Upload policy weights, train config and readme

README.md

@@ -1,21 +1,21 @@
 ---
-datasets: b10902046/il_gym
+datasets: b10902046/so101_test
 library_name: lerobot
 license: apache-2.0
-model_name: act
+model_name: behavior_cloning
 pipeline_tag: robotics
 tags:
-- act
 - robotics
 - lerobot
+- behavior_cloning
 ---
 
-# Model Card for act
+# Model Card for behavior_cloning
 
 <!-- Provide a quick summary of what the model is/does. -->
 
 
-[Action Chunking with Transformers (ACT)](https://huggingface.co/papers/2304.13705) is an imitation-learning method that predicts short action chunks instead of single steps. It learns from teleoperated data and often achieves high success rates.
+_Model type not recognized — please update this template._
 
 
 This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).

config.json

@@ -1,32 +1,15 @@
 {
-    "type": "act",
+    "type": "behavior_cloning",
     "n_obs_steps": 1,
     "normalization_mapping": {
-        "VISUAL": "MEAN_STD",
-        "STATE": "MEAN_STD",
-        "ACTION": "MEAN_STD"
+        "STATE": "MIN_MAX",
+        "ACTION": "MIN_MAX"
     },
     "input_features": {
         "observation.state": {
             "type": "STATE",
             "shape": [
-                18
-            ]
-        },
-        "observation.images.front": {
-            "type": "VISUAL",
-            "shape": [
-                3,
-                128,
-                128
-            ]
-        },
-        "observation.images.wrist": {
-            "type": "VISUAL",
-            "shape": [
-                3,
-                128,
-                128
+                6
             ]
         }
     },
@@ -34,7 +17,7 @@
         "action": {
             "type": "ACTION",
             "shape": [
-                4
+                6
             ]
         }
     },
@@ -45,25 +28,18 @@
     "private": null,
     "tags": null,
     "license": null,
-    "chunk_size": 100,
-    "n_action_steps": 100,
-    "vision_backbone": "resnet18",
-    "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
-    "replace_final_stride_with_dilation": false,
-    "pre_norm": false,
-    "dim_model": 512,
-    "n_heads": 8,
-    "dim_feedforward": 3200,
-    "feedforward_activation": "relu",
-    "n_encoder_layers": 4,
-    "n_decoder_layers": 1,
-    "use_vae": true,
-    "latent_dim": 32,
-    "n_vae_encoder_layers": 4,
-    "temporal_ensemble_coeff": null,
-    "dropout": 0.1,
-    "kl_weight": 10.0,
-    "optimizer_lr": 1e-05,
-    "optimizer_weight_decay": 0.0001,
-    "optimizer_lr_backbone": 1e-05
+    "hidden_dims": [
+        256,
+        256
+    ],
+    "activation": "relu",
+    "optimizer_lr": 0.0001,
+    "optimizer_betas": [
+        0.9,
+        0.999
+    ],
+    "optimizer_eps": 1e-08,
+    "optimizer_weight_decay": 1e-06,
+    "scheduler_name": "cosine",
+    "scheduler_warmup_steps": 500
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f7642c24bdf86fe5b40d246d0af0ff7720c5a1bf0aea9b2bc55174c66ecc07eb
-size 206742080
+oid sha256:67bca0eb7179b50fcfaefdb050bea6feeb46711940e2b6a42e26d6b7a2c62150
+size 277736

train_config.json

@@ -1,6 +1,6 @@
 {
     "dataset": {
-        "repo_id": "b10902046/il_gym",
+        "repo_id": "b10902046/so101_test",
         "root": null,
         "episodes": null,
         "image_transforms": {
@@ -66,34 +66,17 @@
     },
     "env": null,
     "policy": {
-        "type": "act",
+        "type": "behavior_cloning",
         "n_obs_steps": 1,
         "normalization_mapping": {
-            "VISUAL": "MEAN_STD",
-            "STATE": "MEAN_STD",
-            "ACTION": "MEAN_STD"
+            "STATE": "MIN_MAX",
+            "ACTION": "MIN_MAX"
         },
         "input_features": {
             "observation.state": {
                 "type": "STATE",
                 "shape": [
-                    18
-                ]
-            },
-            "observation.images.front": {
-                "type": "VISUAL",
-                "shape": [
-                    3,
-                    128,
-                    128
-                ]
-            },
-            "observation.images.wrist": {
-                "type": "VISUAL",
-                "shape": [
-                    3,
-                    128,
-                    128
+                    6
                 ]
             }
         },
@@ -101,7 +84,7 @@
             "action": {
                 "type": "ACTION",
                 "shape": [
-                    4
+                    6
                 ]
             }
         },
@@ -112,30 +95,23 @@
         "private": null,
         "tags": null,
         "license": null,
-        "chunk_size": 100,
-        "n_action_steps": 100,
-        "vision_backbone": "resnet18",
-        "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
-        "replace_final_stride_with_dilation": false,
-        "pre_norm": false,
-        "dim_model": 512,
-        "n_heads": 8,
-        "dim_feedforward": 3200,
-        "feedforward_activation": "relu",
-        "n_encoder_layers": 4,
-        "n_decoder_layers": 1,
-        "use_vae": true,
-        "latent_dim": 32,
-        "n_vae_encoder_layers": 4,
-        "temporal_ensemble_coeff": null,
-        "dropout": 0.1,
-        "kl_weight": 10.0,
-        "optimizer_lr": 1e-05,
-        "optimizer_weight_decay": 0.0001,
-        "optimizer_lr_backbone": 1e-05
+        "hidden_dims": [
+            256,
+            256
+        ],
+        "activation": "relu",
+        "optimizer_lr": 0.0001,
+        "optimizer_betas": [
+            0.9,
+            0.999
+        ],
+        "optimizer_eps": 1e-08,
+        "optimizer_weight_decay": 1e-06,
+        "scheduler_name": "cosine",
+        "scheduler_warmup_steps": 500
     },
-    "output_dir": "outputs/train/il_sim_test",
-    "job_name": "il_sim_test",
+    "output_dir": "outputs/train/act_so101_test",
+    "job_name": "act_so101_test",
     "resume": false,
     "seed": 1000,
     "num_workers": 4,
@@ -147,9 +123,9 @@
     "save_freq": 20000,
     "use_policy_training_preset": true,
     "optimizer": {
-        "type": "adamw",
-        "lr": 1e-05,
-        "weight_decay": 0.0001,
+        "type": "adam",
+        "lr": 0.0001,
+        "weight_decay": 1e-06,
         "grad_clip_norm": 10.0,
         "betas": [
             0.9,
@@ -157,7 +133,11 @@
         ],
         "eps": 1e-08
     },
-    "scheduler": null,
+    "scheduler": {
+        "type": "diffuser",
+        "num_warmup_steps": 500,
+        "name": "cosine"
+    },
     "eval": {
         "n_episodes": 50,
         "batch_size": 50,