Add files using upload-large-folder tool

.gitattributes

@@ -58,3 +58,5 @@ dit4dit_libero_all_full_finetune_bs256/tokenizer/tokenizer.json filter=lfs diff=
 pi05_paligemma_robocasa_full_data_full_finetune_bs256/pi05_paligemma_robocasa_full_data_full_finetune_2026_08_25_09_11_24.jsonl filter=lfs diff=lfs merge=lfs -text
 pi0_paligemma_robocasa_full_data_full_finetune_bs256/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 pi0_paligemma_robocasa_full_data_full_finetune_bs256/pi0_paligemma_robocasa_full_data_full_finetune_2026_08_26_13_07_33.jsonl filter=lfs diff=lfs merge=lfs -text
+gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_2026_08_28_19_35_25.jsonl filter=lfs diff=lfs merge=lfs -text
+gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/checkpoints/step-100000-epoch-02-loss=0.0054.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b8e72f8f468586186936869b850d1de176727615abf2fd2921fa8678450b876d
+size 8850981096

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/config.json

@@ -0,0 +1,722 @@
+{
+  "_OFFICIAL_GR1_STATS_PATH": "./datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json",
+  "_QWEN3VL_TOKENIZER": "./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/",
+  "_QWEN3VL_VLA_CKPT": "./checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors",
+  "_QWEN3VL_VLA_ROOT": "./checkpoints/gr00t_qwen3vl_0.6b_libero",
+  "_QWEN3VL_VLM_CONFIG": {
+    "architectures": [
+      "Qwen3VLAForConditionalGeneration"
+    ],
+    "dtype": "bfloat16",
+    "eos_token_id": 151645,
+    "image_token_id": 151655,
+    "model_type": "qwen3_vl",
+    "pad_token_id": 151643,
+    "pos_skipping_range": 4096,
+    "text_config": {
+      "attention_bias": false,
+      "attention_dropout": 0.0,
+      "bos_token_id": 151643,
+      "dtype": "bfloat16",
+      "eos_token_id": 151645,
+      "head_dim": 128,
+      "hidden_act": "silu",
+      "hidden_size": 1024,
+      "initializer_range": 0.02,
+      "intermediate_size": 3072,
+      "max_position_embeddings": 262144,
+      "model_type": "qwen3_vl_text",
+      "num_attention_heads": 16,
+      "num_hidden_layers": 28,
+      "num_key_value_heads": 8,
+      "pad_token_id": null,
+      "rms_norm_eps": 1e-06,
+      "rope_parameters": {
+        "mrope_interleaved": true,
+        "mrope_section": [
+          24,
+          20,
+          20
+        ],
+        "rope_theta": 5000000,
+        "rope_type": "default"
+      },
+      "tie_word_embeddings": true,
+      "use_cache": true,
+      "vocab_size": 151936
+    },
+    "tie_word_embeddings": true,
+    "use_another_LLM_path": "",
+    "use_pos_skipping": false,
+    "vision_config": {
+      "deepstack_visual_indexes": [
+        5,
+        11,
+        17
+      ],
+      "depth": 24,
+      "dtype": "bfloat16",
+      "hidden_act": "gelu_pytorch_tanh",
+      "hidden_size": 1024,
+      "in_channels": 3,
+      "initializer_range": 0.02,
+      "intermediate_size": 4096,
+      "model_type": "qwen3_vl",
+      "num_heads": 16,
+      "num_position_embeddings": 2304,
+      "out_hidden_size": 1024,
+      "patch_size": 16,
+      "spatial_merge_size": 2,
+      "temporal_patch_size": 2
+    }
+  },
+  "_ROBOCASA_DATA_ROOT": "./datasets/robocasa_lerobot_V2.1",
+  "_ROBOCASA_ENV_SUFFIX": "_GR1ArmsAndWaistFourierHands_Env",
+  "_ROBOCASA_STATISTIC_NAME": "robocasa_gr1_24tasks_30ep",
+  "_ROBOCASA_TASKS": [
+    "PnPBottleToCabinetClose",
+    "PnPCanToDrawerClose",
+    "PnPCupToDrawerClose",
+    "PnPMilkToMicrowaveClose",
+    "PnPPotatoToMicrowaveClose",
+    "PnPWineToCabinetClose",
+    "PosttrainPnPNovelFromCuttingboardToBasketSplitA",
+    "PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA",
+    "PosttrainPnPNovelFromCuttingboardToPanSplitA",
