Upload policy weights, train config and readme

README.md

@@ -1,21 +1,22 @@
 ---
+base_model: lerobot/smolvla_base
 datasets: mfyu/piper-make-panda
 library_name: lerobot
 license: apache-2.0
-model_name: act
+model_name: smolvla
 pipeline_tag: robotics
 tags:
-- act
 - robotics
+- smolvla
 - lerobot
 ---
 
-# Model Card for act
+# Model Card for smolvla
 
 <!-- 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.
+[SmolVLA](https://huggingface.co/papers/2506.01844) is a compact, efficient vision-language-action model that achieves competitive performance at reduced computational costs and can be deployed on consumer-grade hardware.
 
 
 This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).

config.json

@@ -1,27 +1,35 @@
 {
-    "type": "act",
+    "type": "smolvla",
     "n_obs_steps": 1,
     "input_features": {
         "observation.state": {
             "type": "STATE",
             "shape": [
-                8
+                6
             ]
         },
-        "observation.images.head": {
+        "observation.images.camera1": {
             "type": "VISUAL",
             "shape": [
                 3,
-                480,
-                640
+                256,
+                256
             ]
         },
-        "observation.images.hand": {
+        "observation.images.camera2": {
             "type": "VISUAL",
             "shape": [
                 3,
-                480,
-                640
+                256,
+                256
+            ]
+        },
+        "observation.images.camera3": {
+            "type": "VISUAL",
+            "shape": [
+                3,
+                256,
+                256
             ]
         }
     },
@@ -29,7 +37,7 @@
         "action": {
             "type": "ACTION",
             "shape": [
-                7
+                6
             ]
         }
     },
@@ -40,31 +48,50 @@
     "private": null,
     "tags": null,
     "license": null,
-    "pretrained_path": null,
-    "chunk_size": 100,
-    "n_action_steps": 100,
+    "pretrained_path": "lerobot/smolvla_base",
+    "chunk_size": 50,
+    "n_action_steps": 50,
     "normalization_mapping": {
-        "VISUAL": "MEAN_STD",
+        "VISUAL": "IDENTITY",
         "STATE": "MEAN_STD",
         "ACTION": "MEAN_STD"
     },
-    "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
+    "max_state_dim": 32,
+    "max_action_dim": 32,
+    "resize_imgs_with_padding": [
+        512,
+        512
+    ],
+    "empty_cameras": 0,
+    "adapt_to_pi_aloha": false,
+    "use_delta_joint_actions_aloha": false,
+    "tokenizer_max_length": 48,
+    "num_steps": 10,
+    "use_cache": true,
+    "freeze_vision_encoder": true,
+    "train_expert_only": true,
+    "train_state_proj": true,
+    "optimizer_lr": 0.0001,
+    "optimizer_betas": [
+        0.9,
+        0.95
+    ],
+    "optimizer_eps": 1e-08,
+    "optimizer_weight_decay": 1e-10,
+    "optimizer_grad_clip_norm": 10.0,
+    "scheduler_warmup_steps": 1000,
+    "scheduler_decay_steps": 30000,
+    "scheduler_decay_lr": 2.5e-06,
+    "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
+    "load_vlm_weights": true,
+    "add_image_special_tokens": false,
+    "attention_mode": "cross_attn",
+    "prefix_length": 0,
+    "pad_language_to": "max_length",
+    "num_expert_layers": 0,
+    "num_vlm_layers": 16,
+    "self_attn_every_n_layers": 2,
+    "expert_width_multiplier": 0.75,
+    "min_period": 0.004,
+    "max_period": 4.0
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:45c08d9e8b7573a64c5b91057215781a9c6b6048f3bd7b4331f2dd5717ab15cd
-size 206712028
+oid sha256:a3719e7bb1b08c84ae232f3863bc2366498496c11f7a908a0945f9054b9a9ee3
+size 906712520

