Upload policy weights, train config and readme

README.md

@@ -1,13 +1,13 @@
 ---
 base_model: lerobot/smolvla_base
-datasets: TInkybala/Real_Panda_25eps_Screwdriver
+datasets: TInkybala/Real_Panda_FixedCube
 library_name: lerobot
 license: apache-2.0
 model_name: smolvla
 pipeline_tag: robotics
 tags:
-- smolvla
 - robotics
+- smolvla
 - lerobot
 ---
 

config.json

@@ -12,24 +12,8 @@
             "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
             ]
         }
     },
@@ -49,9 +33,9 @@
     "private": null,
     "tags": null,
     "license": null,
-    "pretrained_path": "lerobot/smolvla_base",
+    "pretrained_path": null,
     "chunk_size": 50,
-    "n_action_steps": 30,
+    "n_action_steps": 50,
     "normalization_mapping": {
         "VISUAL": "IDENTITY",
         "STATE": "MEAN_STD",
@@ -79,7 +63,7 @@
     ],
     "optimizer_eps": 1e-08,
     "optimizer_weight_decay": 1e-10,
-    "optimizer_grad_clip_norm": 10.0,
+    "optimizer_grad_clip_norm": 10,
     "scheduler_warmup_steps": 1000,
     "scheduler_decay_steps": 30000,
     "scheduler_decay_lr": 2.5e-06,
@@ -87,9 +71,9 @@
     "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,
+    "prefix_length": -1,
+    "pad_language_to": "longest",
+    "num_expert_layers": -1,
     "num_vlm_layers": 16,
     "self_attn_every_n_layers": 2,
     "expert_width_multiplier": 0.75,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c61b0fd8ecb8c824c9512d3faaa65bdbf91956e36b78bb9353870ded3dfeb829
+oid sha256:73f5f5a9e876ac27fc959c4481d891a288ebfe93472f7d03c93892d6c3c9658a
 size 906712520

train_config.json

@@ -1,6 +1,6 @@
 {
     "dataset": {
-        "repo_id": "TInkybala/Real_Panda_25eps_Screwdriver",
+        "repo_id": "TInkybala/Real_Panda_FixedCube",
         "root": null,
         "episodes": null,
         "image_transforms": {
@@ -90,28 +90,12 @@
                     8
                 ]
             },
-            "observation.images.camera1": {
+            "observation.images.camera_1": {
                 "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
                 ]
             }
         },
@@ -131,9 +115,9 @@
         "private": null,
         "tags": null,
         "license": null,
-        "pretrained_path": "lerobot/smolvla_base",
+        "pretrained_path": null,
         "chunk_size": 50,
-        "n_action_steps": 30,
+        "n_action_steps": 50,
         "normalization_mapping": {
             "VISUAL": "IDENTITY",
             "STATE": "MEAN_STD",
@@ -161,7 +145,7 @@
         ],
         "optimizer_eps": 1e-08,
         "optimizer_weight_decay": 1e-10,
-        "optimizer_grad_clip_norm": 10.0,
+        "optimizer_grad_clip_norm": 10,
         "scheduler_warmup_steps": 1000,
         "scheduler_decay_steps": 30000,
         "scheduler_decay_lr": 2.5e-06,
@@ -169,9 +153,9 @@
         "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,
+        "prefix_length": -1,
+        "pad_language_to": "longest",
+        "num_expert_layers": -1,
         "num_vlm_layers": 16,
         "self_attn_every_n_layers": 2,
         "expert_width_multiplier": 0.75,
@@ -181,24 +165,24 @@
         "compile_model": false,
         "compile_mode": "max-autotune"
     },
-    "output_dir": "outputs/train/TInkybala_Real_Panda_PickScrewdriver_finetune_smolvla_20260311_112054",
+    "output_dir": "outputs/train/TInkybala_Real_Panda_PickScrewdriver_finetune_smolvla_20260311_144748",
     "job_name": "panda_pick_screwdriver_smolvla",
     "resume": false,
     "seed": 1000,
     "num_workers": 4,
-    "batch_size": 32,
-    "steps": 40000,
+    "batch_size": 16,
+    "steps": 20000,
     "eval_freq": 20000,
     "log_freq": 200,
     "tolerance_s": 0.0001,
     "save_checkpoint": true,
-    "save_freq": 10000,
+    "save_freq": 20000,
     "use_policy_training_preset": true,
     "optimizer": {
         "type": "adamw",
         "lr": 0.0001,
         "weight_decay": 1e-10,
-        "grad_clip_norm": 10.0,
+        "grad_clip_norm": 10,
         "betas": [
             0.9,
             0.95
@@ -223,7 +207,7 @@
         "project": "lerobot",
         "entity": null,
         "notes": null,
-        "run_id": "2z71qcr0",
+        "run_id": "il3flh3f",
         "mode": null
     },
     "peft": null,
@@ -232,8 +216,6 @@
     "rabc_kappa": 0.01,
     "rabc_epsilon": 1e-06,
     "rabc_head_mode": "sparse",
-    "rename_map": {
-        "observation.images.camera_1": "observation.images.camera1"
-    },
+    "rename_map": {},
     "checkpoint_path": null
 }