302.6 1207.5,437.6 1213.9,427.3 1220.3,503.0 1226.7,325.5 1233.1,327.0 1239.5,312.9 1245.9,448.6 1252.3,311.7 1258.6,417.9 1265.0,596.5 1271.4,396.9 1277.8,422.4 1284.2,371.0 1290.6,343.4 1297.0,374.0 1303.4,447.6 1309.8,640.0 1316.2,449.0 1322.6,514.0 1329.0,403.2" fill="none" stroke="#1f77b4" stroke-width="1.8" opacity="1.0"/>
-<polyline points="747.0,412.9 753.4,371.9 759.8,348.5 766.2,325.3 772.6,308.2 779.0,299.2 785.4,300.5 791.8,305.4 798.2,334.7 804.6,356.4 811.0,442.9 817.4,366.9 823.7,341.7 830.1,328.4 836.5,318.8 842.9,310.5 849.3,306.0 855.7,315.8 862.1,326.0 868.5,337.4 874.9,359.3 881.3,391.5 887.7,399.9 894.1,502.0 900.5,464.5 906.9,564.9 913.3,454.5 919.7,393.2 926.1,380.2 932.5,375.3 938.9,353.3 945.3,352.2 951.7,342.8 958.1,351.9 964.5,358.7 970.8,379.5 977.2,402.3 983.6,418.8 990.0,438.0 996.4,431.7 1002.8,426.7 1009.2,406.5 1015.6,442.8 1022.0,458.7 1028.4,415.4 1034.8,407.0 1041.2,581.8 1047.6,551.6 1054.0,459.8 1060.4,385.5 1066.8,362.4 1073.2,365.9 1079.6,376.3 1086.0,404.8 1092.4,450.3 1098.8,486.9 1105.2,443.9 1111.5,409.7 1117.9,397.8 1124.3,393.6 1130.7,374.2 1137.1,374.4 1143.5,379.7 1149.9,396.1 1156.3,477.0 1162.7,386.3 1169.1,351.8 1175.5,344.7 1181.9,319.0 1188.3,303.7 1194.7,292.7 1201.1,291.1 1207.5,290.5 1213.9,293.6 1220.3,304.0 1226.7,333.0 1233.1,342.4 1239.5,379.0 1245.9,388.7 1252.3,338.6 1258.6,333.8 1265.0,322.4 1271.4,337.6 1277.8,334.1 1284.2,317.2 1290.6,288.4 1297.0,281.6 1303.4,274.9 1309.8,274.5 1316.2,285.3 1322.6,294.5 1329.0,304.4" fill="none" stroke="#ff7f0e" stroke-width="1.8" opacity="1.0"/>
-<line x1="747" y1="784.0" x2="1329" y2="784.0" stroke="#e9dcf5"/>
-<line x1="747" y1="634.0" x2="1329" y2="634.0" stroke="#e9dcf5"/>
-<line x1="747" y1="484.0" x2="1329" y2="484.0" stroke="#e9dcf5"/>
-<line x1="747" y1="334.0" x2="1329" y2="334.0" stroke="#e9dcf5"/>
-<line x1="747" y1="184.0" x2="1329" y2="184.0" stroke="#e9dcf5"/>
-<line x1="1329" y1="784" x2="1329" y2="184" stroke="#9467bd"/>
-<line x1="1329" y1="784.0" x2="1333" y2="784.0" stroke="#9467bd"/>
-<text x="1336" y="788.0" font-size="11" fill="#9467bd">0</text>
-<line x1="1329" y1="634.0" x2="1333" y2="634.0" stroke="#9467bd"/>
-<text x="1336" y="638.0" font-size="11" fill="#9467bd">0.25</text>
-<line x1="1329" y1="484.0" x2="1333" y2="484.0" stroke="#9467bd"/>
-<text x="1336" y="488.0" font-size="11" fill="#9467bd">0.5</text>
-<line x1="1329" y1="334.0" x2="1333" y2="334.0" stroke="#9467bd"/>
-<text x="1336" y="338.0" font-size="11" fill="#9467bd">0.75</text>
-<line x1="1329" y1="184.0" x2="1333" y2="184.0" stroke="#9467bd"/>
-<text x="1336" y="188.0" font-size="11" fill="#9467bd">1</text>
-<text x="1346" y="484" text-anchor="middle" font-size="12" fill="#9467bd" transform="rotate(90 1346 484)">alpha</text>
-<polyline points="747.0,184.2 753.4,184.9 759.8,185.6 766.2,186.3 772.6,187.1 779.0,187.8 785.4,188.1 791.8,188.5 798.2,189.2 804.6,189.6 811.0,190.3 817.4,190.9 823.7,190.1 830.1,189.4 836.5,188.7 842.9,188.0 849.3,187.3 855.7,186.6 862.1,185.9 868.5,185.2 874.9,185.7 881.3,186.4 887.7,187.2 894.1,187.9 900.5,188.6 906.9,190.0 913.3,190.7 919.7,191.4 926.1,192.1 932.5,192.5 938.9,192.4 945.3,192.1 951.7,191.7 958.1,191.6 964.5,192.4 970.8,193.1 977.2,193.4 983.6,194.1 990.0,194.5 996.4,195.0 1002.8,195.4 1009.2,196.1 1015.6,196.4 1022.0,197.1 1028.4,197.8 1034.8,198.2 1041.2,198.9 1047.6,199.2 1054.0,199.9 1060.4,200.6 1066.8,200.6 1073.2,200.3 1079.6,200.8 1086.0,201.2 1092.4,201.9 1098.8,202.6 1105.2,203.3 1111.5,204.0 1117.9,204.7 1124.3,205.4 1130.7,206.1 1137.1,206.4 1143.5,207.1 1149.9,207.8 1156.3,208.7 1162.7,209.8 1169.1,209.8 1175.5,209.5 1181.9,208.8 1188.3,208.3 1194.7,207.3 1201.1,206.6 1207.5,205.9 1213.9,205.5 1220.3,204.8 1226.7,204.6 1233.1,205.0 1239.5,205.6 1245.9,206.9 1252.3,207.3 1258.6,206.5 1265.0,205.7 1271.4,205.5 1277.8,205.5 1284.2,204.9 1290.6,203.3 1297.0,202.6 1303.4,201.9 1309.8,201.5 1316.2,200.3 1322.6,199.5 1329.0,198.8" fill="none" stroke="#9467bd" stroke-width="1.8" opacity="1.0"/>
-<line x1="759" y1="836" x2="787" y2="836" stroke="#1f77b4" stroke-width="3"/>
-<text x="793" y="840" font-size="12">critic_loss</text>
-<line x1="759" y1="854" x2="787" y2="854" stroke="#ff7f0e" stroke-width="3"/>
-<text x="793" y="858" font-size="12">|actor_loss|</text>
-<line x1="759" y1="872" x2="787" y2="872" stroke="#9467bd" stroke-width="3"/>
-<text x="793" y="876" font-size="12">alpha</text>
-<text x="70" y="940" font-size="15" font-weight="bold">Throughput - games per hour</text>
-<text x="70" y="955" font-size="11" fill="#555">method: training_metrics.jsonl 'epoch' deltas; 1 row = one 10-game chunk</text>
-<text x="70" y="973" font-size="12" font-style="italic" fill="#555">What: How many games per hour the bot trains (its learning speed).</text>
-<text x="70" y="989" font-size="12" font-style="italic" fill="#555">Axes: X = training progress (10-game chunks); Y = games per hour.</text>
-<text x="70" y="1005" font-size="12" font-style="italic" fill="#555">Better: higher = faster learning; a steady line beats a spiky one.</text>
-<line x1="70" y1="1494.0" x2="697" y2="1494.0" stroke="#dddddd"/>
-<line x1="70" y1="1376.5" x2="697" y2="1376.5" stroke="#dddddd"/>
-<line x1="70" y1="1259.0" x2="697" y2="1259.0" stroke="#dddddd"/>
-<line x1="70" y1="1141.5" x2="697" y2="1141.5" stroke="#dddddd"/>
-<line x1="70" y1="1024.0" x2="697" y2="1024.0" stroke="#dddddd"/>
-<line x1="70" y1="1494" x2="697" y2="1494" stroke="black"/>
-<line x1="70" y1="1494" x2="70" y2="1024" stroke="black"/>
-<line x1="125.7" y1="1494" x2="125.7" y2="1498" stroke="black"/>
-<text x="125.7" y="1511" text-anchor="middle" font-size="11">9</text>
-<line x1="188.4" y1="1494" x2="188.4" y2="1498" stroke="black"/>
-<text x="188.4" y="1511" text-anchor="middle" font-size="11">18</text>
-<line x1="251.1" y1="1494" x2="251.1" y2="1498" stroke="black"/>
-<text x="251.1" y="1511" text-anchor="middle" font-size="11">27</text>
-<line x1="313.8" y1="1494" x2="313.8" y2="1498" stroke="black"/>
-<text x="313.8" y="1511" text-anchor="middle" font-size="11">36</text>
-<line x1="376.5" y1="1494" x2="376.5" y2="1498" stroke="black"/>
-<text x="376.5" y="1511" text-anchor="middle" font-size="11">45</text>
-<line x1="439.2" y1="1494" x2="439.2" y2="1498" stroke="black"/>
-<text x="439.2" y="1511" text-anchor="middle" font-size="11">54</text>
-<line x1="501.9" y1="1494" x2="501.9" y2="1498" stroke="black"/>
-<text x="501.9" y="1511" text-anchor="middle" font-size="11">63</text>
-<line x1="564.6" y1="1494" x2="564.6" y2="1498" stroke="black"/>
-<text x="564.6" y="1511" text-anchor="middle" font-size="11">72</text>
-<line x1="627.3" y1="1494" x2="627.3" y2="1498" stroke="black"/>
-<text x="627.3" y="1511" text-anchor="middle" font-size="11">81</text>
-<line x1="690.0" y1="1494" x2="690.0" y2="1498" stroke="black"/>
-<text x="690.0" y="1511" text-anchor="middle" font-size="11">90</text>
-<line x1="66" y1="1494.0" x2="70" y2="1494.0" stroke="black"/>
-<text x="63" y="1498.0" text-anchor="end" font-size="11">0</text>
-<line x1="66" y1="1376.5" x2="70" y2="1376.5" stroke="black"/>
-<text x="63" y="1380.5" text-anchor="end" font-size="11">1653</text>
-<line x1="66" y1="1259.0" x2="70" y2="1259.0" stroke="black"/>
-<text x="63" y="1263.0" text-anchor="end" font-size="11">3305</text>
-<line x1="66" y1="1141.5" x2="70" y2="1141.5" stroke="black"/>
-<text x="63" y="1145.5" text-anchor="end" font-size="11">4958</text>
-<line x1="66" y1="1024.0" x2="70" y2="1024.0" stroke="black"/>
-<text x="63" y="1028.0" text-anchor="end" font-size="11">6611</text>
-<text x="383" y="1527" text-anchor="middle" font-size="12">chunk interval (10-game chunks)</text>
-<text x="16" y="1259" text-anchor="middle" font-size="12" transform="rotate(-90 16 1259)">games / hour</text>
-<circle cx="70.0" cy="1225.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="77.0" cy="1479.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="83.9" cy="1214.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="90.9" cy="1470.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="97.9" cy="1219.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="104.8" cy="1465.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="111.8" cy="1066.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="118.8" cy="1465.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="125.7" cy="1087.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="132.7" cy="1460.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="139.7" cy="1442.2" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="146.6" cy="1460.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="153.6" cy="1196.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="160.6" cy="1465.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="167.5" cy="1242.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="174.5" cy="1468.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="181.5" cy="1224.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="188.4" cy="1466.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="195.4" cy="1215.2" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="202.4" cy="1466.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="209.3" cy="1237.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="216.3" cy="1471.2" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="223.3" cy="1228.2" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="230.2" cy="1482.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="237.2" cy="1441.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="244.2" cy="1481.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="251.1" cy="1228.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="258.1" cy="1481.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="265.1" cy="1141.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="272.0" cy="1481.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="279.0" cy="1141.2" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="286.0" cy="1480.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="292.9" cy="1164.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="299.9" cy="1483.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="306.9" cy="1222.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="313.8" cy="1474.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="320.8" cy="1196.9" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="327.8" cy="1474.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="334.7" cy="1176.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="341.7" cy="1477.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="348.7" cy="1206.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="355.6" cy="1475.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="362.6" cy="1197.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="369.6" cy="1474.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="376.5" cy="1109.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="383.5" cy="1474.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="390.5" cy="1091.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="397.4" cy="1477.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="404.4" cy="1214.9" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="411.4" cy="1474.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="418.3" cy="1106.2" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="425.3" cy="1474.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="432.3" cy="1082.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="439.2" cy="1474.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="446.2" cy="1206.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="453.2" cy="1474.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="460.1" cy="1226.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="467.1" cy="1474.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="474.1" cy="1217.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="481.0" cy="1479.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="488.0" cy="1087.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="495.0" cy="1474.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="501.9" cy="1209.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="508.9" cy="1475.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="515.9" cy="1444.9" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="522.8" cy="1476.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="529.8" cy="1077.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="536.8" cy="1476.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="543.7" cy="1151.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="550.7" cy="1483.9" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="557.7" cy="1236.2" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="564.6" cy="1477.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="571.6" cy="1096.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="578.6" cy="1474.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="585.5" cy="1241.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="592.5" cy="1474.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="599.5" cy="1245.3" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="606.4" cy="1481.9" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="613.4" cy="1260.9" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="620.4" cy="1474.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="627.3" cy="1242.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="634.3" cy="1491.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="641.3" cy="1088.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="648.2" cy="1478.0" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="655.2" cy="1449.8" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="662.2" cy="1477.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="669.1" cy="1227.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="676.1" cy="1471.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="683.1" cy="1441.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="690.0" cy="1479.1" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<circle cx="697.0" cy="1230.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
-<polyline points="77.0,1321.9 111.8,1309.2 139.7,1361.4 174.5,1367.9 202.4,1372.2 237.2,1362.7 272.0,1298.5 299.9,1359.9 334.7,1292.6 362.6,1346.5 397.4,1268.8 425.3,1348.4 460.1,1312.9 488.0,1368.7 522.8,1336.9 557.7,1322.2 585.5,1365.7 620.4,1324.2 648.2,1408.2 683.1,1405.5" fill="none" stroke="#2ca02c" stroke-width="3.5" opacity="1.0"/>
-<text x="747" y="940" font-size="15" font-weight="bold">Max score per eval cycle</text>
-<text x="747" y="955" font-size="11" fill="#555">campaign_v4_stdout.log - max of the 10 deterministic round scores per eval</text>
-<text x="747" y="973" font-size="12" font-style="italic" fill="#555">What: Best single-round score the bot managed in each test cycle.</text>
-<text x="747" y="989" font-size="12" font-style="italic" fill="#555">Axes: X = eval cycle number; Y = best score achieved.</text>
-<text x="747" y="1005" font-size="12" font-style="italic" fill="#555">Better: higher is better; a rising trend means the bot is improving.</text>
-<line x1="747" y1="1494.0" x2="1375" y2="1494.0" stroke="#dddddd"/>
-<line x1="747" y1="1376.5" x2="1375" y2="1376.5" stroke="#dddddd"/>
-<line x1="747" y1="1259.0" x2="1375" y2="1259.0" stroke="#dddddd"/>
-<line x1="747" y1="1141.5" x2="1375" y2="1141.5" stroke="#dddddd"/>
-<line x1="747" y1="1024.0" x2="1375" y2="1024.0" stroke="#dddddd"/>
-<line x1="747" y1="1494" x2="1375" y2="1494" stroke="black"/>
-<line x1="747" y1="1494" x2="747" y2="1024" stroke="black"/>
-<line x1="747.0" y1="1494" x2="747.0" y2="1498" stroke="black"/>
-<text x="747.0" y="1511" text-anchor="middle" font-size="11">1</text>
-<line x1="904.0" y1="1494" x2="904.0" y2="1498" stroke="black"/>
-<text x="904.0" y="1511" text-anchor="middle" font-size="11">2</text>
-<line x1="1061.0" y1="1494" x2="1061.0" y2="1498" stroke="black"/>
-<text x="1061.0" y="1511" text-anchor="middle" font-size="11">3</text>
-<line x1="1218.0" y1="1494" x2="1218.0" y2="1498" stroke="black"/>
-<text x="1218.0" y="1511" text-anchor="middle" font-size="11">4</text>
-<line x1="1375.0" y1="1494" x2="1375.0" y2="1498" stroke="black"/>
-<text x="1375.0" y="1511" text-anchor="middle" font-size="11">5</text>
-<line x1="743" y1="1494.0" x2="747" y2="1494.0" stroke="black"/>
-<text x="740" y="1498.0" text-anchor="end" font-size="11">0</text>
-<line x1="743" y1="1376.5" x2="747" y2="1376.5" stroke="black"/>
-<text x="740" y="1380.5" text-anchor="end" font-size="11">46</text>
-<line x1="743" y1="1259.0" x2="747" y2="1259.0" stroke="black"/>
-<text x="740" y="1263.0" text-anchor="end" font-size="11">92</text>
-<line x1="743" y1="1141.5" x2="747" y2="1141.5" stroke="black"/>
-<text x="740" y="1145.5" text-anchor="end" font-size="11">138</text>
-<line x1="743" y1="1024.0" x2="747" y2="1024.0" stroke="black"/>
-<text x="740" y="1028.0" text-anchor="end" font-size="11">184</text>
-<text x="1061" y="1527" text-anchor="middle" font-size="12">eval cycle number</text>
-<text x="16" y="1259" text-anchor="middle" font-size="12" transform="rotate(-90 16 1259)">best single-round score</text>
-<circle cx="747.0" cy="1340.5" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="904.0" cy="1481.2" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="1061.0" cy="1207.4" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="1218.0" cy="1340.5" r="2" fill="#d62728" opacity="0.25"/>
-<circle cx="1375.0" cy="1066.7" r="2" fill="#d62728" opacity="0.25"/>
-<polyline points="747.0,1340.5 904.0,1481.2 1061.0,1207.4 1218.0,1340.5 1375.0,1066.7" fill="none" stroke="#d62728" stroke-width="2.5" opacity="1.0"/>
-<circle cx="747.0" cy="1325.1" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="904.0" cy="1256.1" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="1061.0" cy="1320.0" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="1218.0" cy="1312.3" r="2" fill="#1f77b4" opacity="0.25"/>
-<circle cx="1375.0" cy="1463.3" r="2" fill="#1f77b4" opacity="0.25"/>
-<polyline points="747.0,1325.1 904.0,1256.1 1061.0,1320.0 1218.0,1312.3 1375.0,1463.3" fill="none" stroke="#1f77b4" stroke-width="2.5" opacity="1.0"/>
-<circle cx="747.0" cy="1353.3" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="904.0" cy="1312.3" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="1061.0" cy="1207.4" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="1218.0" cy="1376.3" r="2" fill="#2ca02c" opacity="0.25"/>
-<circle cx="1375.0" cy="1309.8" r="2" fill="#2ca02c" opacity="0.25"/>
-<polyline points="747.0,1353.3 904.0,1312.3 1061.0,1207.4 1218.0,1376.3 1375.0,1309.8" fill="none" stroke="#2ca02c" stroke-width="2.5" opacity="1.0"/>
-<polyline points="747.0,1325.1 904.0,1256.1 1061.0,1207.4 1218.0,1312.3 1375.0,1066.7" fill="none" stroke="#555555" stroke-width="2.5" opacity="1.0" stroke-dasharray="6 4"/>
-<line x1="759" y1="1546" x2="787" y2="1546" stroke="#d62728" stroke-width="3"/>
-<text x="793" y="1550" font-size="12">Corners - 5 evals</text>
-<line x1="759" y1="1564" x2="787" y2="1564" stroke="#1f77b4" stroke-width="3"/>
-<text x="793" y="1568" font-size="12">Crazy - 5 evals</text>
-<line x1="759" y1="1582" x2="787" y2="1582" stroke="#2ca02c" stroke-width="3"/>
-<text x="793" y="1586" font-size="12">Target - 5 evals</text>
-<line x1="759" y1="1600" x2="787" y2="1600" stroke="#555555" stroke-width="3"/>
-<text x="793" y="1604" font-size="12">combined max</text>
-<rect x="0" y="1645" width="1400" height="205" fill="#f2f2f2"/>
-<text x="16" y="1666" font-size="15" font-weight="bold">Reading the signs</text>
-<text x="1384" y="1665" text-anchor="end" font-size="11" fill="#666">Regenerate anytime: python3 tools/plot_progress.py</text>
-<text x="16" y="1688" font-size="13" font-weight="bold">Healthy patterns ✅</text>
-<text x="16" y="1706" font-size="12">• Both critic and actor losses trending down over time</text>
-<text x="16" y="1720" font-size="12">• Alpha decaying slowly from ~1.0,</text>
-<text x="16" y="1734" font-size="12">   then plateauing — this is normal</text>
-<text x="16" y="1748" font-size="12">• Win rates appearing and increasing in the eval panel</text>
-<text x="16" y="1762" font-size="12">• Max scores rising in the max-score panel</text>
-<text x="16" y="1776" font-size="12">• Alpha plateauing is NOT a problem — it means</text>
-<text x="16" y="1790" font-size="12">   exploration level is stable</text>
-<text x="500" y="1688" font-size="13" font-weight="bold">Warning signs ⚠️</text>
-<text x="500" y="1706" font-size="12">• Losses exploding (suddenly jumping to</text>
-<text x="500" y="1720" font-size="12">   millions or billions)</text>
-<text x="500" y="1734" font-size="12">• Win rates staying at 0% for a long time</text>
-<text x="500" y="1748" font-size="12">   after the first ~20 eval cycles</text>
-<text x="500" y="1762" font-size="12">• Alpha reaching 0 — bot stops exploring entirely</text>
-<text x="500" y="1776" font-size="12">   (gets stuck)</text>
-<text x="500" y="1790" font-size="12">• Max scores flatlining (no improvement</text>
-<text x="500" y="1804" font-size="12">   over many eval cycles)</text>
-<text x="500" y="1818" font-size="12">• Any single loss value above 1e6</text>
-<text x="985" y="1688" font-size="13" font-weight="bold">What each metric means (brief)</text>
-<text x="985" y="1706" font-size="12">• Critic loss: how wrong the bot's value</text>
-<text x="985" y="1720" font-size="12">   estimates are — should go down</text>
-<text x="985" y="1734" font-size="12">• Actor loss: how well the bot's action policy</text>
-<text x="985" y="1748" font-size="12">   is doing — should go down overall</text>
-<text x="985" y="1762" font-size="12">   (some bumps are normal)</text>
-<text x="985" y="1776" font-size="12">• Alpha: exploration-exploitation tradeoff — starts</text>
-<text x="985" y="1790" font-size="12">   high, settles at a positive value (NOT zero)</text>
-<text x="985" y="1804" font-size="12">• Max score: best score achieved per eval cycle —</text>
-<text x="985" y="1818" font-size="12">   rising trend = learning</text>
-<script type="text/javascript"><![CDATA[ setTimeout(function(){ location.reload(); }, 60000); ]]></script>
-</svg>
diff --git a/SAC_LSTM_Bot_garage/SAC_LSTM_Bot b/SAC_LSTM_Bot_garage/SAC_LSTM_Bot
new file mode 100755
index 0000000..bb118a0
Binary files /dev/null and b/SAC_LSTM_Bot_garage/SAC_LSTM_Bot differ
diff --git a/SAC_LSTM_Bot/SAC_LSTM_Bot.json b/SAC_LSTM_Bot_garage/SAC_LSTM_Bot.json
similarity index 100%
rename from SAC_LSTM_Bot/SAC_LSTM_Bot.json
rename to SAC_LSTM_Bot_garage/SAC_LSTM_Bot.json
diff --git a/SAC_LSTM_Bot/SAC_LSTM_Bot.nimble b/SAC_LSTM_Bot_garage/SAC_LSTM_Bot.nimble
similarity index 80%
rename from SAC_LSTM_Bot/SAC_LSTM_Bot.nimble
rename to SAC_LSTM_Bot_garage/SAC_LSTM_Bot.nimble
index 6a42dc9..21741da 100644
--- a/SAC_LSTM_Bot/SAC_LSTM_Bot.nimble
+++ b/SAC_LSTM_Bot_garage/SAC_LSTM_Bot.nimble
@@ -8,6 +8,6 @@ bin           = @["SAC_LSTM_Bot"]
 
 # Dependencies
 requires "nim >= 2.0.0"
-# tankroyale_botapi is vendored in-tree (libs/tankroyale_botapi) and wired via
+# tankroyale_botapi is vendored in-tree (common_libs/tankroyale_botapi) and wired via
 # config.nims --path; no nimble dependency so builds never touch ~/.nimble/pkgs2.
 requires "arraymancer >= 0.7.0"
diff --git a/SAC_LSTM_Bot/SAC_LSTM_Bot.sh b/SAC_LSTM_Bot_garage/SAC_LSTM_Bot.sh
similarity index 100%
rename from SAC_LSTM_Bot/SAC_LSTM_Bot.sh
rename to SAC_LSTM_Bot_garage/SAC_LSTM_Bot.sh
diff --git a/SAC_LSTM_Bot_garage/campaign_v4_stdout.log.pre-merge b/SAC_LSTM_Bot_garage/campaign_v4_stdout.log.pre-merge
new file mode 100644
index 0000000..06ed4e9
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/campaign_v4_stdout.log.pre-merge
@@ -0,0 +1,173 @@
+=== SAC_LSTM_Bot training harness ===
+Opponents: Corners:3,Crazy:2,RamFire:2,Target:1,SacTwin:1 | budget: 100 rounds in chunks of 10
+Eval: every 2 chunks, 10 rounds vs [Corners,Crazy,Target], MA-5 composite best-gating
+Weights: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/weights/sac_latest.zip
+   Building SAC_LSTM_Bot/SAC_LSTM_Bot using c backend
+      Info: compiling nim package using /nix/store/pq494yyknvhrvkzna9zgcqvma8fq3617-x86_64-unknown-linux-gnu-nim-wrapper-2.2.4/bin/x86_64-unknown-linux-gnu-nim
+=== Chunk 1/10: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:684 score:0 win:false
+Round 2/10 — ticks:631 score:0 win:false
+Round 3/10 — ticks:738 score:0 win:false
+Round 4/10 — ticks:920 score:31 win:false
+Round 5/10 — ticks:682 score:0 win:false
+Round 6/10 — ticks:650 score:0 win:false
+Round 7/10 — ticks:988 score:0 win:false
+Round 8/10 — ticks:365 score:2 win:false
+Round 9/10 — ticks:481 score:32 win:false
+Round 10/10 — ticks:703 score:16 win:false
+Counter check passed: 2291 == expected 2291
+=== Chunk 2/10: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:667 score:17 win:false
+Round 2/10 — ticks:709 score:0 win:false
+Round 3/10 — ticks:856 score:49 win:false
+Round 4/10 — ticks:794 score:2 win:false
+Round 5/10 — ticks:791 score:1 win:false
+Round 6/10 — ticks:936 score:19 win:false
+Round 7/10 — ticks:625 score:3 win:false
+Round 8/10 — ticks:977 score:21 win:false
+Round 9/10 — ticks:658 score:1 win:false
+Round 10/10 — ticks:825 score:1 win:false
+Counter check passed: 2301 == expected 2301
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:3657 score:0 win:false
+Round 2/10 — ticks:3007 score:60 win:false
+Round 3/10 — ticks:3786 score:0 win:false
+Round 4/10 — ticks:1312 score:5 win:false
+Round 5/10 — ticks:1310 score:2 win:false
+Round 6/10 — ticks:1668 score:0 win:false
+Round 7/10 — ticks:1290 score:0 win:false
+Round 8/10 — ticks:1537 score:1 win:false
+Round 9/10 — ticks:1201 score:0 win:false
+Round 10/10 — ticks:4406 score:60 win:false
+Counter check passed: 2311 == expected 2311
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1898 score:1 win:false
+Round 2/10 — ticks:1327 score:11 win:false
+Round 3/10 — ticks:1669 score:5 win:false
+Round 4/10 — ticks:945 score:61 win:false
+Round 5/10 — ticks:1775 score:67 win:false
+Round 6/10 — ticks:1452 score:5 win:false
+Round 7/10 — ticks:1002 score:0 win:false
+Round 8/10 — ticks:2124 score:64 win:false
+Round 9/10 — ticks:1460 score:1 win:false
+Round 10/10 — ticks:935 score:61 win:false
+Counter check passed: 2321 == expected 2321
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:992 score:1 win:false
+Round 2/10 — ticks:1442 score:60 win:true
+Round 3/10 — ticks:1002 score:17 win:false
+Round 4/10 — ticks:1026 score:1 win:false
+Round 5/10 — ticks:992 score:7 win:false
+Round 6/10 — ticks:1422 score:60 win:false
+Round 7/10 — ticks:964 score:9 win:false
+Round 8/10 — ticks:1398 score:1 win:false
+Round 9/10 — ticks:1014 score:4 win:false
+Round 10/10 — ticks:1435 score:60 win:false
+Counter check passed: 2331 == expected 2331
+>>> [eval] win rate: 1/10 (10%) vs Target
+=== Chunk 3/10: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:1012 score:26 win:false
+Round 2/10 — ticks:653 score:0 win:false
+Round 3/10 — ticks:650 score:0 win:false
+Round 4/10 — ticks:676 score:0 win:false
+Round 5/10 — ticks:651 score:0 win:false
+Round 6/10 — ticks:931 score:22 win:false
+Round 7/10 — ticks:954 score:6 win:false
+Round 8/10 — ticks:670 score:0 win:false
+Round 9/10 — ticks:722 score:0 win:false
+Round 10/10 — ticks:897 score:21 win:false
+Counter check passed: 2341 == expected 2341
+=== Chunk 4/10: 10 rounds vs SacTwin ===
+Training: SAC_LSTM_Bot vs SacTwin for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:821 score:0 win:false
+Round 2/10 — ticks:837 score:3 win:false
+Round 3/10 — ticks:858 score:4 win:false
+Round 4/10 — ticks:934 score:14 win:false
+Round 5/10 — ticks:1058 score:36 win:false
+Round 6/10 — ticks:1241 score:76 win:false
+Round 7/10 — ticks:958 score:16 win:false
+Round 8/10 — ticks:940 score:33 win:false
+Round 9/10 — ticks:707 score:0 win:false
+Round 10/10 — ticks:689 score:0 win:false
+Counter check passed: 2351 == expected 2351
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1647 score:0 win:false
+Round 2/10 — ticks:1518 score:0 win:false
+Round 3/10 — ticks:1733 score:0 win:false
+Round 4/10 — ticks:1731 score:0 win:false
+Round 5/10 — ticks:1077 score:17 win:false
+Round 6/10 — ticks:1745 score:0 win:false
+Round 7/10 — ticks:1634 score:0 win:false
+Round 8/10 — ticks:1710 score:1 win:false
+Round 9/10 — ticks:2180 score:60 win:false
+Round 10/10 — ticks:1364 score:0 win:false
+Counter check passed: 2361 == expected 2361
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1426 score:68 win:true
+Round 2/10 — ticks:1649 score:70 win:true
+Round 3/10 — ticks:1456 score:64 win:true
+Round 4/10 — ticks:1032 score:61 win:true
+Round 5/10 — ticks:1710 score:1 win:true
+Round 6/10 — ticks:941 score:61 win:true
+Round 7/10 — ticks:1455 score:61 win:true
+Round 8/10 — ticks:1323 score:66 win:true
+Round 9/10 — ticks:1035 score:5 win:true
+Round 10/10 — ticks:1457 score:65 win:true
+Counter check passed: 2371 == expected 2371
+>>> [eval] win rate: 10/10 (100%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1014 score:14 win:false
+Round 2/10 — ticks:1074 score:20 win:false
+Round 3/10 — ticks:996 score:6 win:false
+Round 4/10 — ticks:1451 score:0 win:false
+Round 5/10 — ticks:970 score:25 win:false
+Round 6/10 — ticks:1451 score:0 win:false
+Round 7/10 — ticks:1044 score:1 win:false
+Round 8/10 — ticks:1119 score:13 win:false
+Round 9/10 — ticks:1184 score:15 win:false
+Round 10/10 — ticks:1028 score:9 win:false
+Counter check passed: 2381 == expected 2381
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 5/10: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:882 score:0 win:false
+Round 2/10 — ticks:865 score:0 win:false
+Round 3/10 — ticks:908 score:0 win:false
+Round 4/10 — ticks:720 score:1 win:false
+Round 5/10 — ticks:832 score:17 win:false
+Round 6/10 — ticks:957 score:16 win:false
+Round 7/10 — ticks:723 score:3 win:false
+Round 8/10 — ticks:645 score:0 win:false
+Round 9/10 — ticks:963 score:0 win:false
+Round 10/10 — ticks:1032 score:16 win:false
+Counter check passed: 2391 == expected 2391
+=== Chunk 6/10: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:818 score:0 win:false
+Round 2/10 — ticks:905 score:0 win:false
+Round 3/10 — ticks:851 score:0 win:false
+Round 4/10 — ticks:381 score:32 win:false
diff --git a/SAC_LSTM_Bot_garage/campaign_v4_stdout.log.prereboot-20260824T1033 b/SAC_LSTM_Bot_garage/campaign_v4_stdout.log.prereboot-20260824T1033
new file mode 100644
index 0000000..5e3fcab
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/campaign_v4_stdout.log.prereboot-20260824T1033
@@ -0,0 +1,2982 @@
+=== SAC_LSTM_Bot training harness ===
+Opponents: Corners:3,Crazy:2,RamFire:2,Target:1,SacTwin:1 | budget: 25000 rounds in chunks of 10
+Eval: every 2 chunks, 10 rounds vs [Corners,Crazy,Target], MA-5 composite best-gating
+Weights: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/weights/sac_latest.zip
+   Building SAC_LSTM_Bot/SAC_LSTM_Bot using c backend
+      Info: compiling nim package using /nix/store/pq494yyknvhrvkzna9zgcqvma8fq3617-x86_64-unknown-linux-gnu-nim-wrapper-2.2.4/bin/x86_64-unknown-linux-gnu-nim
+=== Chunk 1/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:599 score:0 win:false
+Round 2/10 — ticks:849 score:14 win:false
+Round 3/10 — ticks:584 score:18 win:false
+Round 4/10 — ticks:257 score:12 win:false
+Round 5/10 — ticks:701 score:0 win:false
+Round 6/10 — ticks:663 score:0 win:false
+Round 7/10 — ticks:577 score:9 win:false
+Round 8/10 — ticks:623 score:0 win:false
+Round 9/10 — ticks:653 score:0 win:false
+Round 10/10 — ticks:844 score:16 win:false
+Counter check passed: 10 == expected 9
+=== Chunk 2/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:685 score:4 win:false
+Round 2/10 — ticks:620 score:0 win:false
+Round 3/10 — ticks:722 score:7 win:false
+Round 4/10 — ticks:646 score:29 win:false
+Round 5/10 — ticks:634 score:0 win:false
+Round 6/10 — ticks:688 score:12 win:false
+Round 7/10 — ticks:618 score:0 win:false
+Round 8/10 — ticks:604 score:17 win:false
+Round 9/10 — ticks:619 score:11 win:false
+Round 10/10 — ticks:614 score:0 win:false
+Counter check passed: 20 == expected 20
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:927 score:0 win:false
+Round 2/10 — ticks:1483 score:2 win:false
+Round 3/10 — ticks:662 score:4 win:false
+Round 4/10 — ticks:974 score:7 win:false
+Round 5/10 — ticks:648 score:0 win:false
+Round 6/10 — ticks:1135 score:1 win:false
+Round 7/10 — ticks:682 score:0 win:false
+Round 8/10 — ticks:966 score:2 win:false
+Round 9/10 — ticks:260 score:74 win:false
+Round 10/10 — ticks:1676 score:5 win:false
+Counter check passed: 30 == expected 30
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1566 score:6 win:false
+Round 2/10 — ticks:1578 score:6 win:false
+Round 3/10 — ticks:1599 score:3 win:false
+Round 4/10 — ticks:1484 score:6 win:false
+Round 5/10 — ticks:1780 score:7 win:false
+Round 6/10 — ticks:1458 score:7 win:false
+Round 7/10 — ticks:1532 score:5 win:false
+Round 8/10 — ticks:1713 score:4 win:false
+Round 9/10 — ticks:1522 score:2 win:false
+Round 10/10 — ticks:1228 score:11 win:false
+Counter check passed: 40 == expected 40
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1430 score:3 win:false
+Round 2/10 — ticks:3380 score:16 win:false
+Round 3/10 — ticks:3561 score:13 win:false
+Round 4/10 — ticks:1529 score:3 win:false
+Round 5/10 — ticks:2703 score:7 win:false
+Round 6/10 — ticks:992 score:0 win:false
+Round 7/10 — ticks:1001 score:0 win:false
+Round 8/10 — ticks:1320 score:92 win:false
+Round 9/10 — ticks:1036 score:0 win:false
+Round 10/10 — ticks:1030 score:0 win:false
+Counter check passed: 50 == expected 50
+>>> [eval] win rate: 0/10 (0%) vs Target
+>>> [eval] new best composite (0.0000) -> sac_best.zip
+=== Chunk 3/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:729 score:0 win:false
+Round 2/10 — ticks:761 score:0 win:false
+Round 3/10 — ticks:736 score:1 win:false
+Round 4/10 — ticks:721 score:0 win:false
+Round 5/10 — ticks:639 score:0 win:false
+Round 6/10 — ticks:695 score:9 win:false
+Round 7/10 — ticks:689 score:26 win:false
+Round 8/10 — ticks:651 score:0 win:false
+Round 9/10 — ticks:583 score:15 win:false
+Round 10/10 — ticks:656 score:15 win:false
+Counter check passed: 60 == expected 60
+=== Chunk 4/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:563 score:16 win:false
+Round 2/10 — ticks:705 score:5 win:false
+Round 3/10 — ticks:537 score:1 win:false
+Round 4/10 — ticks:613 score:1 win:false
+Round 5/10 — ticks:577 score:1 win:false
+Round 6/10 — ticks:766 score:45 win:false
+Round 7/10 — ticks:717 score:16 win:false
+Round 8/10 — ticks:578 score:1 win:false
+Round 9/10 — ticks:634 score:0 win:false
+Round 10/10 — ticks:664 score:0 win:false
+Counter check passed: 70 == expected 70
