Upload folder using huggingface_hub

.summary/0/events.out.tfevents.1696703324.rhmmedcatt-proliant-ml350-gen10

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c4041c2ffeee80e57f38c4cec57d7949c5ba1e3b443233cf75b2c50ec9c1435
+size 16965

.summary/1/events.out.tfevents.1696703324.rhmmedcatt-proliant-ml350-gen10

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c89b986bfd9e922d86e525a374e070c6428329945df52b7810c5a86bb238a70
+size 11777

README.md

@@ -15,7 +15,7 @@ model-index:
       type: atari_alien
     metrics:
     - type: mean_reward
-      value: 1007.00 +/- 73.36
+      value: 240.00 +/- 54.22
       name: mean_reward
       verified: false
 ---
@@ -40,22 +40,23 @@ This model as with all the others in the benchmarks was trained asynchronously u
 
 The aim is to reach state of the art (SOTA) performance on each atari environment. I will flag the models with SOTA when they reach at or near these levels. 
 
-The hyperparameters used in the model are the ones I have pushed to my fork of sample-factory: https://github.com/MattStammers/sample-factory
+The hyperparameters used in the model are the ones I have pushed to my fork of sample-factory: https://github.com/MattStammers/sample-factory . Given that https://huggingface.co/edbeeching has kindly shared his.
+I saved time by using many of his tuned hyperparameters to maximise performance. However he used 2 billion training steps. I have started at 100 million to see how performance goes at that level:
 ```
 hyperparameters =  {
   "device": "gpu",
-  "seed": null,
+  "seed": 1234,
   "num_policies": 2,
-  "async_rl": false,
+  "async_rl": true,
   "serial_mode": false,
-  "batched_sampling": false,
+  "batched_sampling": true,
   "num_batches_to_accumulate": 2,
   "worker_num_splits": 1,
-  "policy_workers_per_policy": 2,
+  "policy_workers_per_policy": 1,
   "max_policy_lag": 1000,
   "num_workers": 16,
   "num_envs_per_worker": 2,
-  "batch_size": 256,
+  "batch_size": 1024,
   "num_batches_per_epoch": 8,
   "num_epochs": 4,
   "rollout": 128,
@@ -66,7 +67,7 @@ hyperparameters =  {
   "reward_clip": 1000.0,
   "value_bootstrap": false,
   "normalize_returns": true,
-  "exploration_loss_coeff": 0.01,
+  "exploration_loss_coeff": 0.0004677351413,
   "value_loss_coeff": 0.5,
   "kl_loss_coeff": 0.0,
   "exploration_loss": "entropy",
@@ -80,8 +81,8 @@ hyperparameters =  {
   "adam_eps": 1e-05,
   "adam_beta1": 0.9,
   "adam_beta2": 0.999,
-  "max_grad_norm": 0.5,
-  "learning_rate": 0.00025,
+  "max_grad_norm": 0.0,
+  "learning_rate": 0.0003033891184,
   "lr_schedule": "linear_decay",
   "lr_schedule_kl_threshold": 0.008,
   "lr_adaptive_min": 1e-06,
@@ -89,7 +90,9 @@ hyperparameters =  {
   "obs_subtract_mean": 0.0,
   "obs_scale": 255.0,
   "normalize_input": true,
-  "normalize_input_keys": null,
+  "normalize_input_keys": [
+    "obs"
+  ],
   "decorrelate_experience_max_seconds": 0,
   "decorrelate_envs_on_one_worker": true,
   "actor_worker_gpus": [],
@@ -101,14 +104,14 @@ hyperparameters =  {
   "flush_summaries_interval": 30,
   "stats_avg": 100,
   "summaries_use_frameskip": true,
-  "heartbeat_interval": 20,
-  "heartbeat_reporting_interval": 180,
+  "heartbeat_interval": 10,
+  "heartbeat_reporting_interval": 60,
   "train_for_env_steps": 100000000,
   "train_for_seconds": 10000000000,
   "save_every_sec": 120,
   "keep_checkpoints": 2,
   "load_checkpoint_kind": "latest",
-  "save_milestones_sec": -1,
+  "save_milestones_sec": 1200,
   "save_best_every_sec": 5,
   "save_best_metric": "reward",
   "save_best_after": 100000,
@@ -137,7 +140,7 @@ hyperparameters =  {
   "env_gpu_actions": false,
   "env_gpu_observations": true,
   "env_frameskip": 4,
-  "env_framestack": 4
+  "env_framestack": 4,
   }
 
   ```

checkpoint_p0/best_000000128_131072_reward_6.930.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ac7d83347d29e588042f5ec92127d30422115ffbf8a8ad5bd22ee7fe8a1f871
+size 20795763

checkpoint_p0/checkpoint_000000160_163840.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36c71762acac17d24f351e29ca5d927bb415dc020ddb113bb991c81cb1e43d68
+size 20796099

checkpoint_p1/best_000000128_131072_reward_6.290.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a5b0d005f468da50852478485e2d89c0911b6c41189e6b65ae82d6b23377e11
+size 20795763

checkpoint_p1/checkpoint_000000160_163840.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2c6abc927ef4a4ce3f408f9e8f37a1d7f0a4144240f74f2fdba94c43a635c4e
+size 20796099