train_config.json

@@ -81,29 +81,37 @@
     },
     "env": null,
     "policy": {
-        "type": "act",
+        "type": "smolvla",
         "n_obs_steps": 1,
         "input_features": {
             "observation.state": {
                 "type": "STATE",
                 "shape": [
-                    8
+                    6
                 ]
             },
-            "observation.images.head": {
+            "observation.images.camera1": {
                 "type": "VISUAL",
                 "shape": [
                     3,
-                    480,
-                    640
+                    256,
+                    256
                 ]
             },
-            "observation.images.hand": {
+            "observation.images.camera2": {
                 "type": "VISUAL",
                 "shape": [
                     3,
-                    480,
-                    640
+                    256,
+                    256
+                ]
+            },
+            "observation.images.camera3": {
+                "type": "VISUAL",
+                "shape": [
+                    3,
+                    256,
+                    256
                 ]
             }
         },
@@ -111,7 +119,7 @@
             "action": {
                 "type": "ACTION",
                 "shape": [
-                    7
+                    6
                 ]
             }
         },
@@ -122,58 +130,83 @@
         "private": null,
         "tags": null,
         "license": null,
-        "pretrained_path": null,
-        "chunk_size": 100,
-        "n_action_steps": 100,
+        "pretrained_path": "lerobot/smolvla_base",
+        "chunk_size": 50,
+        "n_action_steps": 50,
         "normalization_mapping": {
-            "VISUAL": "MEAN_STD",
+            "VISUAL": "IDENTITY",
             "STATE": "MEAN_STD",
             "ACTION": "MEAN_STD"
         },
-        "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
+        "max_state_dim": 32,
+        "max_action_dim": 32,
+        "resize_imgs_with_padding": [
+            512,
+            512
+        ],
+        "empty_cameras": 0,
+        "adapt_to_pi_aloha": false,
+        "use_delta_joint_actions_aloha": false,
+        "tokenizer_max_length": 48,
+        "num_steps": 10,
+        "use_cache": true,
+        "freeze_vision_encoder": true,
+        "train_expert_only": true,
+        "train_state_proj": true,
+        "optimizer_lr": 0.0001,
+        "optimizer_betas": [
+            0.9,
+            0.95
+        ],
+        "optimizer_eps": 1e-08,
+        "optimizer_weight_decay": 1e-10,
+        "optimizer_grad_clip_norm": 10.0,
+        "scheduler_warmup_steps": 1000,
+        "scheduler_decay_steps": 30000,
+        "scheduler_decay_lr": 2.5e-06,
+        "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
+        "load_vlm_weights": true,
+        "add_image_special_tokens": false,
+        "attention_mode": "cross_attn",
+        "prefix_length": 0,
+        "pad_language_to": "max_length",
+        "num_expert_layers": 0,
+        "num_vlm_layers": 16,
+        "self_attn_every_n_layers": 2,
+        "expert_width_multiplier": 0.75,
+        "min_period": 0.004,
+        "max_period": 4.0
     },
-    "output_dir": "outputs/train/act_piper_make_panda_1028_nums_15",
-    "job_name": "act_piper_make_panda",
+    "output_dir": "outputs/train/my_smolvla",
+    "job_name": "my_smolvla_training",
     "resume": false,
     "seed": 1000,
     "num_workers": 4,
-    "batch_size": 8,
-    "steps": 100000,
+    "batch_size": 64,
+    "steps": 20000,
     "eval_freq": 20000,
     "log_freq": 200,
     "save_checkpoint": true,
-    "save_freq": 20000,
+    "save_freq": 5000,
     "use_policy_training_preset": true,
     "optimizer": {
         "type": "adamw",
-        "lr": 1e-05,
-        "weight_decay": 0.0001,
+        "lr": 0.0001,
+        "weight_decay": 1e-10,
         "grad_clip_norm": 10.0,
         "betas": [
             0.9,
-            0.999
+            0.95
         ],
         "eps": 1e-08
     },
-    "scheduler": null,
+    "scheduler": {
+        "type": "cosine_decay_with_warmup",
+        "num_warmup_steps": 1000,
+        "num_decay_steps": 30000,
+        "peak_lr": 0.0001,
+        "decay_lr": 2.5e-06
+    },
     "eval": {
         "n_episodes": 50,
         "batch_size": 50,
@@ -182,12 +215,15 @@
     "wandb": {
         "enable": true,
         "disable_artifact": false,
-        "project": "lerobot-piper",
+        "project": "lerobot",
         "entity": null,
         "notes": null,
-        "run_id": "xi0lo9wm",
+        "run_id": "wbgf315f",
         "mode": null
     },
     "checkpoint_path": null,
-    "rename_map": {}
+    "rename_map": {
+        "observation.images.head": "observation.images.camera1",
+        "observation.images.hand": "observation.images.camera2"
+    }
 }