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:190 score:100 win:false
+Round 2/10 — ticks:1050 score:2 win:false
+Round 3/10 — ticks:669 score:5 win:false
+Round 4/10 — ticks:552 score:6 win:false
+Round 5/10 — ticks:319 score:23 win:false
+Round 6/10 — ticks:1164 score:0 win:false
+Round 7/10 — ticks:1121 score:0 win:false
+Round 8/10 — ticks:1136 score:8 win:false
+Round 9/10 — ticks:1213 score:5 win:false
+Round 10/10 — ticks:973 score:3 win:false
+Counter check passed: 80 == expected 80
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1020 score:4 win:false
+Round 2/10 — ticks:1116 score:7 win:false
+Round 3/10 — ticks:911 score:0 win:false
+Round 4/10 — ticks:1226 score:4 win:false
+Round 5/10 — ticks:1072 score:8 win:false
+Round 6/10 — ticks:1119 score:5 win:false
+Round 7/10 — ticks:1046 score:4 win:false
+Round 8/10 — ticks:1100 score:5 win:false
+Round 9/10 — ticks:1104 score:4 win:false
+Round 10/10 — ticks:1178 score:9 win:false
+Counter check passed: 90 == expected 90
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:775 score:51 win:false
+Round 2/10 — ticks:972 score:40 win:false
+Round 3/10 — ticks:1075 score:3 win:false
+Round 4/10 — ticks:1165 score:6 win:false
+Round 5/10 — ticks:879 score:90 win:false
+Round 6/10 — ticks:784 score:0 win:false
+Round 7/10 — ticks:914 score:3 win:false
+Round 8/10 — ticks:855 score:85 win:false
+Round 9/10 — ticks:1117 score:25 win:false
+Round 10/10 — ticks:1127 score:53 win:false
+Counter check passed: 100 == expected 100
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 5/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:659 score:16 win:false
+Round 2/10 — ticks:887 score:49 win:false
+Round 3/10 — ticks:567 score:0 win:false
+Round 4/10 — ticks:586 score:0 win:false
+Round 5/10 — ticks:1136 score:54 win:false
+Round 6/10 — ticks:691 score:17 win:false
+Round 7/10 — ticks:522 score:0 win:false
+Round 8/10 — ticks:553 score:0 win:false
+Round 9/10 — ticks:522 score:17 win:false
+Round 10/10 — ticks:737 score:71 win:false
+Counter check passed: 110 == expected 110
+=== Chunk 6/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:441 score:43 win:false
+Round 2/10 — ticks:216 score:50 win:false
+Round 3/10 — ticks:137 score:31 win:false
+Round 4/10 — ticks:292 score:24 win:false
+Round 5/10 — ticks:385 score:33 win:false
+Round 6/10 — ticks:227 score:38 win:false
+Round 7/10 — ticks:257 score:41 win:false
+Round 8/10 — ticks:208 score:17 win:false
+Round 9/10 — ticks:174 score:30 win:false
+Round 10/10 — ticks:215 score:43 win:false
+Counter check passed: 120 == expected 120
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:861 score:0 win:false
+Round 2/10 — ticks:871 score:4 win:false
+Round 3/10 — ticks:1465 score:4 win:false
+Round 4/10 — ticks:761 score:0 win:false
+Round 5/10 — ticks:825 score:5 win:false
+Round 6/10 — ticks:707 score:0 win:false
+Round 7/10 — ticks:1168 score:6 win:false
+Round 8/10 — ticks:754 score:0 win:false
+Round 9/10 — ticks:782 score:30 win:false
+Round 10/10 — ticks:887 score:5 win:false
+Counter check passed: 130 == expected 130
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1138 score:15 win:false
+Round 2/10 — ticks:922 score:9 win:false
+Round 3/10 — ticks:926 score:6 win:false
+Round 4/10 — ticks:703 score:0 win:false
+Round 5/10 — ticks:1212 score:22 win:false
+Round 6/10 — ticks:848 score:2 win:false
+Round 7/10 — ticks:852 score:5 win:false
+Round 8/10 — ticks:1057 score:18 win:false
+Round 9/10 — ticks:907 score:14 win:false
+Round 10/10 — ticks:948 score:15 win:false
+Counter check passed: 140 == expected 140
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:814 score:8 win:false
+Round 2/10 — ticks:715 score:0 win:false
+Round 3/10 — ticks:700 score:7 win:false
+Round 4/10 — ticks:671 score:92 win:false
+Round 5/10 — ticks:937 score:9 win:false
+Round 6/10 — ticks:689 score:0 win:false
+Round 7/10 — ticks:723 score:0 win:false
+Round 8/10 — ticks:1035 score:5 win:false
+Round 9/10 — ticks:798 score:0 win:false
+Round 10/10 — ticks:632 score:76 win:false
+Counter check passed: 150 == expected 150
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 7/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:218 score:74 win:false
+Round 2/10 — ticks:147 score:22 win:false
+Round 3/10 — ticks:354 score:26 win:false
+Round 4/10 — ticks:262 score:49 win:false
+Round 5/10 — ticks:188 score:23 win:false
+Round 6/10 — ticks:287 score:20 win:false
+Round 7/10 — ticks:274 score:27 win:false
+Round 8/10 — ticks:167 score:43 win:false
+Round 9/10 — ticks:215 score:75 win:false
+Round 10/10 — ticks:181 score:33 win:false
+Counter check passed: 160 == expected 160
+=== Chunk 8/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:535 score:0 win:false
+Round 2/10 — ticks:605 score:0 win:false
+Round 3/10 — ticks:656 score:0 win:false
+Round 4/10 — ticks:588 score:0 win:false
+Round 5/10 — ticks:571 score:4 win:false
+Round 6/10 — ticks:438 score:16 win:false
+Round 7/10 — ticks:457 score:1 win:false
+Round 8/10 — ticks:749 score:16 win:false
+Round 9/10 — ticks:501 score:0 win:false
+Round 10/10 — ticks:543 score:0 win:false
+Counter check passed: 170 == expected 170
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:593 score:0 win:false
+Round 2/10 — ticks:713 score:9 win:false
+Round 3/10 — ticks:1009 score:0 win:false
+Round 4/10 — ticks:643 score:0 win:false
+Round 5/10 — ticks:734 score:0 win:false
+Round 6/10 — ticks:597 score:0 win:false
+Round 7/10 — ticks:618 score:0 win:false
+Round 8/10 — ticks:635 score:0 win:false
+Round 9/10 — ticks:690 score:0 win:false
+Round 10/10 — ticks:705 score:14 win:false
+Counter check passed: 180 == expected 180
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:610 score:0 win:false
+Round 2/10 — ticks:673 score:2 win:false
+Round 3/10 — ticks:1270 score:28 win:false
+Round 4/10 — ticks:1103 score:25 win:false
+Round 5/10 — ticks:624 score:2 win:false
+Round 6/10 — ticks:710 score:16 win:false
+Round 7/10 — ticks:755 score:18 win:false
+Round 8/10 — ticks:691 score:9 win:false
+Round 9/10 — ticks:719 score:0 win:false
+Round 10/10 — ticks:977 score:26 win:false
+Counter check passed: 190 == expected 190
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:588 score:0 win:false
+Round 2/10 — ticks:669 score:22 win:false
+Round 3/10 — ticks:729 score:16 win:false
+Round 4/10 — ticks:968 score:35 win:false
+Round 5/10 — ticks:590 score:0 win:false
+Round 6/10 — ticks:781 score:15 win:false
+Round 7/10 — ticks:679 score:19 win:false
+Round 8/10 — ticks:619 score:5 win:false
+Round 9/10 — ticks:827 score:20 win:false
+Round 10/10 — ticks:571 score:15 win:false
+Counter check passed: 200 == expected 200
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 9/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:325 score:44 win:false
+Round 2/10 — ticks:284 score:26 win:false
+Round 3/10 — ticks:198 score:24 win:false
+Round 4/10 — ticks:198 score:13 win:false
+Round 5/10 — ticks:340 score:56 win:false
+Round 6/10 — ticks:235 score:42 win:false
+Round 7/10 — ticks:283 score:12 win:false
+Round 8/10 — ticks:254 score:31 win:false
+Round 9/10 — ticks:191 score:11 win:false
+Round 10/10 — ticks:218 score:31 win:false
+Counter check passed: 210 == expected 210
+=== Chunk 10/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:572 score:0 win:false
+Round 2/10 — ticks:599 score:0 win:false
+Round 3/10 — ticks:560 score:0 win:false
+Round 4/10 — ticks:574 score:0 win:false
+Round 5/10 — ticks:650 score:16 win:false
+Round 6/10 — ticks:673 score:16 win:false
+Round 7/10 — ticks:668 score:2 win:false
+Round 8/10 — ticks:914 score:48 win:false
+Round 9/10 — ticks:673 score:16 win:false
+Round 10/10 — ticks:560 score:0 win:false
+Counter check passed: 220 == expected 220
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:583 score:12 win:false
+Round 2/10 — ticks:697 score:12 win:false
+Round 3/10 — ticks:490 score:7 win:false
+Round 4/10 — ticks:645 score:12 win:false
+Round 5/10 — ticks:371 score:13 win:false
+Round 6/10 — ticks:666 score:12 win:false
+Round 7/10 — ticks:789 score:0 win:false
+Round 8/10 — ticks:637 score:12 win:false
+Round 9/10 — ticks:550 score:2 win:false
+Round 10/10 — ticks:527 score:0 win:false
+Counter check passed: 230 == expected 230
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:531 score:0 win:false
+Round 2/10 — ticks:511 score:0 win:false
+Round 3/10 — ticks:579 score:12 win:false
+Round 4/10 — ticks:512 score:0 win:false
+Round 5/10 — ticks:626 score:24 win:false
+Round 6/10 — ticks:491 score:1 win:false
+Round 7/10 — ticks:750 score:24 win:false
+Round 8/10 — ticks:593 score:4 win:false
+Round 9/10 — ticks:555 score:5 win:false
+Round 10/10 — ticks:532 score:1 win:false
+Counter check passed: 240 == expected 240
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:567 score:12 win:false
+Round 2/10 — ticks:537 score:0 win:false
+Round 3/10 — ticks:630 score:12 win:false
+Round 4/10 — ticks:790 score:11 win:false
+Round 5/10 — ticks:520 score:0 win:false
+Round 6/10 — ticks:1056 score:12 win:false
+Round 7/10 — ticks:534 score:0 win:false
+Round 8/10 — ticks:530 score:0 win:false
+Round 9/10 — ticks:629 score:0 win:false
+Round 10/10 — ticks:587 score:12 win:false
+Counter check passed: 250 == expected 250
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 11/2500: 10 rounds vs SacTwin ===
+Training: SAC_LSTM_Bot vs SacTwin for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:623 score:0 win:false
+Round 2/10 — ticks:682 score:16 win:false
+Round 3/10 — ticks:665 score:32 win:false
+Round 4/10 — ticks:820 score:51 win:false
+Round 5/10 — ticks:652 score:17 win:false
+Round 6/10 — ticks:768 score:27 win:false
+Round 7/10 — ticks:765 score:40 win:false
+Round 8/10 — ticks:780 score:19 win:false
+Round 9/10 — ticks:663 score:16 win:false
+Round 10/10 — ticks:553 score:2 win:false
+Counter check passed: 260 == expected 260
+=== Chunk 12/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:762 score:17 win:false
+Round 2/10 — ticks:719 score:3 win:false
+Round 3/10 — ticks:671 score:0 win:false
+Round 4/10 — ticks:724 score:17 win:false
+Round 5/10 — ticks:715 score:17 win:false
+Round 6/10 — ticks:524 score:5 win:false
+Round 7/10 — ticks:691 score:0 win:false
+Round 8/10 — ticks:759 score:16 win:false
+Round 9/10 — ticks:686 score:23 win:false
+Round 10/10 — ticks:575 score:0 win:false
+Counter check passed: 270 == expected 270
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:887 score:15 win:false
+Round 2/10 — ticks:546 score:31 win:false
+Round 3/10 — ticks:456 score:0 win:false
+Round 4/10 — ticks:482 score:0 win:false
+Round 5/10 — ticks:486 score:0 win:false
+Round 6/10 — ticks:610 score:0 win:false
+Round 7/10 — ticks:488 score:0 win:false
+Round 8/10 — ticks:505 score:5 win:false
+Round 9/10 — ticks:488 score:0 win:false
+Round 10/10 — ticks:792 score:0 win:false
+Counter check passed: 280 == expected 280
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:531 score:15 win:false
+Round 2/10 — ticks:704 score:48 win:false
+Round 3/10 — ticks:754 score:47 win:false
+Round 4/10 — ticks:864 score:63 win:false
+Round 5/10 — ticks:992 score:47 win:false
+Round 6/10 — ticks:696 score:17 win:false
+Round 7/10 — ticks:1124 score:30 win:false
+Round 8/10 — ticks:543 score:16 win:false
+Round 9/10 — ticks:475 score:0 win:false
+Round 10/10 — ticks:490 score:0 win:false
+Counter check passed: 290 == expected 290
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:488 score:0 win:false
+Round 2/10 — ticks:579 score:30 win:false
+Round 3/10 — ticks:489 score:0 win:false
+Round 4/10 — ticks:554 score:0 win:false
+Round 5/10 — ticks:554 score:0 win:false
+Round 6/10 — ticks:579 score:15 win:false
+Round 7/10 — ticks:521 score:16 win:false
+Round 8/10 — ticks:554 score:0 win:false
+Round 9/10 — ticks:489 score:0 win:false
+Round 10/10 — ticks:550 score:21 win:false
+Counter check passed: 300 == expected 300
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 13/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:934 score:0 win:false
+Round 2/10 — ticks:637 score:0 win:false
+Round 3/10 — ticks:591 score:0 win:false
+Round 4/10 — ticks:298 score:34 win:false
+Round 5/10 — ticks:664 score:3 win:false
+Round 6/10 — ticks:439 score:0 win:false
+Round 7/10 — ticks:408 score:3 win:false
+Round 8/10 — ticks:610 score:16 win:false
+Round 9/10 — ticks:543 score:0 win:false
+Round 10/10 — ticks:548 score:0 win:false
+Counter check passed: 310 == expected 310
+=== Chunk 14/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:950 score:32 win:false
+Round 2/10 — ticks:528 score:16 win:false
+Round 3/10 — ticks:559 score:0 win:false
+Round 4/10 — ticks:563 score:16 win:false
+Round 5/10 — ticks:603 score:0 win:false
+Round 6/10 — ticks:543 score:0 win:false
+Round 7/10 — ticks:519 score:0 win:false
+Round 8/10 — ticks:240 score:18 win:false
+Round 9/10 — ticks:575 score:0 win:false
+Round 10/10 — ticks:440 score:0 win:false
+Counter check passed: 320 == expected 320
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:496 score:0 win:false
+Round 2/10 — ticks:505 score:0 win:false
+Round 3/10 — ticks:451 score:0 win:false
+Round 4/10 — ticks:567 score:0 win:false
+Round 5/10 — ticks:478 score:0 win:false
+Round 6/10 — ticks:567 score:0 win:false
+Round 7/10 — ticks:486 score:16 win:false
+Round 8/10 — ticks:737 score:0 win:false
+Round 9/10 — ticks:502 score:0 win:false
+Round 10/10 — ticks:464 score:0 win:false
+Counter check passed: 330 == expected 330
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:994 score:17 win:false
+Round 2/10 — ticks:469 score:7 win:false
+Round 3/10 — ticks:517 score:43 win:false
+Round 4/10 — ticks:567 score:0 win:false
+Round 5/10 — ticks:536 score:3 win:false
+Round 6/10 — ticks:751 score:31 win:false
+Round 7/10 — ticks:1076 score:33 win:false
+Round 8/10 — ticks:475 score:0 win:false
+Round 9/10 — ticks:475 score:0 win:false
+Round 10/10 — ticks:470 score:1 win:false
+Counter check passed: 340 == expected 340
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:583 score:23 win:false
+Round 2/10 — ticks:475 score:0 win:false
+Round 3/10 — ticks:825 score:16 win:false
+Round 4/10 — ticks:567 score:0 win:false
+Round 5/10 — ticks:446 score:188 win:false
+Round 6/10 — ticks:475 score:0 win:false
+Round 7/10 — ticks:463 score:3 win:false
+Round 8/10 — ticks:456 score:8 win:false
+Round 9/10 — ticks:475 score:0 win:false
+Round 10/10 — ticks:599 score:44 win:false
+Counter check passed: 350 == expected 350
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 15/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:569 score:0 win:false
+Round 2/10 — ticks:554 score:0 win:false
+Round 3/10 — ticks:852 score:16 win:false
+Round 4/10 — ticks:561 score:0 win:false
+Round 5/10 — ticks:536 score:0 win:false
+Round 6/10 — ticks:1082 score:32 win:false
+Round 7/10 — ticks:564 score:0 win:false
+Round 8/10 — ticks:753 score:16 win:false
+Round 9/10 — ticks:810 score:16 win:false
+Round 10/10 — ticks:563 score:0 win:false
+Counter check passed: 360 == expected 360
+=== Chunk 16/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:556 score:0 win:false
+Round 2/10 — ticks:634 score:0 win:false
+Round 3/10 — ticks:675 score:16 win:false
+Round 4/10 — ticks:537 score:0 win:false
+Round 5/10 — ticks:606 score:0 win:false
+Round 6/10 — ticks:484 score:17 win:false
+Round 7/10 — ticks:478 score:1 win:false
+Round 8/10 — ticks:766 score:30 win:false
+Round 9/10 — ticks:902 score:21 win:false
+Round 10/10 — ticks:699 score:32 win:false
+Counter check passed: 370 == expected 370
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:511 score:0 win:false
+Round 2/10 — ticks:545 score:0 win:false
+Round 3/10 — ticks:519 score:0 win:false
+Round 4/10 — ticks:690 score:31 win:false
+Round 5/10 — ticks:474 score:0 win:false
+Round 6/10 — ticks:476 score:0 win:false
+Round 7/10 — ticks:474 score:0 win:false
+Round 8/10 — ticks:573 score:0 win:false
+Round 9/10 — ticks:518 score:0 win:false
+Round 10/10 — ticks:545 score:0 win:false
+Counter check passed: 380 == expected 380
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:720 score:47 win:false
+Round 2/10 — ticks:544 score:0 win:false
+Round 3/10 — ticks:439 score:19 win:false
+Round 4/10 — ticks:590 score:19 win:false
+Round 5/10 — ticks:755 score:48 win:false
+Round 6/10 — ticks:632 score:15 win:false
+Round 7/10 — ticks:471 score:0 win:false
+Round 8/10 — ticks:836 score:33 win:false
+Round 9/10 — ticks:677 score:33 win:false
+Round 10/10 — ticks:708 score:1 win:false
+Counter check passed: 390 == expected 390
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:897 score:25 win:false
+Round 2/10 — ticks:482 score:0 win:false
+Round 3/10 — ticks:800 score:16 win:false
+Round 4/10 — ticks:481 score:0 win:false
+Round 5/10 — ticks:574 score:22 win:false
+Round 6/10 — ticks:544 score:0 win:false
+Round 7/10 — ticks:545 score:0 win:false
+Round 8/10 — ticks:578 score:32 win:false
+Round 9/10 — ticks:465 score:10 win:false
+Round 10/10 — ticks:545 score:0 win:false
+Counter check passed: 400 == expected 400
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 17/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:567 score:0 win:false
+Round 2/10 — ticks:550 score:0 win:false
+Round 3/10 — ticks:541 score:0 win:false
+Round 4/10 — ticks:584 score:0 win:false
+Round 5/10 — ticks:545 score:0 win:false
+Round 6/10 — ticks:695 score:16 win:false
+Round 7/10 — ticks:650 score:0 win:false
+Round 8/10 — ticks:471 score:4 win:false
+Round 9/10 — ticks:536 score:0 win:false
+Round 10/10 — ticks:573 score:16 win:false
+Counter check passed: 410 == expected 410
+=== Chunk 18/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:526 score:0 win:false
+Round 2/10 — ticks:547 score:0 win:false
+Round 3/10 — ticks:624 score:16 win:false
+Round 4/10 — ticks:596 score:0 win:false
+Round 5/10 — ticks:553 score:0 win:false
+Round 6/10 — ticks:840 score:54 win:false
+Round 7/10 — ticks:541 score:0 win:false
+Round 8/10 — ticks:806 score:16 win:false
+Round 9/10 — ticks:649 score:16 win:false
+Round 10/10 — ticks:522 score:0 win:false
+Counter check passed: 420 == expected 420
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:448 score:0 win:false
+Round 2/10 — ticks:468 score:16 win:false
+Round 3/10 — ticks:844 score:10 win:false
+Round 4/10 — ticks:453 score:15 win:false
+Round 5/10 — ticks:379 score:7 win:false
+Round 6/10 — ticks:445 score:0 win:false
+Round 7/10 — ticks:444 score:1 win:false
+Round 8/10 — ticks:447 score:0 win:false
+Round 9/10 — ticks:719 score:0 win:false
+Round 10/10 — ticks:804 score:0 win:false
+Counter check passed: 430 == expected 430
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:811 score:2 win:false
+Round 2/10 — ticks:737 score:17 win:false
+Round 3/10 — ticks:593 score:45 win:false
+Round 4/10 — ticks:517 score:15 win:false
+Round 5/10 — ticks:498 score:0 win:false
+Round 6/10 — ticks:728 score:31 win:false
+Round 7/10 — ticks:728 score:32 win:false
+Round 8/10 — ticks:634 score:16 win:false
+Round 9/10 — ticks:468 score:4 win:false
+Round 10/10 — ticks:435 score:0 win:false
+Counter check passed: 440 == expected 440
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:567 score:0 win:false
+Round 2/10 — ticks:567 score:0 win:false
+Round 3/10 — ticks:469 score:0 win:false
+Round 4/10 — ticks:504 score:26 win:false
+Round 5/10 — ticks:567 score:0 win:false
+Round 6/10 — ticks:461 score:0 win:false
+Round 7/10 — ticks:525 score:15 win:false
+Round 8/10 — ticks:469 score:0 win:false
+Round 9/10 — ticks:567 score:0 win:false
+Round 10/10 — ticks:535 score:15 win:false
+Counter check passed: 450 == expected 450
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 19/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:680 score:48 win:false
+Round 2/10 — ticks:552 score:0 win:false
+Round 3/10 — ticks:618 score:16 win:false
+Round 4/10 — ticks:584 score:16 win:false
+Round 5/10 — ticks:533 score:0 win:false
+Round 6/10 — ticks:442 score:0 win:false
+Round 7/10 — ticks:586 score:18 win:false
+Round 8/10 — ticks:707 score:31 win:false
+Round 9/10 — ticks:604 score:0 win:false
+Round 10/10 — ticks:767 score:1 win:false
+Counter check passed: 460 == expected 460
+=== Chunk 20/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:652 score:17 win:false
+Round 2/10 — ticks:681 score:32 win:false
+Round 3/10 — ticks:845 score:5 win:false
+Round 4/10 — ticks:638 score:19 win:false
+Round 5/10 — ticks:580 score:0 win:false
+Round 6/10 — ticks:605 score:16 win:false
+Round 7/10 — ticks:745 score:16 win:false
+Round 8/10 — ticks:543 score:1 win:false
+Round 9/10 — ticks:520 score:0 win:false
+Round 10/10 — ticks:665 score:49 win:false
+Counter check passed: 470 == expected 470
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:567 score:0 win:false
+Round 2/10 — ticks:638 score:1 win:false
+Round 3/10 — ticks:482 score:0 win:false
+Round 4/10 — ticks:446 score:1 win:false
+Round 5/10 — ticks:516 score:0 win:false
+Round 6/10 — ticks:820 score:0 win:false
+Round 7/10 — ticks:553 score:14 win:false
+Round 8/10 — ticks:484 score:0 win:false
+Round 9/10 — ticks:463 score:22 win:false
+Round 10/10 — ticks:703 score:0 win:false
+Counter check passed: 480 == expected 480
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1091 score:29 win:false
+Round 2/10 — ticks:519 score:0 win:false
+Round 3/10 — ticks:521 score:31 win:false
+Round 4/10 — ticks:528 score:15 win:false
+Round 5/10 — ticks:854 score:120 win:false
+Round 6/10 — ticks:488 score:15 win:false
+Round 7/10 — ticks:501 score:0 win:false
+Round 8/10 — ticks:580 score:0 win:false
+Round 9/10 — ticks:829 score:0 win:false
+Round 10/10 — ticks:470 score:1 win:false
+Counter check passed: 490 == expected 490
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:577 score:43 win:false
+Round 2/10 — ticks:580 score:0 win:false
+Round 3/10 — ticks:580 score:0 win:false
+Round 4/10 — ticks:499 score:0 win:false
+Round 5/10 — ticks:501 score:0 win:false
+Round 6/10 — ticks:580 score:0 win:false
+Round 7/10 — ticks:580 score:0 win:false
+Round 8/10 — ticks:541 score:20 win:false
+Round 9/10 — ticks:719 score:66 win:false
+Round 10/10 — ticks:968 score:41 win:false
+Counter check passed: 500 == expected 500
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 21/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:534 score:0 win:false
+Round 2/10 — ticks:754 score:25 win:false
+Round 3/10 — ticks:564 score:0 win:false
+Round 4/10 — ticks:554 score:0 win:false
+Round 5/10 — ticks:554 score:0 win:false
+Round 6/10 — ticks:543 score:0 win:false
+Round 7/10 — ticks:554 score:0 win:false
+Round 8/10 — ticks:716 score:30 win:false
+Round 9/10 — ticks:559 score:31 win:false
+Round 10/10 — ticks:837 score:16 win:false
+Counter check passed: 510 == expected 510
+=== Chunk 22/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:479 score:15 win:false
+Round 2/10 — ticks:535 score:0 win:false
+Round 3/10 — ticks:500 score:16 win:false
+Round 4/10 — ticks:659 score:0 win:false
+Round 5/10 — ticks:561 score:0 win:false
+Round 6/10 — ticks:605 score:0 win:false
+Round 7/10 — ticks:563 score:0 win:false
+Round 8/10 — ticks:568 score:0 win:false
+Round 9/10 — ticks:509 score:0 win:false
+Round 10/10 — ticks:623 score:35 win:false
+Counter check passed: 520 == expected 520
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:507 score:16 win:false
+Round 2/10 — ticks:610 score:0 win:false
+Round 3/10 — ticks:647 score:0 win:false
+Round 4/10 — ticks:629 score:0 win:false
+Round 5/10 — ticks:501 score:6 win:false
+Round 6/10 — ticks:803 score:19 win:false
+Round 7/10 — ticks:647 score:20 win:false
+Round 8/10 — ticks:602 score:10 win:false
+Round 9/10 — ticks:620 score:29 win:false
+Round 10/10 — ticks:1037 score:10 win:false
+Counter check passed: 530 == expected 530
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:594 score:0 win:false
+Round 2/10 — ticks:559 score:1 win:false
+Round 3/10 — ticks:644 score:12 win:false
+Round 4/10 — ticks:623 score:19 win:false
+Round 5/10 — ticks:762 score:3 win:false
+Round 6/10 — ticks:836 score:0 win:false
+Round 7/10 — ticks:764 score:0 win:false
+Round 8/10 — ticks:1021 score:30 win:false
+Round 9/10 — ticks:753 score:20 win:false
+Round 10/10 — ticks:673 score:10 win:false
+Counter check passed: 540 == expected 540
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:548 score:1 win:false
+Round 2/10 — ticks:1147 score:20 win:false
+Round 3/10 — ticks:1099 score:9 win:false
+Round 4/10 — ticks:628 score:20 win:false
+Round 5/10 — ticks:769 score:10 win:false
+Round 6/10 — ticks:634 score:9 win:false
+Round 7/10 — ticks:583 score:3 win:false
+Round 8/10 — ticks:1122 score:29 win:false
+Round 9/10 — ticks:1008 score:53 win:false
+Round 10/10 — ticks:647 score:0 win:false
+Counter check passed: 550 == expected 550
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 23/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:595 score:0 win:false
+Round 2/10 — ticks:585 score:32 win:false
+Round 3/10 — ticks:648 score:0 win:false
+Round 4/10 — ticks:509 score:0 win:false
+Round 5/10 — ticks:540 score:0 win:false
+Round 6/10 — ticks:462 score:0 win:false
+Round 7/10 — ticks:557 score:3 win:false
+Round 8/10 — ticks:608 score:0 win:false
+Round 9/10 — ticks:375 score:0 win:false
+Round 10/10 — ticks:409 score:0 win:false
+Counter check passed: 560 == expected 560
+=== Chunk 24/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:345 score:0 win:false
+Round 2/10 — ticks:626 score:0 win:false
+Round 3/10 — ticks:467 score:8 win:false
+Round 4/10 — ticks:561 score:0 win:false
+Round 5/10 — ticks:514 score:0 win:false
+Round 6/10 — ticks:536 score:0 win:false
+Round 7/10 — ticks:597 score:0 win:false
+Round 8/10 — ticks:790 score:0 win:false
+Round 9/10 — ticks:579 score:0 win:false
+Round 10/10 — ticks:525 score:0 win:false
+Counter check passed: 570 == expected 570
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:3269 score:66 win:true
+Round 2/10 — ticks:1551 score:1 win:false
+Round 3/10 — ticks:2213 score:0 win:false
+Round 4/10 — ticks:4152 score:15 win:false
+Round 5/10 — ticks:2487 score:1 win:false
+Round 6/10 — ticks:1207 score:0 win:false
+Round 7/10 — ticks:1027 score:0 win:false
+Round 8/10 — ticks:1379 score:0 win:false
+Round 9/10 — ticks:1467 score:1 win:false
+Round 10/10 — ticks:1255 score:1 win:false
+Counter check passed: 580 == expected 580
+>>> [eval] win rate: 1/10 (10%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1963 score:70 win:true
+Round 2/10 — ticks:1251 score:66 win:true
+Round 3/10 — ticks:1295 score:2 win:true
+Round 4/10 — ticks:1313 score:5 win:false
+Round 5/10 — ticks:1755 score:68 win:true
+Round 6/10 — ticks:2781 score:11 win:false
+Round 7/10 — ticks:1119 score:2 win:false
+Round 8/10 — ticks:1773 score:5 win:false
+Round 9/10 — ticks:1601 score:4 win:false
+Round 10/10 — ticks:1433 score:5 win:false
+Counter check passed: 590 == expected 590
+>>> [eval] win rate: 4/10 (40%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1273 score:13 win:false
+Round 2/10 — ticks:2224 score:32 win:false
+Round 3/10 — ticks:998 score:16 win:false
+Round 4/10 — ticks:1514 score:50 win:false
+Round 5/10 — ticks:1531 score:27 win:false
+Round 6/10 — ticks:1212 score:1 win:false
+Round 7/10 — ticks:4606 score:27 win:false
+Round 8/10 — ticks:1385 score:15 win:false
+Round 9/10 — ticks:1718 score:52 win:false
+Round 10/10 — ticks:922 score:0 win:false
+Counter check passed: 600 == expected 600
+>>> [eval] win rate: 0/10 (0%) vs Target
+>>> [eval] new best composite (3.3333) -> sac_best.zip
+=== Chunk 25/2500: 10 rounds vs SacTwin ===
+Training: SAC_LSTM_Bot vs SacTwin for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:622 score:0 win:false
+Round 2/10 — ticks:590 score:0 win:false
+Round 3/10 — ticks:533 score:2 win:false
+Round 4/10 — ticks:809 score:144 win:false
+Round 5/10 — ticks:528 score:6 win:false
+Round 6/10 — ticks:658 score:16 win:false
+Round 7/10 — ticks:558 score:0 win:false
+Round 8/10 — ticks:567 score:0 win:false
+Round 9/10 — ticks:909 score:17 win:false
+Round 10/10 — ticks:583 score:10 win:false
+Counter check passed: 610 == expected 610
+=== Chunk 26/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:497 score:1 win:false
+Round 2/10 — ticks:737 score:17 win:false
+Round 3/10 — ticks:699 score:0 win:false
+Round 4/10 — ticks:967 score:13 win:false
+Round 5/10 — ticks:555 score:0 win:false
+Round 6/10 — ticks:580 score:0 win:false
+Round 7/10 — ticks:739 score:0 win:false
+Round 8/10 — ticks:704 score:32 win:false
+Round 9/10 — ticks:632 score:8 win:false
+Round 10/10 — ticks:679 score:0 win:false
+Counter check passed: 620 == expected 620
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1680 score:0 win:false
+Round 2/10 — ticks:1783 score:0 win:false
+Round 3/10 — ticks:1969 score:0 win:false
+Round 4/10 — ticks:2016 score:0 win:false
+Round 5/10 — ticks:1813 score:0 win:false
+Round 6/10 — ticks:1812 score:0 win:false
+Round 7/10 — ticks:1766 score:0 win:false
+Round 8/10 — ticks:1762 score:0 win:false
+Round 9/10 — ticks:1747 score:0 win:false
+Round 10/10 — ticks:1939 score:60 win:false
+Counter check passed: 630 == expected 630
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1858 score:1 win:false
+Round 2/10 — ticks:1956 score:69 win:false
+Round 3/10 — ticks:1356 score:61 win:true
+Round 4/10 — ticks:1497 score:63 win:true
+Round 5/10 — ticks:1761 score:65 win:true
+Round 6/10 — ticks:1423 score:60 win:true
+Round 7/10 — ticks:1006 score:66 win:true
+Round 8/10 — ticks:1960 score:61 win:true
+Round 9/10 — ticks:1800 score:5 win:true
+Round 10/10 — ticks:2243 score:11 win:true
+Counter check passed: 640 == expected 640
+>>> [eval] win rate: 8/10 (80%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1141 score:0 win:false
+Round 2/10 — ticks:1144 score:0 win:false
+Round 3/10 — ticks:1422 score:60 win:false
+Round 4/10 — ticks:1080 score:14 win:false
+Round 5/10 — ticks:1152 score:0 win:false
+Round 6/10 — ticks:1080 score:21 win:false
+Round 7/10 — ticks:931 score:50 win:false
+Round 8/10 — ticks:1122 score:0 win:false
+Round 9/10 — ticks:1063 score:39 win:false
+Round 10/10 — ticks:1038 score:20 win:false
+Counter check passed: 650 == expected 650
+>>> [eval] win rate: 0/10 (0%) vs Target
+>>> [eval] new best composite (8.6667) -> sac_best.zip
+=== Chunk 27/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:522 score:0 win:false
+Round 2/10 — ticks:655 score:16 win:false
+Round 3/10 — ticks:622 score:0 win:false
+Round 4/10 — ticks:567 score:0 win:false
+Round 5/10 — ticks:651 score:32 win:false
+Round 6/10 — ticks:540 score:26 win:false
+Round 7/10 — ticks:490 score:0 win:false
+Round 8/10 — ticks:802 score:17 win:false
+Round 9/10 — ticks:656 score:0 win:false
+Round 10/10 — ticks:561 score:1 win:false
+Counter check passed: 660 == expected 660
+=== Chunk 28/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:798 score:16 win:false
+Round 2/10 — ticks:635 score:34 win:false
+Round 3/10 — ticks:421 score:0 win:false
+Round 4/10 — ticks:493 score:0 win:false
+Round 5/10 — ticks:697 score:0 win:false
+Round 6/10 — ticks:672 score:0 win:false
+Round 7/10 — ticks:613 score:0 win:false
+Round 8/10 — ticks:549 score:1 win:false
+Round 9/10 — ticks:430 score:16 win:false
+Round 10/10 — ticks:920 score:16 win:false
+Counter check passed: 670 == expected 670
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1956 score:0 win:false
+Round 2/10 — ticks:1788 score:0 win:false
+Round 3/10 — ticks:1473 score:0 win:false
+Round 4/10 — ticks:1764 score:0 win:false
+Round 5/10 — ticks:1828 score:0 win:false
+Round 6/10 — ticks:2020 score:0 win:false
+Round 7/10 — ticks:1383 score:0 win:false
+Round 8/10 — ticks:1069 score:1 win:false
+Round 9/10 — ticks:1417 score:0 win:false
+Round 10/10 — ticks:2309 score:0 win:false
+Counter check passed: 680 == expected 680
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1927 score:61 win:true
+Round 2/10 — ticks:2366 score:5 win:false
+Round 3/10 — ticks:1257 score:2 win:false
+Round 4/10 — ticks:1664 score:68 win:false
+Round 5/10 — ticks:1770 score:1 win:false
+Round 6/10 — ticks:1562 score:60 win:false
+Round 7/10 — ticks:1545 score:65 win:false
+Round 8/10 — ticks:1513 score:61 win:false
+Round 9/10 — ticks:1624 score:0 win:false
+Round 10/10 — ticks:1471 score:62 win:false
+Counter check passed: 690 == expected 690
+>>> [eval] win rate: 1/10 (10%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:791 score:101 win:true
+Round 2/10 — ticks:1156 score:5 win:false
+Round 3/10 — ticks:1152 score:0 win:false
+Round 4/10 — ticks:907 score:61 win:false
+Round 5/10 — ticks:1152 score:0 win:false
+Round 6/10 — ticks:966 score:48 win:false
+Round 7/10 — ticks:1135 score:0 win:false
+Round 8/10 — ticks:1128 score:0 win:false
+Round 9/10 — ticks:996 score:37 win:false
+Round 10/10 — ticks:1116 score:7 win:false
+Counter check passed: 700 == expected 700
+>>> [eval] win rate: 1/10 (10%) vs Target
+>>> [eval] new best composite (10.0000) -> sac_best.zip
+=== Chunk 29/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:326 score:13 win:false
+Round 2/10 — ticks:112 score:29 win:false
+Round 3/10 — ticks:414 score:19 win:false
+Round 4/10 — ticks:189 score:93 win:false
+Round 5/10 — ticks:321 score:74 win:false
+Round 6/10 — ticks:268 score:45 win:false
+Round 7/10 — ticks:329 score:41 win:false
+Round 8/10 — ticks:156 score:50 win:false
+Round 9/10 — ticks:204 score:51 win:false
+Round 10/10 — ticks:131 score:46 win:false
+Counter check passed: 710 == expected 710
+=== Chunk 30/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:623 score:0 win:false
+Round 2/10 — ticks:548 score:0 win:false
+Round 3/10 — ticks:748 score:10 win:false
+Round 4/10 — ticks:782 score:0 win:false
+Round 5/10 — ticks:553 score:4 win:false
+Round 6/10 — ticks:550 score:16 win:false
+Round 7/10 — ticks:601 score:0 win:false
+Round 8/10 — ticks:705 score:32 win:false
+Round 9/10 — ticks:473 score:16 win:false
+Round 10/10 — ticks:603 score:0 win:false
+Counter check passed: 720 == expected 720
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1908 score:0 win:false
+Round 2/10 — ticks:1477 score:0 win:false
+Round 3/10 — ticks:3297 score:0 win:false
+Round 4/10 — ticks:1949 score:0 win:false
+Round 5/10 — ticks:1366 score:0 win:false
+Round 6/10 — ticks:1506 score:0 win:false
+Round 7/10 — ticks:2049 score:1 win:false
+Round 8/10 — ticks:1941 score:0 win:false
+Round 9/10 — ticks:1762 score:60 win:false
+Round 10/10 — ticks:1828 score:0 win:false
+Counter check passed: 730 == expected 730
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1407 score:1 win:false
+Round 2/10 — ticks:1817 score:63 win:false
+Round 3/10 — ticks:1052 score:60 win:true
+Round 4/10 — ticks:932 score:60 win:true
+Round 5/10 — ticks:1930 score:67 win:true
+Round 6/10 — ticks:1372 score:7 win:true
+Round 7/10 — ticks:1747 score:64 win:true
+Round 8/10 — ticks:1971 score:2 win:false
+Round 9/10 — ticks:1063 score:60 win:true
+Round 10/10 — ticks:1077 score:60 win:true