+    "PosttrainPnPNovelFromCuttingboardToPotSplitA",
+    "PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA",
+    "PosttrainPnPNovelFromPlacematToBasketSplitA",
+    "PosttrainPnPNovelFromPlacematToBowlSplitA",
+    "PosttrainPnPNovelFromPlacematToPlateSplitA",
+    "PosttrainPnPNovelFromPlacematToTieredshelfSplitA",
+    "PosttrainPnPNovelFromPlateToBowlSplitA",
+    "PosttrainPnPNovelFromPlateToCardboardboxSplitA",
+    "PosttrainPnPNovelFromPlateToPanSplitA",
+    "PosttrainPnPNovelFromPlateToPlateSplitA",
+    "PosttrainPnPNovelFromTrayToCardboardboxSplitA",
+    "PosttrainPnPNovelFromTrayToPlateSplitA",
+    "PosttrainPnPNovelFromTrayToPotSplitA",
+    "PosttrainPnPNovelFromTrayToTieredbasketSplitA",
+    "PosttrainPnPNovelFromTrayToTieredshelfSplitA"
+  ],
+  "_ROBOCASA_TASK_PREFIX": "gr1_unified",
+  "eval": {
+    "action_order": "n15",
+    "benchmark": "robocasa",
+    "dataset": {
+      "transforms": [
+        {
+          "center_crop_scale": 0.95,
+          "embodiment_id": 24,
+          "img_key": "video.ego_view_bg_crop_pad_res256_freq20",
+          "normalize": true,
+          "resize_size": 224,
+          "type": "ProcessRobocasaEvalInputs",
+          "value_range": "unit"
+        },
+        {
+          "type": "RobocasaGR1N15Bridge"
+        },
+        {
+          "action_key": "action",
+          "norm_type": "min_max",
+          "normalize_states": false,
+          "state_dim": 64,
+          "state_key": "proprio",
+          "type": "NormalizeStatesAndActions"
+        },
+        {
+          "do_rescale": false,
+          "image_mean": [
+            0.48145466,
+            0.4578275,
+            0.40821073
+          ],
+          "image_std": [
+            0.26862954,
+            0.26130258,
+            0.27577711
+          ],
+          "img_key": "pixel_values",
+          "max_pixels": 1003520,
+          "merge_size": 2,
+          "min_pixels": 3136,
+          "patch_size": 16,
+          "temporal_patch_size": 2,
+          "to_tensor": true,
+          "type": "QWen2VLImageTransform"
+        },
+        {
+          "tokenizer": {
+            "model_path": "./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/",
+            "type": "PretrainedTokenizer"
+          },
+          "type": "LiberoPromptFromInputs"
+        }
+      ],
+      "type": "RobocasaEvalDataset",
+      "unnorm_key": "robocasa_gr1_24tasks_30ep"
+    },
+    "denormalize_action": {
+      "action_dim": 29,
+      "clip_actions": false,
+      "norm_type": "min_max",
+      "stats_order": "fluxvla",
+      "type": "DenormalizeRobocasaAction"
+    },
+    "enable_mixed_precision_training": true,
+    "eval_chunk_size": 10,
+    "max_episode_steps": 720,
+    "mixed_precision_dtype": "bf16",
+    "model_family": "groot",
+    "num_trials_per_task": 50,
+    "seed": 7,
+    "task_list": [
+      "gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+      "gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env"
+    ],
+    "task_suite_name": "robocasa",
+    "total_tasks": 24,
+    "type": "RobocasaEvalRunner",
+    "unnorm_key": "robocasa_gr1_24tasks_30ep"
+  },
+  "inference_model": {
+    "freeze_projector": false,
+    "freeze_vlm_backbone": false,
+    "name_mapping": null,
+    "pretrained_name_or_path": "./checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors",
+    "strict_mapping": false,
+    "type": "LlavaVLA",
+    "vla_head": {
+      "action_dim": 32,
+      "backbone_embedding_dim": 2048,
+      "diffusion_model_cfg": {
+        "attention_head_dim": 48,
+        "cross_attention_dim": 2048,
+        "dropout": 0.2,
+        "final_dropout": true,
+        "interleave_self_attention": true,
