Upload policy weights, train config and readme

README.md

@@ -0,0 +1,62 @@
+---
+datasets: kodate/record-all-5-fix
+library_name: lerobot
+license: apache-2.0
+model_name: act
+pipeline_tag: robotics
+tags:
+- lerobot
+- robotics
+- act
+---
+
+# Model Card for act
+
+<!-- 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.
+
+
+This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).
+See the full documentation at [LeRobot Docs](https://huggingface.co/docs/lerobot/index).
+
+---
+
+## How to Get Started with the Model
+
+For a complete walkthrough, see the [training guide](https://huggingface.co/docs/lerobot/il_robots#train-a-policy).
+Below is the short version on how to train and run inference/eval:
+
+### Train from scratch
+
+```bash
+lerobot-train \
+  --dataset.repo_id=${HF_USER}/<dataset> \
+  --policy.type=act \
+  --output_dir=outputs/train/<desired_policy_repo_id> \
+  --job_name=lerobot_training \
+  --policy.device=cuda \
+  --policy.repo_id=${HF_USER}/<desired_policy_repo_id>
+  --wandb.enable=true
+```
+
+_Writes checkpoints to `outputs/train/<desired_policy_repo_id>/checkpoints/`._
+
+### Evaluate the policy/run inference
+
+```bash
+lerobot-record \
+  --robot.type=so100_follower \
+  --dataset.repo_id=<hf_user>/eval_<dataset> \
+  --policy.path=<hf_user>/<desired_policy_repo_id> \
+  --episodes=10
+```
+
+Prefix the dataset repo with **eval\_** and supply `--policy.path` pointing to a local or hub checkpoint.
+
+---
+
+## Model Details
+
+- **License:** apache-2.0

config.json

@@ -40,7 +40,7 @@
     "private": null,
     "tags": null,
     "license": null,
-    "pretrained_path": null,
+    "pretrained_path": "outputs/train/gpu_class_test1/checkpoints/last/pretrained_model",
     "chunk_size": 100,
     "n_action_steps": 100,
     "normalization_mapping": {
@@ -50,7 +50,7 @@
     },
     "vision_backbone": "resnet18",
     "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
-    "replace_final_stride_with_dilation": false,
+    "replace_final_stride_with_dilation": 0,
     "pre_norm": false,
     "dim_model": 512,
     "n_heads": 8,

train_config.json

@@ -122,7 +122,7 @@
         "private": null,
         "tags": null,
         "license": null,
-        "pretrained_path": null,
+        "pretrained_path": "outputs/train/gpu_class_test1/checkpoints/last/pretrained_model",
         "chunk_size": 100,
         "n_action_steps": 100,
         "normalization_mapping": {
@@ -132,7 +132,7 @@
         },
         "vision_backbone": "resnet18",
         "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
-        "replace_final_stride_with_dilation": false,
+        "replace_final_stride_with_dilation": 0,
         "pre_norm": false,
         "dim_model": 512,
         "n_heads": 8,
@@ -152,7 +152,7 @@
     },
     "output_dir": "outputs/train/gpu_class_test1",
     "job_name": "gpu_class_test1",
-    "resume": false,
+    "resume": true,
     "seed": 1000,
     "num_workers": 4,
     "batch_size": 8,
@@ -188,6 +188,6 @@
         "run_id": null,
         "mode": null
     },
-    "checkpoint_path": null,
+    "checkpoint_path": "outputs/train/gpu_class_test1/checkpoints/last",
     "rename_map": {}
 }