Upload policy weights, train config and readme

README.md

@@ -0,0 +1,63 @@
+---
+base_model: lerobot/smolvla_base
+datasets: Beable/lerobot-SOARM100-sim2
+library_name: lerobot
+license: apache-2.0
+model_name: smolvla
+pipeline_tag: robotics
+tags:
+- lerobot
+- robotics
+- smolvla
+---
+
+# Model Card for smolvla
+
+<!-- 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.
+
+
+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

@@ -11,8 +11,8 @@
             "type": "VISUAL",
             "shape": [
                 3,
-                360,
-                640
+                720,
+                1280
             ]
         },
         "observation.state": {
@@ -32,8 +32,8 @@
     },
     "device": "cuda",
     "use_amp": false,
-    "push_to_hub": false,
-    "repo_id": null,
+    "push_to_hub": true,
+    "repo_id": "Beable/leroVLA",
     "private": null,
     "tags": null,
     "license": null,
@@ -51,16 +51,16 @@
     "tokenizer_max_length": 48,
     "num_steps": 10,
     "use_cache": true,
-    "freeze_vision_encoder": true,
-    "train_expert_only": true,
+    "freeze_vision_encoder": false,
+    "train_expert_only": false,
     "train_state_proj": true,
-    "optimizer_lr": 0.0001,
+    "optimizer_lr": 1e-05,
     "optimizer_betas": [
         0.9,
         0.95
     ],
     "optimizer_eps": 1e-08,
-    "optimizer_weight_decay": 1e-10,
+    "optimizer_weight_decay": 0.0001,
     "optimizer_grad_clip_norm": 10,
     "scheduler_warmup_steps": 1000,
     "scheduler_decay_steps": 30000,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:de1d0562a61e5a3310a1a486f38aea8707b3493667b62b5bd8afce52d32ca9c6
+oid sha256:41f9cc12e3298863908f85a0dbf8696f88a6f9f8ed256ebbfbd594de9351197a
 size 1197790016

train_config.json

@@ -1,10 +1,10 @@
 {
     "dataset": {
-        "repo_id": "Beable/lerobot-SOARM100-sim",
+        "repo_id": "Beable/lerobot-SOARM100-sim2",
         "root": null,
         "episodes": null,
         "image_transforms": {
-            "enable": false,
+            "enable": true,
             "max_num_transforms": 3,
             "random_order": false,
             "tfs": {
@@ -60,10 +60,10 @@
                 }
             }
         },
-        "revision": null,
+        "revision": "v2.1",
         "use_imagenet_stats": true,
-        "video_backend": "pyav",
-        "tolerance_s": 0.2,
+        "video_backend": "torchcodec",
+        "tolerance_s": 0.02,
         "timestamps_check": "warn"
     },
     "env": null,
@@ -80,8 +80,8 @@
                 "type": "VISUAL",
                 "shape": [
                     3,
-                    360,
-                    640
+                    720,
+                    1280
                 ]
             },
             "observation.state": {
@@ -101,8 +101,8 @@
         },
         "device": "cuda",
         "use_amp": false,
-        "push_to_hub": false,
-        "repo_id": null,
+        "push_to_hub": true,
+        "repo_id": "Beable/leroVLA",
         "private": null,
         "tags": null,
         "license": null,
@@ -120,16 +120,16 @@
         "tokenizer_max_length": 48,
         "num_steps": 10,
         "use_cache": true,
-        "freeze_vision_encoder": true,
-        "train_expert_only": true,
+        "freeze_vision_encoder": false,
+        "train_expert_only": false,
         "train_state_proj": true,
-        "optimizer_lr": 0.0001,
+        "optimizer_lr": 1e-05,
         "optimizer_betas": [
             0.9,
             0.95
         ],
         "optimizer_eps": 1e-08,
-        "optimizer_weight_decay": 1e-10,
+        "optimizer_weight_decay": 0.0001,
         "optimizer_grad_clip_norm": 10,
         "scheduler_warmup_steps": 1000,
         "scheduler_decay_steps": 30000,
@@ -147,22 +147,22 @@
         "min_period": 0.004,
         "max_period": 4.0
     },
-    "output_dir": "outputs/train/vla_ft",
-    "job_name": "smolvla_ft",
+    "output_dir": "outputs/train/2025-08-14/02-35-56_smolvla",
+    "job_name": "smolvla",
     "resume": false,
     "seed": 1000,
     "num_workers": 4,
-    "batch_size": 32,
-    "steps": 20000,
+    "batch_size": 4,
+    "steps": 15000,
     "eval_freq": 20000,
-    "log_freq": 50,
+    "log_freq": 100,
     "save_checkpoint": true,
-    "save_freq": 20000,
+    "save_freq": 2000,
     "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,
         "betas": [
             0.9,
@@ -174,12 +174,12 @@
         "type": "cosine_decay_with_warmup",
         "num_warmup_steps": 1000,
         "num_decay_steps": 30000,
-        "peak_lr": 0.0001,
+        "peak_lr": 1e-05,
         "decay_lr": 2.5e-06
     },
     "eval": {
-        "n_episodes": 0,
-        "batch_size": 0,
+        "n_episodes": 50,
+        "batch_size": 50,
         "use_async_envs": false
     },
     "wandb": {