Upload policy weights, train config and readme

README.md

@@ -1,22 +1,21 @@
 ---
-base_model: lerobot/smolvla_base
 datasets: ThavT/red_block_in_tape
 library_name: lerobot
 license: apache-2.0
-model_name: smolvla
+model_name: act
 pipeline_tag: robotics
 tags:
+- act
 - lerobot
 - robotics
-- smolvla
 ---
 
-# Model Card for smolvla
+# Model Card for act
 
 <!-- Provide a quick summary of what the model is/does. -->
 
 
-[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.
+[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.
 
 
 This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).

config.json

@@ -1,5 +1,5 @@
 {
-    "type": "smolvla",
+    "type": "act",
     "n_obs_steps": 1,
     "input_features": {
         "observation.state": {
@@ -8,31 +8,15 @@
                 6
             ]
         },
-        "observation.images.camera1": {
+        "observation.images.cam_0": {
             "type": "VISUAL",
             "shape": [
                 3,
-                256,
-                256
-            ]
-        },
-        "observation.images.camera2": {
-            "type": "VISUAL",
-            "shape": [
-                3,
-                256,
-                256
-            ]
-        },
-        "observation.images.camera3": {
-            "type": "VISUAL",
-            "shape": [
-                3,
-                256,
-                256
+                480,
+                640
             ]
         },
-        "observation.images.empty_camera_0": {
+        "observation.images.cam_1": {
             "type": "VISUAL",
             "shape": [
                 3,
@@ -57,53 +41,31 @@
     "private": null,
     "tags": null,
     "license": null,
-    "pretrained_path": "lerobot/smolvla_base",
-    "chunk_size": 50,
-    "n_action_steps": 50,
+    "pretrained_path": null,
+    "chunk_size": 100,
+    "n_action_steps": 100,
     "normalization_mapping": {
-        "VISUAL": "IDENTITY",
+        "VISUAL": "MEAN_STD",
         "STATE": "MEAN_STD",
         "ACTION": "MEAN_STD"
     },
-    "max_state_dim": 32,
-    "max_action_dim": 32,
-    "resize_imgs_with_padding": [
-        512,
-        512
-    ],
-    "empty_cameras": 1,
-    "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,
-    "rtc_config": null,
-    "compile_model": false,
-    "compile_mode": "max-autotune"
+    "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
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3f26002ff1b0f42cc9392fa6604471fdbefcc425694617556f51d2e75904bed
-size 906712520
+oid sha256:82b9741edd0f5c3b5ed1eb6e2c68792a435fe3dc21b3a69293ce158ed9dff73b
+size 206699736

train_config.json

@@ -81,7 +81,7 @@
     },
     "env": null,
     "policy": {
-        "type": "smolvla",
+        "type": "act",
         "n_obs_steps": 1,
         "input_features": {
             "observation.state": {
@@ -90,31 +90,15 @@
                     6
                 ]
             },
-            "observation.images.camera1": {
+            "observation.images.cam_0": {
                 "type": "VISUAL",
                 "shape": [
                     3,
-                    256,
-                    256
-                ]
-            },
-            "observation.images.camera2": {
-                "type": "VISUAL",
-                "shape": [
-                    3,
-                    256,
-                    256
-                ]
-            },
-            "observation.images.camera3": {
-                "type": "VISUAL",
-                "shape": [
-                    3,
-                    256,
-                    256
+                    480,
+                    640
                 ]
             },
-            "observation.images.empty_camera_0": {
+            "observation.images.cam_1": {
                 "type": "VISUAL",
                 "shape": [
                     3,
@@ -139,55 +123,33 @@
         "private": null,
         "tags": null,
         "license": null,
-        "pretrained_path": "lerobot/smolvla_base",
-        "chunk_size": 50,
-        "n_action_steps": 50,
+        "pretrained_path": null,
+        "chunk_size": 100,
+        "n_action_steps": 100,
         "normalization_mapping": {
-            "VISUAL": "IDENTITY",
+            "VISUAL": "MEAN_STD",
             "STATE": "MEAN_STD",
             "ACTION": "MEAN_STD"
         },
-        "max_state_dim": 32,
-        "max_action_dim": 32,
-        "resize_imgs_with_padding": [
-            512,
-            512
-        ],
-        "empty_cameras": 1,
-        "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,
-        "rtc_config": null,
-        "compile_model": false,
-        "compile_mode": "max-autotune"
+        "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
     },
     "output_dir": "outputs/train/red_block_in_tape",
     "job_name": "smolvla_red_block_training",
@@ -205,22 +167,16 @@
     "use_policy_training_preset": true,
     "optimizer": {
         "type": "adamw",
-        "lr": 0.0001,
-        "weight_decay": 1e-10,
+        "lr": 1e-05,
+        "weight_decay": 0.0001,
         "grad_clip_norm": 10.0,
         "betas": [
             0.9,
-            0.95
+            0.999
         ],
         "eps": 1e-08
     },
-    "scheduler": {
-        "type": "cosine_decay_with_warmup",
-        "num_warmup_steps": 1000,
-        "num_decay_steps": 30000,
-        "peak_lr": 0.0001,
-        "decay_lr": 2.5e-06
-    },
+    "scheduler": null,
     "eval": {
         "n_episodes": 50,
         "batch_size": 50,
@@ -232,7 +188,7 @@
         "project": "lerobot_red_block",
         "entity": null,
         "notes": null,
-        "run_id": "tl6jkmrr",
+        "run_id": "26obbwcp",
         "mode": null,
         "add_tags": true
     },
@@ -242,9 +198,6 @@
     "rabc_kappa": 0.01,
     "rabc_epsilon": 1e-06,
     "rabc_head_mode": "sparse",
-    "rename_map": {
-        "observation.images.cam_0": "observation.images.camera1",
-        "observation.images.cam_1": "observation.images.camera2"
-    },
+    "rename_map": {},
     "checkpoint_path": null
 }