Upload model checkpoint from pretrained_model

config.json

@@ -33,7 +33,7 @@
     "private": null,
     "tags": null,
     "license": null,
-    "pretrained_path": "outputs/train/so101_sort_so101_sort_smolvla_20251107_2019/checkpoints/last/pretrained_model",
+    "pretrained_path": null,
     "chunk_size": 50,
     "n_action_steps": 50,
     "normalization_mapping": {
@@ -56,16 +56,16 @@
     "freeze_vision_encoder": true,
     "train_expert_only": true,
     "train_state_proj": true,
-    "optimizer_lr": 0.0003,
+    "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": 350,
-    "scheduler_decay_steps": 7000,
+    "optimizer_grad_clip_norm": 10,
+    "scheduler_warmup_steps": 10000,
+    "scheduler_decay_steps": 200000,
     "scheduler_decay_lr": 2.5e-06,
     "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
     "load_vlm_weights": false,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6265a9d5b6b0b093f931192c8a92657d6aeabd8425f0081922d71b6fce508d72
+oid sha256:ef6debe8b189c220f6251cc5bf14390893588620adb0dabd1dd82a3ab8e272b6
 size 1197789224

train_config.json

@@ -101,7 +101,7 @@
         "private": null,
         "tags": null,
         "license": null,
-        "pretrained_path": "outputs/train/so101_sort_so101_sort_smolvla_20251107_2019/checkpoints/last/pretrained_model",
+        "pretrained_path": null,
         "chunk_size": 50,
         "n_action_steps": 50,
         "normalization_mapping": {
@@ -124,16 +124,16 @@
         "freeze_vision_encoder": true,
         "train_expert_only": true,
         "train_state_proj": true,
-        "optimizer_lr": 0.0003,
+        "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": 350,
-        "scheduler_decay_steps": 7000,
+        "optimizer_grad_clip_norm": 10,
+        "scheduler_warmup_steps": 10000,
+        "scheduler_decay_steps": 200000,
         "scheduler_decay_lr": 2.5e-06,
         "vlm_model_name": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct",
         "load_vlm_weights": false,
@@ -148,13 +148,13 @@
         "min_period": 0.004,
         "max_period": 4.0
     },
-    "output_dir": "outputs/train/so101_sort_so101_sort_smolvla_20251107_2019",
-    "job_name": "so101_sort_so101_sort_smolvla_20251107_2019",
-    "resume": true,
+    "output_dir": "outputs/train/so101_sort_so101_sort_smolvla_20251110_0139",
+    "job_name": "so101_sort_so101_sort_smolvla_20251110_0139",
+    "resume": false,
     "seed": 1000,
     "num_workers": 4,
-    "batch_size": 4,
-    "steps": 7000,
+    "batch_size": 64,
+    "steps": 200000,
     "eval_freq": 200,
     "log_freq": 100,
     "save_checkpoint": true,
@@ -162,9 +162,9 @@
     "use_policy_training_preset": true,
     "optimizer": {
         "type": "adamw",
-        "lr": 0.0003,
+        "lr": 0.0001,
         "weight_decay": 1e-10,
-        "grad_clip_norm": 10.0,
+        "grad_clip_norm": 10,
         "betas": [
             0.9,
             0.95
@@ -173,9 +173,9 @@
     },
     "scheduler": {
         "type": "cosine_decay_with_warmup",
-        "num_warmup_steps": 250,
-        "num_decay_steps": 5000,
-        "peak_lr": 0.0003,
+        "num_warmup_steps": 10000,
+        "num_decay_steps": 200000,
+        "peak_lr": 0.0001,
         "decay_lr": 2.5e-06
     },
     "eval": {
@@ -184,12 +184,12 @@
         "use_async_envs": false
     },
     "wandb": {
-        "enable": false,
+        "enable": true,
         "disable_artifact": false,
         "project": "lerobot-shabby",
         "entity": null,
         "notes": null,
-        "run_id": null,
+        "run_id": "dzsh8gnl",
         "mode": null
     },
     "peft": null