+Counter check passed: 740 == expected 740
+>>> [eval] win rate: 7/10 (70%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1447 score:60 win:true
+Round 2/10 — ticks:1064 score:16 win:true
+Round 3/10 — ticks:1422 score:60 win:true
+Round 4/10 — ticks:1326 score:9 win:true
+Round 5/10 — ticks:1074 score:6 win:false
+Round 6/10 — ticks:1020 score:25 win:false
+Round 7/10 — ticks:1030 score:15 win:false
+Round 8/10 — ticks:1076 score:0 win:false
+Round 9/10 — ticks:1080 score:15 win:false
+Round 10/10 — ticks:1092 score:0 win:false
+Counter check passed: 750 == expected 750
+>>> [eval] win rate: 4/10 (40%) vs Target
+>>> [eval] new best composite (17.3333) -> sac_best.zip
+=== Chunk 31/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:187 score:80 win:false
+Round 2/10 — ticks:375 score:46 win:false
+Round 3/10 — ticks:160 score:35 win:false
+Round 4/10 — ticks:202 score:54 win:false
+Round 5/10 — ticks:245 score:29 win:false
+Round 6/10 — ticks:212 score:38 win:false
+Round 7/10 — ticks:429 score:34 win:false
+Round 8/10 — ticks:282 score:41 win:false
+Round 9/10 — ticks:438 score:36 win:false
+Round 10/10 — ticks:180 score:26 win:false
+Counter check passed: 760 == expected 760
+=== Chunk 32/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:129 score:35 win:false
+Round 2/10 — ticks:303 score:11 win:false
+Round 3/10 — ticks:369 score:30 win:false
+Round 4/10 — ticks:293 score:26 win:false
+Round 5/10 — ticks:280 score:45 win:false
+Round 6/10 — ticks:142 score:28 win:false
+Round 7/10 — ticks:171 score:32 win:false
+Round 8/10 — ticks:482 score:10 win:false
+Round 9/10 — ticks:259 score:20 win:false
+Round 10/10 — ticks:231 score:56 win:false
+Counter check passed: 770 == expected 770
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:3017 score:60 win:true
+Round 2/10 — ticks:1607 score:0 win:false
+Round 3/10 — ticks:1575 score:60 win:false
+Round 4/10 — ticks:2330 score:0 win:false
+Round 5/10 — ticks:1399 score:60 win:false
+Round 6/10 — ticks:1869 score:0 win:false
+Round 7/10 — ticks:1982 score:0 win:false
+Round 8/10 — ticks:2094 score:0 win:false
+Round 9/10 — ticks:1772 score:0 win:false
+Round 10/10 — ticks:2159 score:0 win:false
+Counter check passed: 780 == expected 780
+>>> [eval] win rate: 1/10 (10%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1469 score:61 win:true
+Round 2/10 — ticks:1689 score:75 win:true
+Round 3/10 — ticks:1142 score:63 win:true
+Round 4/10 — ticks:1294 score:20 win:true
+Round 5/10 — ticks:1688 score:67 win:true
+Round 6/10 — ticks:1281 score:64 win:true
+Round 7/10 — ticks:1750 score:74 win:true
+Round 8/10 — ticks:1686 score:3 win:true
+Round 9/10 — ticks:979 score:62 win:true
+Round 10/10 — ticks:1667 score:63 win:true
+Counter check passed: 790 == expected 790
+>>> [eval] win rate: 10/10 (100%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1098 score:28 win:false
+Round 2/10 — ticks:1074 score:20 win:false
+Round 3/10 — ticks:1126 score:18 win:false
+Round 4/10 — ticks:1134 score:14 win:false
+Round 5/10 — ticks:1146 score:2 win:false
+Round 6/10 — ticks:1074 score:26 win:false
+Round 7/10 — ticks:778 score:89 win:false
+Round 8/10 — ticks:1100 score:13 win:false
+Round 9/10 — ticks:1044 score:34 win:false
+Round 10/10 — ticks:1153 score:6 win:false
+Counter check passed: 800 == expected 800
+>>> [eval] win rate: 0/10 (0%) vs Target
+>>> [eval] new best composite (24.6667) -> sac_best.zip
+=== Chunk 33/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:149 score:26 win:false
+Round 2/10 — ticks:245 score:37 win:false
+Round 3/10 — ticks:275 score:76 win:false
+Round 4/10 — ticks:321 score:35 win:false
+Round 5/10 — ticks:177 score:59 win:false
+Round 6/10 — ticks:569 score:0 win:false
+Round 7/10 — ticks:598 score:41 win:false
+Round 8/10 — ticks:279 score:31 win:false
+Round 9/10 — ticks:421 score:7 win:false
+Round 10/10 — ticks:253 score:55 win:false
+Counter check passed: 810 == expected 810
+=== Chunk 34/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:663 score:0 win:false
+Round 2/10 — ticks:613 score:16 win:false
+Round 3/10 — ticks:394 score:1 win:false
+Round 4/10 — ticks:573 score:0 win:false
+Round 5/10 — ticks:563 score:0 win:false
+Round 6/10 — ticks:553 score:0 win:false
+Round 7/10 — ticks:591 score:0 win:false
+Round 8/10 — ticks:568 score:0 win:false
+Round 9/10 — ticks:598 score:16 win:false
+Round 10/10 — ticks:785 score:0 win:false
+Counter check passed: 820 == expected 820
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2672 score:0 win:false
+Round 2/10 — ticks:3961 score:0 win:false
+Round 3/10 — ticks:2300 score:0 win:false
+Round 4/10 — ticks:1065 score:0 win:false
+Round 5/10 — ticks:1914 score:0 win:false
+Round 6/10 — ticks:2637 score:0 win:false
+Round 7/10 — ticks:1863 score:0 win:false
+Round 8/10 — ticks:337 score:18 win:false
+Round 9/10 — ticks:2549 score:0 win:false
+Round 10/10 — ticks:2886 score:0 win:false
+Counter check passed: 830 == expected 830
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2429 score:7 win:false
+Round 2/10 — ticks:3436 score:64 win:false
+Round 3/10 — ticks:2296 score:0 win:false
+Round 4/10 — ticks:3287 score:0 win:false
+Round 5/10 — ticks:3146 score:62 win:false
+Round 6/10 — ticks:2322 score:61 win:false
+Round 7/10 — ticks:2685 score:5 win:false
+Round 8/10 — ticks:2183 score:7 win:false
+Round 9/10 — ticks:2209 score:2 win:false
+Round 10/10 — ticks:2739 score:2 win:false
+Counter check passed: 840 == expected 840
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1451 score:0 win:false
+Round 2/10 — ticks:1422 score:60 win:true
+Round 3/10 — ticks:1451 score:0 win:true
+Round 4/10 — ticks:1432 score:60 win:true
+Round 5/10 — ticks:1429 score:61 win:true
+Round 6/10 — ticks:1422 score:60 win:true
+Round 7/10 — ticks:1451 score:0 win:true
+Round 8/10 — ticks:1433 score:3 win:true
+Round 9/10 — ticks:1437 score:60 win:true
+Round 10/10 — ticks:1422 score:60 win:true
+Counter check passed: 850 == expected 850
+>>> [eval] win rate: 9/10 (90%) vs Target
+>>> [eval] new best composite (27.3333) -> sac_best.zip
+=== Chunk 35/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:609 score:0 win:false
+Round 2/10 — ticks:578 score:1 win:false
+Round 3/10 — ticks:800 score:57 win:false
+Round 4/10 — ticks:555 score:0 win:false
+Round 5/10 — ticks:554 score:0 win:false
+Round 6/10 — ticks:556 score:0 win:false
+Round 7/10 — ticks:633 score:16 win:false
+Round 8/10 — ticks:643 score:15 win:false
+Round 9/10 — ticks:580 score:0 win:false
+Round 10/10 — ticks:564 score:0 win:false
+Counter check passed: 860 == expected 860
+=== Chunk 36/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:184 score:20 win:false
+Round 2/10 — ticks:616 score:20 win:false
+Round 3/10 — ticks:272 score:17 win:false
+Round 4/10 — ticks:179 score:36 win:false
+Round 5/10 — ticks:133 score:22 win:false
+Round 6/10 — ticks:272 score:34 win:false
+Round 7/10 — ticks:179 score:44 win:false
+Round 8/10 — ticks:300 score:21 win:false
+Round 9/10 — ticks:226 score:26 win:false
+Round 10/10 — ticks:232 score:17 win:false
+Counter check passed: 870 == expected 870
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:536 score:54 win:false
+Round 2/10 — ticks:2874 score:0 win:false
+Round 3/10 — ticks:2926 score:0 win:false
+Round 4/10 — ticks:2347 score:0 win:false
+Round 5/10 — ticks:3366 score:0 win:false
+Round 6/10 — ticks:466 score:104 win:false
+Round 7/10 — ticks:1140 score:0 win:false
+Round 8/10 — ticks:3411 score:0 win:false
+Round 9/10 — ticks:2370 score:0 win:false
+Round 10/10 — ticks:432 score:93 win:false
+Counter check passed: 880 == expected 880
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2239 score:1 win:false
+Round 2/10 — ticks:1186 score:65 win:false
+Round 3/10 — ticks:1415 score:1 win:false
+Round 4/10 — ticks:2279 score:60 win:false
+Round 5/10 — ticks:1906 score:61 win:false
+Round 6/10 — ticks:1083 score:63 win:false
+Round 7/10 — ticks:2139 score:0 win:false
+Round 8/10 — ticks:2460 score:3 win:false
+Round 9/10 — ticks:1069 score:60 win:false
+Round 10/10 — ticks:1900 score:6 win:false
+Counter check passed: 890 == expected 890
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:626 score:143 win:true
+Round 2/10 — ticks:962 score:56 win:true
+Round 3/10 — ticks:1228 score:0 win:true
+Round 4/10 — ticks:1182 score:0 win:false
+Round 5/10 — ticks:1214 score:0 win:false
+Round 6/10 — ticks:1015 score:64 win:false
+Round 7/10 — ticks:1009 score:49 win:false
+Round 8/10 — ticks:1242 score:0 win:false
+Round 9/10 — ticks:1227 score:0 win:false
+Round 10/10 — ticks:1272 score:0 win:false
+Counter check passed: 900 == expected 900
+>>> [eval] win rate: 3/10 (30%) vs Target
+=== Chunk 37/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:535 score:0 win:false
+Round 2/10 — ticks:782 score:32 win:false
+Round 3/10 — ticks:624 score:0 win:false
+Round 4/10 — ticks:551 score:0 win:false
+Round 5/10 — ticks:532 score:16 win:false
+Round 6/10 — ticks:597 score:0 win:false
+Round 7/10 — ticks:599 score:0 win:false
+Round 8/10 — ticks:528 score:0 win:false
+Round 9/10 — ticks:606 score:40 win:false
+Round 10/10 — ticks:639 score:16 win:false
+Counter check passed: 910 == expected 910
+=== Chunk 38/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:269 score:22 win:false
+Round 2/10 — ticks:273 score:14 win:false
+Round 3/10 — ticks:171 score:46 win:false
+Round 4/10 — ticks:189 score:14 win:false
+Round 5/10 — ticks:195 score:66 win:false
+Round 6/10 — ticks:208 score:64 win:false
+Round 7/10 — ticks:280 score:37 win:false
+Round 8/10 — ticks:261 score:25 win:false
+Round 9/10 — ticks:214 score:44 win:false
+Round 10/10 — ticks:147 score:12 win:false
+Counter check passed: 920 == expected 920
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1312 score:0 win:false
+Round 2/10 — ticks:1096 score:0 win:false
+Round 3/10 — ticks:1479 score:0 win:false
+Round 4/10 — ticks:856 score:0 win:false
+Round 5/10 — ticks:1184 score:0 win:false
+Round 6/10 — ticks:715 score:0 win:false
+Round 7/10 — ticks:743 score:0 win:false
+Round 8/10 — ticks:709 score:0 win:false
+Round 9/10 — ticks:1480 score:0 win:false
+Round 10/10 — ticks:791 score:0 win:false
+Counter check passed: 930 == expected 930
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2554 score:0 win:false
+Round 2/10 — ticks:2287 score:0 win:false
+Round 3/10 — ticks:2582 score:0 win:false
+Round 4/10 — ticks:1559 score:22 win:false
+Round 5/10 — ticks:2094 score:0 win:false
+Round 6/10 — ticks:2333 score:0 win:false
+Round 7/10 — ticks:2771 score:6 win:false
+Round 8/10 — ticks:2162 score:0 win:false
+Round 9/10 — ticks:2228 score:0 win:false
+Round 10/10 — ticks:2581 score:0 win:false
+Counter check passed: 940 == expected 940
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1362 score:0 win:false
+Round 2/10 — ticks:1332 score:0 win:false
+Round 3/10 — ticks:1332 score:0 win:false
+Round 4/10 — ticks:1332 score:0 win:false
+Round 5/10 — ticks:540 score:170 win:false
+Round 6/10 — ticks:540 score:171 win:false
+Round 7/10 — ticks:1332 score:0 win:false
+Round 8/10 — ticks:1332 score:0 win:false
+Round 9/10 — ticks:1332 score:0 win:false
+Round 10/10 — ticks:1332 score:0 win:false
+Counter check passed: 950 == expected 950
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 39/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:701 score:16 win:false
+Round 2/10 — ticks:397 score:0 win:false
+Round 3/10 — ticks:188 score:0 win:false
+Round 4/10 — ticks:573 score:0 win:false
+Round 5/10 — ticks:551 score:30 win:false
+Round 6/10 — ticks:306 score:21 win:false
+Round 7/10 — ticks:656 score:26 win:false
+Round 8/10 — ticks:411 score:18 win:false
+Round 9/10 — ticks:356 score:0 win:false
+Round 10/10 — ticks:713 score:60 win:false
+Counter check passed: 960 == expected 960
+=== Chunk 40/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:137 score:40 win:false
+Round 2/10 — ticks:236 score:53 win:false
+Round 3/10 — ticks:322 score:40 win:false
+Round 4/10 — ticks:239 score:56 win:false
+Round 5/10 — ticks:119 score:56 win:false
+Round 6/10 — ticks:538 score:5 win:false
+Round 7/10 — ticks:372 score:29 win:false
+Round 8/10 — ticks:234 score:59 win:false
+Round 9/10 — ticks:146 score:52 win:false
+Round 10/10 — ticks:239 score:18 win:false
+Counter check passed: 970 == expected 970
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2402 score:0 win:false
+Round 2/10 — ticks:2975 score:0 win:false
+Round 3/10 — ticks:2026 score:0 win:false
+Round 4/10 — ticks:1800 score:0 win:false
+Round 5/10 — ticks:2198 score:0 win:false
+Round 6/10 — ticks:2937 score:0 win:false
+Round 7/10 — ticks:2965 score:0 win:false
+Round 8/10 — ticks:2250 score:0 win:false
+Round 9/10 — ticks:2474 score:0 win:false
+Round 10/10 — ticks:3066 score:0 win:false
+Counter check passed: 980 == expected 980
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2214 score:4 win:false
+Round 2/10 — ticks:1852 score:67 win:false
+Round 3/10 — ticks:1500 score:60 win:false
+Round 4/10 — ticks:1948 score:61 win:true
+Round 5/10 — ticks:2311 score:10 win:false
+Round 6/10 — ticks:2119 score:15 win:false
+Round 7/10 — ticks:1957 score:4 win:false
+Round 8/10 — ticks:2139 score:6 win:false
+Round 9/10 — ticks:2680 score:2 win:false
+Round 10/10 — ticks:1880 score:7 win:false
+Counter check passed: 990 == expected 990
+>>> [eval] win rate: 1/10 (10%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1182 score:0 win:false
+Round 2/10 — ticks:1182 score:0 win:false
+Round 3/10 — ticks:1212 score:0 win:false
+Round 4/10 — ticks:1212 score:0 win:false
+Round 5/10 — ticks:1236 score:0 win:false
+Round 6/10 — ticks:1134 score:16 win:false
+Round 7/10 — ticks:876 score:91 win:false
+Round 8/10 — ticks:1111 score:26 win:false
+Round 9/10 — ticks:1020 score:43 win:false
+Round 10/10 — ticks:857 score:87 win:false
+Counter check passed: 1000 == expected 1000
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 41/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:568 score:0 win:false
+Round 2/10 — ticks:561 score:14 win:false
+Round 3/10 — ticks:568 score:0 win:false
+Round 4/10 — ticks:727 score:15 win:false
+Round 5/10 — ticks:737 score:63 win:false
+Round 6/10 — ticks:633 score:4 win:false
+Round 7/10 — ticks:645 score:16 win:false
+Round 8/10 — ticks:558 score:0 win:false
+Round 9/10 — ticks:568 score:0 win:false
+Round 10/10 — ticks:562 score:0 win:false
+Counter check passed: 1010 == expected 1010
+=== Chunk 42/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:126 score:22 win:false
+Round 2/10 — ticks:187 score:21 win:false
+Round 3/10 — ticks:262 score:55 win:false
+Round 4/10 — ticks:370 score:42 win:false
+Round 5/10 — ticks:286 score:15 win:false
+Round 6/10 — ticks:322 score:17 win:false
+Round 7/10 — ticks:182 score:64 win:false
+Round 8/10 — ticks:172 score:71 win:false
+Round 9/10 — ticks:566 score:35 win:false
+Round 10/10 — ticks:317 score:37 win:false
+Counter check passed: 1020 == expected 1020
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2002 score:0 win:false
+Round 2/10 — ticks:2052 score:0 win:false
+Round 3/10 — ticks:2296 score:0 win:false
+Round 4/10 — ticks:2444 score:0 win:false
+Round 5/10 — ticks:1955 score:0 win:false
+Round 6/10 — ticks:1482 score:6 win:false
+Round 7/10 — ticks:1669 score:0 win:false
+Round 8/10 — ticks:1445 score:20 win:false
+Round 9/10 — ticks:2076 score:0 win:false
+Round 10/10 — ticks:1836 score:0 win:false
+Counter check passed: 1030 == expected 1030
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2070 score:4 win:false
+Round 2/10 — ticks:1722 score:64 win:false
+Round 3/10 — ticks:1649 score:60 win:false
+Round 4/10 — ticks:1210 score:60 win:true
+Round 5/10 — ticks:1587 score:63 win:true
+Round 6/10 — ticks:1209 score:69 win:true
+Round 7/10 — ticks:1330 score:60 win:true
+Round 8/10 — ticks:2200 score:6 win:true
+Round 9/10 — ticks:1033 score:60 win:true
+Round 10/10 — ticks:1485 score:61 win:true
+Counter check passed: 1040 == expected 1040
+>>> [eval] win rate: 7/10 (70%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:817 score:89 win:true
+Round 2/10 — ticks:1151 score:0 win:false
+Round 3/10 — ticks:1152 score:0 win:false
+Round 4/10 — ticks:1152 score:0 win:false
+Round 5/10 — ticks:1040 score:19 win:false
+Round 6/10 — ticks:1074 score:22 win:false
+Round 7/10 — ticks:1122 score:0 win:false
+Round 8/10 — ticks:1138 score:0 win:false
+Round 9/10 — ticks:1086 score:25 win:false
+Round 10/10 — ticks:887 score:65 win:false
+Counter check passed: 1050 == expected 1050
+>>> [eval] win rate: 1/10 (10%) vs Target
+=== Chunk 43/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:497 score:38 win:false
+Round 2/10 — ticks:861 score:57 win:false
+Round 3/10 — ticks:673 score:23 win:false
+Round 4/10 — ticks:612 score:0 win:false
+Round 5/10 — ticks:599 score:0 win:false
+Round 6/10 — ticks:588 score:0 win:false
+Round 7/10 — ticks:592 score:0 win:false
+Round 8/10 — ticks:589 score:0 win:false
+Round 9/10 — ticks:629 score:15 win:false
+Round 10/10 — ticks:684 score:14 win:false
+Counter check passed: 1060 == expected 1060
+=== Chunk 44/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:610 score:0 win:false
+Round 2/10 — ticks:594 score:0 win:false
+Round 3/10 — ticks:607 score:0 win:false
+Round 4/10 — ticks:1024 score:44 win:false
+Round 5/10 — ticks:759 score:32 win:false
+Round 6/10 — ticks:583 score:0 win:false
+Round 7/10 — ticks:765 score:16 win:false
+Round 8/10 — ticks:741 score:14 win:false
+Round 9/10 — ticks:598 score:0 win:false
+Round 10/10 — ticks:581 score:0 win:false
+Counter check passed: 1070 == expected 1070
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1499 score:0 win:false
+Round 2/10 — ticks:1824 score:0 win:false
+Round 3/10 — ticks:1964 score:0 win:false
+Round 4/10 — ticks:1522 score:0 win:false
+Round 5/10 — ticks:1420 score:0 win:false
+Round 6/10 — ticks:1762 score:0 win:false
+Round 7/10 — ticks:2529 score:0 win:false
+Round 8/10 — ticks:3851 score:0 win:false
+Round 9/10 — ticks:1606 score:0 win:false
+Round 10/10 — ticks:1659 score:0 win:false
+Counter check passed: 1080 == expected 1080
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1832 score:4 win:false
+Round 2/10 — ticks:1562 score:2 win:false
+Round 3/10 — ticks:992 score:60 win:false
+Round 4/10 — ticks:1283 score:62 win:false
+Round 5/10 — ticks:1523 score:3 win:false
+Round 6/10 — ticks:1276 score:2 win:false
+Round 7/10 — ticks:981 score:60 win:false
+Round 8/10 — ticks:1391 score:64 win:false
+Round 9/10 — ticks:1112 score:1 win:false
+Round 10/10 — ticks:1366 score:61 win:false
+Counter check passed: 1090 == expected 1090
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1056 score:12 win:false
+Round 2/10 — ticks:1026 score:12 win:false
+Round 3/10 — ticks:1026 score:12 win:false
+Round 4/10 — ticks:1326 score:7 win:false
+Round 5/10 — ticks:1080 score:7 win:false
+Round 6/10 — ticks:1432 score:60 win:false
+Round 7/10 — ticks:898 score:62 win:false
+Round 8/10 — ticks:1116 score:18 win:false
+Round 9/10 — ticks:1048 score:15 win:false
+Round 10/10 — ticks:1060 score:16 win:false
+Counter check passed: 1100 == expected 1100
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 45/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:397 score:42 win:false
+Round 2/10 — ticks:217 score:45 win:false
+Round 3/10 — ticks:158 score:79 win:false
+Round 4/10 — ticks:317 score:85 win:false
+Round 5/10 — ticks:362 score:45 win:false
+Round 6/10 — ticks:530 score:24 win:false
+Round 7/10 — ticks:333 score:42 win:false
+Round 8/10 — ticks:254 score:11 win:false
+Round 9/10 — ticks:210 score:48 win:false
+Round 10/10 — ticks:305 score:32 win:false
+Counter check passed: 1110 == expected 1110
+=== Chunk 46/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:622 score:0 win:false
+Round 2/10 — ticks:855 score:27 win:false
+Round 3/10 — ticks:607 score:0 win:false
+Round 4/10 — ticks:611 score:0 win:false
+Round 5/10 — ticks:611 score:0 win:false
+Round 6/10 — ticks:611 score:0 win:false
+Round 7/10 — ticks:838 score:16 win:false
+Round 8/10 — ticks:633 score:1 win:false
+Round 9/10 — ticks:1065 score:61 win:false
+Round 10/10 — ticks:1054 score:53 win:false
+Counter check passed: 1120 == expected 1120
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1082 score:60 win:true
+Round 2/10 — ticks:1183 score:0 win:false
+Round 3/10 — ticks:1763 score:0 win:false
+Round 4/10 — ticks:1567 score:0 win:false
+Round 5/10 — ticks:1381 score:20 win:false
+Round 6/10 — ticks:1577 score:0 win:false
+Round 7/10 — ticks:2082 score:0 win:false
+Round 8/10 — ticks:1513 score:60 win:false
+Round 9/10 — ticks:1913 score:0 win:false
+Round 10/10 — ticks:2200 score:0 win:false
+Counter check passed: 1130 == expected 1130
+>>> [eval] win rate: 1/10 (10%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1036 score:60 win:true
+Round 2/10 — ticks:1734 score:64 win:true
+Round 3/10 — ticks:2129 score:10 win:true
+Round 4/10 — ticks:1239 score:62 win:true
+Round 5/10 — ticks:1509 score:62 win:true
+Round 6/10 — ticks:1598 score:66 win:true
+Round 7/10 — ticks:1538 score:60 win:true
+Round 8/10 — ticks:1750 score:63 win:true
+Round 9/10 — ticks:2103 score:3 win:true
+Round 10/10 — ticks:1902 score:66 win:true
+Counter check passed: 1140 == expected 1140
+>>> [eval] win rate: 10/10 (100%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1074 score:8 win:false
+Round 2/10 — ticks:925 score:46 win:false
+Round 3/10 — ticks:1122 score:0 win:false
+Round 4/10 — ticks:1113 score:8 win:false
+Round 5/10 — ticks:1332 score:84 win:false
+Round 6/10 — ticks:1019 score:23 win:false
+Round 7/10 — ticks:1451 score:0 win:false
+Round 8/10 — ticks:1032 score:28 win:false
+Round 9/10 — ticks:1050 score:32 win:false
+Round 10/10 — ticks:1164 score:32 win:false
+Counter check passed: 1150 == expected 1150
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 47/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:270 score:22 win:false
+Round 2/10 — ticks:271 score:25 win:false
+Round 3/10 — ticks:151 score:37 win:false
+Round 4/10 — ticks:204 score:30 win:false
+Round 5/10 — ticks:219 score:5 win:false
+Round 6/10 — ticks:293 score:34 win:false
+Round 7/10 — ticks:280 score:8 win:false
+Round 8/10 — ticks:257 score:65 win:false
+Round 9/10 — ticks:209 score:41 win:false
+Round 10/10 — ticks:285 score:41 win:false
+Counter check passed: 1160 == expected 1160
+=== Chunk 48/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:804 score:1 win:false
+Round 2/10 — ticks:724 score:2 win:false
+Round 3/10 — ticks:609 score:0 win:false
+Round 4/10 — ticks:818 score:17 win:false
+Round 5/10 — ticks:721 score:15 win:false
+Round 6/10 — ticks:527 score:0 win:false
+Round 7/10 — ticks:823 score:33 win:false
+Round 8/10 — ticks:712 score:19 win:false
+Round 9/10 — ticks:637 score:2 win:false
+Round 10/10 — ticks:530 score:1 win:false
+Counter check passed: 1170 == expected 1170
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1539 score:0 win:false
+Round 2/10 — ticks:1019 score:11 win:false
+Round 3/10 — ticks:2642 score:60 win:false
+Round 4/10 — ticks:1667 score:0 win:false
+Round 5/10 — ticks:2928 score:0 win:false
+Round 6/10 — ticks:2801 score:60 win:false
+Round 7/10 — ticks:1868 score:60 win:false
+Round 8/10 — ticks:2326 score:60 win:false
+Round 9/10 — ticks:1793 score:0 win:false
+Round 10/10 — ticks:761 score:11 win:false
+Counter check passed: 1180 == expected 1180
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1436 score:62 win:true
+Round 2/10 — ticks:1262 score:60 win:true
+Round 3/10 — ticks:2391 score:0 win:false
+Round 4/10 — ticks:2608 score:0 win:false
+Round 5/10 — ticks:1579 score:60 win:false
+Round 6/10 — ticks:1482 score:62 win:false
+Round 7/10 — ticks:2841 score:0 win:false
+Round 8/10 — ticks:2507 score:0 win:false
+Round 9/10 — ticks:2599 score:1 win:false
+Round 10/10 — ticks:2660 score:0 win:false
+Counter check passed: 1190 == expected 1190
+>>> [eval] win rate: 2/10 (20%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1422 score:60 win:true
+Round 2/10 — ticks:1421 score:0 win:true
+Round 3/10 — ticks:1449 score:0 win:true
+Round 4/10 — ticks:1451 score:0 win:true
+Round 5/10 — ticks:1451 score:0 win:true
+Round 6/10 — ticks:1407 score:60 win:true
+Round 7/10 — ticks:1422 score:60 win:true
+Round 8/10 — ticks:1422 score:0 win:true
+Round 9/10 — ticks:1368 score:70 win:true
+Round 10/10 — ticks:1422 score:60 win:true
+Counter check passed: 1200 == expected 1200
+>>> [eval] win rate: 10/10 (100%) vs Target
+=== Chunk 49/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:544 score:1 win:false
+Round 2/10 — ticks:650 score:1 win:false
+Round 3/10 — ticks:857 score:107 win:false
+Round 4/10 — ticks:619 score:0 win:false
+Round 5/10 — ticks:744 score:3 win:false
+Round 6/10 — ticks:917 score:62 win:false
+Round 7/10 — ticks:761 score:16 win:false
+Round 8/10 — ticks:614 score:2 win:false
+Round 9/10 — ticks:637 score:16 win:false
+Round 10/10 — ticks:615 score:0 win:false
+Counter check passed: 1210 == expected 1210
+=== Chunk 50/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:606 score:0 win:false
+Round 2/10 — ticks:857 score:17 win:false
+Round 3/10 — ticks:667 score:0 win:false
+Round 4/10 — ticks:641 score:0 win:false
+Round 5/10 — ticks:599 score:0 win:false
+Round 6/10 — ticks:845 score:1 win:false
+Round 7/10 — ticks:642 score:19 win:false
+Round 8/10 — ticks:709 score:13 win:false
+Round 9/10 — ticks:667 score:0 win:false
+Round 10/10 — ticks:551 score:15 win:false
+Counter check passed: 1220 == expected 1220
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1731 score:1 win:false
+Round 2/10 — ticks:1427 score:0 win:false
+Round 3/10 — ticks:2541 score:63 win:false
+Round 4/10 — ticks:1358 score:0 win:false
+Round 5/10 — ticks:1339 score:0 win:false
+Round 6/10 — ticks:2909 score:0 win:false
+Round 7/10 — ticks:1542 score:0 win:false
+Round 8/10 — ticks:3481 score:60 win:false
+Round 9/10 — ticks:1272 score:0 win:false
+Round 10/10 — ticks:2327 score:60 win:false
+Counter check passed: 1230 == expected 1230
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1672 score:2 win:false
+Round 2/10 — ticks:1261 score:63 win:false
+Round 3/10 — ticks:1682 score:62 win:true
+Round 4/10 — ticks:1410 score:63 win:true
+Round 5/10 — ticks:1435 score:1 win:false
+Round 6/10 — ticks:1392 score:3 win:false
+Round 7/10 — ticks:1039 score:62 win:false
+Round 8/10 — ticks:1436 score:61 win:true
+Round 9/10 — ticks:1468 score:0 win:false
+Round 10/10 — ticks:1096 score:61 win:true
+Counter check passed: 1240 == expected 1240
+>>> [eval] win rate: 4/10 (40%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1451 score:0 win:false
+Round 2/10 — ticks:1140 score:2 win:false
+Round 3/10 — ticks:972 score:12 win:false
+Round 4/10 — ticks:1197 score:6 win:false
+Round 5/10 — ticks:1007 score:0 win:false
+Round 6/10 — ticks:1007 score:0 win:false
+Round 7/10 — ticks:1020 score:3 win:false
+Round 8/10 — ticks:1021 score:0 win:false
+Round 9/10 — ticks:956 score:36 win:false
+Round 10/10 — ticks:1451 score:0 win:false
+Counter check passed: 1250 == expected 1250
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 51/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:210 score:7 win:false
+Round 2/10 — ticks:349 score:25 win:false
+Round 3/10 — ticks:331 score:11 win:false
+Round 4/10 — ticks:215 score:16 win:false
+Round 5/10 — ticks:183 score:39 win:false
+Round 6/10 — ticks:326 score:41 win:false
+Round 7/10 — ticks:376 score:7 win:false
+Round 8/10 — ticks:324 score:46 win:false
+Round 9/10 — ticks:340 score:48 win:false
+Round 10/10 — ticks:490 score:22 win:false
+Counter check passed: 1260 == expected 1260
+=== Chunk 52/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:931 score:24 win:false
+Round 2/10 — ticks:603 score:0 win:false
+Round 3/10 — ticks:633 score:0 win:false
+Round 4/10 — ticks:652 score:0 win:false
+Round 5/10 — ticks:620 score:0 win:false
+Round 6/10 — ticks:629 score:0 win:false
+Round 7/10 — ticks:633 score:0 win:false
+Round 8/10 — ticks:641 score:0 win:false
+Round 9/10 — ticks:667 score:16 win:false
+Round 10/10 — ticks:586 score:0 win:false
+Counter check passed: 1270 == expected 1270
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1067 score:60 win:true
+Round 2/10 — ticks:1317 score:60 win:true
+Round 3/10 — ticks:1577 score:0 win:true
+Round 4/10 — ticks:1650 score:0 win:false
+Round 5/10 — ticks:2328 score:60 win:false
+Round 6/10 — ticks:1325 score:0 win:false
+Round 7/10 — ticks:1451 score:0 win:false
+Round 8/10 — ticks:4437 score:0 win:false
+Round 9/10 — ticks:2531 score:0 win:false
+Round 10/10 — ticks:1314 score:0 win:false
+Counter check passed: 1280 == expected 1280
+>>> [eval] win rate: 3/10 (30%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1036 score:61 win:true
+Round 2/10 — ticks:1378 score:0 win:false
+Round 3/10 — ticks:1563 score:60 win:true
+Round 4/10 — ticks:1175 score:64 win:true
+Round 5/10 — ticks:1407 score:5 win:true
+Round 6/10 — ticks:1159 score:1 win:false
+Round 7/10 — ticks:906 score:60 win:true
+Round 8/10 — ticks:1856 score:62 win:true
+Round 9/10 — ticks:1039 score:60 win:true
+Round 10/10 — ticks:1185 score:60 win:true
+Counter check passed: 1290 == expected 1290
+>>> [eval] win rate: 8/10 (80%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1422 score:60 win:true
+Round 2/10 — ticks:1187 score:5 win:true
+Round 3/10 — ticks:1422 score:60 win:true
+Round 4/10 — ticks:1002 score:0 win:true
+Round 5/10 — ticks:984 score:8 win:false
+Round 6/10 — ticks:980 score:5 win:false
+Round 7/10 — ticks:1002 score:0 win:false
+Round 8/10 — ticks:1451 score:0 win:false
+Round 9/10 — ticks:1011 score:14 win:false
+Round 10/10 — ticks:1002 score:0 win:false
+Counter check passed: 1300 == expected 1300
+>>> [eval] win rate: 4/10 (40%) vs Target
+>>> [eval] new best composite (28.0000) -> sac_best.zip
+=== Chunk 53/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:203 score:58 win:false
+Round 2/10 — ticks:177 score:40 win:false
+Round 3/10 — ticks:410 score:37 win:false
+Round 4/10 — ticks:167 score:47 win:false
+Round 5/10 — ticks:181 score:49 win:false
+Round 6/10 — ticks:524 score:52 win:false
+Round 7/10 — ticks:324 score:27 win:false
+Round 8/10 — ticks:208 score:16 win:false
+Round 9/10 — ticks:133 score:41 win:false
+Round 10/10 — ticks:183 score:50 win:false
+Counter check passed: 1310 == expected 1310
+=== Chunk 54/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:787 score:0 win:false
+Round 2/10 — ticks:698 score:16 win:false
+Round 3/10 — ticks:349 score:15 win:false
+Round 4/10 — ticks:696 score:12 win:false
+Round 5/10 — ticks:389 score:41 win:false
+Round 6/10 — ticks:591 score:0 win:false
+Round 7/10 — ticks:709 score:12 win:false
+Round 8/10 — ticks:715 score:16 win:false
+Round 9/10 — ticks:675 score:0 win:false
+Round 10/10 — ticks:540 score:4 win:false
+Counter check passed: 1320 == expected 1320
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1722 score:0 win:false
+Round 2/10 — ticks:2227 score:0 win:false
+Round 3/10 — ticks:2094 score:0 win:false
+Round 4/10 — ticks:1921 score:0 win:false
+Round 5/10 — ticks:1719 score:0 win:false
+Round 6/10 — ticks:2049 score:0 win:false
+Round 7/10 — ticks:1966 score:0 win:false
+Round 8/10 — ticks:2061 score:0 win:false
+Round 9/10 — ticks:1941 score:0 win:false
+Round 10/10 — ticks:1640 score:60 win:false
+Counter check passed: 1330 == expected 1330
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1814 score:1 win:false
+Round 2/10 — ticks:1289 score:64 win:false
+Round 3/10 — ticks:1387 score:63 win:true
+Round 4/10 — ticks:1670 score:60 win:true
+Round 5/10 — ticks:979 score:62 win:true
+Round 6/10 — ticks:1352 score:60 win:true
+Round 7/10 — ticks:1478 score:2 win:true
+Round 8/10 — ticks:981 score:62 win:true
+Round 9/10 — ticks:1968 score:66 win:true
+Round 10/10 — ticks:1768 score:11 win:true
+Counter check passed: 1340 == expected 1340
+>>> [eval] win rate: 8/10 (80%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1020 score:32 win:false
+Round 2/10 — ticks:1128 score:5 win:false
+Round 3/10 — ticks:1092 score:0 win:false
+Round 4/10 — ticks:1140 score:9 win:false
+Round 5/10 — ticks:1451 score:0 win:false
+Round 6/10 — ticks:1047 score:26 win:false
+Round 7/10 — ticks:1179 score:23 win:false
+Round 8/10 — ticks:1116 score:18 win:false
+Round 9/10 — ticks:1123 score:0 win:false
+Round 10/10 — ticks:1122 score:0 win:false
+Counter check passed: 1350 == expected 1350
+>>> [eval] win rate: 0/10 (0%) vs Target
+>>> [eval] new best composite (33.3333) -> sac_best.zip
+=== Chunk 55/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:608 score:0 win:false
+Round 2/10 — ticks:577 score:10 win:false
+Round 3/10 — ticks:696 score:19 win:false
+Round 4/10 — ticks:637 score:0 win:false
+Round 5/10 — ticks:536 score:0 win:false
+Round 6/10 — ticks:617 score:16 win:false
+Round 7/10 — ticks:877 score:84 win:false
+Round 8/10 — ticks:628 score:16 win:false
+Round 9/10 — ticks:668 score:0 win:false
+Round 10/10 — ticks:651 score:0 win:false
+Counter check passed: 1360 == expected 1360
+=== Chunk 56/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:706 score:48 win:false
+Round 2/10 — ticks:636 score:12 win:false
+Round 3/10 — ticks:698 score:6 win:false
+Round 4/10 — ticks:333 score:3 win:false
+Round 5/10 — ticks:611 score:0 win:false
+Round 6/10 — ticks:782 score:29 win:false
+Round 7/10 — ticks:735 score:0 win:false
+Round 8/10 — ticks:778 score:1 win:false
+Round 9/10 — ticks:503 score:0 win:false
+Round 10/10 — ticks:925 score:0 win:false
+Counter check passed: 1370 == expected 1370
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2582 score:0 win:false
+Round 2/10 — ticks:1373 score:60 win:false
+Round 3/10 — ticks:2126 score:0 win:false
+Round 4/10 — ticks:1978 score:0 win:false
+Round 5/10 — ticks:2021 score:0 win:false
+Round 6/10 — ticks:1072 score:60 win:false
+Round 7/10 — ticks:2536 score:0 win:false
+Round 8/10 — ticks:2483 score:0 win:false
+Round 9/10 — ticks:1669 score:0 win:false
+Round 10/10 — ticks:1794 score:0 win:false
+Counter check passed: 1380 == expected 1380
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1443 score:64 win:true
+Round 2/10 — ticks:2319 score:62 win:true
+Round 3/10 — ticks:1980 score:6 win:false
+Round 4/10 — ticks:1870 score:66 win:true
+Round 5/10 — ticks:1168 score:61 win:true
+Round 6/10 — ticks:1439 score:3 win:false
+Round 7/10 — ticks:955 score:61 win:true
+Round 8/10 — ticks:1869 score:8 win:false
+Round 9/10 — ticks:1961 score:1 win:false
+Round 10/10 — ticks:1487 score:60 win:false
+Counter check passed: 1390 == expected 1390
+>>> [eval] win rate: 5/10 (50%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1067 score:20 win:false
+Round 2/10 — ticks:935 score:50 win:false
+Round 3/10 — ticks:893 score:56 win:false
+Round 4/10 — ticks:1116 score:13 win:false
+Round 5/10 — ticks:1050 score:25 win:false
+Round 6/10 — ticks:1074 score:21 win:false
+Round 7/10 — ticks:1032 score:35 win:false