config.json

@@ -8,16 +8,16 @@
   "device": "gpu",
   "seed": 1234,
   "num_policies": 2,
-  "async_rl": false,
+  "async_rl": true,
   "serial_mode": false,
-  "batched_sampling": false,
+  "batched_sampling": true,
   "num_batches_to_accumulate": 2,
   "worker_num_splits": 1,
   "policy_workers_per_policy": 1,
   "max_policy_lag": 1000,
   "num_workers": 16,
   "num_envs_per_worker": 2,
-  "batch_size": 256,
+  "batch_size": 1024,
   "num_batches_per_epoch": 8,
   "num_epochs": 4,
   "rollout": 128,
@@ -28,7 +28,7 @@
   "reward_clip": 1000.0,
   "value_bootstrap": false,
   "normalize_returns": true,
-  "exploration_loss_coeff": 0.01,
+  "exploration_loss_coeff": 0.0004677351413,
   "value_loss_coeff": 0.5,
   "kl_loss_coeff": 0.0,
   "exploration_loss": "entropy",
@@ -42,8 +42,8 @@
   "adam_eps": 1e-05,
   "adam_beta1": 0.9,
   "adam_beta2": 0.999,
-  "max_grad_norm": 0.5,
-  "learning_rate": 0.00025,
+  "max_grad_norm": 0.0,
+  "learning_rate": 0.0003033891184,
   "lr_schedule": "linear_decay",
   "lr_schedule_kl_threshold": 0.008,
   "lr_adaptive_min": 1e-06,
@@ -51,7 +51,9 @@
   "obs_subtract_mean": 0.0,
   "obs_scale": 255.0,
   "normalize_input": true,
-  "normalize_input_keys": null,
+  "normalize_input_keys": [
+    "obs"
+  ],
   "decorrelate_experience_max_seconds": 0,
   "decorrelate_envs_on_one_worker": true,
   "actor_worker_gpus": [],
@@ -63,14 +65,14 @@
   "flush_summaries_interval": 30,
   "stats_avg": 100,
   "summaries_use_frameskip": true,
-  "heartbeat_interval": 20,
-  "heartbeat_reporting_interval": 180,
+  "heartbeat_interval": 10,
+  "heartbeat_reporting_interval": 60,
   "train_for_env_steps": 100000000,
   "train_for_seconds": 10000000000,
   "save_every_sec": 120,
   "keep_checkpoints": 2,
   "load_checkpoint_kind": "latest",
-  "save_milestones_sec": -1,
+  "save_milestones_sec": 1200,
   "save_best_every_sec": 5,
   "save_best_metric": "reward",
   "save_best_after": 100000,
@@ -122,7 +124,7 @@
   "pbt_target_objective": "true_objective",
   "pbt_perturb_min": 1.1,
   "pbt_perturb_max": 1.5,
-  "command_line": "--algo=APPO --env=atari_alien --experiment=atari_alien_APPO --num_policies=2 --restart_behavior=restart --train_dir=./train_atari --train_for_env_steps=100000000 --seed=1234 --num_workers=16 --num_envs_per_worker=2 --num_batches_per_epoch=8 --with_wandb=true --wandb_user=matt-stammers --wandb_project=atari_APPO --wandb_group=atari_alien --wandb_job_type=SF --wandb_tags=atari",
+  "command_line": "--algo=APPO --env=atari_alien --experiment=atari_alien_APPO --num_policies=2 --restart_behavior=restart --train_dir=./train_atari --train_for_env_steps=100000000 --seed=1234 --num_workers=16 --num_envs_per_worker=2 --num_batches_per_epoch=8 --async_rl=true --batched_sampling=true --batch_size=1024 --max_grad_norm=0 --learning_rate=0.0003033891184 --heartbeat_interval=10 --heartbeat_reporting_interval=60 --save_milestones_sec=1200 --num_epochs=4 --exploration_loss_coeff=0.0004677351413 --with_wandb=true --wandb_user=matt-stammers --wandb_project=atari_APPO --wandb_group=atari_alien --wandb_job_type=SF --wandb_tags=atari",
   "cli_args": {
     "algo": "APPO",
     "env": "atari_alien",
@@ -131,10 +133,20 @@
     "restart_behavior": "restart",
     "seed": 1234,
     "num_policies": 2,
+    "async_rl": true,
+    "batched_sampling": true,
     "num_workers": 16,
     "num_envs_per_worker": 2,
+    "batch_size": 1024,
     "num_batches_per_epoch": 8,
+    "num_epochs": 4,
+    "exploration_loss_coeff": 0.0004677351413,
+    "max_grad_norm": 0.0,
+    "learning_rate": 0.0003033891184,
+    "heartbeat_interval": 10,
+    "heartbeat_reporting_interval": 60,
     "train_for_env_steps": 100000000,
+    "save_milestones_sec": 1200,
     "with_wandb": true,
     "wandb_user": "matt-stammers",
     "wandb_project": "atari_APPO",
@@ -146,5 +158,5 @@
   },
   "git_hash": "5fff97c2f535da5987d358cdbe6927cccd43621e",
   "git_repo_name": "not a git repository",
-  "wandb_unique_id": "atari_alien_APPO_20231007_171858_185155"
+  "wandb_unique_id": "atari_alien_APPO_20231007_192842_088132"
 }

replay.mp4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4803d6281edfc5aedefef94d87dadad9713cd5209c976ae1ba2f81937ce4209d
-size 2057580
+oid sha256:90a4e8f39d42032eb384f1984cbf359b91df904281209911175b568634c681b9
+size 1394763