+        "norm_type": "ada_norm",
+        "num_attention_heads": 32,
+        "num_layers": 16,
+        "output_dim": 1024,
+        "positional_embeddings": null
+      },
+      "hidden_size": 1024,
+      "input_embedding_dim": 1536,
+      "num_inference_timesteps": 4,
+      "num_steps": 10,
+      "ori_action_dim": 29,
+      "state_dim": 64,
+      "type": "FlowMatchingHead",
+      "vl_self_attention_cfg": {
+        "attention_head_dim": 64,
+        "dropout": 0.2,
+        "final_dropout": true,
+        "num_attention_heads": 32,
+        "num_layers": 4,
+        "positional_embeddings": null
+      }
+    },
+    "vlm_backbone": {
+      "attn_implementation": "sdpa",
+      "projection_output_dim": 2048,
+      "projection_type": "linear",
+      "type": "Qwen3VL",
+      "use_projection": true,
+      "vlm_backbone_id": "qwen3_0.6b_vl_pt",
+      "vlm_config": {
+        "architectures": [
+          "Qwen3VLAForConditionalGeneration"
+        ],
+        "dtype": "bfloat16",
+        "eos_token_id": 151645,
+        "image_token_id": 151655,
+        "model_type": "qwen3_vl",
+        "pad_token_id": 151643,
+        "pos_skipping_range": 4096,
+        "text_config": {
+          "attention_bias": false,
+          "attention_dropout": 0.0,
+          "bos_token_id": 151643,
+          "dtype": "bfloat16",
+          "eos_token_id": 151645,
+          "head_dim": 128,
+          "hidden_act": "silu",
+          "hidden_size": 1024,
+          "initializer_range": 0.02,
+          "intermediate_size": 3072,
+          "max_position_embeddings": 262144,
+          "model_type": "qwen3_vl_text",
+          "num_attention_heads": 16,
+          "num_hidden_layers": 28,
+          "num_key_value_heads": 8,
+          "pad_token_id": null,
+          "rms_norm_eps": 1e-06,
+          "rope_parameters": {
+            "mrope_interleaved": true,
+            "mrope_section": [
+              24,
+              20,
+              20
+            ],
+            "rope_theta": 5000000,
+            "rope_type": "default"
+          },
+          "tie_word_embeddings": true,
+          "use_cache": true,
+          "vocab_size": 151936
+        },
+        "tie_word_embeddings": true,
+        "use_another_LLM_path": "",
+        "use_pos_skipping": false,
+        "vision_config": {
+          "deepstack_visual_indexes": [
+            5,
+            11,
+            17
+          ],
+          "depth": 24,
+          "dtype": "bfloat16",
+          "hidden_act": "gelu_pytorch_tanh",
+          "hidden_size": 1024,
+          "in_channels": 3,
+          "initializer_range": 0.02,
+          "intermediate_size": 4096,
+          "model_type": "qwen3_vl",
+          "num_heads": 16,
+          "num_position_embeddings": 2304,
+          "out_hidden_size": 1024,
+          "patch_size": 16,
+          "spatial_merge_size": 2,
+          "temporal_patch_size": 2
+        }
+      },
+      "vlm_path": null
+    }
+  },
+  "model": {
+    "freeze_projector": false,
+    "freeze_vlm_backbone": false,
+    "name_mapping": null,
+    "pretrained_name_or_path": "./checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors",
+    "strict_mapping": false,
+    "type": "LlavaVLA",
+    "vla_head": {
+      "action_dim": 32,
+      "backbone_embedding_dim": 2048,
+      "diffusion_model_cfg": {
+        "attention_head_dim": 48,
+        "cross_attention_dim": 2048,
+        "dropout": 0.2,
+        "final_dropout": true,
+        "interleave_self_attention": true,
+        "norm_type": "ada_norm",
+        "num_attention_heads": 32,
+        "num_layers": 16,
+        "output_dim": 1024,
+        "positional_embeddings": null
+      },
+      "hidden_size": 1024,
+      "input_embedding_dim": 1536,