+Round 8/10 — ticks:1088 score:12 win:false
+Round 9/10 — ticks:943 score:54 win:false
+Round 10/10 — ticks:1032 score:34 win:false
+Counter check passed: 1400 == expected 1400
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 57/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:676 score:0 win:false
+Round 2/10 — ticks:781 score:1 win:false
+Round 3/10 — ticks:667 score:16 win:false
+Round 4/10 — ticks:716 score:17 win:false
+Round 5/10 — ticks:725 score:16 win:false
+Round 6/10 — ticks:793 score:5 win:false
+Round 7/10 — ticks:979 score:27 win:false
+Round 8/10 — ticks:620 score:0 win:false
+Round 9/10 — ticks:831 score:30 win:false
+Round 10/10 — ticks:620 score:1 win:false
+Counter check passed: 1410 == expected 1410
+=== Chunk 58/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:633 score:19 win:false
+Round 2/10 — ticks:626 score:0 win:false
+Round 3/10 — ticks:760 score:0 win:false
+Round 4/10 — ticks:622 score:16 win:false
+Round 5/10 — ticks:561 score:0 win:false
+Round 6/10 — ticks:419 score:32 win:false
+Round 7/10 — ticks:669 score:0 win:false
+Round 8/10 — ticks:294 score:4 win:false
+Round 9/10 — ticks:858 score:41 win:false
+Round 10/10 — ticks:708 score:0 win:false
+Counter check passed: 1420 == expected 1420
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1460 score:0 win:false
+Round 2/10 — ticks:1856 score:0 win:false
+Round 3/10 — ticks:781 score:0 win:false
+Round 4/10 — ticks:928 score:0 win:false
+Round 5/10 — ticks:1312 score:0 win:false
+Round 6/10 — ticks:1335 score:0 win:false
+Round 7/10 — ticks:1530 score:0 win:false
+Round 8/10 — ticks:2448 score:0 win:false
+Round 9/10 — ticks:949 score:22 win:false
+Round 10/10 — ticks:2095 score:0 win:false
+Counter check passed: 1430 == expected 1430
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2098 score:29 win:false
+Round 2/10 — ticks:1459 score:63 win:false
+Round 3/10 — ticks:1301 score:60 win:true
+Round 4/10 — ticks:2197 score:60 win:true
+Round 5/10 — ticks:1997 score:5 win:false
+Round 6/10 — ticks:2208 score:5 win:false
+Round 7/10 — ticks:2024 score:7 win:false
+Round 8/10 — ticks:1168 score:4 win:false
+Round 9/10 — ticks:2178 score:1 win:false
+Round 10/10 — ticks:1482 score:71 win:false
+Counter check passed: 1440 == expected 1440
+>>> [eval] win rate: 2/10 (20%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1152 score:0 win:false
+Round 2/10 — ticks:1152 score:0 win:false
+Round 3/10 — ticks:921 score:46 win:false
+Round 4/10 — ticks:1152 score:0 win:false
+Round 5/10 — ticks:1212 score:0 win:false
+Round 6/10 — ticks:1167 score:0 win:false
+Round 7/10 — ticks:983 score:38 win:false
+Round 8/10 — ticks:915 score:49 win:false
+Round 9/10 — ticks:1062 score:24 win:false
+Round 10/10 — ticks:1122 score:0 win:false
+Counter check passed: 1450 == expected 1450
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 59/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:668 score:16 win:false
+Round 2/10 — ticks:631 score:16 win:false
+Round 3/10 — ticks:745 score:16 win:false
+Round 4/10 — ticks:654 score:1 win:false
+Round 5/10 — ticks:800 score:50 win:false
+Round 6/10 — ticks:779 score:17 win:false
+Round 7/10 — ticks:654 score:0 win:false
+Round 8/10 — ticks:586 score:0 win:false
+Round 9/10 — ticks:803 score:17 win:false
+Round 10/10 — ticks:629 score:0 win:false
+Counter check passed: 1460 == expected 1460
+=== Chunk 60/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:767 score:18 win:false
+Round 2/10 — ticks:704 score:0 win:false
+Round 3/10 — ticks:707 score:16 win:false
+Round 4/10 — ticks:756 score:34 win:false
+Round 5/10 — ticks:654 score:1 win:false
+Round 6/10 — ticks:687 score:0 win:false
+Round 7/10 — ticks:772 score:16 win:false
+Round 8/10 — ticks:723 score:16 win:false
+Round 9/10 — ticks:764 score:16 win:false
+Round 10/10 — ticks:1061 score:50 win:false
+Counter check passed: 1470 == expected 1470
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2913 score:0 win:false
+Round 2/10 — ticks:2461 score:0 win:false
+Round 3/10 — ticks:910 score:65 win:false
+Round 4/10 — ticks:2585 score:0 win:false
+Round 5/10 — ticks:2659 score:0 win:false
+Round 6/10 — ticks:2074 score:0 win:false
+Round 7/10 — ticks:1352 score:0 win:false
+Round 8/10 — ticks:3046 score:0 win:false
+Round 9/10 — ticks:3251 score:0 win:false
+Round 10/10 — ticks:3020 score:0 win:false
+Counter check passed: 1480 == expected 1480
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1698 score:1 win:false
+Round 2/10 — ticks:2013 score:3 win:false
+Round 3/10 — ticks:1649 score:2 win:false
+Round 4/10 — ticks:2198 score:1 win:false
+Round 5/10 — ticks:2061 score:7 win:false
+Round 6/10 — ticks:1186 score:60 win:false
+Round 7/10 — ticks:1300 score:60 win:false
+Round 8/10 — ticks:2062 score:4 win:false
+Round 9/10 — ticks:1506 score:62 win:false
+Round 10/10 — ticks:1471 score:3 win:false
+Counter check passed: 1490 == expected 1490
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:807 score:98 win:true
+Round 2/10 — ticks:1019 score:56 win:true
+Round 3/10 — ticks:971 score:48 win:true
+Round 4/10 — ticks:709 score:122 win:true
+Round 5/10 — ticks:911 score:67 win:true
+Round 6/10 — ticks:912 score:66 win:true
+Round 7/10 — ticks:1212 score:0 win:true
+Round 8/10 — ticks:1157 score:18 win:false
+Round 9/10 — ticks:1188 score:0 win:false
+Round 10/10 — ticks:965 score:49 win:false
+Counter check passed: 1500 == expected 1500
+>>> [eval] win rate: 7/10 (70%) vs Target
+=== Chunk 61/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:185 score:23 win:false
+Round 2/10 — ticks:283 score:11 win:false
+Round 3/10 — ticks:131 score:26 win:false
+Round 4/10 — ticks:380 score:32 win:false
+Round 5/10 — ticks:243 score:13 win:false
+Round 6/10 — ticks:251 score:53 win:false
+Round 7/10 — ticks:313 score:25 win:false
+Round 8/10 — ticks:299 score:9 win:false
+Round 9/10 — ticks:131 score:29 win:false
+Round 10/10 — ticks:329 score:107 win:false
+Counter check passed: 1510 == expected 1510
+=== Chunk 62/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:647 score:22 win:false
+Round 2/10 — ticks:700 score:1 win:false
+Round 3/10 — ticks:764 score:16 win:false
+Round 4/10 — ticks:602 score:1 win:false
+Round 5/10 — ticks:646 score:16 win:false
+Round 6/10 — ticks:696 score:0 win:false
+Round 7/10 — ticks:727 score:16 win:false
+Round 8/10 — ticks:781 score:29 win:false
+Round 9/10 — ticks:642 score:1 win:false
+Round 10/10 — ticks:630 score:1 win:false
+Counter check passed: 1520 == expected 1520
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2549 score:60 win:true
+Round 2/10 — ticks:1631 score:60 win:true
+Round 3/10 — ticks:1687 score:0 win:true
+Round 4/10 — ticks:1452 score:0 win:false
+Round 5/10 — ticks:1451 score:0 win:false
+Round 6/10 — ticks:1222 score:0 win:false
+Round 7/10 — ticks:1571 score:0 win:false
+Round 8/10 — ticks:800 score:32 win:false
+Round 9/10 — ticks:1255 score:0 win:false
+Round 10/10 — ticks:1176 score:0 win:false
+Counter check passed: 1530 == expected 1530
+>>> [eval] win rate: 3/10 (30%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1324 score:64 win:true
+Round 2/10 — ticks:1788 score:3 win:false
+Round 3/10 — ticks:1620 score:5 win:false
+Round 4/10 — ticks:1808 score:61 win:false
+Round 5/10 — ticks:2175 score:4 win:false
+Round 6/10 — ticks:1693 score:3 win:false
+Round 7/10 — ticks:2081 score:5 win:false
+Round 8/10 — ticks:911 score:63 win:false
+Round 9/10 — ticks:1389 score:61 win:false
+Round 10/10 — ticks:850 score:61 win:false
+Counter check passed: 1540 == expected 1540
+>>> [eval] win rate: 1/10 (10%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1038 score:0 win:false
+Round 2/10 — ticks:1050 score:14 win:false
+Round 3/10 — ticks:1442 score:60 win:false
+Round 4/10 — ticks:1032 score:6 win:false
+Round 5/10 — ticks:1049 score:0 win:false
+Round 6/10 — ticks:1035 score:0 win:false
+Round 7/10 — ticks:1032 score:0 win:false
+Round 8/10 — ticks:936 score:40 win:false
+Round 9/10 — ticks:1032 score:0 win:false
+Round 10/10 — ticks:1032 score:0 win:false
+Counter check passed: 1550 == expected 1550
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 63/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:542 score:0 win:false
+Round 2/10 — ticks:708 score:10 win:false
+Round 3/10 — ticks:593 score:0 win:false
+Round 4/10 — ticks:683 score:0 win:false
+Round 5/10 — ticks:610 score:0 win:false
+Round 6/10 — ticks:726 score:0 win:false
+Round 7/10 — ticks:751 score:0 win:false
+Round 8/10 — ticks:817 score:0 win:false
+Round 9/10 — ticks:590 score:9 win:false
+Round 10/10 — ticks:470 score:0 win:false
+Counter check passed: 1560 == expected 1560
+=== Chunk 64/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:847 score:32 win:false
+Round 2/10 — ticks:766 score:16 win:false
+Round 3/10 — ticks:633 score:0 win:false
+Round 4/10 — ticks:801 score:2 win:false
+Round 5/10 — ticks:547 score:0 win:false
+Round 6/10 — ticks:643 score:0 win:false
+Round 7/10 — ticks:661 score:1 win:false
+Round 8/10 — ticks:713 score:16 win:false
+Round 9/10 — ticks:781 score:0 win:false
+Round 10/10 — ticks:565 score:0 win:false
+Counter check passed: 1570 == expected 1570
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2570 score:0 win:false
+Round 2/10 — ticks:1503 score:0 win:false
+Round 3/10 — ticks:1701 score:2 win:false
+Round 4/10 — ticks:2054 score:0 win:false
+Round 5/10 — ticks:2252 score:0 win:false
+Round 6/10 — ticks:1997 score:0 win:false
+Round 7/10 — ticks:194 score:129 win:false
+Round 8/10 — ticks:2080 score:0 win:false
+Round 9/10 — ticks:2085 score:0 win:false
+Round 10/10 — ticks:230 score:44 win:false
+Counter check passed: 1580 == expected 1580
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1992 score:2 win:false
+Round 2/10 — ticks:1930 score:60 win:false
+Round 3/10 — ticks:1502 score:60 win:false
+Round 4/10 — ticks:1872 score:62 win:false
+Round 5/10 — ticks:2341 score:2 win:false
+Round 6/10 — ticks:2153 score:1 win:false
+Round 7/10 — ticks:2438 score:6 win:false
+Round 8/10 — ticks:2580 score:0 win:false
+Round 9/10 — ticks:2205 score:0 win:false
+Round 10/10 — ticks:1096 score:61 win:false
+Counter check passed: 1590 == expected 1590
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:828 score:94 win:true
+Round 2/10 — ticks:1212 score:0 win:false
+Round 3/10 — ticks:1086 score:37 win:false
+Round 4/10 — ticks:604 score:153 win:true
+Round 5/10 — ticks:1134 score:22 win:true
+Round 6/10 — ticks:1086 score:30 win:false
+Round 7/10 — ticks:1212 score:0 win:false
+Round 8/10 — ticks:1182 score:0 win:false
+Round 9/10 — ticks:1212 score:0 win:false
+Round 10/10 — ticks:1036 score:49 win:false
+Counter check passed: 1600 == expected 1600
+>>> [eval] win rate: 3/10 (30%) vs Target
+=== Chunk 65/2500: 10 rounds vs SacTwin ===
+Training: SAC_LSTM_Bot vs SacTwin for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:649 score:0 win:false
+Round 2/10 — ticks:744 score:0 win:false
+Round 3/10 — ticks:691 score:16 win:false
+Round 4/10 — ticks:948 score:16 win:false
+Round 5/10 — ticks:748 score:15 win:false
+Round 6/10 — ticks:652 score:0 win:false
+Round 7/10 — ticks:751 score:0 win:false
+Round 8/10 — ticks:592 score:0 win:false
+Round 9/10 — ticks:835 score:32 win:false
+Round 10/10 — ticks:777 score:16 win:false
+Counter check passed: 1610 == expected 1610
+=== Chunk 66/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:904 score:48 win:false
+Round 2/10 — ticks:642 score:2 win:false
+Round 3/10 — ticks:783 score:16 win:false
+Round 4/10 — ticks:727 score:0 win:false
+Round 5/10 — ticks:638 score:1 win:false
+Round 6/10 — ticks:844 score:33 win:false
+Round 7/10 — ticks:644 score:0 win:false
+Round 8/10 — ticks:803 score:46 win:false
+Round 9/10 — ticks:732 score:0 win:false
+Round 10/10 — ticks:878 score:0 win:false
+Counter check passed: 1620 == expected 1620
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2288 score:0 win:false
+Round 2/10 — ticks:2006 score:0 win:false
+Round 3/10 — ticks:1949 score:0 win:false
+Round 4/10 — ticks:2169 score:0 win:false
+Round 5/10 — ticks:1976 score:0 win:false
+Round 6/10 — ticks:2006 score:0 win:false
+Round 7/10 — ticks:1839 score:0 win:false
+Round 8/10 — ticks:2075 score:0 win:false
+Round 9/10 — ticks:1771 score:0 win:false
+Round 10/10 — ticks:2192 score:0 win:false
+Counter check passed: 1630 == expected 1630
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2490 score:0 win:false
+Round 2/10 — ticks:1868 score:1 win:false
+Round 3/10 — ticks:2121 score:3 win:false
+Round 4/10 — ticks:1803 score:9 win:false
+Round 5/10 — ticks:2162 score:4 win:false
+Round 6/10 — ticks:2348 score:0 win:false
+Round 7/10 — ticks:2069 score:1 win:false
+Round 8/10 — ticks:1872 score:2 win:false
+Round 9/10 — ticks:2524 score:9 win:false
+Round 10/10 — ticks:1677 score:60 win:false
+Counter check passed: 1640 == expected 1640
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1013 score:47 win:false
+Round 2/10 — ticks:1232 score:0 win:false
+Round 3/10 — ticks:1018 score:57 win:false
+Round 4/10 — ticks:857 score:82 win:false
+Round 5/10 — ticks:1182 score:0 win:false
+Round 6/10 — ticks:1140 score:14 win:false
+Round 7/10 — ticks:1212 score:0 win:false
+Round 8/10 — ticks:1234 score:0 win:false
+Round 9/10 — ticks:1025 score:50 win:false
+Round 10/10 — ticks:1044 score:38 win:false
+Counter check passed: 1650 == expected 1650
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 67/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:161 score:55 win:false
+Round 2/10 — ticks:166 score:51 win:false
+Round 3/10 — ticks:307 score:16 win:false
+Round 4/10 — ticks:175 score:19 win:false
+Round 5/10 — ticks:254 score:35 win:false
+Round 6/10 — ticks:183 score:41 win:false
+Round 7/10 — ticks:316 score:21 win:false
+Round 8/10 — ticks:455 score:29 win:false
+Round 9/10 — ticks:248 score:51 win:false
+Round 10/10 — ticks:157 score:26 win:false
+Counter check passed: 1660 == expected 1660
+=== Chunk 68/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:575 score:0 win:false
+Round 2/10 — ticks:529 score:0 win:false
+Round 3/10 — ticks:692 score:16 win:false
+Round 4/10 — ticks:589 score:0 win:false
+Round 5/10 — ticks:549 score:0 win:false
+Round 6/10 — ticks:603 score:0 win:false
+Round 7/10 — ticks:638 score:0 win:false
+Round 8/10 — ticks:676 score:0 win:false
+Round 9/10 — ticks:632 score:0 win:false
+Round 10/10 — ticks:921 score:7 win:false
+Counter check passed: 1670 == expected 1670
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2300 score:0 win:false
+Round 2/10 — ticks:2092 score:0 win:false
+Round 3/10 — ticks:1956 score:0 win:false
+Round 4/10 — ticks:1782 score:0 win:false
+Round 5/10 — ticks:2126 score:0 win:false
+Round 6/10 — ticks:1462 score:0 win:false
+Round 7/10 — ticks:2130 score:0 win:false
+Round 8/10 — ticks:2254 score:60 win:false
+Round 9/10 — ticks:1974 score:0 win:false
+Round 10/10 — ticks:1443 score:0 win:false
+Counter check passed: 1680 == expected 1680
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1524 score:8 win:false
+Round 2/10 — ticks:1111 score:62 win:false
+Round 3/10 — ticks:1931 score:61 win:true
+Round 4/10 — ticks:1378 score:62 win:true
+Round 5/10 — ticks:1774 score:3 win:false
+Round 6/10 — ticks:2465 score:64 win:true
+Round 7/10 — ticks:1121 score:66 win:true
+Round 8/10 — ticks:2525 score:68 win:true
+Round 9/10 — ticks:1711 score:67 win:true
+Round 10/10 — ticks:1437 score:65 win:true
+Counter check passed: 1690 == expected 1690
+>>> [eval] win rate: 7/10 (70%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1122 score:0 win:false
+Round 2/10 — ticks:1092 score:18 win:false
+Round 3/10 — ticks:1422 score:60 win:false
+Round 4/10 — ticks:1422 score:60 win:true
+Round 5/10 — ticks:1019 score:32 win:false
+Round 6/10 — ticks:1123 score:14 win:false
+Round 7/10 — ticks:1075 score:21 win:false
+Round 8/10 — ticks:1171 score:0 win:false
+Round 9/10 — ticks:966 score:48 win:false
+Round 10/10 — ticks:1150 score:22 win:false
+Counter check passed: 1700 == expected 1700
+>>> [eval] win rate: 1/10 (10%) vs Target
+=== Chunk 69/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:276 score:32 win:false
+Round 2/10 — ticks:266 score:20 win:false
+Round 3/10 — ticks:210 score:56 win:false
+Round 4/10 — ticks:814 score:71 win:false
+Round 5/10 — ticks:465 score:29 win:false
+Round 6/10 — ticks:478 score:38 win:false
+Round 7/10 — ticks:280 score:16 win:false
+Round 8/10 — ticks:572 score:20 win:false
+Round 9/10 — ticks:253 score:66 win:false
+Round 10/10 — ticks:262 score:21 win:false
+Counter check passed: 1710 == expected 1710
+=== Chunk 70/2500: 10 rounds vs SacTwin ===
+Training: SAC_LSTM_Bot vs SacTwin for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:673 score:0 win:false
+Round 2/10 — ticks:614 score:0 win:false
+Round 3/10 — ticks:592 score:0 win:false
+Round 4/10 — ticks:697 score:0 win:false
+Round 5/10 — ticks:567 score:8 win:false
+Round 6/10 — ticks:797 score:12 win:false
+Round 7/10 — ticks:613 score:0 win:false
+Round 8/10 — ticks:747 score:0 win:false
+Round 9/10 — ticks:503 score:64 win:false
+Round 10/10 — ticks:600 score:0 win:false
+Counter check passed: 1720 == expected 1720
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1991 score:0 win:false
+Round 2/10 — ticks:1012 score:60 win:false
+Round 3/10 — ticks:5534 score:0 win:false
+Round 4/10 — ticks:2077 score:14 win:false
+Round 5/10 — ticks:1984 score:0 win:false
+Round 6/10 — ticks:1851 score:0 win:false
+Round 7/10 — ticks:2193 score:0 win:false
+Round 8/10 — ticks:2023 score:0 win:false
+Round 9/10 — ticks:1490 score:60 win:false
+Round 10/10 — ticks:1931 score:0 win:false
+Counter check passed: 1730 == expected 1730
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2187 score:12 win:false
+Round 2/10 — ticks:1532 score:10 win:false
+Round 3/10 — ticks:1560 score:7 win:false
+Round 4/10 — ticks:2005 score:3 win:false
+Round 5/10 — ticks:2040 score:5 win:false
+Round 6/10 — ticks:1172 score:61 win:false
+Round 7/10 — ticks:1082 score:60 win:false
+Round 8/10 — ticks:1652 score:3 win:false
+Round 9/10 — ticks:1792 score:3 win:false
+Round 10/10 — ticks:2060 score:75 win:false
+Counter check passed: 1740 == expected 1740
+>>> [eval] win rate: 0/10 (0%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1146 score:1 win:false
+Round 2/10 — ticks:1152 score:0 win:false
+Round 3/10 — ticks:1122 score:0 win:false
+Round 4/10 — ticks:1022 score:39 win:false
+Round 5/10 — ticks:1068 score:15 win:false
+Round 6/10 — ticks:1152 score:0 win:false
+Round 7/10 — ticks:1050 score:23 win:false
+Round 8/10 — ticks:1116 score:7 win:false
+Round 9/10 — ticks:1085 score:22 win:false
+Round 10/10 — ticks:1177 score:0 win:false
+Counter check passed: 1750 == expected 1750
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 71/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:620 score:5 win:false
+Round 2/10 — ticks:594 score:0 win:false
+Round 3/10 — ticks:618 score:64 win:false
+Round 4/10 — ticks:646 score:0 win:false
+Round 5/10 — ticks:618 score:0 win:false
+Round 6/10 — ticks:623 score:0 win:false
+Round 7/10 — ticks:619 score:0 win:false
+Round 8/10 — ticks:633 score:40 win:false
+Round 9/10 — ticks:620 score:44 win:false
+Round 10/10 — ticks:583 score:0 win:false
+Counter check passed: 1760 == expected 1760
+=== Chunk 72/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:778 score:16 win:false
+Round 2/10 — ticks:593 score:0 win:false
+Round 3/10 — ticks:670 score:16 win:false
+Round 4/10 — ticks:657 score:15 win:false
+Round 5/10 — ticks:628 score:16 win:false
+Round 6/10 — ticks:651 score:1 win:false
+Round 7/10 — ticks:630 score:3 win:false
+Round 8/10 — ticks:654 score:0 win:false
+Round 9/10 — ticks:575 score:18 win:false
+Round 10/10 — ticks:601 score:0 win:false
+Counter check passed: 1770 == expected 1770
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1373 score:60 win:true
+Round 2/10 — ticks:1810 score:0 win:false
+Round 3/10 — ticks:2070 score:0 win:false
+Round 4/10 — ticks:2489 score:0 win:false
+Round 5/10 — ticks:2298 score:0 win:false
+Round 6/10 — ticks:1072 score:60 win:false
+Round 7/10 — ticks:2182 score:0 win:false
+Round 8/10 — ticks:1627 score:20 win:false
+Round 9/10 — ticks:1900 score:0 win:false
+Round 10/10 — ticks:1467 score:0 win:false
+Counter check passed: 1780 == expected 1780
+>>> [eval] win rate: 1/10 (10%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1564 score:64 win:true
+Round 2/10 — ticks:1784 score:9 win:false
+Round 3/10 — ticks:1898 score:70 win:true
+Round 4/10 — ticks:1409 score:68 win:true
+Round 5/10 — ticks:1570 score:7 win:false
+Round 6/10 — ticks:1608 score:0 win:false
+Round 7/10 — ticks:1006 score:60 win:false
+Round 8/10 — ticks:2004 score:4 win:false
+Round 9/10 — ticks:1371 score:61 win:false
+Round 10/10 — ticks:1466 score:70 win:false
+Counter check passed: 1790 == expected 1790
+>>> [eval] win rate: 3/10 (30%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1107 score:10 win:false
+Round 2/10 — ticks:1422 score:60 win:true
+Round 3/10 — ticks:984 score:31 win:false
+Round 4/10 — ticks:1152 score:0 win:false
+Round 5/10 — ticks:1008 score:38 win:false
+Round 6/10 — ticks:1166 score:1 win:false
+Round 7/10 — ticks:1178 score:0 win:false
+Round 8/10 — ticks:1051 score:38 win:false
+Round 9/10 — ticks:1144 score:7 win:false
+Round 10/10 — ticks:1236 score:6 win:false
+Counter check passed: 1800 == expected 1800
+>>> [eval] win rate: 1/10 (10%) vs Target
+=== Chunk 73/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:492 score:20 win:false
+Round 2/10 — ticks:295 score:17 win:false
+Round 3/10 — ticks:187 score:30 win:false
+Round 4/10 — ticks:296 score:23 win:false
+Round 5/10 — ticks:292 score:36 win:false
+Round 6/10 — ticks:290 score:12 win:false
+Round 7/10 — ticks:327 score:71 win:false
+Round 8/10 — ticks:216 score:32 win:false
+Round 9/10 — ticks:159 score:43 win:false
+Round 10/10 — ticks:204 score:36 win:false
+Counter check passed: 1810 == expected 1810
+=== Chunk 74/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:673 score:18 win:false
+Round 2/10 — ticks:663 score:0 win:false
+Round 3/10 — ticks:714 score:2 win:false
+Round 4/10 — ticks:590 score:3 win:false
+Round 5/10 — ticks:676 score:13 win:false
+Round 6/10 — ticks:888 score:59 win:false
+Round 7/10 — ticks:607 score:16 win:false
+Round 8/10 — ticks:729 score:34 win:false
+Round 9/10 — ticks:684 score:1 win:false
+Round 10/10 — ticks:645 score:0 win:false
+Counter check passed: 1820 == expected 1820
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2775 score:60 win:true
+Round 2/10 — ticks:1146 score:0 win:false
+Round 3/10 — ticks:1209 score:20 win:false
+Round 4/10 — ticks:1256 score:60 win:false
+Round 5/10 — ticks:1293 score:0 win:false
+Round 6/10 — ticks:1403 score:60 win:false
+Round 7/10 — ticks:3261 score:60 win:false
+Round 8/10 — ticks:1121 score:0 win:false
+Round 9/10 — ticks:1685 score:0 win:false
+Round 10/10 — ticks:1226 score:60 win:false
+Counter check passed: 1830 == expected 1830
+>>> [eval] win rate: 1/10 (10%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1497 score:64 win:true
+Round 2/10 — ticks:1234 score:61 win:true
+Round 3/10 — ticks:1659 score:5 win:true
+Round 4/10 — ticks:1578 score:63 win:true
+Round 5/10 — ticks:1268 score:66 win:true
+Round 6/10 — ticks:1309 score:60 win:true
+Round 7/10 — ticks:1781 score:60 win:true
+Round 8/10 — ticks:1492 score:63 win:true
+Round 9/10 — ticks:1006 score:60 win:true
+Round 10/10 — ticks:1121 score:0 win:true
+Counter check passed: 1840 == expected 1840
+>>> [eval] win rate: 10/10 (100%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1048 score:0 win:false
+Round 2/10 — ticks:953 score:19 win:false
+Round 3/10 — ticks:960 score:19 win:false
+Round 4/10 — ticks:1451 score:0 win:false
+Round 5/10 — ticks:918 score:29 win:false
+Round 6/10 — ticks:1032 score:12 win:false
+Round 7/10 — ticks:1248 score:4 win:false
+Round 8/10 — ticks:1100 score:19 win:false
+Round 9/10 — ticks:1052 score:0 win:false
+Round 10/10 — ticks:1132 score:12 win:false
+Counter check passed: 1850 == expected 1850
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 75/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:664 score:0 win:false
+Round 2/10 — ticks:664 score:0 win:false
+Round 3/10 — ticks:375 score:16 win:false
+Round 4/10 — ticks:875 score:0 win:false
+Round 5/10 — ticks:475 score:16 win:false
+Round 6/10 — ticks:722 score:7 win:false
+Round 7/10 — ticks:803 score:16 win:false
+Round 8/10 — ticks:691 score:0 win:false
+Round 9/10 — ticks:1032 score:0 win:false
+Round 10/10 — ticks:853 score:0 win:false
+Counter check passed: 1860 == expected 1860
+=== Chunk 76/2500: 10 rounds vs RamFire ===
+Training: SAC_LSTM_Bot vs RamFire for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:160 score:25 win:false
+Round 2/10 — ticks:318 score:18 win:false
+Round 3/10 — ticks:297 score:22 win:false
+Round 4/10 — ticks:152 score:27 win:false
+Round 5/10 — ticks:233 score:51 win:false
+Round 6/10 — ticks:132 score:21 win:false
+Round 7/10 — ticks:275 score:32 win:false
+Round 8/10 — ticks:167 score:17 win:false
+Round 9/10 — ticks:136 score:36 win:false
+Round 10/10 — ticks:463 score:39 win:false
+Counter check passed: 1870 == expected 1870
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1514 score:0 win:false
+Round 2/10 — ticks:1809 score:0 win:false
+Round 3/10 — ticks:1052 score:60 win:false
+Round 4/10 — ticks:1629 score:1 win:false
+Round 5/10 — ticks:1279 score:0 win:false
+Round 6/10 — ticks:1885 score:0 win:false
+Round 7/10 — ticks:1724 score:0 win:false
+Round 8/10 — ticks:2090 score:0 win:false
+Round 9/10 — ticks:2050 score:0 win:false
+Round 10/10 — ticks:1745 score:0 win:false
+Counter check passed: 1880 == expected 1880
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2109 score:64 win:true
+Round 2/10 — ticks:1451 score:61 win:true
+Round 3/10 — ticks:1111 score:0 win:true
+Round 4/10 — ticks:1919 score:63 win:true
+Round 5/10 — ticks:1303 score:64 win:true
+Round 6/10 — ticks:1166 score:61 win:true
+Round 7/10 — ticks:1010 score:61 win:true
+Round 8/10 — ticks:1508 score:64 win:true
+Round 9/10 — ticks:1708 score:62 win:true
+Round 10/10 — ticks:933 score:60 win:true
+Counter check passed: 1890 == expected 1890
+>>> [eval] win rate: 10/10 (100%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1064 score:16 win:false
+Round 2/10 — ticks:1026 score:16 win:false
+Round 3/10 — ticks:1056 score:23 win:false
+Round 4/10 — ticks:1074 score:0 win:false
+Round 5/10 — ticks:1086 score:0 win:false
+Round 6/10 — ticks:1078 score:1 win:false
+Round 7/10 — ticks:1102 score:2 win:false
+Round 8/10 — ticks:1267 score:24 win:false
+Round 9/10 — ticks:949 score:32 win:false
+Round 10/10 — ticks:1044 score:20 win:false
+Counter check passed: 1900 == expected 1900
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 77/2500: 10 rounds vs Target ===
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:601 score:0 win:false
+Round 2/10 — ticks:668 score:0 win:false
+Round 3/10 — ticks:641 score:0 win:false
+Round 4/10 — ticks:983 score:34 win:false
+Round 5/10 — ticks:988 score:30 win:false
+Round 6/10 — ticks:667 score:0 win:false
+Round 7/10 — ticks:621 score:0 win:false
+Round 8/10 — ticks:794 score:26 win:false
+Round 9/10 — ticks:742 score:8 win:false
+Round 10/10 — ticks:794 score:15 win:false
+Counter check passed: 1910 == expected 1910
+=== Chunk 78/2500: 10 rounds vs SacTwin ===
+Training: SAC_LSTM_Bot vs SacTwin for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:787 score:16 win:false
+Round 2/10 — ticks:615 score:1 win:false
+Round 3/10 — ticks:946 score:32 win:false
+Round 4/10 — ticks:746 score:6 win:false
+Round 5/10 — ticks:764 score:16 win:false
+Round 6/10 — ticks:758 score:16 win:false
+Round 7/10 — ticks:767 score:16 win:false
+Round 8/10 — ticks:545 score:6 win:false
+Round 9/10 — ticks:599 score:0 win:false
+Round 10/10 — ticks:752 score:12 win:false
+Counter check passed: 1920 == expected 1920
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:3592 score:0 win:false
+Round 2/10 — ticks:2422 score:0 win:false
+Round 3/10 — ticks:3670 score:0 win:false
+Round 4/10 — ticks:2942 score:0 win:false
+Round 5/10 — ticks:3241 score:0 win:false
+Round 6/10 — ticks:2701 score:0 win:false
+Round 7/10 — ticks:3118 score:0 win:false
+Round 8/10 — ticks:3351 score:0 win:false
+Round 9/10 — ticks:2493 score:0 win:false
+Round 10/10 — ticks:2943 score:0 win:false
+Counter check passed: 1930 == expected 1930
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:2193 score:6 win:false
+Round 2/10 — ticks:1409 score:61 win:false
+Round 3/10 — ticks:1547 score:61 win:false
+Round 4/10 — ticks:1844 score:62 win:true
+Round 5/10 — ticks:2623 score:3 win:false
+Round 6/10 — ticks:2058 score:6 win:false
+Round 7/10 — ticks:1853 score:60 win:false
+Round 8/10 — ticks:2144 score:3 win:false
+Round 9/10 — ticks:1382 score:60 win:false
+Round 10/10 — ticks:1547 score:1 win:false
+Counter check passed: 1940 == expected 1940
+>>> [eval] win rate: 1/10 (10%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:995 score:55 win:false
+Round 2/10 — ticks:1182 score:0 win:false
+Round 3/10 — ticks:1243 score:0 win:false
+Round 4/10 — ticks:1239 score:0 win:false
+Round 5/10 — ticks:1033 score:29 win:false
+Round 6/10 — ticks:1203 score:0 win:false
+Round 7/10 — ticks:1070 score:23 win:false
+Round 8/10 — ticks:941 score:53 win:false
+Round 9/10 — ticks:1071 score:34 win:false
+Round 10/10 — ticks:995 score:50 win:false
+Counter check passed: 1950 == expected 1950
+>>> [eval] win rate: 0/10 (0%) vs Target
+=== Chunk 79/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:680 score:20 win:false
+Round 2/10 — ticks:563 score:1 win:false
+Round 3/10 — ticks:972 score:113 win:false
+Round 4/10 — ticks:1019 score:31 win:false
+Round 5/10 — ticks:734 score:16 win:false
+Round 6/10 — ticks:772 score:0 win:false
+Round 7/10 — ticks:804 score:0 win:false
+Round 8/10 — ticks:887 score:33 win:false
+Round 9/10 — ticks:703 score:30 win:false
+Round 10/10 — ticks:764 score:12 win:false
+Counter check passed: 1960 == expected 1960
+=== Chunk 80/2500: 10 rounds vs Crazy ===
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:851 score:50 win:false
+Round 2/10 — ticks:714 score:0 win:false
+Round 3/10 — ticks:711 score:0 win:false
+Round 4/10 — ticks:768 score:15 win:false
+Round 5/10 — ticks:678 score:0 win:false
+Round 6/10 — ticks:1010 score:29 win:false
+Round 7/10 — ticks:744 score:16 win:false
+Round 8/10 — ticks:1004 score:0 win:false
+Round 9/10 — ticks:932 score:40 win:false
+Round 10/10 — ticks:706 score:16 win:false
+Counter check passed: 1970 == expected 1970
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:1457 score:4 win:false
+Round 2/10 — ticks:2061 score:60 win:false
+Round 3/10 — ticks:1411 score:0 win:false
+Round 4/10 — ticks:1158 score:0 win:false
+Round 5/10 — ticks:1452 score:0 win:false
+Round 6/10 — ticks:1112 score:60 win:false
+Round 7/10 — ticks:1295 score:6 win:false
+Round 8/10 — ticks:1293 score:0 win:false
+Round 9/10 — ticks:1263 score:0 win:false
+Round 10/10 — ticks:1694 score:0 win:false
+Counter check passed: 1980 == expected 1980
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1642 score:1 win:false
+Round 2/10 — ticks:1572 score:63 win:false
+Round 3/10 — ticks:2368 score:64 win:true
+Round 4/10 — ticks:1652 score:4 win:false
+Round 5/10 — ticks:1734 score:64 win:false
+Round 6/10 — ticks:1051 score:0 win:false
+Round 7/10 — ticks:1259 score:64 win:false
+Round 8/10 — ticks:1057 score:2 win:false
+Round 9/10 — ticks:2175 score:64 win:false
+Round 10/10 — ticks:1410 score:3 win:false
+Counter check passed: 1990 == expected 1990
+>>> [eval] win rate: 1/10 (10%) vs Crazy
+>>> [eval] 10 deterministic rounds vs Target
+Training: SAC_LSTM_Bot vs Target for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Target.tmp
+Round 1/10 — ticks:1425 score:60 win:true
+Round 2/10 — ticks:1405 score:60 win:true
+Round 3/10 — ticks:1092 score:5 win:true
+Round 4/10 — ticks:1058 score:1 win:true
+Round 5/10 — ticks:878 score:35 win:false
+Round 6/10 — ticks:1422 score:60 win:true
+Round 7/10 — ticks:1038 score:15 win:false
+Round 8/10 — ticks:942 score:34 win:false
+Round 9/10 — ticks:1020 score:8 win:false
+Round 10/10 — ticks:1402 score:60 win:false
+Counter check passed: 2000 == expected 2000
+>>> [eval] win rate: 5/10 (50%) vs Target
+=== Chunk 81/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:647 score:0 win:false
+Round 2/10 — ticks:794 score:16 win:false
+Round 3/10 — ticks:936 score:0 win:false
+Round 4/10 — ticks:254 score:0 win:false
+Round 5/10 — ticks:803 score:0 win:false
+Round 6/10 — ticks:676 score:0 win:false
+Round 7/10 — ticks:984 score:16 win:false
+Round 8/10 — ticks:693 score:0 win:false
+Round 9/10 — ticks:369 score:3 win:false
+Round 10/10 — ticks:869 score:0 win:false
+Counter check passed: 2010 == expected 2010
+=== Chunk 82/2500: 10 rounds vs Corners ===
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/training_log.jsonl
+Round 1/10 — ticks:494 score:0 win:false
+Round 2/10 — ticks:770 score:5 win:false
+Round 3/10 — ticks:687 score:0 win:false
+Round 4/10 — ticks:511 score:0 win:false
+Round 5/10 — ticks:795 score:0 win:false
+Round 6/10 — ticks:705 score:0 win:false
+Round 7/10 — ticks:755 score:16 win:false
+Round 8/10 — ticks:739 score:16 win:false
+Round 9/10 — ticks:778 score:0 win:false
+Round 10/10 — ticks:485 score:16 win:false
+Counter check passed: 2020 == expected 2020
+>>> [eval] 10 deterministic rounds vs Corners
+Training: SAC_LSTM_Bot vs Corners for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Corners.tmp
+Round 1/10 — ticks:2338 score:0 win:false
+Round 2/10 — ticks:2499 score:0 win:false
+Round 3/10 — ticks:3018 score:0 win:false
+Round 4/10 — ticks:2781 score:0 win:false
+Round 5/10 — ticks:2637 score:0 win:false
+Round 6/10 — ticks:2546 score:0 win:false
+Round 7/10 — ticks:2749 score:0 win:false
+Round 8/10 — ticks:3267 score:0 win:false
+Round 9/10 — ticks:2666 score:0 win:false
+Round 10/10 — ticks:2666 score:0 win:false
+Counter check passed: 2030 == expected 2030
+>>> [eval] win rate: 0/10 (0%) vs Corners
+>>> [eval] 10 deterministic rounds vs Crazy
+Training: SAC_LSTM_Bot vs Crazy for 10 rounds
+Log: /home/davide/Projects/SirRoboGarage/SAC_LSTM_Bot/eval_log.jsonl.Crazy.tmp
+Round 1/10 — ticks:1696 score:60 win:true
diff --git a/SAC_LSTM_Bot/config.nims b/SAC_LSTM_Bot_garage/config.nims
similarity index 79%
rename from SAC_LSTM_Bot/config.nims
rename to SAC_LSTM_Bot_garage/config.nims
index 7bcea66..eb6d48f 100644
--- a/SAC_LSTM_Bot/config.nims
+++ b/SAC_LSTM_Bot_garage/config.nims
@@ -12,7 +12,7 @@ switch("path", thisDir() & "/src")
 when withDir(thisDir(), system.fileExists("nimble.paths")):
   include "nimble.paths"
 # end Nimble config
-# Use the repo-vendored Tank Royale bot API (libs/) instead of the nimble pkg.
+# Use the repo-vendored Tank Royale bot API (common_libs/) instead of the nimble pkg.
