Initial model upload

config.json

@@ -35,7 +35,7 @@
     "pretrained_path": null,
     "paligemma_variant": "gemma_2b",
     "action_expert_variant": "gemma_300m",
-    "dtype": "float32",
+    "dtype": "bfloat16",
     "chunk_size": 50,
     "n_action_steps": 50,
     "max_state_dim": 32,
@@ -57,7 +57,7 @@
         "STATE": "MEAN_STD",
         "ACTION": "MEAN_STD"
     },
-    "gradient_checkpointing": false,
+    "gradient_checkpointing": true,
     "compile_model": false,
     "compile_mode": "max-autotune",
     "optimizer_lr": 2.5e-05,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:220ed4edfbaad285f056ecc5becb21e2522a91d2911f322fa9de43648c1d168b
-size 14005623256
+oid sha256:2c44c15268282487f85d19869e2b39cf8be7c78a8b8e34f6399a5f545f41dcb4
+size 7011548640

policy_postprocessor_step_0_unnormalizer_processor.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1eb191c579eeeccd6fe9ddc5da07838d3aa3fa3a9d535542fd8f75dddc87280d
+oid sha256:c937ac28296b94e2a585ba8931aedc3fdb90e879f31fd263482485434ceb3c33
 size 8528

policy_preprocessor_step_5_normalizer_processor.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1eb191c579eeeccd6fe9ddc5da07838d3aa3fa3a9d535542fd8f75dddc87280d
+oid sha256:c937ac28296b94e2a585ba8931aedc3fdb90e879f31fd263482485434ceb3c33
 size 8528

train_config.json

@@ -1,7 +1,7 @@
 {
     "dataset": {
         "repo_id": "rh5_sample_dataset",
-        "root": "../datasets/rh5_sample_dataset/",
+        "root": "../datasets/RH5_dataset/",
         "episodes": null,
         "image_transforms": {
             "enable": false,
@@ -117,7 +117,7 @@
         "pretrained_path": null,
         "paligemma_variant": "gemma_2b",
         "action_expert_variant": "gemma_300m",
-        "dtype": "float32",
+        "dtype": "bfloat16",
         "chunk_size": 50,
         "n_action_steps": 50,
         "max_state_dim": 32,
@@ -139,7 +139,7 @@
             "STATE": "MEAN_STD",
             "ACTION": "MEAN_STD"
         },
-        "gradient_checkpointing": false,
+        "gradient_checkpointing": true,
         "compile_model": false,
         "compile_mode": "max-autotune",
         "optimizer_lr": 2.5e-05,
@@ -155,17 +155,17 @@
         "scheduler_decay_lr": 2.5e-06,
         "tokenizer_max_length": 48
     },
-    "output_dir": "outputs/train/2025-11-11/01-06-49_pi0",
-    "job_name": "pi0",
+    "output_dir": "outputs/train/2025-11-14/01-13-16_pi0_training",
+    "job_name": "pi0_training",
     "resume": false,
     "seed": 1000,
     "num_workers": 4,
-    "batch_size": 1,
-    "steps": 100000,
+    "batch_size": 32,
+    "steps": 200000,
     "eval_freq": 20000,
-    "log_freq": 100,
+    "log_freq": 200,
     "save_checkpoint": true,
-    "save_freq": 10000,
+    "save_freq": 20000,
     "use_policy_training_preset": true,
     "optimizer": {
         "type": "adamw",
@@ -196,7 +196,7 @@
         "project": "lerobot",
         "entity": null,
         "notes": null,
-        "run_id": "ng36roi7",
+        "run_id": "k9pdbqh0",
         "mode": null
     },
     "checkpoint_path": null,