+      "num_inference_timesteps": 4,
+      "num_steps": 10,
+      "ori_action_dim": 29,
+      "state_dim": 64,
+      "type": "FlowMatchingHead",
+      "vl_self_attention_cfg": {
+        "attention_head_dim": 64,
+        "dropout": 0.2,
+        "final_dropout": true,
+        "num_attention_heads": 32,
+        "num_layers": 4,
+        "positional_embeddings": null
+      }
+    },
+    "vlm_backbone": {
+      "attn_implementation": "sdpa",
+      "projection_output_dim": 2048,
+      "projection_type": "linear",
+      "type": "Qwen3VL",
+      "use_projection": true,
+      "vlm_backbone_id": "qwen3_0.6b_vl_pt",
+      "vlm_config": {
+        "architectures": [
+          "Qwen3VLAForConditionalGeneration"
+        ],
+        "dtype": "bfloat16",
+        "eos_token_id": 151645,
+        "image_token_id": 151655,
+        "model_type": "qwen3_vl",
+        "pad_token_id": 151643,
+        "pos_skipping_range": 4096,
+        "text_config": {
+          "attention_bias": false,
+          "attention_dropout": 0.0,
+          "bos_token_id": 151643,
+          "dtype": "bfloat16",
+          "eos_token_id": 151645,
+          "head_dim": 128,
+          "hidden_act": "silu",
+          "hidden_size": 1024,
+          "initializer_range": 0.02,
+          "intermediate_size": 3072,
+          "max_position_embeddings": 262144,
+          "model_type": "qwen3_vl_text",
+          "num_attention_heads": 16,
+          "num_hidden_layers": 28,
+          "num_key_value_heads": 8,
+          "pad_token_id": null,
+          "rms_norm_eps": 1e-06,
+          "rope_parameters": {
+            "mrope_interleaved": true,
+            "mrope_section": [
+              24,
+              20,
+              20
+            ],
+            "rope_theta": 5000000,
+            "rope_type": "default"
+          },
+          "tie_word_embeddings": true,
+          "use_cache": true,
+          "vocab_size": 151936
+        },
+        "tie_word_embeddings": true,
+        "use_another_LLM_path": "",
+        "use_pos_skipping": false,
+        "vision_config": {
+          "deepstack_visual_indexes": [
+            5,
+            11,
+            17
+          ],
+          "depth": 24,
+          "dtype": "bfloat16",
+          "hidden_act": "gelu_pytorch_tanh",
+          "hidden_size": 1024,
+          "in_channels": 3,
+          "initializer_range": 0.02,
+          "intermediate_size": 4096,
+          "model_type": "qwen3_vl",
+          "num_heads": 16,
+          "num_position_embeddings": 2304,
+          "out_hidden_size": 1024,
+          "patch_size": 16,
+          "spatial_merge_size": 2,
+          "temporal_patch_size": 2
+        }
+      },
+      "vlm_path": null
+    }
+  },
+  "runner": {
+    "change_key_name": false,
+    "collator": {
+      "keys": [
+        "states",
+        "observation.eepose",
+        "timestamp",
+        "images",
+        "img_masks",
+        "lang_tokens",
+        "lang_masks",
+        "actions",
+        "action_masks",
+        "embodiment_ids",
+        "image_grid_thw"
+      ],
+      "meta_keys": [
+        "task_description",
+        "prompt",
+        "info",
+        "stats"
+      ],
+      "type": "DictCollator"
+    },
+    "enable_gradient_checkpointing": false,
+    "enable_mixed_precision_training": true,
+    "grad_accumulation_steps": 1,
+    "lr_scheduler": {
+      "type": "linear-warmup+cosine-decay",
+      "warmup_ratio": 0.03
+    },
+    "max_epochs": null,
+    "max_grad_norm": 1.0,
+    "max_keep_ckpts": 8,
+    "max_steps": 100000,
+    "metric": {
+      "active_trackers": [
+        "jsonl",
+        "wandb"
+      ],
+      "grad_accumulation_steps": 1,
+      "run_dir": "work_dirs",