 # Must come AFTER the nimble.paths include: later --path wins the import search.
-switch("path", thisDir() & "/../libs/tankroyale_botapi")
-switch("path", thisDir() & "/../libs/radar_lock")
+switch("path", thisDir() & "/../common_libs/tankroyale_botapi")
+switch("path", thisDir() & "/../common_libs/radar_lock")
diff --git a/SAC_LSTM_Bot_garage/docs/bozza_requisiti.md b/SAC_LSTM_Bot_garage/docs/bozza_requisiti.md
new file mode 100644
index 0000000..a91d3c4
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/docs/bozza_requisiti.md
@@ -0,0 +1,98 @@
+# Implementazione Agente DRL Recurrent-SAC (SAC-GRU) Nativo per Robocode TankRoyale
+
+ Bot nativo basato sull'algoritmo **Recurrent Soft Actor-Critic (SAC-GRU)** addestrato per combattere nell'ambiente **Robocode TankRoyale**.
+
+Il bot deve connettersi direttamente al server WebSocket di TankRoyale tramite un'interfaccia client API nativa, senza l'uso di middleware o socket intermediari extra. Il sistema deve garantire il rispetto del limite rigido di **30 ms per turno** imposto dal simulatore, eseguendo l'inferenza e l'addestramento in modo concorrente all'interno dello stesso processo.
+
+---
+
+## 1. Architettura Multi-Thread e Concorrenza (30 ms Tick Constraint)
+
+Per evitare il fenomeno degli "skipped turns", l'applicazione deve essere suddivisa in due thread principali coordinati nello stesso processo:
+
+1. **Thread 1: Realtime Async Event Loop (WebSocket Client)**
+   * Riceve i messaggi di evento dal server TankRoyale via WebSocket ad ogni tick.
+   * Estragga e normalizza il vettore di stato a 17 dimensioni ($S_t$).
+   * Esegue l'inferenza rapida dell'Actor ($< 2 \text{ ms}$) fornendo lo stato $S_t$ e lo stato nascosto corrente della GRU ($h_{t-1}$).
+   * Mappa le azioni restituite e invia immediatamente il `BotIntent` al server WebSocket.
+   * Invia la transizione $(S_t, A_t, R_t, S_{t+1}, D_t)$ a una coda/canale thread-safe (`Thread-safe Channel`).
+
+2. **Thread 2: Background Trainer Thread (SAC-GRU Engine)**
+   * Preleva le transizioni dal canale e le accumula in un **Sequential Replay Buffer**.
+   * Quando il buffer contiene un numero sufficiente di esperienze, estrae mini-batch di sequenze temporali.
+   * Esegue l'addestramento in background (Forward/Backward Pass dell'Actor, dei Dual Critic e dell'Alpha autotuning) sfruttando l'accelerazione hardware (GPU/CPU).
+   * Aggiorna periodicamente i pesi della rete Actor usata dal Thread 1 in modo thread-safe (es. scambio atomico di puntatori o mutua esclusione leggera).
+
+---
+
+## 2. Specifiche dell'Ambiente DRL (POMDP)
+
+### 2.1 Vettore di Stato ($S \in \mathbb{R}^{17}$, Normalizzato in $[-1, 1]$)
+
+* `s[0]`: Posizione X propria ($X / \text{width}$)
+* `s[1]`: Posizione Y propria ($Y / \text{height}$)
+* `s[2]`: Orientamento scafo ($[-\pi, \pi] / \pi$)
+* `s[3]`: Velocità lineare propria ($[-8, 8] / 8$)
+* `s[4]`: Orientamento cannone ($[-\pi, \pi] / \pi$)
+* `s[5]`: Orientamento radar ($[-\pi, \pi] / \pi$)
+* `s[6]`: Temperatura del cannone ($[0, 3] / 3$)
+* `s[7]`: Energia propria ($[0, 100] / 100$)
+* `s[8]`: Distanza dal muro NORD ($(\text{height} - Y) / \text{height}$)
+* `s[9]`: Distanza dal muro SUD ($Y / \text{height}$)
+* `s[10]`: Distanza dal muro EST ($(\text{width} - X) / \text{width}$)
+* `s[11]`: Distanza dal muro OVEST ($X / \text{width}$)
+* `s[12]`: Ultima distanza rilevata del nemico ($[0, \text{max\_dist}] / \text{max\_dist}$)
+* `s[13]`: Angolo relativo (bearing) del nemico ($[-\pi, \pi] / \pi$)
+* `s[14]`: Orientamento del nemico ($[-\pi, \pi] / \pi$)
+* `s[15]`: Velocità del nemico ($[-8, 8] / 8$)
+* `s[16]`: Energia residua del nemico ($[0, 100] / 100$)
+
+### 2.2 Vettore delle Azioni Continuo ($A \in \mathbb{R}^4$, Output $[-1, 1]$)
+
+* `a[0]`: Rotazione scafo $\rightarrow$ Mappato su $[-10^\circ, +10^\circ]$ per tick.
+* `a[1]`: Traslazione $\rightarrow$ Mappato su $[-8, +8]$ px/tick.
+* `a[2]`: Rotazione cannone $\rightarrow$ Mappato su $[-20^\circ, +20^\circ]$ per tick.
+* `a[3]`: Potenza di sparo $\rightarrow$ $\text{ReLU}(a_3) \times 3.0$ (Spara solo se $> 0.1$).
+
+### 2.3 Reward Function
+
+$$R_t = R_{\text{danno\_inflitto}} - R_{\text{danno\_subito}} + R_{\text{vittoria/sconfitta}} - R_{\text{muri}} - R_{\text{sparo\_vuoto}}$$
+
+* Danno inflitto: $+ (4p + 2(p - 1))$ con $p \le 3$.
+* Danno subito: $- (4p_{\text{nemico}} + 2(p_{\text{nemico}} - 1))$.
+* Muri: $-5.0$ per tick di impatto.
+* Sparo a vuoto: $-0.1 \times p$.
+* Vittoria/Sconfitta: $+20.0$ / $-10.0$.
+
+---
+
+## 3. Modelli Neurali e Meccanismi di Addestramento
+
+Tutte le reti devono integrare una cella ricorsiva **GRU (Gated Recurrent Unit)** per gestire la parziale osservabilità dell'arena (radar in rotazione):
+
+* **Recurrent Actor Network $\pi_\phi(a_t | s_{:t}, h_{t-1})$:**
+  * Feature Extractor: Linear/Dense ($17 \rightarrow 128$) + Attivazione
+  * Memory: GRU Layer (Input: $128$, Hidden: $128$)
+  * Heads: Linear/Dense ($128 \rightarrow 64$) $\rightarrow$ Outputs: $\mu \in \mathbb{R}^4$ e $\log \sigma \in \mathbb{R}^4$ (Clamped $[-20, 2]$)
+  * Sampling: Reparameterization Trick con squashing $\tanh$.
+
+* **Recurrent Dual-Critic Networks $Q_{\theta_1, \theta_2}(s_{:t}, a_{:t}, h_{t-1})$:**
+  * Fusion: Concatenazione di Stato e Azione ($17 + 4 = 21$)
+  * Feature Extractor: Linear/Dense ($21 \rightarrow 128$) + Attivazione
+  * Memory: GRU Layer (Input: $128$, Hidden: $128$)
+  * Q-Head: Linear/Dense ($128 \rightarrow 64$) $\rightarrow$ Output Q-Value scalare.
+
+* **Sequential Replay Buffer & Burn-in Strategy:**
+  * Campionamento di sequenze temporali contigue di lunghezza $L = 24$.
+  * **Burn-in ($L_{\text{burn}} = 8$ step):** I primi 8 step vengono usati unicamente per aggiornare lo stato nascosto $h$ della GRU, senza calcolo di loss o backpropagation.
+  * **Training ($L_{\text{train}} = 16$ step):** I successivi 16 step calcolano le loss dell'Actor, dei Critic e l'autotuning del parametro di temperatura $\alpha$ (con target $H_{\text{target}} = -4.0$).
+
+---
+
+## 4. Requisiti di Strutturazione del Codice
+
+Fornisci il codice sorgente completo, modulare e pronto all'uso articolato nelle seguenti componenti:
+
+1. **Model Definitions:** Classi/Strutture per Actor, Critic e Target Networks basate su GRU.
+2. **Sequential Replay Buffer:** Struttura dati thread-safe con supporto per estrazione di sequenze e gestione Burn-in.
+3. **SAC-GRU Trainer Engine:** Algoritmo di aggiornamento, calcolo loss, gradient clipping ($1.0$) e soft update ($\tau = 0.005$).
diff --git a/SAC_LSTM_Bot_garage/docs/campaign_dashboard.svg b/SAC_LSTM_Bot_garage/docs/campaign_dashboard.svg
new file mode 100644
index 0000000..ac9a4ed
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/docs/campaign_dashboard.svg
@@ -0,0 +1,229 @@
+<svg xmlns="http://www.w3.org/2000/svg" width="1400" height="1850" viewBox="0 0 1400 1850" font-family="sans-serif">
+<rect width="1400" height="1850" fill="white"/>
+<text x="700" y="32" text-anchor="middle" font-size="21" font-weight="bold">SAC-LSTM campaign dashboard - live run (current only)</text>
+<text x="700" y="56" text-anchor="middle" font-size="12" fill="#555">generated 2026-08-24 12:15:51 - auto-reloads every 60 s (open this file in Chrome)</text>
+<text x="70" y="100" font-size="15" font-weight="bold">Test matches - win % vs opponents</text>
+<text x="70" y="115" font-size="11" fill="#555">raw dots = single test matches, thick = rolling-mean-10</text>
+<text x="70" y="133" font-size="12" font-style="italic" fill="#555">What: How often the bot wins against each opponent in test battles.</text>
+<text x="70" y="149" font-size="12" font-style="italic" fill="#555">Axes: X = training progress (match number); Y = win rate, 0-100%.</text>
+<text x="70" y="165" font-size="12" font-style="italic" fill="#555">Better: higher is better.</text>
+<line x1="70" y1="784.0" x2="697" y2="784.0" stroke="#dddddd"/>
+<line x1="70" y1="664.0" x2="697" y2="664.0" stroke="#dddddd"/>
+<line x1="70" y1="544.0" x2="697" y2="544.0" stroke="#dddddd"/>
+<line x1="70" y1="424.0" x2="697" y2="424.0" stroke="#dddddd"/>
+<line x1="70" y1="304.0" x2="697" y2="304.0" stroke="#dddddd"/>
+<line x1="70" y1="184.0" x2="697" y2="184.0" stroke="#dddddd"/>
+<line x1="70" y1="784" x2="697" y2="784" stroke="black"/>
+<line x1="70" y1="784" x2="70" y2="184" stroke="black"/>
+<line x1="70.0" y1="784" x2="70.0" y2="788" stroke="black"/>
+<text x="70.0" y="801" text-anchor="middle" font-size="11">1</text>
+<line x1="697.0" y1="784" x2="697.0" y2="788" stroke="black"/>
+<text x="697.0" y="801" text-anchor="middle" font-size="11">2</text>
+<line x1="66" y1="784.0" x2="70" y2="784.0" stroke="black"/>
+<text x="63" y="788.0" text-anchor="end" font-size="11">0</text>
+<line x1="66" y1="664.0" x2="70" y2="664.0" stroke="black"/>
+<text x="63" y="668.0" text-anchor="end" font-size="11">20</text>
+<line x1="66" y1="544.0" x2="70" y2="544.0" stroke="black"/>
+<text x="63" y="548.0" text-anchor="end" font-size="11">40</text>
+<line x1="66" y1="424.0" x2="70" y2="424.0" stroke="black"/>
+<text x="63" y="428.0" text-anchor="end" font-size="11">60</text>
+<line x1="66" y1="304.0" x2="70" y2="304.0" stroke="black"/>
+<text x="63" y="308.0" text-anchor="end" font-size="11">80</text>
+<line x1="66" y1="184.0" x2="70" y2="184.0" stroke="black"/>
+<text x="63" y="188.0" text-anchor="end" font-size="11">100</text>
+<text x="383" y="817" text-anchor="middle" font-size="12">test match number (each opponent)</text>
+<text x="16" y="484" text-anchor="middle" font-size="12" transform="rotate(-90 16 484)">win rate (%)</text>
+<circle cx="70.0" cy="784.0" r="2" fill="#d62728" opacity="0.25"/>
+<circle cx="697.0" cy="784.0" r="2" fill="#d62728" opacity="0.25"/>
+<polyline points="70.0,784.0 697.0,784.0" fill="none" stroke="#d62728" stroke-width="3.5" opacity="1.0"/>
+<circle cx="70.0" cy="784.0" r="2" fill="#1f77b4" opacity="0.25"/>
+<circle cx="697.0" cy="184.0" r="2" fill="#1f77b4" opacity="0.25"/>
+<polyline points="70.0,784.0 697.0,484.0" fill="none" stroke="#1f77b4" stroke-width="3.5" opacity="1.0"/>
+<circle cx="70.0" cy="724.0" r="2" fill="#2ca02c" opacity="0.25"/>
+<circle cx="697.0" cy="784.0" r="2" fill="#2ca02c" opacity="0.25"/>
+<polyline points="70.0,724.0 697.0,754.0" fill="none" stroke="#2ca02c" stroke-width="3.5" opacity="1.0"/>
+<line x1="82" y1="836" x2="110" y2="836" stroke="#d62728" stroke-width="3"/>
+<text x="116" y="840" font-size="12">Corners - 2 evals</text>
+<line x1="82" y1="854" x2="110" y2="854" stroke="#1f77b4" stroke-width="3"/>
+<text x="116" y="858" font-size="12">Crazy - 2 evals</text>
+<line x1="82" y1="872" x2="110" y2="872" stroke="#2ca02c" stroke-width="3"/>
+<text x="116" y="876" font-size="12">Target - 2 evals</text>
+<text x="747" y="100" font-size="15" font-weight="bold">Training losses (log) &amp; alpha (linear)</text>
+<text x="747" y="115" font-size="11" fill="#555">training_metrics.jsonl - big early spikes are normal</text>
+<text x="747" y="133" font-size="12" font-style="italic" fill="#555">What: How well the brain is learning: losses should fall; alpha sets explore/exploit.</text>
+<text x="747" y="149" font-size="12" font-style="italic" fill="#555">Axes: X = training progress; left Y (log) = losses; right Y (linear) = alpha.</text>
+<text x="747" y="165" font-size="12" font-style="italic" fill="#555">Better: lower losses = better; alpha falls over time as the bot gets confident.</text>
+<line x1="747" y1="784.0" x2="1329" y2="784.0" stroke="#dddddd"/>
+<line x1="747" y1="634.0" x2="1329" y2="634.0" stroke="#dddddd"/>
+<line x1="747" y1="484.0" x2="1329" y2="484.0" stroke="#dddddd"/>
+<line x1="747" y1="334.0" x2="1329" y2="334.0" stroke="#dddddd"/>
+<line x1="747" y1="184.0" x2="1329" y2="184.0" stroke="#dddddd"/>
+<line x1="747" y1="784" x2="1329" y2="784" stroke="black"/>
+<line x1="747" y1="784" x2="747" y2="184" stroke="black"/>
+<line x1="747.0" y1="784" x2="747.0" y2="788" stroke="black"/>
+<text x="747.0" y="801" text-anchor="middle" font-size="11">1</text>
+<line x1="892.5" y1="784" x2="892.5" y2="788" stroke="black"/>
+<text x="892.5" y="801" text-anchor="middle" font-size="11">2</text>
+<line x1="1038.0" y1="784" x2="1038.0" y2="788" stroke="black"/>
+<text x="1038.0" y="801" text-anchor="middle" font-size="11">4</text>
+<line x1="1183.5" y1="784" x2="1183.5" y2="788" stroke="black"/>
+<text x="1183.5" y="801" text-anchor="middle" font-size="11">5</text>
+<line x1="1329.0" y1="784" x2="1329.0" y2="788" stroke="black"/>
+<text x="1329.0" y="801" text-anchor="middle" font-size="11">6</text>
+<line x1="743" y1="784.0" x2="747" y2="784.0" stroke="black"/>
+<text x="740" y="788.0" text-anchor="end" font-size="11">0.01</text>
+<line x1="743" y1="634.0" x2="747" y2="634.0" stroke="black"/>
+<text x="740" y="638.0" text-anchor="end" font-size="11">0.1</text>
+<line x1="743" y1="484.0" x2="747" y2="484.0" stroke="black"/>
+<text x="740" y="488.0" text-anchor="end" font-size="11">1</text>
+<line x1="743" y1="334.0" x2="747" y2="334.0" stroke="black"/>
+<text x="740" y="338.0" text-anchor="end" font-size="11">10</text>
+<line x1="743" y1="184.0" x2="747" y2="184.0" stroke="black"/>
+<text x="740" y="188.0" text-anchor="end" font-size="11">100</text>
+<text x="1038" y="817" text-anchor="middle" font-size="12">metric line number</text>
+<text x="16" y="484" text-anchor="middle" font-size="12" transform="rotate(-90 16 484)">loss (log)</text>
+<polyline points="747.0,357.9 863.4,677.5 979.8,539.5 1096.2,560.5 1212.6,408.6 1329.0,434.1" fill="none" stroke="#1f77b4" stroke-width="1.8" opacity="1.0"/>
+<polyline points="747.0,314.8 863.4,317.5 979.8,308.2 1096.2,305.6 1212.6,281.0 1329.0,282.6" fill="none" stroke="#ff7f0e" stroke-width="1.8" opacity="1.0"/>
+<line x1="747" y1="784.0" x2="1329" y2="784.0" stroke="#e9dcf5"/>
+<line x1="747" y1="634.0" x2="1329" y2="634.0" stroke="#e9dcf5"/>
+<line x1="747" y1="484.0" x2="1329" y2="484.0" stroke="#e9dcf5"/>
+<line x1="747" y1="334.0" x2="1329" y2="334.0" stroke="#e9dcf5"/>
+<line x1="747" y1="184.0" x2="1329" y2="184.0" stroke="#e9dcf5"/>
+<line x1="1329" y1="784" x2="1329" y2="184" stroke="#9467bd"/>
+<line x1="1329" y1="784.0" x2="1333" y2="784.0" stroke="#9467bd"/>
+<text x="1336" y="788.0" font-size="11" fill="#9467bd">0</text>
+<line x1="1329" y1="634.0" x2="1333" y2="634.0" stroke="#9467bd"/>
+<text x="1336" y="638.0" font-size="11" fill="#9467bd">0.25</text>
+<line x1="1329" y1="484.0" x2="1333" y2="484.0" stroke="#9467bd"/>
+<text x="1336" y="488.0" font-size="11" fill="#9467bd">0.5</text>
+<line x1="1329" y1="334.0" x2="1333" y2="334.0" stroke="#9467bd"/>
+<text x="1336" y="338.0" font-size="11" fill="#9467bd">0.75</text>
+<line x1="1329" y1="184.0" x2="1333" y2="184.0" stroke="#9467bd"/>
+<text x="1336" y="188.0" font-size="11" fill="#9467bd">1</text>
+<text x="1346" y="484" text-anchor="middle" font-size="12" fill="#9467bd" transform="rotate(90 1346 484)">alpha</text>
+<polyline points="747.0,198.8 863.4,198.2 979.8,197.6 1096.2,196.7 1212.6,195.1 1329.0,194.2" fill="none" stroke="#9467bd" stroke-width="1.8" opacity="1.0"/>
+<line x1="759" y1="836" x2="787" y2="836" stroke="#1f77b4" stroke-width="3"/>
+<text x="793" y="840" font-size="12">critic_loss</text>
+<line x1="759" y1="854" x2="787" y2="854" stroke="#ff7f0e" stroke-width="3"/>
+<text x="793" y="858" font-size="12">|actor_loss|</text>
+<line x1="759" y1="872" x2="787" y2="872" stroke="#9467bd" stroke-width="3"/>
+<text x="793" y="876" font-size="12">alpha</text>
+<text x="70" y="940" font-size="15" font-weight="bold">Throughput - games per hour</text>
+<text x="70" y="955" font-size="11" fill="#555">method: training_metrics.jsonl 'epoch' deltas; 1 row = one 10-game chunk</text>
+<text x="70" y="973" font-size="12" font-style="italic" fill="#555">What: How many games per hour the bot trains (its learning speed).</text>
+<text x="70" y="989" font-size="12" font-style="italic" fill="#555">Axes: X = training progress (10-game chunks); Y = games per hour.</text>
+<text x="70" y="1005" font-size="12" font-style="italic" fill="#555">Better: higher = faster learning; a steady line beats a spiky one.</text>
+<line x1="70" y1="1494.0" x2="697" y2="1494.0" stroke="#dddddd"/>
+<line x1="70" y1="1376.5" x2="697" y2="1376.5" stroke="#dddddd"/>
+<line x1="70" y1="1259.0" x2="697" y2="1259.0" stroke="#dddddd"/>
+<line x1="70" y1="1141.5" x2="697" y2="1141.5" stroke="#dddddd"/>
+<line x1="70" y1="1024.0" x2="697" y2="1024.0" stroke="#dddddd"/>
+<line x1="70" y1="1494" x2="697" y2="1494" stroke="black"/>
+<line x1="70" y1="1494" x2="70" y2="1024" stroke="black"/>
+<line x1="70.0" y1="1494" x2="70.0" y2="1498" stroke="black"/>
+<text x="70.0" y="1511" text-anchor="middle" font-size="11">1</text>
+<line x1="226.8" y1="1494" x2="226.8" y2="1498" stroke="black"/>
+<text x="226.8" y="1511" text-anchor="middle" font-size="11">2</text>
+<line x1="383.5" y1="1494" x2="383.5" y2="1498" stroke="black"/>
+<text x="383.5" y="1511" text-anchor="middle" font-size="11">3</text>
+<line x1="540.2" y1="1494" x2="540.2" y2="1498" stroke="black"/>
+<text x="540.2" y="1511" text-anchor="middle" font-size="11">4</text>
+<line x1="697.0" y1="1494" x2="697.0" y2="1498" stroke="black"/>
+<text x="697.0" y="1511" text-anchor="middle" font-size="11">5</text>
+<line x1="66" y1="1494.0" x2="70" y2="1494.0" stroke="black"/>
+<text x="63" y="1498.0" text-anchor="end" font-size="11">0</text>
+<line x1="66" y1="1376.5" x2="70" y2="1376.5" stroke="black"/>
+<text x="63" y="1380.5" text-anchor="end" font-size="11">1083</text>
+<line x1="66" y1="1259.0" x2="70" y2="1259.0" stroke="black"/>
+<text x="63" y="1263.0" text-anchor="end" font-size="11">2166</text>
+<line x1="66" y1="1141.5" x2="70" y2="1141.5" stroke="black"/>
+<text x="63" y="1145.5" text-anchor="end" font-size="11">3249</text>
+<line x1="66" y1="1024.0" x2="70" y2="1024.0" stroke="black"/>
+<text x="63" y="1028.0" text-anchor="end" font-size="11">4332</text>
+<text x="383" y="1527" text-anchor="middle" font-size="12">chunk interval (10-game chunks)</text>
+<text x="16" y="1259" text-anchor="middle" font-size="12" transform="rotate(-90 16 1259)">games / hour</text>
+<circle cx="70.0" cy="1066.7" r="1.6" fill="#7f7f7f" opacity="0.25"/>
+<circle cx="226.8" cy="1467.5" r="1.6" fill="#7f7f7f" opacity="0.25"/>
+<circle cx="383.5" cy="1105.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
+<circle cx="540.2" cy="1474.6" r="1.6" fill="#7f7f7f" opacity="0.25"/>
+<circle cx="697.0" cy="1161.4" r="1.6" fill="#7f7f7f" opacity="0.25"/>
+<polyline points="70.0,1066.7 226.8,1467.5 226.8,1105.4 540.2,1474.6 540.2,1161.4" fill="none" stroke="#2ca02c" stroke-width="3.5" opacity="1.0"/>
+<text x="747" y="940" font-size="15" font-weight="bold">Max score per eval cycle</text>
+<text x="747" y="955" font-size="11" fill="#555">campaign_v4_stdout.log - max of the 10 deterministic round scores per eval</text>
+<text x="747" y="973" font-size="12" font-style="italic" fill="#555">What: Best single-round score the bot managed in each test cycle.</text>
+<text x="747" y="989" font-size="12" font-style="italic" fill="#555">Axes: X = eval cycle number; Y = best score achieved.</text>
+<text x="747" y="1005" font-size="12" font-style="italic" fill="#555">Better: higher is better; a rising trend means the bot is improving.</text>
+<line x1="747" y1="1494.0" x2="1375" y2="1494.0" stroke="#dddddd"/>
+<line x1="747" y1="1376.5" x2="1375" y2="1376.5" stroke="#dddddd"/>
+<line x1="747" y1="1259.0" x2="1375" y2="1259.0" stroke="#dddddd"/>
+<line x1="747" y1="1141.5" x2="1375" y2="1141.5" stroke="#dddddd"/>
+<line x1="747" y1="1024.0" x2="1375" y2="1024.0" stroke="#dddddd"/>
+<line x1="747" y1="1494" x2="1375" y2="1494" stroke="black"/>
+<line x1="747" y1="1494" x2="747" y2="1024" stroke="black"/>
+<line x1="747.0" y1="1494" x2="747.0" y2="1498" stroke="black"/>
+<text x="747.0" y="1511" text-anchor="middle" font-size="11">1</text>
+<line x1="1375.0" y1="1494" x2="1375.0" y2="1498" stroke="black"/>
+<text x="1375.0" y="1511" text-anchor="middle" font-size="11">2</text>
+<line x1="743" y1="1494.0" x2="747" y2="1494.0" stroke="black"/>
+<text x="740" y="1498.0" text-anchor="end" font-size="11">0</text>
+<line x1="743" y1="1376.5" x2="747" y2="1376.5" stroke="black"/>
+<text x="740" y="1380.5" text-anchor="end" font-size="11">19</text>
+<line x1="743" y1="1259.0" x2="747" y2="1259.0" stroke="black"/>
+<text x="740" y="1263.0" text-anchor="end" font-size="11">38</text>
+<line x1="743" y1="1141.5" x2="747" y2="1141.5" stroke="black"/>
+<text x="740" y="1145.5" text-anchor="end" font-size="11">58</text>
+<line x1="743" y1="1024.0" x2="747" y2="1024.0" stroke="black"/>
+<text x="740" y="1028.0" text-anchor="end" font-size="11">77</text>
+<text x="1061" y="1527" text-anchor="middle" font-size="12">eval cycle number</text>
+<text x="16" y="1259" text-anchor="middle" font-size="12" transform="rotate(-90 16 1259)">best single-round score</text>
+<circle cx="747.0" cy="1127.8" r="2" fill="#d62728" opacity="0.25"/>
+<circle cx="1375.0" cy="1127.8" r="2" fill="#d62728" opacity="0.25"/>
+<polyline points="747.0,1127.8 1375.0,1127.8" fill="none" stroke="#d62728" stroke-width="2.5" opacity="1.0"/>
+<circle cx="747.0" cy="1085.0" r="2" fill="#1f77b4" opacity="0.25"/>
+<circle cx="1375.0" cy="1066.7" r="2" fill="#1f77b4" opacity="0.25"/>
+<polyline points="747.0,1085.0 1375.0,1066.7" fill="none" stroke="#1f77b4" stroke-width="2.5" opacity="1.0"/>
+<circle cx="747.0" cy="1127.8" r="2" fill="#2ca02c" opacity="0.25"/>
+<circle cx="1375.0" cy="1341.4" r="2" fill="#2ca02c" opacity="0.25"/>
+<polyline points="747.0,1127.8 1375.0,1341.4" fill="none" stroke="#2ca02c" stroke-width="2.5" opacity="1.0"/>
+<polyline points="747.0,1085.0 1375.0,1066.7" fill="none" stroke="#555555" stroke-width="2.5" opacity="1.0" stroke-dasharray="6 4"/>
+<line x1="759" y1="1546" x2="787" y2="1546" stroke="#d62728" stroke-width="3"/>
+<text x="793" y="1550" font-size="12">Corners - 2 evals</text>
+<line x1="759" y1="1564" x2="787" y2="1564" stroke="#1f77b4" stroke-width="3"/>
+<text x="793" y="1568" font-size="12">Crazy - 2 evals</text>
+<line x1="759" y1="1582" x2="787" y2="1582" stroke="#2ca02c" stroke-width="3"/>
+<text x="793" y="1586" font-size="12">Target - 2 evals</text>
+<line x1="759" y1="1600" x2="787" y2="1600" stroke="#555555" stroke-width="3"/>
+<text x="793" y="1604" font-size="12">combined max</text>
+<rect x="0" y="1645" width="1400" height="205" fill="#f2f2f2"/>
+<text x="16" y="1666" font-size="15" font-weight="bold">Reading the signs</text>
+<text x="1384" y="1665" text-anchor="end" font-size="11" fill="#666">Regenerate anytime: python3 tools/plot_progress.py</text>
+<text x="16" y="1688" font-size="13" font-weight="bold">Healthy patterns ✅</text>
+<text x="16" y="1706" font-size="12">• Both critic and actor losses trending down over time</text>
+<text x="16" y="1720" font-size="12">• Alpha decaying slowly from ~1.0,</text>
+<text x="16" y="1734" font-size="12">   then plateauing — this is normal</text>
+<text x="16" y="1748" font-size="12">• Win rates appearing and increasing in the eval panel</text>
+<text x="16" y="1762" font-size="12">• Max scores rising in the max-score panel</text>
+<text x="16" y="1776" font-size="12">• Alpha plateauing is NOT a problem — it means</text>
+<text x="16" y="1790" font-size="12">   exploration level is stable</text>
+<text x="500" y="1688" font-size="13" font-weight="bold">Warning signs ⚠️</text>
+<text x="500" y="1706" font-size="12">• Losses exploding (suddenly jumping to</text>
+<text x="500" y="1720" font-size="12">   millions or billions)</text>
+<text x="500" y="1734" font-size="12">• Win rates staying at 0% for a long time</text>
+<text x="500" y="1748" font-size="12">   after the first ~20 eval cycles</text>
+<text x="500" y="1762" font-size="12">• Alpha reaching 0 — bot stops exploring entirely</text>
+<text x="500" y="1776" font-size="12">   (gets stuck)</text>
+<text x="500" y="1790" font-size="12">• Max scores flatlining (no improvement</text>
+<text x="500" y="1804" font-size="12">   over many eval cycles)</text>
+<text x="500" y="1818" font-size="12">• Any single loss value above 1e6</text>
+<text x="985" y="1688" font-size="13" font-weight="bold">What each metric means (brief)</text>
+<text x="985" y="1706" font-size="12">• Critic loss: how wrong the bot's value</text>
+<text x="985" y="1720" font-size="12">   estimates are — should go down</text>
+<text x="985" y="1734" font-size="12">• Actor loss: how well the bot's action policy</text>
+<text x="985" y="1748" font-size="12">   is doing — should go down overall</text>
+<text x="985" y="1762" font-size="12">   (some bumps are normal)</text>
+<text x="985" y="1776" font-size="12">• Alpha: exploration-exploitation tradeoff — starts</text>
+<text x="985" y="1790" font-size="12">   high, settles at a positive value (NOT zero)</text>
+<text x="985" y="1804" font-size="12">• Max score: best score achieved per eval cycle —</text>
+<text x="985" y="1818" font-size="12">   rising trend = learning</text>
+<script type="text/javascript"><![CDATA[ setTimeout(function(){ location.reload(); }, 60000); ]]></script>
+</svg>
diff --git a/SAC_LSTM_Bot/docs/campaign_notebook.md b/SAC_LSTM_Bot_garage/docs/campaign_notebook.md
similarity index 100%
rename from SAC_LSTM_Bot/docs/campaign_notebook.md
rename to SAC_LSTM_Bot_garage/docs/campaign_notebook.md
diff --git a/SAC_LSTM_Bot_garage/eval_log.jsonl.pre-merge b/SAC_LSTM_Bot_garage/eval_log.jsonl.pre-merge
new file mode 100644
index 0000000..d1b006e
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/eval_log.jsonl.pre-merge
@@ -0,0 +1,30 @@
+{"type":"game","round":1,"ticks":1647,"score":0,"total_score":0,"win":false,"opponent":"Corners"}
+{"type":"game","round":2,"ticks":1518,"score":0,"total_score":0,"win":false,"opponent":"Corners"}
+{"type":"game","round":3,"ticks":1733,"score":0,"total_score":0,"win":false,"opponent":"Corners"}
+{"type":"game","round":4,"ticks":1731,"score":0,"total_score":0,"win":false,"opponent":"Corners"}
+{"type":"game","round":5,"ticks":1077,"score":17,"total_score":17,"win":false,"opponent":"Corners"}
+{"type":"game","round":6,"ticks":1745,"score":0,"total_score":17,"win":false,"opponent":"Corners"}
+{"type":"game","round":7,"ticks":1634,"score":0,"total_score":17,"win":false,"opponent":"Corners"}
+{"type":"game","round":8,"ticks":1710,"score":1,"total_score":18,"win":false,"opponent":"Corners"}
+{"type":"game","round":9,"ticks":2180,"score":60,"total_score":78,"win":false,"opponent":"Corners"}
+{"type":"game","round":10,"ticks":1364,"score":0,"total_score":78,"win":false,"opponent":"Corners"}
+{"type":"game","round":1,"ticks":1426,"score":68,"total_score":68,"win":true,"opponent":"Crazy"}
+{"type":"game","round":2,"ticks":1649,"score":70,"total_score":138,"win":true,"opponent":"Crazy"}
+{"type":"game","round":3,"ticks":1456,"score":64,"total_score":202,"win":true,"opponent":"Crazy"}
+{"type":"game","round":4,"ticks":1032,"score":61,"total_score":263,"win":true,"opponent":"Crazy"}
+{"type":"game","round":5,"ticks":1710,"score":1,"total_score":264,"win":true,"opponent":"Crazy"}
+{"type":"game","round":6,"ticks":941,"score":61,"total_score":325,"win":true,"opponent":"Crazy"}
+{"type":"game","round":7,"ticks":1455,"score":61,"total_score":386,"win":true,"opponent":"Crazy"}
+{"type":"game","round":8,"ticks":1323,"score":66,"total_score":452,"win":true,"opponent":"Crazy"}
+{"type":"game","round":9,"ticks":1035,"score":5,"total_score":457,"win":true,"opponent":"Crazy"}
+{"type":"game","round":10,"ticks":1457,"score":65,"total_score":522,"win":true,"opponent":"Crazy"}
+{"type":"game","round":1,"ticks":1014,"score":14,"total_score":14,"win":false,"opponent":"Target"}
+{"type":"game","round":2,"ticks":1074,"score":20,"total_score":34,"win":false,"opponent":"Target"}
+{"type":"game","round":3,"ticks":996,"score":6,"total_score":40,"win":false,"opponent":"Target"}
+{"type":"game","round":4,"ticks":1451,"score":0,"total_score":40,"win":false,"opponent":"Target"}
+{"type":"game","round":5,"ticks":970,"score":25,"total_score":65,"win":false,"opponent":"Target"}
+{"type":"game","round":6,"ticks":1451,"score":0,"total_score":65,"win":false,"opponent":"Target"}
+{"type":"game","round":7,"ticks":1044,"score":1,"total_score":66,"win":false,"opponent":"Target"}
+{"type":"game","round":8,"ticks":1119,"score":13,"total_score":79,"win":false,"opponent":"Target"}
+{"type":"game","round":9,"ticks":1184,"score":15,"total_score":94,"win":false,"opponent":"Target"}
+{"type":"game","round":10,"ticks":1028,"score":9,"total_score":103,"win":false,"opponent":"Target"}
diff --git a/SAC_LSTM_Bot/make_twin.sh b/SAC_LSTM_Bot_garage/make_twin.sh
similarity index 100%
rename from SAC_LSTM_Bot/make_twin.sh
rename to SAC_LSTM_Bot_garage/make_twin.sh
diff --git a/SAC_LSTM_Bot/sac_train.sh b/SAC_LSTM_Bot_garage/sac_train.sh
similarity index 100%
rename from SAC_LSTM_Bot/sac_train.sh
rename to SAC_LSTM_Bot_garage/sac_train.sh
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot.json b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot.json
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot.json
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot.json
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot.nim
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/.gitkeep b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/.gitkeep
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/.gitkeep
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/.gitkeep
diff --git a/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/actions b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/actions
new file mode 100755
index 0000000..5b03ab0
Binary files /dev/null and b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/actions differ
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/actions.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/actions.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/actions.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/actions.nim
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/integration.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/integration.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/integration.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/integration.nim
diff --git a/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/network b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/network
new file mode 100755
index 0000000..0322edc
Binary files /dev/null and b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/network differ
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/network.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/network.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/network.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/network.nim
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/replay_buffer.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/replay_buffer.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/replay_buffer.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/replay_buffer.nim
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/rewards.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/rewards.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/rewards.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/rewards.nim
diff --git a/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/state b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/state
new file mode 100755
index 0000000..ddc9aa0
Binary files /dev/null and b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/state differ
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/state.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/state.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/state.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/state.nim
diff --git a/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/training b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/training
new file mode 100755
index 0000000..c4e7e0e
Binary files /dev/null and b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/training differ
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/training.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/training.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/training.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/training.nim
diff --git a/SAC_LSTM_Bot/src/SAC_LSTM_Bot/weights.nim b/SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/weights.nim
similarity index 100%
rename from SAC_LSTM_Bot/src/SAC_LSTM_Bot/weights.nim
rename to SAC_LSTM_Bot_garage/src/SAC_LSTM_Bot/weights.nim
diff --git a/SAC_LSTM_Bot/tests/config.nims b/SAC_LSTM_Bot_garage/tests/config.nims
similarity index 100%
rename from SAC_LSTM_Bot/tests/config.nims
