Upload folder using huggingface_hub

.summary/0/events.out.tfevents.1715928454.f186c3f8af36

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b703eebcfbe830ecbbc87c2b6342e1b51cd12991a83a0c0eb90fc97ede7891f9
+size 439460

README.md

@@ -15,7 +15,7 @@ model-index:
       type: doom_health_gathering_supreme
     metrics:
     - type: mean_reward
-      value: 8.36 +/- 4.79
+      value: 9.46 +/- 5.33
       name: mean_reward
       verified: false
 ---

checkpoint_p0/best_000000727_2977792_reward_22.923.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edb709545039827fac47c97b0e58f7096f5d310bfe3f1856f3b6c7f4c215589d
+size 34929051

checkpoint_p0/checkpoint_000000934_3825664.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a08a1930c15f41cf3d62ee404714b54d1845d0a261573ac48d837ec71341264
+size 34929477

checkpoint_p0/checkpoint_000000978_4005888.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab8d153277fc0035347bc05b18fa6f49391f445404ed649821d601f9bf260a51
+oid sha256:0fc3c91f45dddb59a50df15bcdedf79b96730793505c3ea1539061166caac640
 size 34929477

config.json

@@ -6,7 +6,7 @@
   "train_dir": "/kaggle/working/train_dir",
   "restart_behavior": "resume",
   "device": "gpu",
-  "seed": null,
+  "seed": 0,
   "num_policies": 1,
   "async_rl": true,
   "serial_mode": false,
@@ -16,7 +16,7 @@
   "policy_workers_per_policy": 1,
   "max_policy_lag": 1000,
   "num_workers": 8,
-  "num_envs_per_worker": 4,
+  "num_envs_per_worker": 8,
   "batch_size": 1024,
   "num_batches_per_epoch": 1,
   "num_epochs": 1,
@@ -30,7 +30,7 @@
   "normalize_returns": true,
   "exploration_loss_coeff": 0.001,
   "value_loss_coeff": 0.5,
-  "kl_loss_coeff": 0.0,
+  "kl_loss_coeff": 0.3,
   "exploration_loss": "symmetric_kl",
   "gae_lambda": 0.95,
   "ppo_clip_ratio": 0.1,
@@ -130,11 +130,13 @@
   "wide_aspect_ratio": false,
   "eval_env_frameskip": 1,
   "fps": 35,
-  "command_line": "--env=doom_health_gathering_supreme --num_workers=8 --num_envs_per_worker=4 --train_for_env_steps=4000000",
+  "command_line": "--env=doom_health_gathering_supreme --seed=0 --kl_loss_coeff=0.3 --num_workers=8 --num_envs_per_worker=8 --train_for_env_steps=4000000",
   "cli_args": {
     "env": "doom_health_gathering_supreme",
+    "seed": 0,
     "num_workers": 8,
-    "num_envs_per_worker": 4,
+    "num_envs_per_worker": 8,
+    "kl_loss_coeff": 0.3,
     "train_for_env_steps": 4000000
   },
   "git_hash": "unknown",

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aefb90d9d9fc0a9d21e8d66140b485b21cd5696a30c9d972a2b2c01f8c381788
-size 16378745
+oid sha256:f5a3b35f488a265d644abd68809e7dce7747cf58e06d9b565a3b67aeb10a55c6
+size 18095784