+      "type": "VLAMetric",
+      "window_size": 1
+    },
+    "mixed_precision_dtype": "bf16",
+    "optimizer": {
+      "lr": 1.5e-05,
+      "type": "AdamW",
+      "weight_decay": 0.0
+    },
+    "sampler": null,
+    "save_epoch_interval": 1,
+    "save_iter_interval": 5000,
+    "sharding_strategy": "full-shard",
+    "tokenizer": {
+      "model_path": "./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/",
+      "type": "PretrainedTokenizer"
+    },
+    "type": "FSDPTrainRunner"
+  },
+  "themis": {
+    "ros_server": {
+      "dataset_section": "eval",
+      "denormalize_context": {},
+      "denormalize_per_action": true,
+      "device": "cuda:0",
+      "enable_mixed_precision": true,
+      "evaluation_reporting": {
+        "report_kind": "robocasa",
+        "result_output_dir": "work_dirs/fluxthemis"
+      },
+      "forward_seed": false,
+      "mixed_precision_dtype": "bf16",
+      "model_outputs_environment_actions": false,
+      "ros_version": 1,
+      "workers": {
+        "lease_timeout_s": 900.0,
+        "request_timeout_s": 120.0,
+        "startup_timeout_s": 900.0
+      }
+    },
+    "runner": {
+      "environment": {
+        "action_order": "n15",
+        "deterministic_env": true,
+        "prompt_key": "annotation.human.coarse_action",
+        "render_key": "video.ego_view_pad_res256_freq20",
+        "task_list": [
+          "gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env",
+          "gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env"
+        ],
+        "type": "RoboCasaEnvironment"
+      },
+      "episodes_per_task": 50,
+      "evaluator": {
+        "type": "SuccessRateEvaluator"
+      },
+      "execute_horizon": 10,
+      "max_episode_steps": 720,
+      "model_client": {
+        "type": "FluxVLAROSModelClient"
+      },
+      "parallel_workers": 1,
+      "seed": 7,
+      "simulator_gpu_ids": null,
+      "stop_on_success": true,
+      "type": "EvalRunner",
+      "work_dir": "work_dirs/fluxthemis"
+    },
+    "transport": {
+      "image_encoding": "rgb8",
+      "image_keys": [
+        "video.ego_view_bg_crop_pad_res256_freq20"
+      ],
+      "report_service_name": "/fluxvla/report_evaluation",
+      "service_name": "/fluxvla/predict_action",
+      "state_keys": [
+        "state.left_arm",
+        "state.left_hand",
+        "state.right_arm",
+        "state.right_hand",
+        "state.waist"
+      ],
+      "timeout_s": 30.0,
+      "unnorm_key": "robocasa_gr1_24tasks_30ep"
+    }
+  },
+  "train_dataloader": {
+    "dataset": {
+      "dataset_statistics_path": "./datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json",
+      "datasets": {
+        "action_key": "action",
+        "action_window_size": 10,
+        "data_root_path": [
+          "./datasets/robocasa_lerobot_V2.1/PnPBottleToCabinetClose",
+          "./datasets/robocasa_lerobot_V2.1/PnPCanToDrawerClose",
+          "./datasets/robocasa_lerobot_V2.1/PnPCupToDrawerClose",
+          "./datasets/robocasa_lerobot_V2.1/PnPMilkToMicrowaveClose",
+          "./datasets/robocasa_lerobot_V2.1/PnPPotatoToMicrowaveClose",
+          "./datasets/robocasa_lerobot_V2.1/PnPWineToCabinetClose",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToBasketSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPanSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPotSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBasketSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBowlSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToPlateSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToTieredshelfSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToBowlSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToCardboardboxSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPanSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPlateSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToCardboardboxSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPlateSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPotSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredbasketSplitA",