rename to SAC_LSTM_Bot_garage/tests/config.nims
diff --git a/SAC_LSTM_Bot/tests/test_actions.nim b/SAC_LSTM_Bot_garage/tests/test_actions.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_actions.nim
rename to SAC_LSTM_Bot_garage/tests/test_actions.nim
diff --git a/SAC_LSTM_Bot/tests/test_integration.nim b/SAC_LSTM_Bot_garage/tests/test_integration.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_integration.nim
rename to SAC_LSTM_Bot_garage/tests/test_integration.nim
diff --git a/SAC_LSTM_Bot/tests/test_network.nim b/SAC_LSTM_Bot_garage/tests/test_network.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_network.nim
rename to SAC_LSTM_Bot_garage/tests/test_network.nim
diff --git a/SAC_LSTM_Bot/tests/test_replay_buffer.nim b/SAC_LSTM_Bot_garage/tests/test_replay_buffer.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_replay_buffer.nim
rename to SAC_LSTM_Bot_garage/tests/test_replay_buffer.nim
diff --git a/SAC_LSTM_Bot/tests/test_rewards.nim b/SAC_LSTM_Bot_garage/tests/test_rewards.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_rewards.nim
rename to SAC_LSTM_Bot_garage/tests/test_rewards.nim
diff --git a/SAC_LSTM_Bot/tests/test_state.nim b/SAC_LSTM_Bot_garage/tests/test_state.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_state.nim
rename to SAC_LSTM_Bot_garage/tests/test_state.nim
diff --git a/SAC_LSTM_Bot/tests/test_training.nim b/SAC_LSTM_Bot_garage/tests/test_training.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_training.nim
rename to SAC_LSTM_Bot_garage/tests/test_training.nim
diff --git a/SAC_LSTM_Bot/tests/test_weights.nim b/SAC_LSTM_Bot_garage/tests/test_weights.nim
similarity index 100%
rename from SAC_LSTM_Bot/tests/test_weights.nim
rename to SAC_LSTM_Bot_garage/tests/test_weights.nim
diff --git a/SAC_LSTM_Bot_garage/tools/.watch_dashboard.lock b/SAC_LSTM_Bot_garage/tools/.watch_dashboard.lock
new file mode 100644
index 0000000..e69de29
diff --git a/SAC_LSTM_Bot/tools/plot_progress.py b/SAC_LSTM_Bot_garage/tools/plot_progress.py
similarity index 100%
rename from SAC_LSTM_Bot/tools/plot_progress.py
rename to SAC_LSTM_Bot_garage/tools/plot_progress.py
diff --git a/SAC_LSTM_Bot/tools/watch_dashboard.sh b/SAC_LSTM_Bot_garage/tools/watch_dashboard.sh
similarity index 100%
rename from SAC_LSTM_Bot/tools/watch_dashboard.sh
rename to SAC_LSTM_Bot_garage/tools/watch_dashboard.sh
diff --git a/SAC_LSTM_Bot_garage/training_metrics.jsonl.pre-merge b/SAC_LSTM_Bot_garage/training_metrics.jsonl.pre-merge
new file mode 100644
index 0000000..d6aac04
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/training_metrics.jsonl.pre-merge
@@ -0,0 +1,6 @@
+{"epoch":1787565804.965847,"steps":1,"buffer_size":24,"drained":1,"grad_steps":1,"critic_loss":6.929736,"actor_loss":13.433716,"alpha_loss":-0.0044489726,"alpha":0.9753562}
+{"epoch":1787565814.1075847,"steps":1,"buffer_size":24,"drained":1,"grad_steps":1,"critic_loss":0.05124796,"actor_loss":12.885783,"alpha_loss":0.016930042,"alpha":0.9762861}
+{"epoch":1787565961.245973,"steps":1,"buffer_size":24,"drained":1,"grad_steps":1,"critic_loss":0.4263701,"actor_loss":14.863115,"alpha_loss":0.03347406,"alpha":0.97734594}
+{"epoch":1787565971.2967772,"steps":1,"buffer_size":24,"drained":1,"grad_steps":1,"critic_loss":0.30882663,"actor_loss":15.45638,"alpha_loss":0.047085714,"alpha":0.97881573}
+{"epoch":1787566172.4240468,"steps":1,"buffer_size":24,"drained":1,"grad_steps":1,"critic_loss":3.1835616,"actor_loss":22.56434,"alpha_loss":0.117146604,"alpha":0.9814878}
+{"epoch":1787566184.1664138,"steps":1,"buffer_size":24,"drained":1,"grad_steps":1,"critic_loss":2.150265,"actor_loss":22.027954,"alpha_loss":0.07355834,"alpha":0.98294044}
diff --git a/SAC_LSTM_Bot_garage/weights/best_score.txt b/SAC_LSTM_Bot_garage/weights/best_score.txt
new file mode 100644
index 0000000..20bce2f
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights/best_score.txt
@@ -0,0 +1 @@
+33.3333
diff --git a/SAC_LSTM_Bot_garage/weights/ma_history_Corners.txt b/SAC_LSTM_Bot_garage/weights/ma_history_Corners.txt
new file mode 100644
index 0000000..56c9579
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights/ma_history_Corners.txt
@@ -0,0 +1 @@
+0 10 0 0 0 
\ No newline at end of file
diff --git a/SAC_LSTM_Bot_garage/weights/ma_history_Crazy.txt b/SAC_LSTM_Bot_garage/weights/ma_history_Crazy.txt
new file mode 100644
index 0000000..80ef797
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights/ma_history_Crazy.txt
@@ -0,0 +1 @@
+10 100 0 0 100 
\ No newline at end of file
diff --git a/SAC_LSTM_Bot_garage/weights/ma_history_Target.txt b/SAC_LSTM_Bot_garage/weights/ma_history_Target.txt
new file mode 100644
index 0000000..7a46dce
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights/ma_history_Target.txt
@@ -0,0 +1 @@
+0 0 0 10 0 
\ No newline at end of file
diff --git a/SAC_LSTM_Bot_garage/weights/round_counter.txt b/SAC_LSTM_Bot_garage/weights/round_counter.txt
new file mode 100644
index 0000000..e2d43aa
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights/round_counter.txt
@@ -0,0 +1 @@
+2395
\ No newline at end of file
diff --git a/SAC_LSTM_Bot_garage/weights/sac_best.zip b/SAC_LSTM_Bot_garage/weights/sac_best.zip
new file mode 100644
index 0000000..d9098d2
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights/sac_best.zip differ
diff --git a/SAC_LSTM_Bot_garage/weights/sac_latest.zip b/SAC_LSTM_Bot_garage/weights/sac_latest.zip
new file mode 100644
index 0000000..5768c40
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights/sac_latest.zip differ
diff --git a/SAC_LSTM_Bot_garage/weights/sac_latest.zip.tmp.e0f21a.part b/SAC_LSTM_Bot_garage/weights/sac_latest.zip.tmp.e0f21a.part
new file mode 100644
index 0000000..7a9fb02
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights/sac_latest.zip.tmp.e0f21a.part differ
diff --git a/SAC_LSTM_Bot_garage/weights/sac_latest.zip.tmp.l542ex.part b/SAC_LSTM_Bot_garage/weights/sac_latest.zip.tmp.l542ex.part
new file mode 100644
index 0000000..e04ca9d
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights/sac_latest.zip.tmp.l542ex.part differ
diff --git a/SAC_LSTM_Bot_garage/weights_smoke49_backup/best_score.txt b/SAC_LSTM_Bot_garage/weights_smoke49_backup/best_score.txt
new file mode 100644
index 0000000..573541a
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights_smoke49_backup/best_score.txt
@@ -0,0 +1 @@
+0
diff --git a/SAC_LSTM_Bot_garage/weights_smoke49_backup/round_counter.txt b/SAC_LSTM_Bot_garage/weights_smoke49_backup/round_counter.txt
new file mode 100644
index 0000000..b5db9c4
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights_smoke49_backup/round_counter.txt
@@ -0,0 +1 @@
+139
\ No newline at end of file
diff --git a/SAC_LSTM_Bot_garage/weights_smoke49_backup/sac_best.zip b/SAC_LSTM_Bot_garage/weights_smoke49_backup/sac_best.zip
new file mode 100644
index 0000000..412c051
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights_smoke49_backup/sac_best.zip differ
diff --git a/SAC_LSTM_Bot_garage/weights_smoke49_backup/sac_latest.zip b/SAC_LSTM_Bot_garage/weights_smoke49_backup/sac_latest.zip
new file mode 100644
index 0000000..ebdca94
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights_smoke49_backup/sac_latest.zip differ
diff --git a/SAC_LSTM_Bot_garage/weights_v1_archive/best_score.txt b/SAC_LSTM_Bot_garage/weights_v1_archive/best_score.txt
new file mode 100644
index 0000000..29d6383
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights_v1_archive/best_score.txt
@@ -0,0 +1 @@
+100
diff --git a/SAC_LSTM_Bot_garage/weights_v1_archive/round_counter.txt b/SAC_LSTM_Bot_garage/weights_v1_archive/round_counter.txt
new file mode 100644
index 0000000..b042c25
--- /dev/null
+++ b/SAC_LSTM_Bot_garage/weights_v1_archive/round_counter.txt
@@ -0,0 +1 @@
+38912
\ No newline at end of file
diff --git a/SAC_LSTM_Bot_garage/weights_v1_archive/sac_best.zip b/SAC_LSTM_Bot_garage/weights_v1_archive/sac_best.zip
new file mode 100644
index 0000000..bc58c8a
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights_v1_archive/sac_best.zip differ
diff --git a/SAC_LSTM_Bot_garage/weights_v1_archive/sac_latest.zip b/SAC_LSTM_Bot_garage/weights_v1_archive/sac_latest.zip
new file mode 100644
index 0000000..98b692b
Binary files /dev/null and b/SAC_LSTM_Bot_garage/weights_v1_archive/sac_latest.zip differ
diff --git a/libs/radar_lock/radar_lock.nim b/common_libs/radar_lock/radar_lock.nim
similarity index 100%
rename from libs/radar_lock/radar_lock.nim
rename to common_libs/radar_lock/radar_lock.nim
diff --git a/libs/radar_lock/radar_lock.nimble b/common_libs/radar_lock/radar_lock.nimble
similarity index 100%
rename from libs/radar_lock/radar_lock.nimble
rename to common_libs/radar_lock/radar_lock.nimble
diff --git a/libs/radar_lock/tests/test_radar_lock b/common_libs/radar_lock/tests/test_radar_lock
similarity index 100%
rename from libs/radar_lock/tests/test_radar_lock
rename to common_libs/radar_lock/tests/test_radar_lock
diff --git a/libs/radar_lock/tests/test_radar_lock.nim b/common_libs/radar_lock/tests/test_radar_lock.nim
similarity index 100%
rename from libs/radar_lock/tests/test_radar_lock.nim
rename to common_libs/radar_lock/tests/test_radar_lock.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi.nim b/common_libs/tankroyale_botapi/tankroyale_botapi.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi.nimble b/common_libs/tankroyale_botapi/tankroyale_botapi.nimble
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi.nimble
rename to common_libs/tankroyale_botapi/tankroyale_botapi.nimble
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/bot.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/bot.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/bot.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/bot.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/bot_info.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/bot_info.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/bot_info.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/bot_info.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/color.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/color.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/color.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/color.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/constants.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/constants.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/constants.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/constants.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/event_queue.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/event_queue.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/event_queue.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/event_queue.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/graphics.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/graphics.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/graphics.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/graphics.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/json_parse.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/json_parse.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/json_parse.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/json_parse.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/schemas.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/schemas.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/schemas.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/schemas.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/utils.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/utils.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/utils.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/utils.nim
diff --git a/libs/tankroyale_botapi/tankroyale_botapi/ws_client.nim b/common_libs/tankroyale_botapi/tankroyale_botapi/ws_client.nim
similarity index 100%
rename from libs/tankroyale_botapi/tankroyale_botapi/ws_client.nim
rename to common_libs/tankroyale_botapi/tankroyale_botapi/ws_client.nim
diff --git a/docs/papers/neuroevolution/2011_competitive_coevolution_neat.pdf b/docs/papers/neuroevolution/2011_competitive_coevolution_neat.pdf
new file mode 100644
index 0000000..2989629
Binary files /dev/null and b/docs/papers/neuroevolution/2011_competitive_coevolution_neat.pdf differ
diff --git a/docs/papers/neuroevolution/2017_deep_neuroevolution_uber.pdf b/docs/papers/neuroevolution/2017_deep_neuroevolution_uber.pdf
new file mode 100644
index 0000000..dc498e5
Binary files /dev/null and b/docs/papers/neuroevolution/2017_deep_neuroevolution_uber.pdf differ
diff --git a/docs/papers/neuroevolution/2017_es_scalable_alternative_rl_openai.pdf b/docs/papers/neuroevolution/2017_es_scalable_alternative_rl_openai.pdf
new file mode 100644
index 0000000..b1f6427
Binary files /dev/null and b/docs/papers/neuroevolution/2017_es_scalable_alternative_rl_openai.pdf differ
diff --git a/docs/papers/neuroevolution/2018_back_to_basics_es.pdf b/docs/papers/neuroevolution/2018_back_to_basics_es.pdf
new file mode 100644
index 0000000..19b6811
Binary files /dev/null and b/docs/papers/neuroevolution/2018_back_to_basics_es.pdf differ
diff --git a/docs/papers/neuroevolution/2018_challenges_high_dim_rl_es.pdf b/docs/papers/neuroevolution/2018_challenges_high_dim_rl_es.pdf
new file mode 100644
index 0000000..1f3505d
Binary files /dev/null and b/docs/papers/neuroevolution/2018_challenges_high_dim_rl_es.pdf differ
diff --git a/docs/papers/neuroevolution/2018_playing_atari_six_neurons.pdf b/docs/papers/neuroevolution/2018_playing_atari_six_neurons.pdf
new file mode 100644
index 0000000..39503aa
Binary files /dev/null and b/docs/papers/neuroevolution/2018_playing_atari_six_neurons.pdf differ
diff --git a/docs/papers/neuroevolution/2018_recurrent_world_models.pdf b/docs/papers/neuroevolution/2018_recurrent_world_models.pdf
new file mode 100644
index 0000000..0ffc8f6
Binary files /dev/null and b/docs/papers/neuroevolution/2018_recurrent_world_models.pdf differ
diff --git a/docs/papers/neuroevolution/2018_world_models.pdf b/docs/papers/neuroevolution/2018_world_models.pdf
new file mode 100644
index 0000000..19639ce
Binary files /dev/null and b/docs/papers/neuroevolution/2018_world_models.pdf differ
diff --git a/docs/papers/neuroevolution/2021_deep_neuroevolution_squeezes_small_networks.pdf b/docs/papers/neuroevolution/2021_deep_neuroevolution_squeezes_small_networks.pdf
new file mode 100644
index 0000000..eaa2cbe
Binary files /dev/null and b/docs/papers/neuroevolution/2021_deep_neuroevolution_squeezes_small_networks.pdf differ
diff --git a/docs/papers/neuroevolution/2023_evolving_generalist_controllers.pdf b/docs/papers/neuroevolution/2023_evolving_generalist_controllers.pdf
new file mode 100644
index 0000000..604d166
Binary files /dev/null and b/docs/papers/neuroevolution/2023_evolving_generalist_controllers.pdf differ
diff --git a/docs/papers/neuroevolution/2023_neuroevolution_recurrent_architectures.pdf b/docs/papers/neuroevolution/2023_neuroevolution_recurrent_architectures.pdf
new file mode 100644
index 0000000..1c910f0
Binary files /dev/null and b/docs/papers/neuroevolution/2023_neuroevolution_recurrent_architectures.pdf differ
diff --git a/docs/papers/neuroevolution/2025_echo_networks.pdf b/docs/papers/neuroevolution/2025_echo_networks.pdf
new file mode 100644
index 0000000..343c663
Binary files /dev/null and b/docs/papers/neuroevolution/2025_echo_networks.pdf differ
diff --git a/docs/papers/neuroevolution/2025_seq103.pdf b/docs/papers/neuroevolution/2025_seq103.pdf
new file mode 100644
index 0000000..f9bed8d
Binary files /dev/null and b/docs/papers/neuroevolution/2025_seq103.pdf differ
diff --git a/tools/battle_runner/run.sh b/tools/battle_runner/run.sh
index e827a59..fbbb812 100755
--- a/tools/battle_runner/run.sh
+++ b/tools/battle_runner/run.sh
@@ -5,7 +5,7 @@ SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
 REPO_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)"
 
 JAR="${TANK_ROYALE_JAR:-/home/davide/Projects/tank-royale/runner/examples/lib/robocode-tankroyale-runner.jar}"
-PPO_BOT_SRC="$REPO_ROOT/PPO_Bot"
+PPO_BOT_SRC="$REPO_ROOT/PPO_Bot_garage"
 
 export PPO_BOT_DIR="$PPO_BOT_SRC"
 export SAMPLE_BOTS_DIR="${SAMPLE_BOTS_DIR:-/home/davide/Projects/tank-royale/sample-bots/java/build/archive}"
diff --git a/tools/training_runner/CURRICULUM_STATE.md b/tools/training_runner/CURRICULUM_STATE.md
new file mode 100644
index 0000000..0f2b9a6
--- /dev/null
+++ b/tools/training_runner/CURRICULUM_STATE.md
@@ -0,0 +1,187 @@
+# PPO Curriculum — Running State File
+Chunk 1 of larger campaign. Source of truth for resuming.
+Bots (easy->hard order): Fire, MyFirstDroid, Target, MyFirstLeader, Crazy, MyFirstBot, PaintingBot, Corners.
+Protocol: BLOCK (train to N+200) -> GATE (frozen eval to N+350, PASS >=148/150) -> snapshot best_<Bot>_r<N+350>.
+Maintenance from 3rd mastered bot on: frozen eval 100 rounds per mastered bot; bar >=97/100; else replay.
+GATE bar: >=148/150. MAINT bar: >=97/100. Max 3 attempts per bot.
+
+## Baseline (verified pre-chunk)
+- round_counter.txt = 10905
+- Weights proven (frozen eval): Fire/Target/Crazy/MyFirstDroid/MyFirstLeader 100%,
+  MyFirstBot/PaintingBot 99%, Corners 93%, VelocityBot 10%, TrackFire/RamFire/SpinBot/Walls 0%,
+  MyFirstTeam 0% (1v5, flagged).
+- Bot launchers chmod-fixed. Logs rotate per phase. Snapshots = cp -a of weights/latest.
+
+---
+
+## Bot 1: Fire
+- block1: 2026-08-19 16:41, counter 10905 -> 11105 (200 train rounds, 200 train lines, 199W/1L, 0 NaN, no crash)
+  log: logs/train_Fire_block1_20260819_164245.jsonl
+- gate1: 2026-08-19 16:42, eval to 11255 (150 rounds, 0 train lines, 149W) -> PASS >=148
+  log: logs/eval_Fire_gate_20260819_164311.jsonl
+- SNAPSHOT best_Fire_r11255: MISSING (operator error: snapshots/ dir absent; cp failed, training overwrote weights before retry). RECOVERY below.
+- block2 (REDO): 2026-08-19 16:46, counter 11605 -> 11805 (200 train, 200W, 0 NaN). Log: logs/train_Fire_block2_20260819_164631.jsonl
+- gate2 (REDO): 2026-08-19 16:47, eval to 11955 (150 rounds, 0 train lines, 150W) -> PASS. Log: logs/eval_Fire_gate2_20260819_164740.jsonl
+- SNAPSHOT: snapshots/best_Fire_r11955 (40 files)
+- STATUS: PASS (150/150)
+- maint (post-Target): 100/100 PASS. Log: logs/eval_Fire_maint_20260819_164849.jsonl
+
+## Bot 2: MyFirstDroid
+- block1: 2026-08-19 16:43, counter 11255 -> 11455 (200 train rounds, 200 train lines, 200W, 0 NaN)
+  ANOMALY: operator Ctrl-C at round ~127 triggered run.sh crash-restart (">>> Restart #2, 73 rounds remaining"); round sequence in log continuous (11256..11455, no gaps), slice valid. Log: logs/train_MyFirstDroid_block1_20260819_164349.jsonl
+- gate1: 2026-08-19 16:44, eval to 11605 (150 rounds, 0 train lines, 150W) -> PASS. Log: logs/eval_MyFirstDroid_gate_20260819_164506.jsonl
+- SNAPSHOT: snapshots/best_MyFirstDroid_r11605 (40 files)
+- STATUS: PASS (150/150)
+- maint (post-Target): 100/100 PASS. Log: logs/eval_MyFirstDroid_maint_20260819_164919.jsonl
+
+## Bot 3: Target
+- block1: 2026-08-19 16:48, counter 11955 -> 12155 (200 train, 200W, 0 NaN). Log: logs/train_Target_block1_20260819_164823.jsonl
+- gate1: 2026-08-19 16:49, eval to 12305 (150 rounds, 0 train lines, 150W) -> PASS. Log: logs/eval_Target_gate_20260819_164928.jsonl
+- SNAPSHOT: snapshots/best_Target_r12305 (40 files)
+- STATUS: PASS (150/150)
+- maint (post-self): 100/100 PASS. Log: logs/eval_Target_maint_20260819_165033.jsonl
+- MAINTENANCE ROUND 1 (post-Target): Fire 100/100, MyFirstDroid 100/100, Target 100/100. NO regressions. No replays.
+
+## Bot 4: MyFirstLeader
+- block1: 2026-08-19 17:06, counter 12605 -> 12805 (200 train, 200W, 0 NaN). Log: logs/train_MyFirstLeader_block1_20260819_170632.jsonl
+- gate1: 2026-08-19 17:07, eval to 12955 (150 rounds, 0 train, 150W) -> PASS. Log: logs/eval_MyFirstLeader_gate_20260819_170720.jsonl
+- SNAPSHOT: snapshots/best_MyFirstLeader_r12955 (40 files)
+- STATUS: PASS (150/150)
+- maint r2 (post-self): 100/100 PASS. Log: logs/eval_MyFirstLeader_maint_20260819_171104.jsonl
+- maint r3 (post-Crazy): 100/100 PASS. Log: logs/eval_MyFirstLeader_maint_20260819_171246.jsonl
+- MAINTENANCE ROUND 2 (post-Leader): Fire 100, MyFirstDroid 100, Target 100, MyFirstLeader 100 -> NO regressions, no replays.
+
+## Bot 5: Crazy
+- block1: 2026-08-19 17:12, counter 13355 -> 13555 (200 train, 200W, 0 NaN). Log: logs/train_Crazy_block1_20260819_171246.jsonl
+- gate1: 2026-08-19 17:13, eval to 13705 (150 rounds, 0 train, 150W) -> PASS. Log: logs/eval_Crazy_gate_20260819_171346.jsonl
+- SNAPSHOT: snapshots/best_Crazy_r13705 (40 files)
+- STATUS: PASS (150/150)
+- maint r3 (post-self): 99/100 PASS. Log: logs/eval_Crazy_maint_20260819_171623.jsonl
+- MAINTENANCE ROUND 3 (post-Crazy): Fire 99, MyFirstDroid 100, Target 100, MyFirstLeader 100, Crazy 99 -> NO regressions, no replays.
+  ANOMALY (logs): mid-run log rename at 17:00:44 (before Droid maint eval finished) split Droid/Target round lines across files; win evidence intact
+  (Droid 100 game lines win, Target 100 game lines win). Logs: eval_MyFirstDroid_maint_20260819_170044.jsonl (85 round+100 game),
+  eval_Target_maint_20260819_170055.jsonl (100 game). No crash; counter continuity verified (13805->13905->14005).
+
+## Bot 6: MyFirstBot
+- block1: 2026-08-19 17:18, counter 14205 -> 14405 (200 train, 0W/200L, 0 NaN) — REGRESSION vs baseline 99%; policy loses every round.
+  Log: logs/train_MyFirstBot_block1_20260819_171827.jsonl
+- gate1: eval to 14555: 0/150 FAIL. Log: logs/eval_MyFirstBot_gate1_20260819_171920.jsonl
+- block2: 2026-08-19 17:22, counter 14555 -> 14955 (400 train, 0W/400L). Log: logs/train_MyFirstBot_block2_20260819_172215.jsonl
+- gate2: eval to 15105: 0/150 FAIL. Log: logs/eval_MyFirstBot_gate2_20260819_172327.jsonl
+- block3: 2026-08-19 17:24, counter 15105 -> 15505 (400 train, 0W/400L). Log: logs/train_MyFirstBot_block3_20260819_172448.jsonl
+- gate3: eval to 15655: 0/150 FAIL. Log: logs/eval_MyFirstBot_gate3_20260819_172625.jsonl
+- STATUS: OPEN (3 attempts, 1000 train + 450 gate rounds, 0 wins total; eval mechanicals verified: 0 train lines in gates, counter advances, no NaN). Likely policy basin shift (aggressive brawler vs stationary out-damager). Snapshot anyway: snapshots/best_MyFirstBot_r15655.
+- Chunk 2/3 to revisit (possibly via PaintingBot/Corners transfer).
+
+## Bot 7: PaintingBot
+- CATASTROPHE (previous session overshoot): run.sh ran unbounded from 15655 to 31694 (16039 rounds)
+  vs PaintingBot; 3/16027 wins (0%). Policy collapsed. Log: logs/train_PaintingBot_catastrophe_20260819_191718.jsonl
+  (0 NaN, no coredump, but weights destroyed by repeated loss signal).
+- RECOVERY: 2026-08-19 19:17 — restored weights from snapshots/best_Crazy_r13705; round_counter stays 31694.
+  ROOT CAUSE: Crazy-snapshot weights cannot beat PaintingBot (chunk training caused catastrophic forgetting;
+  baseline 99% was from pre-chunk Fire-campaign weights). best_Fire_r11955 (closest to baseline) needed.
+- block1: 2026-08-19 19:18, counter 31694 -> 31894 (200 train rounds from Crazy restore; 0/200 wins)
+  log: logs/train_PaintingBot_block1_20260819_192027.jsonl
+- gate1: 2026-08-19 19:21, eval to 32044 (150 rounds, 0 train lines, 0/150) -> FAIL
+  log: logs/eval_PaintingBot_gate1_20260819_192129.jsonl
+- RESTORE: best_Fire_r11955 (pre-chunk-adjacent, known 99% vs PaintingBot). Counter stays 32044.
+- block2: 2026-08-19 19:23, counter 32044 -> 32444 (400 train rounds, 400/400 wins, 0 NaN)
+  log: logs/train_PaintingBot_block2_20260819_192325.jsonl
+- gate2: 2026-08-19 19:24, eval to 32594 (150 rounds, 0 train lines, 146/150) -> FAIL (bar >=148)
+  log: logs/eval_PaintingBot_gate2_20260819_192401.jsonl
+- block3: 2026-08-19 19:24, counter 32594 -> 32994 (400 train rounds, 400/400 wins, 0 NaN)
+  log: logs/train_PaintingBot_block3_20260819_192507.jsonl
+- gate3: 2026-08-19 19:26, eval to 33144 (150 rounds, 0 train lines, 147/150) -> FAIL (bar >=148)
+  log: logs/eval_PaintingBot_gate3_20260819_192630.jsonl
+- STATUS: OPEN (3 attempts: block1 0/200 from Crazy, block2 400/400 from Fire, block3 400/400 from Fire;
+  gate results 0/150, 146/150, 147/150 — high variance, bar missed by 1-4 in gates 2+3).
+  Likely issue: gate runs single JVM with Corners-like random-state dependency (initial corner selection for
+  PaintingBot = random; policy may be sensitive to specific PaintingBot seed/state at eval time).
+  Snapshot: snapshots/best_PaintingBot_r33144.
+- MAINTENANCE (post-OPEN, 2026-08-19 19:27): Fire 100, MyFirstDroid 100, Target 100, MyFirstLeader 100,
+  Crazy 99 — NO regressions. No replays.
+  Logs: eval_Fire_maint_20260819_192658.jsonl, eval_MyFirstDroid_maint_20260819_XXXXXX.jsonl, etc.
+
+## Bot 8: Corners
+- NOTE: Corners bot switches corner on death when losing. Single-JVM gate has fixed initial corner;
+  policy must handle all 4 corner positions. Training (each JVM = new random corner) cycles all 4.
+- block1: 2026-08-19 19:29, counter 33644 -> 33844 (200 train rounds from PaintingBot weights; 78/200 wins,
+  declining trend: 1/50 in last 50). Training degraded policy.
+  log: logs/train_Corners_block1_20260819_192947.jsonl
+- gate1: 2026-08-19 19:30, eval to 33994 (150 rounds, 0 train, 0/150) -> FAIL
+  log: logs/eval_Corners_gate1_20260819_193050.jsonl
+- RESTORE: best_Fire_r11955 (same as PaintingBot recovery). Counter stays 33994.
+- block2: 2026-08-19 19:32, counter 33994 -> 34394 (400 train rounds from Fire restore; 231/400 wins,
+  improving: 0/50 early, 50/50 last 50; 0 NaN). Policy converges during training.
+  log: logs/train_Corners_block2_20260819_193239.jsonl
+- gate2: 2026-08-19 19:32, eval to 34544 (150 rounds, 0 train, 17/150) -> FAIL.
+  Gate starts wins then fails: 7 wins (favorable initial corner), then losses (unfavorable corners).
+  log: logs/eval_Corners_gate2_20260819_193510.jsonl
+- block3: 2026-08-19 19:35, counter 34544 -> 34944 (400 train rounds; 343/400 wins, 0/50 early,
+  50/50 last 50; 0 NaN). log: logs/train_Corners_block3_20260819_193622.jsonl
+- gate3: 2026-08-19 19:37, eval to 35094 (150 rounds, 0 train, 148/150) -> PASS (exactly at bar)
+  log: logs/eval_Corners_gate3_20260819_193705.jsonl
+- SNAPSHOT: snapshots/best_Corners_r35094 (40 files)
+- STATUS: PASS (148/150, gate3)
+- MAINTENANCE ROUND 5 (post-Corners PASS): Fire 100, MyFirstDroid 100, Target 100, MyFirstLeader 100,
+  Crazy 99 — NO regressions. No replays.
+  Final counter: 35594.
+
+---
+
+## Chunk 1 Summary
+- Mastered: Fire, MyFirstDroid, Target, MyFirstLeader, Crazy, Corners (6 bots)
+- OPEN: MyFirstBot, PaintingBot
+- Final round_counter: 35594
+- Active weights: best_Corners_r35094 restore point
+- Snapshots: best_Fire_r11955, best_MyFirstDroid_r11605, best_Target_r12305,
+  best_MyFirstLeader_r12955, best_Crazy_r13705, best_PaintingBot_r33144, best_Corners_r35094
+  (best_MyFirstBot_r15655 = OPEN snapshot)
+- Key findings: chunk training causes forgetting of PaintingBot/Corners — pre-chunk weights
+  (best_Fire_r11955 as proxy) needed to re-learn these. Corners gate is corner-position-variance
+  sensitive (single JVM, single random corner for full eval run).
+
+---
+
+## Chunk 2 (session 2026-08-19, agent: claude-code)
+
+### PaintingBot retry
+- Restored best_PaintingBot_r33144 weights
+- Training block: 200 rounds, 15.5% win rate (exploration noise; 57% in last 50 rounds)
+- Gate eval 1 (frozen, stochastic): 145/150 — FAIL (losses rounds 1-5, cold-start)
+- Training block 2: 200 more rounds, 200/200 (100%) in training mode
+- Gate eval 2: 147/150 — FAIL (losses rounds 1, 2, 6)
+- Gate eval 3: 146/150 — FAIL (losses rounds 1, 2, 5, 6)
+- Deterministic eval fix attempted (mean action instead of sampling): 0/150 — REVERTED
+  - Root cause: policy trained with Gaussian noise; mean action is degenerate
+  - PPOB_EVAL_ONLY already correct: freezes weights but keeps stochastic sampling
+- PaintingBot: OPEN (cold-start EnemyTracker history buffer = ~3-5 early-round losses; 97%+ once warmed up)
+- Snapshot: best_PaintingBot_r36603
+
+### Maintenance evals (from best_Corners_r35094 restore)
+- Fire: 100/100 PASS
+- MyFirstDroid: 100/100 PASS
+- Target: 100/100 PASS
+- MyFirstLeader: 100/100 PASS
+- Crazy: 100/100 PASS
+- Corners: 93/100 FAIL → replay 200 rounds (200/200 training) → re-eval 100/100 PASS
+
+### MyFirstBot investigation
+- 500 training rounds: 17 wins in rounds 1-18 (carry-over), then 0/482
+- Total across chunks: ~2400 rounds, near-zero learning signal
+- Root cause analysis — architectural hard counter:
+  1. No bullet-in-flight state representation → can't correlate fire decisions with dodge response
+  2. Aim space too tight (±12.5% arena around enemy) → can't aim at predicted post-dodge position
+  3. Credit assignment chain too long: fire → bullet travel → hit → dodge → needed different aim
+  4. State[15] (enemy turn rate) CAN detect the dodge, but can't ANTICIPATE it
+  5. MyFirstBot's key mechanic: 90° perpendicular turn on every bullet hit + constant gun sweep
+- MyFirstBot: OPEN (needs state/action space architecture changes, not more training)
+
+### Chunk 2 Summary
+- Mastered: Fire, MyFirstDroid, Target, MyFirstLeader, Crazy, Corners (6/8 — unchanged)
+- OPEN: PaintingBot (cold-start variance, 97%+ capable), MyFirstBot (architectural gap)
+- Final round_counter: ~38003
+- Active weights: post-maintenance with Corners replay
+- Next steps:
+  - PaintingBot: add EnemyTracker pre-population or warm-up rounds to fix cold-start
+  - MyFirstBot: add bullet-in-flight features to state vector; widen aim space; or accept as beyond current architecture
diff --git a/tools/training_runner/RunTraining.class b/tools/training_runner/RunTraining.class
new file mode 100644
index 0000000..eb4df4f
Binary files /dev/null and b/tools/training_runner/RunTraining.class differ
diff --git a/tools/training_runner/curriculum_retrain.sh b/tools/training_runner/curriculum_retrain.sh
index 3383e20..0044c59 100755
--- a/tools/training_runner/curriculum_retrain.sh
+++ b/tools/training_runner/curriculum_retrain.sh
@@ -5,7 +5,7 @@ set -uo pipefail
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
 REPO_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)"
-PPO_BOT_SRC="$REPO_ROOT/PPO_Bot"
+PPO_BOT_SRC="$REPO_ROOT/PPO_Bot_garage"
 
 # Source hyperparams (but we'll override opponent/rounds/log per phase)
 ENV_FILE="$SCRIPT_DIR/training.env"
diff --git a/tools/training_runner/generalist_train.sh b/tools/training_runner/generalist_train.sh
index 3d9fbbc..f359be3 100755
--- a/tools/training_runner/generalist_train.sh
+++ b/tools/training_runner/generalist_train.sh
@@ -32,7 +32,7 @@ export LC_ALL=C
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
 REPO_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)"
-PPO_BOT_SRC="$REPO_ROOT/PPO_Bot"
+PPO_BOT_SRC="$REPO_ROOT/PPO_Bot_garage"
 
 ENV_FILE="$SCRIPT_DIR/training.env"
 [ -f "$ENV_FILE" ] && set -a && . "$ENV_FILE" && set +a
diff --git a/tools/training_runner/logs/curriculum_retrain_summary.txt b/tools/training_runner/logs/curriculum_retrain_summary.txt
new file mode 100644
index 0000000..820b27c
--- /dev/null
+++ b/tools/training_runner/logs/curriculum_retrain_summary.txt
@@ -0,0 +1,74 @@
+=== PPO_Bot v2 curriculum retrain — mer 19 ago 2026, 20:56:46, CEST ===
+
+══ Fire  (train=200 gate=150) ══
+  [train] 200 rounds...
+=== PPO_Bot v2 curriculum retrain — mer 19 ago 2026, 21:13:27, CEST ===
+
+══ Fire  (train=200 gate=150) ══
+  [train] 200 rounds...
+=== PPO_Bot v2 curriculum retrain — mer 19 ago 2026, 21:36:42, CEST ===
+
+══ Fire  (train=200 gate=150) ══
+  [train] 200 rounds...
+=== PPO_Bot v2 curriculum retrain — gio 20 ago 2026, 09:52:58, CEST ===
+
+══ Fire  (train=200 gate=150) ══
+  [train] 200 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (0
+0/150)
+
+══ MyFirstDroid  (train=200 gate=150) ══
+  [train] 200 rounds...
+=== PPO_Bot v2 curriculum retrain — gio 20 ago 2026, 10:01:06, CEST ===
+
+══ Fire  (train=200 gate=150) ══
+  [train] 200 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (0/150)
+
+══ MyFirstDroid  (train=200 gate=150) ══
+  [train] 200 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (141/150)
+
+══ Target  (train=200 gate=150) ══
+  [train] 200 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (130/150)
+
+══ MyFirstLeader  (train=200 gate=150) ══
+  [train] 200 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  PASS (150/150)
+
+══ Crazy  (train=200 gate=150) ══
+  [train] 200 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (13/150)
+
+══ Corners  (train=400 gate=150) ══
+  [train] 400 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (0/150)
+
+══ PaintingBot  (train=400 gate=150) ══
+  [train] 400 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (1/150)
+
+══ MyFirstBot  (train=1000 gate=150) ══
+  [train] 1000 rounds...
+  [train] done
+  [eval]  150 rounds...
+  [gate]  FAIL (0/150)
+
+=== curriculum complete — gio 20 ago 2026, 10:24:24, CEST ===
diff --git a/tools/training_runner/logs/myfirstbot_train1_20260819_202905.jsonl.stdout.txt b/tools/training_runner/logs/myfirstbot_train1_20260819_202905.jsonl.stdout.txt
new file mode 100644
index 0000000..34b44d4
--- /dev/null
+++ b/tools/training_runner/logs/myfirstbot_train1_20260819_202905.jsonl.stdout.txt
@@ -0,0 +1 @@
+Error: PPO_BOT_DIR env var not set
diff --git a/tools/training_runner/logs/myfirstbot_train1_20260819_202913.jsonl.stdout.txt b/tools/training_runner/logs/myfirstbot_train1_20260819_202913.jsonl.stdout.txt
new file mode 100644
index 0000000..62f5293
--- /dev/null
+++ b/tools/training_runner/logs/myfirstbot_train1_20260819_202913.jsonl.stdout.txt
@@ -0,0 +1,503 @@
+Training: PPO_Bot vs MyFirstBot for 500 rounds
+Log: /home/davide/Projects/SirRoboGarage/tools/training_runner/logs/myfirstbot_train1_20260819_202913.jsonl
+Round 1/500 — ticks:533 score:182 win:true
+Round 2/500 — ticks:414 score:207 win:true
+Round 3/500 — ticks:604 score:201 win:true
+Round 4/500 — ticks:561 score:86 win:true
+Round 5/500 — ticks:588 score:48 win:true
+Round 6/500 — ticks:757 score:62 win:true
+Round 7/500 — ticks:526 score:18 win:true
+Round 8/500 — ticks:754 score:55 win:true
+Round 9/500 — ticks:560 score:201 win:true
+Round 10/500 — ticks:636 score:16 win:true
+Round 11/500 — ticks:603 score:64 win:true
+Round 12/500 — ticks:568 score:1 win:true
+Round 13/500 — ticks:622 score:33 win:true
+Round 14/500 — ticks:523 score:64 win:false
+Round 15/500 — ticks:454 score:181 win:true
+Round 16/500 — ticks:746 score:112 win:true
+Round 17/500 — ticks:604 score:64 win:true
+Round 18/500 — ticks:677 score:80 win:true
+Round 19/500 — ticks:663 score:80 win:false
+Round 20/500 — ticks:638 score:33 win:false
+Round 21/500 — ticks:595 score:49 win:false
+Round 22/500 — ticks:599 score:32 win:false
+Round 23/500 — ticks:607 score:40 win:false
+Round 24/500 — ticks:551 score:3 win:false
+Round 25/500 — ticks:466 score:48 win:false
+Round 26/500 — ticks:673 score:65 win:false
+Round 27/500 — ticks:409 score:77 win:false
+Round 28/500 — ticks:604 score:64 win:false
+Round 29/500 — ticks:635 score:48 win:false
+Round 30/500 — ticks:635 score:79 win:false
+Round 31/500 — ticks:792 score:98 win:false
+Round 32/500 — ticks:747 score:108 win:false
+Round 33/500 — ticks:657 score:197 win:false
+Round 34/500 — ticks:764 score:88 win:false
+Round 35/500 — ticks:635 score:51 win:false
+Round 36/500 — ticks:416 score:194 win:false
+Round 37/500 — ticks:534 score:177 win:false
+Round 38/500 — ticks:663 score:67 win:false
+Round 39/500 — ticks:597 score:66 win:false
+Round 40/500 — ticks:786 score:89 win:false
+Round 41/500 — ticks:749 score:78 win:false
+Round 42/500 — ticks:574 score:32 win:false
+Round 43/500 — ticks:569 score:70 win:false
+Round 44/500 — ticks:518 score:98 win:false
+Round 45/500 — ticks:680 score:57 win:false
+Round 46/500 — ticks:805 score:103 win:false
+Round 47/500 — ticks:678 score:64 win:false
+Round 48/500 — ticks:749 score:66 win:false
+Round 49/500 — ticks:812 score:96 win:false
+Round 50/500 — ticks:679 score:73 win:false
+Round 51/500 — ticks:661 score:48 win:false
+Round 52/500 — ticks:684 score:49 win:false
+Round 53/500 — ticks:558 score:32 win:false
+Round 54/500 — ticks:767 score:81 win:false
+Round 55/500 — ticks:694 score:50 win:false
+Round 56/500 — ticks:631 score:64 win:false
+Round 57/500 — ticks:639 score:64 win:false
+Round 58/500 — ticks:630 score:201 win:false
+Round 59/500 — ticks:589 score:196 win:false
+Round 60/500 — ticks:586 score:51 win:false
+Round 61/500 — ticks:709 score:64 win:false
+Round 62/500 — ticks:694 score:192 win:false
+Round 63/500 — ticks:529 score:200 win:false
+Round 64/500 — ticks:538 score:204 win:false
+Round 65/500 — ticks:611 score:50 win:false
+Round 66/500 — ticks:609 score:214 win:false
+Round 67/500 — ticks:479 score:0 win:false
+Round 68/500 — ticks:669 score:100 win:false
+Round 69/500 — ticks:788 score:112 win:false
+Round 70/500 — ticks:640 score:85 win:false
+Round 71/500 — ticks:423 score:62 win:false
+Round 72/500 — ticks:786 score:81 win:false
+Round 73/500 — ticks:509 score:49 win:false
+Round 74/500 — ticks:685 score:47 win:false
+Round 75/500 — ticks:310 score:55 win:false
+Round 76/500 — ticks:695 score:81 win:false
+Round 77/500 — ticks:668 score:63 win:false
+Round 78/500 — ticks:710 score:52 win:false
+Round 79/500 — ticks:611 score:9 win:false
+Round 80/500 — ticks:465 score:170 win:false
+Round 81/500 — ticks:521 score:49 win:false
+Round 82/500 — ticks:634 score:178 win:false
+Round 83/500 — ticks:787 score:97 win:false
+Round 84/500 — ticks:660 score:60 win:false
+Round 85/500 — ticks:390 score:158 win:false
+Round 86/500 — ticks:756 score:59 win:false
+Round 87/500 — ticks:607 score:33 win:false
+Round 88/500 — ticks:607 score:32 win:false
+Round 89/500 — ticks:613 score:68 win:false
+Round 90/500 — ticks:770 score:112 win:false
+Round 91/500 — ticks:483 score:48 win:false
+Round 92/500 — ticks:665 score:64 win:false
+Round 93/500 — ticks:708 score:80 win:false
+Round 94/500 — ticks:526 score:17 win:false
+Round 95/500 — ticks:641 score:39 win:false
+Round 96/500 — ticks:709 score:47 win:false
+Round 97/500 — ticks:676 score:80 win:false
+Round 98/500 — ticks:593 score:33 win:false
+Round 99/500 — ticks:839 score:97 win:false
+Round 100/500 — ticks:637 score:65 win:false
+Round 101/500 — ticks:792 score:80 win:false
+Round 102/500 — ticks:611 score:32 win:false
+Round 103/500 — ticks:639 score:32 win:false
+Round 104/500 — ticks:622 score:43 win:false
+Round 105/500 — ticks:396 score:171 win:false
+Round 106/500 — ticks:674 score:85 win:false
+Round 107/500 — ticks:410 score:43 win:false
+Round 108/500 — ticks:634 score:33 win:false
+Round 109/500 — ticks:734 score:48 win:false
+Round 110/500 — ticks:707 score:207 win:false
+Round 111/500 — ticks:700 score:66 win:false
+Round 112/500 — ticks:754 score:96 win:false
+Round 113/500 — ticks:703 score:66 win:false
+Round 114/500 — ticks:637 score:49 win:false
+Round 115/500 — ticks:677 score:51 win:false
+Round 116/500 — ticks:656 score:35 win:false
+Round 117/500 — ticks:441 score:206 win:false
+Round 118/500 — ticks:653 score:64 win:false
+Round 119/500 — ticks:517 score:121 win:false
+Round 120/500 — ticks:750 score:60 win:false
+Round 121/500 — ticks:609 score:32 win:false
+Round 122/500 — ticks:807 score:48 win:false
+Round 123/500 — ticks:640 score:33 win:false
+Round 124/500 — ticks:385 score:24 win:false
+Round 125/500 — ticks:860 score:113 win:false
+Round 126/500 — ticks:734 score:49 win:false
+Round 127/500 — ticks:757 score:81 win:false
+Round 128/500 — ticks:630 score:49 win:false
+Round 129/500 — ticks:587 score:32 win:false
+Round 130/500 — ticks:702 score:73 win:false
+Round 131/500 — ticks:564 score:16 win:false
+Round 132/500 — ticks:632 score:93 win:false
+Round 133/500 — ticks:692 score:82 win:false
+Round 134/500 — ticks:743 score:116 win:false
+Round 135/500 — ticks:758 score:76 win:false
+Round 136/500 — ticks:575 score:46 win:false
+Round 137/500 — ticks:386 score:88 win:false
+Round 138/500 — ticks:626 score:17 win:false
+Round 139/500 — ticks:821 score:85 win:false
+Round 140/500 — ticks:640 score:169 win:false
+Round 141/500 — ticks:735 score:49 win:false
+Round 142/500 — ticks:558 score:18 win:false
+Round 143/500 — ticks:654 score:30 win:false
+Round 144/500 — ticks:604 score:32 win:false
+Round 145/500 — ticks:710 score:56 win:false
+Round 146/500 — ticks:688 score:196 win:false
+Round 147/500 — ticks:714 score:50 win:false
+Round 148/500 — ticks:690 score:42 win:false
+Round 149/500 — ticks:387 score:186 win:false
+Round 150/500 — ticks:911 score:84 win:false
+Round 151/500 — ticks:653 score:33 win:false
+Round 152/500 — ticks:681 score:67 win:false
+Round 153/500 — ticks:605 score:17 win:false
+Round 154/500 — ticks:734 score:58 win:false
+Round 155/500 — ticks:793 score:105 win:false
+Round 156/500 — ticks:784 score:78 win:false
+Round 157/500 — ticks:620 score:32 win:false
+Round 158/500 — ticks:713 score:48 win:false
+Round 159/500 — ticks:669 score:67 win:false
+Round 160/500 — ticks:594 score:30 win:false
+Round 161/500 — ticks:558 score:17 win:false
+Round 162/500 — ticks:609 score:48 win:false
+Round 163/500 — ticks:658 score:50 win:false
+Round 164/500 — ticks:596 score:32 win:false
+Round 165/500 — ticks:684 score:64 win:false
+Round 166/500 — ticks:579 score:234 win:false
+Round 167/500 — ticks:827 score:100 win:false
+Round 168/500 — ticks:562 score:80 win:false
+Round 169/500 — ticks:727 score:64 win:false
+Round 170/500 — ticks:865 score:17 win:false
+Round 171/500 — ticks:704 score:48 win:false
+Round 172/500 — ticks:533 score:16 win:false
+Round 173/500 — ticks:684 score:91 win:false
+Round 174/500 — ticks:600 score:33 win:false
+Round 175/500 — ticks:786 score:64 win:false
+Round 176/500 — ticks:767 score:65 win:false
+Round 177/500 — ticks:877 score:98 win:false
+Round 178/500 — ticks:734 score:64 win:false
+Round 179/500 — ticks:592 score:34 win:false
+Round 180/500 — ticks:730 score:53 win:false
+Round 181/500 — ticks:672 score:176 win:false
+Round 182/500 — ticks:484 score:33 win:false
+Round 183/500 — ticks:600 score:18 win:false
+Round 184/500 — ticks:628 score:21 win:false
+Round 185/500 — ticks:729 score:36 win:false
+Round 186/500 — ticks:822 score:61 win:false
+Round 187/500 — ticks:764 score:64 win:false
+Round 188/500 — ticks:715 score:199 win:false
+Round 189/500 — ticks:556 score:77 win:false
+Round 190/500 — ticks:574 score:21 win:false
+Round 191/500 — ticks:708 score:65 win:false
+Round 192/500 — ticks:580 score:2 win:false
+Round 193/500 — ticks:771 score:66 win:false
+Round 194/500 — ticks:692 score:33 win:false
+Round 195/500 — ticks:759 score:48 win:false
+Round 196/500 — ticks:829 score:63 win:false
+Round 197/500 — ticks:767 score:65 win:false
+Round 198/500 — ticks:246 score:115 win:false
+Round 199/500 — ticks:642 score:47 win:false
+Round 200/500 — ticks:712 score:27 win:false
+Round 201/500 — ticks:711 score:51 win:false
+Round 202/500 — ticks:752 score:49 win:false
+Round 203/500 — ticks:654 score:48 win:false
+Round 204/500 — ticks:602 score:33 win:false
+Round 205/500 — ticks:556 score:20 win:false
+Round 206/500 — ticks:799 score:73 win:false
+Round 207/500 — ticks:874 score:81 win:false
+Round 208/500 — ticks:917 score:95 win:false
+Round 209/500 — ticks:615 score:16 win:false
+Round 210/500 — ticks:776 score:86 win:false
+Round 211/500 — ticks:869 score:113 win:false
+Round 212/500 — ticks:887 score:81 win:false
+Round 213/500 — ticks:675 score:66 win:false
+Round 214/500 — ticks:541 score:2 win:false
+Round 215/500 — ticks:717 score:45 win:false
+Round 216/500 — ticks:817 score:55 win:false
+Round 217/500 — ticks:760 score:66 win:false
+Round 218/500 — ticks:688 score:33 win:false
+Round 219/500 — ticks:658 score:44 win:false
+Round 220/500 — ticks:492 score:205 win:false
+Round 221/500 — ticks:753 score:50 win:false
+Round 222/500 — ticks:882 score:97 win:false
+Round 223/500 — ticks:782 score:100 win:false
+Round 224/500 — ticks:650 score:49 win:false
+Round 225/500 — ticks:712 score:44 win:false
+Round 226/500 — ticks:734 score:57 win:false
+Round 227/500 — ticks:580 score:201 win:false
+Round 228/500 — ticks:708 score:212 win:false
+Round 229/500 — ticks:710 score:4 win:false
+Round 230/500 — ticks:655 score:37 win:false
+Round 231/500 — ticks:677 score:31 win:false
+Round 232/500 — ticks:549 score:132 win:false
+Round 233/500 — ticks:474 score:22 win:false
+Round 234/500 — ticks:819 score:80 win:false
+Round 235/500 — ticks:872 score:127 win:false
+Round 236/500 — ticks:691 score:70 win:false
+Round 237/500 — ticks:680 score:30 win:false
+Round 238/500 — ticks:645 score:31 win:false
+Round 239/500 — ticks:628 score:76 win:false
+Round 240/500 — ticks:681 score:52 win:false
+Round 241/500 — ticks:555 score:2 win:false
+Round 242/500 — ticks:811 score:78 win:false
+Round 243/500 — ticks:777 score:115 win:false
+Round 244/500 — ticks:649 score:33 win:false
+Round 245/500 — ticks:500 score:55 win:false
+Round 246/500 — ticks:751 score:49 win:false
+Round 247/500 — ticks:847 score:62 win:false
+Round 248/500 — ticks:726 score:17 win:false
+Round 249/500 — ticks:639 score:64 win:false
+Round 250/500 — ticks:696 score:54 win:false
+Round 251/500 — ticks:757 score:83 win:false
+Round 252/500 — ticks:860 score:193 win:false
+Round 253/500 — ticks:596 score:1 win:false
+Round 254/500 — ticks:765 score:91 win:false
+Round 255/500 — ticks:748 score:96 win:false
+Round 256/500 — ticks:841 score:90 win:false
+Round 257/500 — ticks:675 score:66 win:false
+Round 258/500 — ticks:573 score:92 win:false
+Round 259/500 — ticks:762 score:69 win:false
+Round 260/500 — ticks:678 score:64 win:false
+Round 261/500 — ticks:647 score:64 win:false
+Round 262/500 — ticks:668 score:33 win:false
+Round 263/500 — ticks:752 score:20 win:false
+Round 264/500 — ticks:774 score:66 win:false
+Round 265/500 — ticks:642 score:72 win:false
+Round 266/500 — ticks:668 score:32 win:false
+Round 267/500 — ticks:678 score:38 win:false
+Round 268/500 — ticks:603 score:230 win:false
+Round 269/500 — ticks:711 score:46 win:false
+Round 270/500 — ticks:937 score:65 win:false
+Round 271/500 — ticks:683 score:66 win:false
+Round 272/500 — ticks:755 score:65 win:false
+Round 273/500 — ticks:554 score:0 win:false
+Round 274/500 — ticks:750 score:33 win:false
+Round 275/500 — ticks:901 score:113 win:false
+Round 276/500 — ticks:812 score:109 win:false
+Round 277/500 — ticks:704 score:82 win:false
+Round 278/500 — ticks:663 score:101 win:false
+Round 279/500 — ticks:636 score:48 win:false
+Round 280/500 — ticks:484 score:201 win:false
+Round 281/500 — ticks:634 score:68 win:false
+Round 282/500 — ticks:766 score:58 win:false
+Round 283/500 — ticks:886 score:112 win:false
+Round 284/500 — ticks:676 score:115 win:false
+Round 285/500 — ticks:530 score:200 win:false
+Round 286/500 — ticks:531 score:17 win:false
+Round 287/500 — ticks:463 score:203 win:false
+Round 288/500 — ticks:599 score:17 win:false
+Round 289/500 — ticks:636 score:46 win:false
+Round 290/500 — ticks:832 score:79 win:false
+Round 291/500 — ticks:557 score:69 win:false
+Round 292/500 — ticks:763 score:81 win:false
+Round 293/500 — ticks:566 score:192 win:false
+Round 294/500 — ticks:807 score:54 win:false
+Round 295/500 — ticks:698 score:81 win:false
+Round 296/500 — ticks:694 score:185 win:false
+Round 297/500 — ticks:806 score:56 win:false
+Round 298/500 — ticks:852 score:113 win:false
+Round 299/500 — ticks:681 score:209 win:false
+Round 300/500 — ticks:530 score:1 win:false
+Round 301/500 — ticks:508 score:17 win:false
+Round 302/500 — ticks:220 score:79 win:false
+Round 303/500 — ticks:712 score:65 win:false
+Round 304/500 — ticks:709 score:97 win:false
+Round 305/500 — ticks:792 score:198 win:false
+Round 306/500 — ticks:825 score:105 win:false
+Round 307/500 — ticks:710 score:188 win:false
+Round 308/500 — ticks:655 score:72 win:false
+Round 309/500 — ticks:807 score:199 win:false
+Round 310/500 — ticks:659 score:17 win:false
+Round 311/500 — ticks:744 score:50 win:false
+Round 312/500 — ticks:674 score:65 win:false
+Round 313/500 — ticks:713 score:112 win:false
+Round 314/500 — ticks:699 score:48 win:false
+Round 315/500 — ticks:753 score:212 win:false
+Round 316/500 — ticks:832 score:185 win:false
+Round 317/500 — ticks:617 score:46 win:false
+Round 318/500 — ticks:839 score:95 win:false
+Round 319/500 — ticks:718 score:96 win:false
+Round 320/500 — ticks:369 score:101 win:false
+Round 321/500 — ticks:532 score:18 win:false
+Round 322/500 — ticks:570 score:130 win:false
+Round 323/500 — ticks:476 score:33 win:false
+Round 324/500 — ticks:548 score:17 win:false
+Round 325/500 — ticks:693 score:83 win:false
+Round 326/500 — ticks:498 score:96 win:false
+Round 327/500 — ticks:782 score:200 win:false
+Round 328/500 — ticks:500 score:210 win:false
+Round 329/500 — ticks:685 score:52 win:false
+Round 330/500 — ticks:579 score:16 win:false
+Round 331/500 — ticks:751 score:50 win:false
+Round 332/500 — ticks:649 score:20 win:false
+Round 333/500 — ticks:512 score:63 win:false
+Round 334/500 — ticks:681 score:90 win:false
+Round 335/500 — ticks:656 score:50 win:false
+Round 336/500 — ticks:647 score:80 win:false
+Round 337/500 — ticks:639 score:82 win:false
+Round 338/500 — ticks:592 score:51 win:false
+Round 339/500 — ticks:552 score:59 win:false
+Round 340/500 — ticks:738 score:108 win:false
+Round 341/500 — ticks:677 score:63 win:false
+Round 342/500 — ticks:793 score:96 win:false
+Round 343/500 — ticks:907 score:95 win:false
+Round 344/500 — ticks:644 score:208 win:false
+Round 345/500 — ticks:438 score:62 win:false
+Round 346/500 — ticks:760 score:98 win:false
+Round 347/500 — ticks:605 score:40 win:false
+Round 348/500 — ticks:639 score:86 win:false
+Round 349/500 — ticks:510 score:32 win:false
+Round 350/500 — ticks:184 score:213 win:false
+Round 351/500 — ticks:590 score:213 win:false
+Round 352/500 — ticks:977 score:68 win:false
+Round 353/500 — ticks:639 score:50 win:false
+Round 354/500 — ticks:547 score:33 win:false
+Round 355/500 — ticks:597 score:65 win:false
+Round 356/500 — ticks:828 score:80 win:false
+Round 357/500 — ticks:640 score:38 win:false
+Round 358/500 — ticks:663 score:206 win:false
+Round 359/500 — ticks:678 score:176 win:false
+Round 360/500 — ticks:561 score:33 win:false
+Round 361/500 — ticks:614 score:59 win:false
+Round 362/500 — ticks:744 score:40 win:false
+Round 363/500 — ticks:653 score:208 win:false
+Round 364/500 — ticks:665 score:49 win:false
+Round 365/500 — ticks:593 score:47 win:false
+Round 366/500 — ticks:782 score:65 win:false
+Round 367/500 — ticks:665 score:82 win:false
+Round 368/500 — ticks:647 score:65 win:false
+Round 369/500 — ticks:629 score:96 win:false
+Round 370/500 — ticks:534 score:196 win:false
+Round 371/500 — ticks:715 score:33 win:false
+Round 372/500 — ticks:635 score:49 win:false
+Round 373/500 — ticks:627 score:71 win:false
+Round 374/500 — ticks:792 score:74 win:false
+Round 375/500 — ticks:599 score:80 win:false
+Round 376/500 — ticks:788 score:205 win:false
+Round 377/500 — ticks:729 score:65 win:false
+Round 378/500 — ticks:484 score:192 win:false
+Round 379/500 — ticks:562 score:40 win:false
+Round 380/500 — ticks:367 score:175 win:false
+Round 381/500 — ticks:638 score:33 win:false
+Round 382/500 — ticks:636 score:2 win:false
+Round 383/500 — ticks:663 score:99 win:false
+Round 384/500 — ticks:860 score:48 win:false
+Round 385/500 — ticks:672 score:65 win:false
+Round 386/500 — ticks:347 score:71 win:false
+Round 387/500 — ticks:684 score:92 win:false
+Round 388/500 — ticks:907 score:96 win:false
+Round 389/500 — ticks:690 score:32 win:false
+Round 390/500 — ticks:876 score:100 win:false
+Round 391/500 — ticks:451 score:1 win:false
+Round 392/500 — ticks:478 score:0 win:false
+Round 393/500 — ticks:666 score:94 win:false
+Round 394/500 — ticks:734 score:64 win:false
+Round 395/500 — ticks:681 score:66 win:false
+Round 396/500 — ticks:682 score:79 win:false
+Round 397/500 — ticks:588 score:73 win:false
+Round 398/500 — ticks:638 score:46 win:false
+Round 399/500 — ticks:619 score:45 win:false
+Round 400/500 — ticks:369 score:204 win:false
+Round 401/500 — ticks:614 score:49 win:false
+Round 402/500 — ticks:786 score:49 win:false
+Round 403/500 — ticks:1280 score:103 win:false
+Round 404/500 — ticks:683 score:96 win:false
+Round 405/500 — ticks:692 score:51 win:false
+Round 406/500 — ticks:753 score:77 win:false
+Round 407/500 — ticks:547 score:48 win:false
+Round 408/500 — ticks:467 score:48 win:false
+Round 409/500 — ticks:605 score:146 win:false
+Round 410/500 — ticks:601 score:72 win:false
+Round 411/500 — ticks:531 score:187 win:false
+Round 412/500 — ticks:592 score:64 win:false
+Round 413/500 — ticks:715 score:68 win:false
+Round 414/500 — ticks:646 score:175 win:false
+Round 415/500 — ticks:259 score:235 win:false
+Round 416/500 — ticks:667 score:81 win:false
+Round 417/500 — ticks:531 score:35 win:false
+Round 418/500 — ticks:974 score:97 win:false
+Round 419/500 — ticks:636 score:37 win:false
+Round 420/500 — ticks:769 score:33 win:false
+Round 421/500 — ticks:667 score:64 win:false
+Round 422/500 — ticks:565 score:64 win:false
+Round 423/500 — ticks:904 score:101 win:false
+Round 424/500 — ticks:621 score:63 win:false
+Round 425/500 — ticks:798 score:97 win:false
+Round 426/500 — ticks:805 score:97 win:false
+Round 427/500 — ticks:660 score:17 win:false
+Round 428/500 — ticks:720 score:48 win:false
+Round 429/500 — ticks:560 score:32 win:false
+Round 430/500 — ticks:901 score:66 win:false
+Round 431/500 — ticks:718 score:65 win:false
+Round 432/500 — ticks:824 score:62 win:false
+Round 433/500 — ticks:904 score:95 win:false
+Round 434/500 — ticks:680 score:49 win:false
+Round 435/500 — ticks:838 score:103 win:false
+Round 436/500 — ticks:761 score:52 win:false
+Round 437/500 — ticks:561 score:34 win:false
+Round 438/500 — ticks:625 score:59 win:false
+Round 439/500 — ticks:789 score:69 win:false
+Round 440/500 — ticks:602 score:31 win:false
+Round 441/500 — ticks:547 score:39 win:false
+Round 442/500 — ticks:790 score:84 win:false
+Round 443/500 — ticks:641 score:33 win:false
+Round 444/500 — ticks:585 score:54 win:false
+Round 445/500 — ticks:578 score:179 win:false
+Round 446/500 — ticks:654 score:81 win:false
+Round 447/500 — ticks:635 score:6 win:false
+Round 448/500 — ticks:697 score:35 win:false
+Round 449/500 — ticks:807 score:192 win:false
+Round 450/500 — ticks:518 score:0 win:false
+Round 451/500 — ticks:834 score:86 win:false
+Round 452/500 — ticks:702 score:49 win:false
+Round 453/500 — ticks:671 score:86 win:false
+Round 454/500 — ticks:525 score:189 win:false
+Round 455/500 — ticks:750 score:80 win:false
+Round 456/500 — ticks:641 score:32 win:false
+Round 457/500 — ticks:600 score:33 win:false
+Round 458/500 — ticks:565 score:16 win:false
+Round 459/500 — ticks:557 score:2 win:false
+Round 460/500 — ticks:686 score:31 win:false
+Round 461/500 — ticks:606 score:18 win:false
+Round 462/500 — ticks:581 score:33 win:false
+Round 463/500 — ticks:707 score:125 win:false
+Round 464/500 — ticks:828 score:83 win:false
+Round 465/500 — ticks:601 score:17 win:false
+Round 466/500 — ticks:644 score:64 win:false
+Round 467/500 — ticks:787 score:48 win:false
+Round 468/500 — ticks:827 score:81 win:false
+Round 469/500 — ticks:601 score:13 win:false
+Round 470/500 — ticks:818 score:75 win:false
+Round 471/500 — ticks:605 score:34 win:false
+Round 472/500 — ticks:748 score:48 win:false
+Round 473/500 — ticks:484 score:0 win:false
+Round 474/500 — ticks:753 score:48 win:false
+Round 475/500 — ticks:675 score:50 win:false
+Round 476/500 — ticks:623 score:35 win:false
+Round 477/500 — ticks:906 score:95 win:false
+Round 478/500 — ticks:683 score:66 win:false
+Round 479/500 — ticks:783 score:81 win:false
+Round 480/500 — ticks:748 score:110 win:false
+Round 481/500 — ticks:690 score:64 win:false
+Round 482/500 — ticks:694 score:16 win:false
+Round 483/500 — ticks:656 score:30 win:false
+Round 484/500 — ticks:688 score:192 win:false
+Round 485/500 — ticks:512 score:46 win:false
+Round 486/500 — ticks:694 score:77 win:false
+Round 487/500 — ticks:474 score:178 win:false
+Round 488/500 — ticks:403 score:182 win:false
+Round 489/500 — ticks:598 score:36 win:false
+Round 490/500 — ticks:199 score:132 win:false
+Round 491/500 — ticks:378 score:158 win:false
+Round 492/500 — ticks:531 score:33 win:false
+Round 493/500 — ticks:749 score:84 win:false
+Round 494/500 — ticks:720 score:14 win:false
+Round 495/500 — ticks:653 score:79 win:false
+Round 496/500 — ticks:828 score:113 win:false
+Round 497/500 — ticks:595 score:5 win:false
+Round 498/500 — ticks:717 score:198 win:false
+Round 499/500 — ticks:794 score:193 win:false
+Round 500/500 — ticks:487 score:0 win:false
+Counter check passed: 38003 == expected 38003
diff --git a/tools/training_runner/mixed_train.sh b/tools/training_runner/mixed_train.sh
index abc621e..fa97c8e 100755
--- a/tools/training_runner/mixed_train.sh
+++ b/tools/training_runner/mixed_train.sh
@@ -5,7 +5,7 @@ set -uo pipefail
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
 REPO_ROOT="$(cd "$SCRIPT_DIR/../.." && pwd)"
-PPO_BOT_SRC="$REPO_ROOT/PPO_Bot"
+PPO_BOT_SRC="$REPO_ROOT/PPO_Bot_garage"
 
 ENV_FILE="$SCRIPT_DIR/training.env"
 [ -f "$ENV_FILE" ] && set -a && . "$ENV_FILE" && set +a
diff --git a/tools/training_runner/run.sh b/tools/training_runner/run.sh
index 41affe2..273c0c8 100755
--- a/tools/training_runner/run.sh
+++ b/tools/training_runner/run.sh
@@ -27,7 +27,7 @@ OPPONENT="${1:-${TRAINING_OPPONENT:-Target}}"
 TOTAL_ROUNDS="${2:-${TRAINING_ROUNDS:-100}}"
 
 JAR="${TANK_ROYALE_JAR:-/home/davide/Projects/tank-royale/runner/examples/lib/robocode-tankroyale-runner.jar}"
-PPO_BOT_SRC="$REPO_ROOT/PPO_Bot"
+PPO_BOT_SRC="$REPO_ROOT/PPO_Bot_garage"
 
 export PPO_BOT_DIR="$PPO_BOT_SRC"
 export SAMPLE_BOTS_DIR="${SAMPLE_BOTS_DIR:-/home/davide/Projects/tank-royale/sample-bots/java/build/archive}"
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/actor_b1.npy b/tools/training_runner/snapshots/best_Corners_r35094/actor_b1.npy
new file mode 100644
index 0000000..309e7c7
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/actor_b2.npy b/tools/training_runner/snapshots/best_Corners_r35094/actor_b2.npy
new file mode 100644
index 0000000..c61ff05
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/actor_b3.npy b/tools/training_runner/snapshots/best_Corners_r35094/actor_b3.npy
new file mode 100644
index 0000000..9355c9e
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/actor_w1.npy b/tools/training_runner/snapshots/best_Corners_r35094/actor_w1.npy
new file mode 100644
index 0000000..6e2d3d4
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/actor_w2.npy b/tools/training_runner/snapshots/best_Corners_r35094/actor_w2.npy
new file mode 100644
index 0000000..7f98b0d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/actor_w3.npy b/tools/training_runner/snapshots/best_Corners_r35094/actor_w3.npy
new file mode 100644
index 0000000..2934e7e
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_ab1_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab1_m.npy
new file mode 100644
index 0000000..640386d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_ab1_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab1_v.npy
new file mode 100644
index 0000000..70cada6
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_ab2_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab2_m.npy
new file mode 100644
index 0000000..0569eb2
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_ab2_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab2_v.npy
new file mode 100644
index 0000000..9386f6d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_ab3_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab3_m.npy
new file mode 100644
index 0000000..cf570ca
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_ab3_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab3_v.npy
new file mode 100644
index 0000000..a046fab
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_aw1_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw1_m.npy
new file mode 100644
index 0000000..8e8799d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_aw1_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw1_v.npy
new file mode 100644
index 0000000..881d1e2
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_aw2_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw2_m.npy
new file mode 100644
index 0000000..d7b2b60
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_aw2_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw2_v.npy
new file mode 100644
index 0000000..92112a1
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_aw3_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw3_m.npy
new file mode 100644
index 0000000..cc5c57b
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_aw3_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw3_v.npy
new file mode 100644
index 0000000..3e9a1cc
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cb1_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb1_m.npy
new file mode 100644
index 0000000..1867780
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cb1_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb1_v.npy
new file mode 100644
index 0000000..9615c7a
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cb2_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb2_m.npy
new file mode 100644
index 0000000..fc190ef
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cb2_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb2_v.npy
new file mode 100644
index 0000000..9f17553
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cb3_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb3_m.npy
new file mode 100644
index 0000000..82ba8c8
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cb3_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb3_v.npy
new file mode 100644
index 0000000..d507295
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cw1_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw1_m.npy
new file mode 100644
index 0000000..a028fd2
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cw1_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw1_v.npy
new file mode 100644
index 0000000..c0a5803
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cw2_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw2_m.npy
new file mode 100644
index 0000000..312d35e
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cw2_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw2_v.npy
new file mode 100644
index 0000000..005ff1f
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cw3_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw3_m.npy
new file mode 100644
index 0000000..a93939a
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_cw3_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw3_v.npy
new file mode 100644
index 0000000..8a61186
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_logstd_m.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_logstd_m.npy
new file mode 100644
index 0000000..6271249
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_logstd_v.npy b/tools/training_runner/snapshots/best_Corners_r35094/adam_logstd_v.npy
new file mode 100644
index 0000000..e1cf079
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/adam_t.txt b/tools/training_runner/snapshots/best_Corners_r35094/adam_t.txt
new file mode 100644
index 0000000..73f4098
--- /dev/null
+++ b/tools/training_runner/snapshots/best_Corners_r35094/adam_t.txt
@@ -0,0 +1,13 @@
+374720
+374720
+374720
+374720
+374720
+374720
+374720
+374720
+374720
+374720
+374720
+374720
+374720
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/critic_b1.npy b/tools/training_runner/snapshots/best_Corners_r35094/critic_b1.npy
new file mode 100644
index 0000000..25f41df
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/critic_b2.npy b/tools/training_runner/snapshots/best_Corners_r35094/critic_b2.npy
new file mode 100644
index 0000000..44794d6
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/critic_b3.npy b/tools/training_runner/snapshots/best_Corners_r35094/critic_b3.npy
new file mode 100644
index 0000000..a5b434a
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/critic_w1.npy b/tools/training_runner/snapshots/best_Corners_r35094/critic_w1.npy
new file mode 100644
index 0000000..df7e9cf
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/critic_w2.npy b/tools/training_runner/snapshots/best_Corners_r35094/critic_w2.npy
new file mode 100644
index 0000000..60e4886
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/critic_w3.npy b/tools/training_runner/snapshots/best_Corners_r35094/critic_w3.npy
new file mode 100644
index 0000000..6c7cce1
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Corners_r35094/log_std.npy b/tools/training_runner/snapshots/best_Corners_r35094/log_std.npy
new file mode 100644
index 0000000..e971017
Binary files /dev/null and b/tools/training_runner/snapshots/best_Corners_r35094/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/actor_b1.npy b/tools/training_runner/snapshots/best_Crazy_r13705/actor_b1.npy
new file mode 100644
index 0000000..b077156
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/actor_b2.npy b/tools/training_runner/snapshots/best_Crazy_r13705/actor_b2.npy
new file mode 100644
index 0000000..d800446
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/actor_b3.npy b/tools/training_runner/snapshots/best_Crazy_r13705/actor_b3.npy
new file mode 100644
index 0000000..03aefaa
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/actor_w1.npy b/tools/training_runner/snapshots/best_Crazy_r13705/actor_w1.npy
new file mode 100644
index 0000000..b32dedc
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/actor_w2.npy b/tools/training_runner/snapshots/best_Crazy_r13705/actor_w2.npy
new file mode 100644
index 0000000..633f27a
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/actor_w3.npy b/tools/training_runner/snapshots/best_Crazy_r13705/actor_w3.npy
new file mode 100644
index 0000000..1477712
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab1_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab1_m.npy
new file mode 100644
index 0000000..05a4d70
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab1_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab1_v.npy
new file mode 100644
index 0000000..59500be
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab2_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab2_m.npy
new file mode 100644
index 0000000..96b1827
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab2_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab2_v.npy
new file mode 100644
index 0000000..d062454
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab3_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab3_m.npy
new file mode 100644
index 0000000..0770bb0
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab3_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab3_v.npy
new file mode 100644
index 0000000..7976179
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw1_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw1_m.npy
new file mode 100644
index 0000000..45d9a42
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw1_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw1_v.npy
new file mode 100644
index 0000000..fdd2732
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw2_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw2_m.npy
new file mode 100644
index 0000000..0f7d8dd
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw2_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw2_v.npy
new file mode 100644
index 0000000..1b5d63c
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw3_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw3_m.npy
new file mode 100644
index 0000000..f5ff0ce
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw3_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw3_v.npy
new file mode 100644
index 0000000..584422d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb1_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb1_m.npy
new file mode 100644
index 0000000..b875924
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb1_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb1_v.npy
new file mode 100644
index 0000000..2aff0d5
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb2_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb2_m.npy
new file mode 100644
index 0000000..aa66fae
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb2_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb2_v.npy
new file mode 100644
index 0000000..1fe3c77
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb3_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb3_m.npy
new file mode 100644
index 0000000..1ff4a68
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb3_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb3_v.npy
new file mode 100644
index 0000000..9d740d0
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw1_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw1_m.npy
new file mode 100644
index 0000000..25cecd5
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw1_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw1_v.npy
new file mode 100644
index 0000000..37ecde7
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw2_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw2_m.npy
new file mode 100644
index 0000000..0097c3c
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw2_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw2_v.npy
new file mode 100644
index 0000000..58d7678
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw3_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw3_m.npy
new file mode 100644
index 0000000..f2929df
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw3_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw3_v.npy
new file mode 100644
index 0000000..86fdb9d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_logstd_m.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_logstd_m.npy
new file mode 100644
index 0000000..dffb137
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_logstd_v.npy b/tools/training_runner/snapshots/best_Crazy_r13705/adam_logstd_v.npy
new file mode 100644
index 0000000..9595181
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/adam_t.txt b/tools/training_runner/snapshots/best_Crazy_r13705/adam_t.txt
new file mode 100644
index 0000000..663a8c4
--- /dev/null
+++ b/tools/training_runner/snapshots/best_Crazy_r13705/adam_t.txt
@@ -0,0 +1,13 @@
+381996
+381996
+381996
+381996
+381996
+381996
+381996
+381996
+381996
+381996
+381996
+381996
+381996
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/critic_b1.npy b/tools/training_runner/snapshots/best_Crazy_r13705/critic_b1.npy
new file mode 100644
index 0000000..d34a849
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/critic_b2.npy b/tools/training_runner/snapshots/best_Crazy_r13705/critic_b2.npy
new file mode 100644
index 0000000..62b4e31
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/critic_b3.npy b/tools/training_runner/snapshots/best_Crazy_r13705/critic_b3.npy
new file mode 100644
index 0000000..8befe54
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/critic_w1.npy b/tools/training_runner/snapshots/best_Crazy_r13705/critic_w1.npy
new file mode 100644
index 0000000..57bb4d8
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/critic_w2.npy b/tools/training_runner/snapshots/best_Crazy_r13705/critic_w2.npy
new file mode 100644
index 0000000..ea0ad48
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/critic_w3.npy b/tools/training_runner/snapshots/best_Crazy_r13705/critic_w3.npy
new file mode 100644
index 0000000..007f3ca
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Crazy_r13705/log_std.npy b/tools/training_runner/snapshots/best_Crazy_r13705/log_std.npy
new file mode 100644
index 0000000..36e89df
Binary files /dev/null and b/tools/training_runner/snapshots/best_Crazy_r13705/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/actor_b1.npy b/tools/training_runner/snapshots/best_Fire_r11955/actor_b1.npy
new file mode 100644
index 0000000..4c1a573
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/actor_b2.npy b/tools/training_runner/snapshots/best_Fire_r11955/actor_b2.npy
new file mode 100644
index 0000000..a507759
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/actor_b3.npy b/tools/training_runner/snapshots/best_Fire_r11955/actor_b3.npy
new file mode 100644
index 0000000..46d293a
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/actor_w1.npy b/tools/training_runner/snapshots/best_Fire_r11955/actor_w1.npy
new file mode 100644
index 0000000..15059f9
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/actor_w2.npy b/tools/training_runner/snapshots/best_Fire_r11955/actor_w2.npy
new file mode 100644
index 0000000..cc406fd
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/actor_w3.npy b/tools/training_runner/snapshots/best_Fire_r11955/actor_w3.npy
new file mode 100644
index 0000000..104aabe
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_ab1_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab1_m.npy
new file mode 100644
index 0000000..4446837
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_ab1_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab1_v.npy
new file mode 100644
index 0000000..df79b55
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_ab2_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab2_m.npy
new file mode 100644
index 0000000..3b1a639
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_ab2_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab2_v.npy
new file mode 100644
index 0000000..b901b52
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_ab3_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab3_m.npy
new file mode 100644
index 0000000..c8fece4
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_ab3_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab3_v.npy
new file mode 100644
index 0000000..3bd4b8c
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_aw1_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw1_m.npy
new file mode 100644
index 0000000..5e099ef
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_aw1_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw1_v.npy
new file mode 100644
index 0000000..43b3109
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_aw2_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw2_m.npy
new file mode 100644
index 0000000..036ec57
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_aw2_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw2_v.npy
new file mode 100644
index 0000000..e4f9405
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_aw3_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw3_m.npy
new file mode 100644
index 0000000..49371a5
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_aw3_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw3_v.npy
new file mode 100644
index 0000000..11c5ef5
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cb1_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb1_m.npy
new file mode 100644
index 0000000..10dc9a8
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cb1_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb1_v.npy
new file mode 100644
index 0000000..0cac7e4
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cb2_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb2_m.npy
new file mode 100644
index 0000000..488b455
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cb2_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb2_v.npy
new file mode 100644
index 0000000..8b0751e
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cb3_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb3_m.npy
new file mode 100644
index 0000000..6c5d731
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cb3_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb3_v.npy
new file mode 100644
index 0000000..b093d16
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cw1_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw1_m.npy
new file mode 100644
index 0000000..18bf20f
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cw1_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw1_v.npy
new file mode 100644
index 0000000..42c32b0
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cw2_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw2_m.npy
new file mode 100644
index 0000000..f3a1175
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cw2_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw2_v.npy
new file mode 100644
index 0000000..653a70f
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cw3_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw3_m.npy
new file mode 100644
index 0000000..aef2948
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_cw3_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw3_v.npy
new file mode 100644
index 0000000..880da6f
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_logstd_m.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_logstd_m.npy
new file mode 100644
index 0000000..02e8df8
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_logstd_v.npy b/tools/training_runner/snapshots/best_Fire_r11955/adam_logstd_v.npy
new file mode 100644
index 0000000..ac21059
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/adam_t.txt b/tools/training_runner/snapshots/best_Fire_r11955/adam_t.txt
new file mode 100644
index 0000000..68dd77d
--- /dev/null
+++ b/tools/training_runner/snapshots/best_Fire_r11955/adam_t.txt
@@ -0,0 +1,13 @@
+347532
+347532
+347532
+347532
+347532
+347532
+347532
+347532
+347532
+347532
+347532
+347532
+347532
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/critic_b1.npy b/tools/training_runner/snapshots/best_Fire_r11955/critic_b1.npy
new file mode 100644
index 0000000..f0bb4b4
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/critic_b2.npy b/tools/training_runner/snapshots/best_Fire_r11955/critic_b2.npy
new file mode 100644
index 0000000..0cb5b58
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/critic_b3.npy b/tools/training_runner/snapshots/best_Fire_r11955/critic_b3.npy
new file mode 100644
index 0000000..0086182
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/critic_w1.npy b/tools/training_runner/snapshots/best_Fire_r11955/critic_w1.npy
new file mode 100644
index 0000000..dd15db0
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/critic_w2.npy b/tools/training_runner/snapshots/best_Fire_r11955/critic_w2.npy
new file mode 100644
index 0000000..56caa5b
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/critic_w3.npy b/tools/training_runner/snapshots/best_Fire_r11955/critic_w3.npy
new file mode 100644
index 0000000..6dbd67a
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Fire_r11955/log_std.npy b/tools/training_runner/snapshots/best_Fire_r11955/log_std.npy
new file mode 100644
index 0000000..540fa6e
Binary files /dev/null and b/tools/training_runner/snapshots/best_Fire_r11955/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b1.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b1.npy
new file mode 100644
index 0000000..e1e359e
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b2.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b2.npy
new file mode 100644
index 0000000..205f4f0
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b3.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b3.npy
new file mode 100644
index 0000000..a837540
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w1.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w1.npy
new file mode 100644
index 0000000..06b8428
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w2.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w2.npy
new file mode 100644
index 0000000..68aa4f8
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w3.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w3.npy
new file mode 100644
index 0000000..8458ff0
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab1_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab1_m.npy
new file mode 100644
index 0000000..6b5512f
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab1_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab1_v.npy
new file mode 100644
index 0000000..5659f35
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab2_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab2_m.npy
new file mode 100644
index 0000000..863f1b2
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab2_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab2_v.npy
new file mode 100644
index 0000000..e8f088f
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab3_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab3_m.npy
new file mode 100644
index 0000000..5c2c781
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab3_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab3_v.npy
new file mode 100644
index 0000000..a4827d0
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw1_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw1_m.npy
new file mode 100644
index 0000000..304eea4
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw1_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw1_v.npy
new file mode 100644
index 0000000..ed3ac28
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw2_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw2_m.npy
new file mode 100644
index 0000000..638b7fd
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw2_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw2_v.npy
new file mode 100644
index 0000000..04267c2
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw3_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw3_m.npy
new file mode 100644
index 0000000..1c8e5f7
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw3_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw3_v.npy
new file mode 100644
index 0000000..2e55084
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb1_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb1_m.npy
new file mode 100644
index 0000000..3b58612
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb1_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb1_v.npy
new file mode 100644
index 0000000..859eb94
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb2_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb2_m.npy
new file mode 100644
index 0000000..90d5e43
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb2_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb2_v.npy
new file mode 100644
index 0000000..73f755b
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb3_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb3_m.npy
new file mode 100644
index 0000000..37b5872
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb3_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb3_v.npy
new file mode 100644
index 0000000..fe3db06
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw1_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw1_m.npy
new file mode 100644
index 0000000..7eea56c
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw1_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw1_v.npy
new file mode 100644
index 0000000..79e3642
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw2_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw2_m.npy
new file mode 100644
index 0000000..82b57e9
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw2_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw2_v.npy
new file mode 100644
index 0000000..c79238e
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw3_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw3_m.npy
new file mode 100644
index 0000000..3dd60d5
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw3_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw3_v.npy
new file mode 100644
index 0000000..f535301
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_logstd_m.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_logstd_m.npy
new file mode 100644
index 0000000..dd8d043
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_logstd_v.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_logstd_v.npy
new file mode 100644
index 0000000..39e6032
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_t.txt b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_t.txt
new file mode 100644
index 0000000..8ecd1eb
--- /dev/null
+++ b/tools/training_runner/snapshots/best_MyFirstBot_r15655/adam_t.txt
@@ -0,0 +1,13 @@
+444844
+444844
+444844
+444844
+444844
+444844
+444844
+444844
+444844
+444844
+444844
+444844
+444844
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b1.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b1.npy
new file mode 100644
index 0000000..09e0f47
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b2.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b2.npy
new file mode 100644
index 0000000..faaff3f
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b3.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b3.npy
new file mode 100644
index 0000000..98cecb9
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w1.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w1.npy
new file mode 100644
index 0000000..05afc9a
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w2.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w2.npy
new file mode 100644
index 0000000..94778c3
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w3.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w3.npy
new file mode 100644
index 0000000..1c5a230
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstBot_r15655/log_std.npy b/tools/training_runner/snapshots/best_MyFirstBot_r15655/log_std.npy
new file mode 100644
index 0000000..15dad8f
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstBot_r15655/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b1.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b1.npy
new file mode 100644
index 0000000..078ec39
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b2.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b2.npy
new file mode 100644
index 0000000..ff18f16
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b3.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b3.npy
new file mode 100644
index 0000000..adea926
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w1.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w1.npy
new file mode 100644
index 0000000..d0dd043
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w2.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w2.npy
new file mode 100644
index 0000000..4ef5c2e
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w3.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w3.npy
new file mode 100644
index 0000000..de863cd
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab1_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab1_m.npy
new file mode 100644
index 0000000..c7e012d
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab1_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab1_v.npy
new file mode 100644
index 0000000..fd63192
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab2_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab2_m.npy
new file mode 100644
index 0000000..48487d9
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab2_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab2_v.npy
new file mode 100644
index 0000000..72a7ffc
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab3_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab3_m.npy
new file mode 100644
index 0000000..aa1313b
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab3_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab3_v.npy
new file mode 100644
index 0000000..4c2b224
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw1_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw1_m.npy
new file mode 100644
index 0000000..8373bea
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw1_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw1_v.npy
new file mode 100644
index 0000000..9300196
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw2_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw2_m.npy
new file mode 100644
index 0000000..4a1c0df
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw2_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw2_v.npy
new file mode 100644
index 0000000..bb9a550
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw3_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw3_m.npy
new file mode 100644
index 0000000..6aa37f3
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw3_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw3_v.npy
new file mode 100644
index 0000000..3430028
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb1_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb1_m.npy
new file mode 100644
index 0000000..7d132ad
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb1_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb1_v.npy
new file mode 100644
index 0000000..c86423e
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb2_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb2_m.npy
new file mode 100644
index 0000000..a2f85da
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb2_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb2_v.npy
new file mode 100644
index 0000000..12ddc2d
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb3_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb3_m.npy
new file mode 100644
index 0000000..68cffbb
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb3_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb3_v.npy
new file mode 100644
index 0000000..a38f2ca
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw1_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw1_m.npy
new file mode 100644
index 0000000..49d3270
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw1_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw1_v.npy
new file mode 100644
index 0000000..eecf332
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw2_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw2_m.npy
new file mode 100644
index 0000000..652109a
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw2_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw2_v.npy
new file mode 100644
index 0000000..e7cf997
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw3_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw3_m.npy
new file mode 100644
index 0000000..14fbbbc
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw3_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw3_v.npy
new file mode 100644
index 0000000..505da7a
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_logstd_m.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_logstd_m.npy
new file mode 100644
index 0000000..775634d
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_logstd_v.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_logstd_v.npy
new file mode 100644
index 0000000..76946a2
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_t.txt b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_t.txt
new file mode 100644
index 0000000..99b597b
--- /dev/null
+++ b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/adam_t.txt
@@ -0,0 +1,13 @@
+340440
+340440
+340440
+340440
+340440
+340440
+340440
+340440
+340440
+340440
+340440
+340440
+340440
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b1.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b1.npy
new file mode 100644
index 0000000..b72d048
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b2.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b2.npy
new file mode 100644
index 0000000..8ce0cab
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b3.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b3.npy
new file mode 100644
index 0000000..4427dc3
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w1.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w1.npy
new file mode 100644
index 0000000..1212a64
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w2.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w2.npy
new file mode 100644
index 0000000..a97ac82
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w3.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w3.npy
new file mode 100644
index 0000000..4738238
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstDroid_r11605/log_std.npy b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/log_std.npy
new file mode 100644
index 0000000..688baea
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstDroid_r11605/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b1.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b1.npy
new file mode 100644
index 0000000..c9b2b8b
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b2.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b2.npy
new file mode 100644
index 0000000..10532d7
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b3.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b3.npy
new file mode 100644
index 0000000..23377e4
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w1.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w1.npy
new file mode 100644
index 0000000..56ac045
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w2.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w2.npy
new file mode 100644
index 0000000..f4d021f
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w3.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w3.npy
new file mode 100644
index 0000000..b900c3c
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab1_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab1_m.npy
new file mode 100644
index 0000000..1a4a38e
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab1_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab1_v.npy
new file mode 100644
index 0000000..63f6103
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab2_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab2_m.npy
new file mode 100644
index 0000000..a1b58d1
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab2_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab2_v.npy
new file mode 100644
index 0000000..442725c
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab3_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab3_m.npy
new file mode 100644
index 0000000..83f5ba0
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab3_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab3_v.npy
new file mode 100644
index 0000000..a0f610a
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw1_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw1_m.npy
new file mode 100644
index 0000000..8ed07d8
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw1_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw1_v.npy
new file mode 100644
index 0000000..185b033
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw2_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw2_m.npy
new file mode 100644
index 0000000..4224b9b
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw2_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw2_v.npy
new file mode 100644
index 0000000..d3a347c
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw3_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw3_m.npy
new file mode 100644
index 0000000..acf9a29
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw3_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw3_v.npy
new file mode 100644
index 0000000..48e0786
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb1_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb1_m.npy
new file mode 100644
index 0000000..0cca616
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb1_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb1_v.npy
new file mode 100644
index 0000000..b1ae9fe
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb2_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb2_m.npy
new file mode 100644
index 0000000..9c3ef9c
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb2_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb2_v.npy
new file mode 100644
index 0000000..a22932c
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb3_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb3_m.npy
new file mode 100644
index 0000000..b8b7c9d
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb3_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb3_v.npy
new file mode 100644
index 0000000..875c5fc
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw1_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw1_m.npy
new file mode 100644
index 0000000..4f82eb7
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw1_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw1_v.npy
new file mode 100644
index 0000000..8d37543
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw2_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw2_m.npy
new file mode 100644
index 0000000..4d3af7b
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw2_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw2_v.npy
new file mode 100644
index 0000000..ba77cae
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw3_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw3_m.npy
new file mode 100644
index 0000000..ddac336
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw3_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw3_v.npy
new file mode 100644
index 0000000..2fa7098
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_logstd_m.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_logstd_m.npy
new file mode 100644
index 0000000..6a00103
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_logstd_v.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_logstd_v.npy
new file mode 100644
index 0000000..27c63bc
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_t.txt b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_t.txt
new file mode 100644
index 0000000..def7cd6
--- /dev/null
+++ b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/adam_t.txt
@@ -0,0 +1,13 @@
+368144
+368144
+368144
+368144
+368144
+368144
+368144
+368144
+368144
+368144
+368144
+368144
+368144
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b1.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b1.npy
new file mode 100644
index 0000000..33a60bb
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b2.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b2.npy
new file mode 100644
index 0000000..f23031b
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b3.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b3.npy
new file mode 100644
index 0000000..aad4106
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w1.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w1.npy
new file mode 100644
index 0000000..df3ca3b
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w2.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w2.npy
new file mode 100644
index 0000000..c957eec
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w3.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w3.npy
new file mode 100644
index 0000000..1e758e3
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_MyFirstLeader_r12955/log_std.npy b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/log_std.npy
new file mode 100644
index 0000000..7a2488f
Binary files /dev/null and b/tools/training_runner/snapshots/best_MyFirstLeader_r12955/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b1.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b1.npy
new file mode 100644
index 0000000..cd136db
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b2.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b2.npy
new file mode 100644
index 0000000..b1cdd49
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b3.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b3.npy
new file mode 100644
index 0000000..5397b09
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w1.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w1.npy
new file mode 100644
index 0000000..8e94833
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w2.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w2.npy
new file mode 100644
index 0000000..693243e
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w3.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w3.npy
new file mode 100644
index 0000000..c1e571a
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab1_m.npy
new file mode 100644
index 0000000..b6ca639
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab1_v.npy
new file mode 100644
index 0000000..b498b6d
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab2_m.npy
new file mode 100644
index 0000000..eee3e53
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab2_v.npy
new file mode 100644
index 0000000..f1df341
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab3_m.npy
new file mode 100644
index 0000000..cc9f0ce
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab3_v.npy
new file mode 100644
index 0000000..2c5fc07
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw1_m.npy
new file mode 100644
index 0000000..cc786c6
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw1_v.npy
new file mode 100644
index 0000000..8a9ea89
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw2_m.npy
new file mode 100644
index 0000000..979abef
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw2_v.npy
new file mode 100644
index 0000000..a7d194c
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw3_m.npy
new file mode 100644
index 0000000..69cc9a9
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw3_v.npy
new file mode 100644
index 0000000..f95325c
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb1_m.npy
new file mode 100644
index 0000000..da003b5
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb1_v.npy
new file mode 100644
index 0000000..8e5dd76
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb2_m.npy
new file mode 100644
index 0000000..ed85a14
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb2_v.npy
new file mode 100644
index 0000000..deaab93
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb3_m.npy
new file mode 100644
index 0000000..fc02b1a
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb3_v.npy
new file mode 100644
index 0000000..2909648
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw1_m.npy
new file mode 100644
index 0000000..b856b44
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw1_v.npy
new file mode 100644
index 0000000..843bc74
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw2_m.npy
new file mode 100644
index 0000000..939efa8
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw2_v.npy
new file mode 100644
index 0000000..dc79c90
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw3_m.npy
new file mode 100644
index 0000000..b397bb2
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw3_v.npy
new file mode 100644
index 0000000..911f407
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_logstd_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_logstd_m.npy
new file mode 100644
index 0000000..512443b
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_logstd_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_logstd_v.npy
new file mode 100644
index 0000000..82bc32a
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_t.txt b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_t.txt
new file mode 100644
index 0000000..85bbe48
--- /dev/null
+++ b/tools/training_runner/snapshots/best_PaintingBot_r33144/adam_t.txt
@@ -0,0 +1,13 @@
+376972
+376972
+376972
+376972
+376972
+376972
+376972
+376972
+376972
+376972
+376972
+376972
+376972
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b1.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b1.npy
new file mode 100644
index 0000000..84f66bf
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b2.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b2.npy
new file mode 100644
index 0000000..5242d32
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b3.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b3.npy
new file mode 100644
index 0000000..86910dc
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w1.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w1.npy
new file mode 100644
index 0000000..c025d6f
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w2.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w2.npy
new file mode 100644
index 0000000..c8e2498
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w3.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w3.npy
new file mode 100644
index 0000000..f190a97
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r33144/log_std.npy b/tools/training_runner/snapshots/best_PaintingBot_r33144/log_std.npy
new file mode 100644
index 0000000..eb65db2
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r33144/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b1.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b1.npy
new file mode 100644
index 0000000..f476fcc
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b2.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b2.npy
new file mode 100644
index 0000000..08e77f2
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b3.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b3.npy
new file mode 100644
index 0000000..55777ed
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w1.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w1.npy
new file mode 100644
index 0000000..f4daf1e
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w2.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w2.npy
new file mode 100644
index 0000000..4dbc14a
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w3.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w3.npy
new file mode 100644
index 0000000..f37b14c
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab1_m.npy
new file mode 100644
index 0000000..cffd5c3
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab1_v.npy
new file mode 100644
index 0000000..8caf765
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab2_m.npy
new file mode 100644
index 0000000..1520349
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab2_v.npy
new file mode 100644
index 0000000..d204fd0
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab3_m.npy
new file mode 100644
index 0000000..a86582b
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab3_v.npy
new file mode 100644
index 0000000..2831b25
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw1_m.npy
new file mode 100644
index 0000000..ba6181d
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw1_v.npy
new file mode 100644
index 0000000..0965696
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw2_m.npy
new file mode 100644
index 0000000..532bc40
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw2_v.npy
new file mode 100644
index 0000000..30b3310
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw3_m.npy
new file mode 100644
index 0000000..63108bb
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw3_v.npy
new file mode 100644
index 0000000..3aac35a
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb1_m.npy
new file mode 100644
index 0000000..501ba58
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb1_v.npy
new file mode 100644
index 0000000..3103849
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb2_m.npy
new file mode 100644
index 0000000..ce11a82
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb2_v.npy
new file mode 100644
index 0000000..1f21ce1
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb3_m.npy
new file mode 100644
index 0000000..e8fdc7a
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb3_v.npy
new file mode 100644
index 0000000..8d3593f
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw1_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw1_m.npy
new file mode 100644
index 0000000..9108b80
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw1_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw1_v.npy
new file mode 100644
index 0000000..14d9856
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw2_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw2_m.npy
new file mode 100644
index 0000000..12f54d9
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw2_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw2_v.npy
new file mode 100644
index 0000000..6944686
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw3_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw3_m.npy
new file mode 100644
index 0000000..0b185f6
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw3_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw3_v.npy
new file mode 100644
index 0000000..8c53b6c
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_logstd_m.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_logstd_m.npy
new file mode 100644
index 0000000..f68bb48
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_logstd_v.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_logstd_v.npy
new file mode 100644
index 0000000..7d80962
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_t.txt b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_t.txt
new file mode 100644
index 0000000..7981571
--- /dev/null
+++ b/tools/training_runner/snapshots/best_PaintingBot_r36603/adam_t.txt
@@ -0,0 +1,13 @@
+392204
+392204
+392204
+392204
+392204
+392204
+392204
+392204
+392204
+392204
+392204
+392204
+392204
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b1.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b1.npy
new file mode 100644
index 0000000..26c0a79
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b2.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b2.npy
new file mode 100644
index 0000000..f0dcc75
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b3.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b3.npy
new file mode 100644
index 0000000..f969918
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w1.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w1.npy
new file mode 100644
index 0000000..f851c58
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w2.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w2.npy
new file mode 100644
index 0000000..df5fcce
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w3.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w3.npy
new file mode 100644
index 0000000..a90677a
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_PaintingBot_r36603/log_std.npy b/tools/training_runner/snapshots/best_PaintingBot_r36603/log_std.npy
new file mode 100644
index 0000000..d7ebae0
Binary files /dev/null and b/tools/training_runner/snapshots/best_PaintingBot_r36603/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/actor_b1.npy b/tools/training_runner/snapshots/best_Target_r12305/actor_b1.npy
new file mode 100644
index 0000000..54c27bc
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/actor_b2.npy b/tools/training_runner/snapshots/best_Target_r12305/actor_b2.npy
new file mode 100644
index 0000000..31c2213
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/actor_b3.npy b/tools/training_runner/snapshots/best_Target_r12305/actor_b3.npy
new file mode 100644
index 0000000..3c99525
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/actor_w1.npy b/tools/training_runner/snapshots/best_Target_r12305/actor_w1.npy
new file mode 100644
index 0000000..fcf91df
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/actor_w2.npy b/tools/training_runner/snapshots/best_Target_r12305/actor_w2.npy
new file mode 100644
index 0000000..e735a7d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/actor_w3.npy b/tools/training_runner/snapshots/best_Target_r12305/actor_w3.npy
new file mode 100644
index 0000000..8f26d5d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_ab1_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_ab1_m.npy
new file mode 100644
index 0000000..bc16796
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_ab1_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_ab1_v.npy
new file mode 100644
index 0000000..ab6d104
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_ab2_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_ab2_m.npy
new file mode 100644
index 0000000..e693d2b
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_ab2_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_ab2_v.npy
new file mode 100644
index 0000000..5b34089
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_ab3_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_ab3_m.npy
new file mode 100644
index 0000000..8b52220
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_ab3_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_ab3_v.npy
new file mode 100644
index 0000000..78ff0ea
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_aw1_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_aw1_m.npy
new file mode 100644
index 0000000..6ce14cd
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_aw1_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_aw1_v.npy
new file mode 100644
index 0000000..cbf1aa1
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_aw2_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_aw2_m.npy
new file mode 100644
index 0000000..8596b40
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_aw2_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_aw2_v.npy
new file mode 100644
index 0000000..2644a64
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_aw3_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_aw3_m.npy
new file mode 100644
index 0000000..db13041
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_aw3_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_aw3_v.npy
new file mode 100644
index 0000000..9c2141f
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cb1_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cb1_m.npy
new file mode 100644
index 0000000..be13f4c
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cb1_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cb1_v.npy
new file mode 100644
index 0000000..24656c5
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cb2_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cb2_m.npy
new file mode 100644
index 0000000..1b7df7c
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cb2_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cb2_v.npy
new file mode 100644
index 0000000..77abd5b
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cb3_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cb3_m.npy
new file mode 100644
index 0000000..b98a66d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cb3_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cb3_v.npy
new file mode 100644
index 0000000..bf67f94
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cw1_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cw1_m.npy
new file mode 100644
index 0000000..5153905
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cw1_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cw1_v.npy
new file mode 100644
index 0000000..ec120ff
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cw2_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cw2_m.npy
new file mode 100644
index 0000000..2b5d76d
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cw2_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cw2_v.npy
new file mode 100644
index 0000000..ff48448
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cw3_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cw3_m.npy
new file mode 100644
index 0000000..2e6d679
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_cw3_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_cw3_v.npy
new file mode 100644
index 0000000..53300cf
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_logstd_m.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_logstd_m.npy
new file mode 100644
index 0000000..7f1e450
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_logstd_v.npy b/tools/training_runner/snapshots/best_Target_r12305/adam_logstd_v.npy
new file mode 100644
index 0000000..987e701
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/adam_t.txt b/tools/training_runner/snapshots/best_Target_r12305/adam_t.txt
new file mode 100644
index 0000000..b84d676
--- /dev/null
+++ b/tools/training_runner/snapshots/best_Target_r12305/adam_t.txt
@@ -0,0 +1,13 @@
+358020
+358020
+358020
+358020
+358020
+358020
+358020
+358020
+358020
+358020
+358020
+358020
+358020
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_Target_r12305/critic_b1.npy b/tools/training_runner/snapshots/best_Target_r12305/critic_b1.npy
new file mode 100644
index 0000000..ff8347e
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/critic_b2.npy b/tools/training_runner/snapshots/best_Target_r12305/critic_b2.npy
new file mode 100644
index 0000000..e781b90
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/critic_b3.npy b/tools/training_runner/snapshots/best_Target_r12305/critic_b3.npy
new file mode 100644
index 0000000..c309cf3
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/critic_w1.npy b/tools/training_runner/snapshots/best_Target_r12305/critic_w1.npy
new file mode 100644
index 0000000..0c008d7
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/critic_w2.npy b/tools/training_runner/snapshots/best_Target_r12305/critic_w2.npy
new file mode 100644
index 0000000..b8a8ab7
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/critic_w3.npy b/tools/training_runner/snapshots/best_Target_r12305/critic_w3.npy
new file mode 100644
index 0000000..63700ec
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_Target_r12305/log_std.npy b/tools/training_runner/snapshots/best_Target_r12305/log_std.npy
new file mode 100644
index 0000000..e853594
Binary files /dev/null and b/tools/training_runner/snapshots/best_Target_r12305/log_std.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/actor_b1.npy b/tools/training_runner/snapshots/best_post_maint_r38003/actor_b1.npy
new file mode 100644
index 0000000..f39c3fc
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/actor_b2.npy b/tools/training_runner/snapshots/best_post_maint_r38003/actor_b2.npy
new file mode 100644
index 0000000..7b40e85
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/actor_b3.npy b/tools/training_runner/snapshots/best_post_maint_r38003/actor_b3.npy
new file mode 100644
index 0000000..e771ba8
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/actor_w1.npy b/tools/training_runner/snapshots/best_post_maint_r38003/actor_w1.npy
new file mode 100644
index 0000000..4b2387b
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/actor_w2.npy b/tools/training_runner/snapshots/best_post_maint_r38003/actor_w2.npy
new file mode 100644
index 0000000..680b23a
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/actor_w3.npy b/tools/training_runner/snapshots/best_post_maint_r38003/actor_w3.npy
new file mode 100644
index 0000000..5c9c63b
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab1_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab1_m.npy
new file mode 100644
index 0000000..8f3a01a
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab1_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab1_v.npy
new file mode 100644
index 0000000..8487eed
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab2_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab2_m.npy
new file mode 100644
index 0000000..61e0629
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab2_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab2_v.npy
new file mode 100644
index 0000000..7f05a94
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab3_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab3_m.npy
new file mode 100644
index 0000000..a25e638
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab3_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab3_v.npy
new file mode 100644
index 0000000..862a224
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw1_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw1_m.npy
new file mode 100644
index 0000000..5ca6ceb
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw1_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw1_v.npy
new file mode 100644
index 0000000..c6da504
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw2_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw2_m.npy
new file mode 100644
index 0000000..1b897ab
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw2_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw2_v.npy
new file mode 100644
index 0000000..a67d3f4
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw3_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw3_m.npy
new file mode 100644
index 0000000..29673e7
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw3_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw3_v.npy
new file mode 100644
index 0000000..7b941a8
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb1_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb1_m.npy
new file mode 100644
index 0000000..850d8a7
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb1_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb1_v.npy
new file mode 100644
index 0000000..cd600ee
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb2_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb2_m.npy
new file mode 100644
index 0000000..be977fe
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb2_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb2_v.npy
new file mode 100644
index 0000000..b7f3f74
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb3_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb3_m.npy
new file mode 100644
index 0000000..82a030b
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb3_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb3_v.npy
new file mode 100644
index 0000000..f478816
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw1_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw1_m.npy
new file mode 100644
index 0000000..00d36df
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw1_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw1_v.npy
new file mode 100644
index 0000000..528fc6f
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw2_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw2_m.npy
new file mode 100644
index 0000000..6a517ed
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw2_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw2_v.npy
new file mode 100644
index 0000000..65abed7
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw3_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw3_m.npy
new file mode 100644
index 0000000..4b7dbbf
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw3_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw3_v.npy
new file mode 100644
index 0000000..356d0e9
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_logstd_m.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_logstd_m.npy
new file mode 100644
index 0000000..5767380
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_logstd_v.npy b/tools/training_runner/snapshots/best_post_maint_r38003/adam_logstd_v.npy
new file mode 100644
index 0000000..5105c14
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/adam_t.txt b/tools/training_runner/snapshots/best_post_maint_r38003/adam_t.txt
new file mode 100644
index 0000000..ab0eb3f
--- /dev/null
+++ b/tools/training_runner/snapshots/best_post_maint_r38003/adam_t.txt
@@ -0,0 +1,13 @@
+402796
+402796
+402796
+402796
+402796
+402796
+402796
+402796
+402796
+402796
+402796
+402796
+402796
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/critic_b1.npy b/tools/training_runner/snapshots/best_post_maint_r38003/critic_b1.npy
new file mode 100644
index 0000000..3075ad4
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/critic_b2.npy b/tools/training_runner/snapshots/best_post_maint_r38003/critic_b2.npy
new file mode 100644
index 0000000..90f668a
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/critic_b3.npy b/tools/training_runner/snapshots/best_post_maint_r38003/critic_b3.npy
new file mode 100644
index 0000000..0c7bcf9
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/critic_w1.npy b/tools/training_runner/snapshots/best_post_maint_r38003/critic_w1.npy
new file mode 100644
index 0000000..e77858d
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/critic_w2.npy b/tools/training_runner/snapshots/best_post_maint_r38003/critic_w2.npy
new file mode 100644
index 0000000..d3c3de2
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/critic_w3.npy b/tools/training_runner/snapshots/best_post_maint_r38003/critic_w3.npy
new file mode 100644
index 0000000..4a51cde
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/best_post_maint_r38003/log_std.npy b/tools/training_runner/snapshots/best_post_maint_r38003/log_std.npy
new file mode 100644
index 0000000..aa77ea2
Binary files /dev/null and b/tools/training_runner/snapshots/best_post_maint_r38003/log_std.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/actor_b1.npy b/tools/training_runner/snapshots/generalist_best_r1000/actor_b1.npy
new file mode 100644
index 0000000..02acd51
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/actor_b2.npy b/tools/training_runner/snapshots/generalist_best_r1000/actor_b2.npy
new file mode 100644
index 0000000..f863a32
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/actor_b3.npy b/tools/training_runner/snapshots/generalist_best_r1000/actor_b3.npy
new file mode 100644
index 0000000..131cab8
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/actor_w1.npy b/tools/training_runner/snapshots/generalist_best_r1000/actor_w1.npy
new file mode 100644
index 0000000..c8e20b8
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/actor_w2.npy b/tools/training_runner/snapshots/generalist_best_r1000/actor_w2.npy
new file mode 100644
index 0000000..b48ff0b
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/actor_w3.npy b/tools/training_runner/snapshots/generalist_best_r1000/actor_w3.npy
new file mode 100644
index 0000000..562af28
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_ab1_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab1_m.npy
new file mode 100644
index 0000000..758357d
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_ab1_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab1_v.npy
new file mode 100644
index 0000000..d288a6c
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_ab2_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab2_m.npy
new file mode 100644
index 0000000..df44393
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_ab2_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab2_v.npy
new file mode 100644
index 0000000..edd9885
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_ab3_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab3_m.npy
new file mode 100644
index 0000000..01a1857
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_ab3_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab3_v.npy
new file mode 100644
index 0000000..1ad331e
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_aw1_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw1_m.npy
new file mode 100644
index 0000000..de7b8b8
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_aw1_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw1_v.npy
new file mode 100644
index 0000000..69a01a1
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_aw2_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw2_m.npy
new file mode 100644
index 0000000..3feac21
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_aw2_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw2_v.npy
new file mode 100644
index 0000000..2cbc003
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_aw3_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw3_m.npy
new file mode 100644
index 0000000..2d66af5
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_aw3_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw3_v.npy
new file mode 100644
index 0000000..add5ec4
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cb1_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb1_m.npy
new file mode 100644
index 0000000..4db34de
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cb1_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb1_v.npy
new file mode 100644
index 0000000..858b85a
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cb2_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb2_m.npy
new file mode 100644
index 0000000..5c9e4be
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cb2_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb2_v.npy
new file mode 100644
index 0000000..4eaefa9
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cb3_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb3_m.npy
new file mode 100644
index 0000000..57c0a72
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cb3_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb3_v.npy
new file mode 100644
index 0000000..1431264
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cw1_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw1_m.npy
new file mode 100644
index 0000000..c6a168a
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cw1_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw1_v.npy
new file mode 100644
index 0000000..cb88a3c
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cw2_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw2_m.npy
new file mode 100644
index 0000000..67c4e33
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cw2_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw2_v.npy
new file mode 100644
index 0000000..78ee31d
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cw3_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw3_m.npy
new file mode 100644
index 0000000..21fe820
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_cw3_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw3_v.npy
new file mode 100644
index 0000000..3840630
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_logstd_m.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_logstd_m.npy
new file mode 100644
index 0000000..b82fe2f
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_logstd_v.npy b/tools/training_runner/snapshots/generalist_best_r1000/adam_logstd_v.npy
new file mode 100644
index 0000000..de5ffee
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/adam_t.txt b/tools/training_runner/snapshots/generalist_best_r1000/adam_t.txt
new file mode 100644
index 0000000..f289ac0
--- /dev/null
+++ b/tools/training_runner/snapshots/generalist_best_r1000/adam_t.txt
@@ -0,0 +1,13 @@
+1556
+1556
+1556
+1556
+1556
+1556
+1556
+1556
+1556
+1556
+1556
+1556
+1556
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/critic_b1.npy b/tools/training_runner/snapshots/generalist_best_r1000/critic_b1.npy
new file mode 100644
index 0000000..ba92f5f
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/critic_b2.npy b/tools/training_runner/snapshots/generalist_best_r1000/critic_b2.npy
new file mode 100644
index 0000000..6c7bd48
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/critic_b3.npy b/tools/training_runner/snapshots/generalist_best_r1000/critic_b3.npy
new file mode 100644
index 0000000..bb11bb9
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/critic_w1.npy b/tools/training_runner/snapshots/generalist_best_r1000/critic_w1.npy
new file mode 100644
index 0000000..f447748
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/critic_w2.npy b/tools/training_runner/snapshots/generalist_best_r1000/critic_w2.npy
new file mode 100644
index 0000000..93a20e5
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/critic_w3.npy b/tools/training_runner/snapshots/generalist_best_r1000/critic_w3.npy
new file mode 100644
index 0000000..0ec81b7
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/generalist_best_r1000/log_std.npy b/tools/training_runner/snapshots/generalist_best_r1000/log_std.npy
new file mode 100644
index 0000000..83d4a5d
Binary files /dev/null and b/tools/training_runner/snapshots/generalist_best_r1000/log_std.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b1.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b1.npy
new file mode 100644
index 0000000..27a9d95
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b2.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b2.npy
new file mode 100644
index 0000000..27bb520
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b3.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b3.npy
new file mode 100644
index 0000000..2bf5bbf
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w1.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w1.npy
new file mode 100644
index 0000000..d210d4e
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w2.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w2.npy
new file mode 100644
index 0000000..5663e13
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w3.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w3.npy
new file mode 100644
index 0000000..2a6c043
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab1_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab1_m.npy
new file mode 100644
index 0000000..33dddeb
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab1_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab1_v.npy
new file mode 100644
index 0000000..544fbdc
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab2_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab2_m.npy
new file mode 100644
index 0000000..1185f32
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab2_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab2_v.npy
new file mode 100644
index 0000000..9464665
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab3_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab3_m.npy
new file mode 100644
index 0000000..524ffa8
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab3_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab3_v.npy
new file mode 100644
index 0000000..de3f070
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw1_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw1_m.npy
new file mode 100644
index 0000000..bb69063
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw1_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw1_v.npy
new file mode 100644
index 0000000..de2b9ba
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw2_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw2_m.npy
new file mode 100644
index 0000000..7c7d608
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw2_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw2_v.npy
new file mode 100644
index 0000000..713ee35
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw3_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw3_m.npy
new file mode 100644
index 0000000..98308de
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw3_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw3_v.npy
new file mode 100644
index 0000000..9da9bbd
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb1_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb1_m.npy
new file mode 100644
index 0000000..ef89469
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb1_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb1_v.npy
new file mode 100644
index 0000000..be0dc0e
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb2_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb2_m.npy
new file mode 100644
index 0000000..b6387b2
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb2_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb2_v.npy
new file mode 100644
index 0000000..8824e94
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb3_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb3_m.npy
new file mode 100644
index 0000000..04a9580
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb3_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb3_v.npy
new file mode 100644
index 0000000..fcc69f6
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw1_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw1_m.npy
new file mode 100644
index 0000000..b07d020
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw1_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw1_v.npy
new file mode 100644
index 0000000..7d6b734
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw2_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw2_m.npy
new file mode 100644
index 0000000..9acbfcf
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw2_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw2_v.npy
new file mode 100644
index 0000000..81af40d
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw3_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw3_m.npy
new file mode 100644
index 0000000..4ac3e84
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw3_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw3_v.npy
new file mode 100644
index 0000000..d6dd904
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_logstd_m.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_logstd_m.npy
new file mode 100644
index 0000000..34db1c3
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_logstd_v.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_logstd_v.npy
new file mode 100644
index 0000000..724d6f2
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_t.txt b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_t.txt
new file mode 100644
index 0000000..e443c0d
--- /dev/null
+++ b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/adam_t.txt
@@ -0,0 +1,13 @@
+381256
+381256
+381256
+381256
+381256
+381256
+381256
+381256
+381256
+381256
+381256
+381256
+381256
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b1.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b1.npy
new file mode 100644
index 0000000..8270306
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b2.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b2.npy
new file mode 100644
index 0000000..feeac09
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b3.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b3.npy
new file mode 100644
index 0000000..aa46798
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w1.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w1.npy
new file mode 100644
index 0000000..f35022c
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w2.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w2.npy
new file mode 100644
index 0000000..cc3fef7
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w3.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w3.npy
new file mode 100644
index 0000000..71451b4
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/pre_MyFirstBot_r37503/log_std.npy b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/log_std.npy
new file mode 100644
index 0000000..031e1c4
Binary files /dev/null and b/tools/training_runner/snapshots/pre_MyFirstBot_r37503/log_std.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b1.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b1.npy
new file mode 100644
index 0000000..f075004
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b1.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b2.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b2.npy
new file mode 100644
index 0000000..a1f7fb2
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b2.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b3.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b3.npy
new file mode 100644
index 0000000..07174fc
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_b3.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w1.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w1.npy
new file mode 100644
index 0000000..367f7c6
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w1.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w2.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w2.npy
new file mode 100644
index 0000000..5a64b64
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w2.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w3.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w3.npy
new file mode 100644
index 0000000..2e59f37
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/actor_w3.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab1_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab1_m.npy
new file mode 100644
index 0000000..ad898d5
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab1_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab1_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab1_v.npy
new file mode 100644
index 0000000..5c74bf0
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab1_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab2_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab2_m.npy
new file mode 100644
index 0000000..85bda51
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab2_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab2_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab2_v.npy
new file mode 100644
index 0000000..5200f0c
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab2_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab3_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab3_m.npy
new file mode 100644
index 0000000..9f75f5d
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab3_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab3_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab3_v.npy
new file mode 100644
index 0000000..b663bb8
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_ab3_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw1_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw1_m.npy
new file mode 100644
index 0000000..3d0fce8
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw1_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw1_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw1_v.npy
new file mode 100644
index 0000000..200c6ff
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw1_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw2_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw2_m.npy
new file mode 100644
index 0000000..547bf42
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw2_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw2_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw2_v.npy
new file mode 100644
index 0000000..708170a
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw2_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw3_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw3_m.npy
new file mode 100644
index 0000000..41d7225
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw3_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw3_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw3_v.npy
new file mode 100644
index 0000000..d386edb
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_aw3_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb1_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb1_m.npy
new file mode 100644
index 0000000..28b3a8a
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb1_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb1_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb1_v.npy
new file mode 100644
index 0000000..30a341f
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb1_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb2_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb2_m.npy
new file mode 100644
index 0000000..774ab80
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb2_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb2_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb2_v.npy
new file mode 100644
index 0000000..eea901f
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb2_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb3_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb3_m.npy
new file mode 100644
index 0000000..f0872a9
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb3_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb3_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb3_v.npy
new file mode 100644
index 0000000..1893f30
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cb3_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw1_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw1_m.npy
new file mode 100644
index 0000000..b5671ea
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw1_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw1_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw1_v.npy
new file mode 100644
index 0000000..0d7f600
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw1_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw2_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw2_m.npy
new file mode 100644
index 0000000..2353059
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw2_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw2_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw2_v.npy
new file mode 100644
index 0000000..921bbc2
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw2_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw3_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw3_m.npy
new file mode 100644
index 0000000..eee5fc4
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw3_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw3_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw3_v.npy
new file mode 100644
index 0000000..48f86cd
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_cw3_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_logstd_m.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_logstd_m.npy
new file mode 100644
index 0000000..e0480ca
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_logstd_m.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_logstd_v.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_logstd_v.npy
new file mode 100644
index 0000000..f2559f3
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_logstd_v.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_t.txt b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_t.txt
new file mode 100644
index 0000000..db44821
--- /dev/null
+++ b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/adam_t.txt
@@ -0,0 +1,13 @@
+30756
+30756
+30756
+30756
+30756
+30756
+30756
+30756
+30756
+30756
+30756
+30756
+30756
\ No newline at end of file
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b1.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b1.npy
new file mode 100644
index 0000000..4ab65ab
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b1.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b2.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b2.npy
new file mode 100644
index 0000000..564cec6
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b2.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b3.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b3.npy
new file mode 100644
index 0000000..86a66a5
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_b3.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w1.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w1.npy
new file mode 100644
index 0000000..2d422ec
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w1.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w2.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w2.npy
new file mode 100644
index 0000000..d101a54
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w2.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w3.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w3.npy
new file mode 100644
index 0000000..b1e8435
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/critic_w3.npy differ
diff --git a/tools/training_runner/snapshots/v2_MyFirstLeader_r150/log_std.npy b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/log_std.npy
new file mode 100644
index 0000000..09bb96a
Binary files /dev/null and b/tools/training_runner/snapshots/v2_MyFirstLeader_r150/log_std.npy differ