+          "./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredshelfSplitA"
+        ],
+        "statistic_name": "robocasa_gr1_24tasks_30ep",
+        "transforms": [
+          {
+            "embodiment_id": 24,
+            "name_mappings": {
+              "actions": [
+                "actions"
+              ],
+              "observation.state": [
+                "states"
+              ]
+            },
+            "parquet_keys": [
+              "observation.state",
+              "timestamp",
+              "actions",
+              "info",
+              "stats",
+              "action_masks"
+            ],
+            "type": "ProcessParquetInputs",
+            "video_keys": [
+              "observation.images.ego_view"
+            ]
+          },
+          {
+            "type": "RobocasaGR1N15Bridge"
+          },
+          {
+            "type": "ParquetPrompter"
+          },
+          {
+            "tokenizer": {
+              "model_path": "./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/",
+              "type": "PretrainedTokenizer"
+            },
+            "type": "ProcessPrompts"
+          },
+          {
+            "scale": 0.95,
+            "type": "RandomCropImages"
+          },
+          {
+            "height": 224,
+            "type": "ResizeImages",
+            "width": 224
+          },
+          {
+            "brightness": 0.3,
+            "contrast": 0.4,
+            "hue": 0.08,
+            "saturation": 0.5,
+            "type": "ColorJitterImages"
+          },
+          {
+            "image_mean": [
+              0.48145466,
+              0.4578275,
+              0.40821073
+            ],
+            "image_std": [
+              0.26862954,
+              0.26130258,
+              0.27577711
+            ],
+            "max_pixels": 1003520,
+            "merge_size": 2,
+            "min_pixels": 3136,
+            "patch_size": 16,
+            "temporal_patch_size": 2,
+            "type": "QWen2VLImageTransform"
+          },
+          {
+            "action_dim": 32,
+            "action_key": "action",
+            "norm_type": "min_max",
+            "normalize_states": false,
+            "state_dim": 64,
+            "state_key": "proprio",
+            "type": "NormalizeStatesAndActions"
+          }
+        ],
+        "type": "ParquetDataset",
+        "use_delta": false,
+        "window_start_idx": 0
+      },
+      "name_mappings": {
+        "action": [
+          "action"
+        ],
+        "observation.state": [
+          "proprio"
+        ]
+      },
+      "statistic_keys": [
+        "observation.state",
+        "timestamp",
+        "action"
+      ],
+      "statistic_name": "robocasa_gr1_24tasks_30ep",
+      "type": "DistributedRepeatingDataset"
+    },
+    "per_device_batch_size": 8,
+    "per_device_num_workers": 4
+  }
+}

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/config.yaml

@@ -0,0 +1,614 @@
+_OFFICIAL_GR1_STATS_PATH: ./datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json
+_QWEN3VL_TOKENIZER: ./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/
+_QWEN3VL_VLA_CKPT: ./checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors
+_QWEN3VL_VLA_ROOT: ./checkpoints/gr00t_qwen3vl_0.6b_libero
+_QWEN3VL_VLM_CONFIG:
+  architectures:
+  - Qwen3VLAForConditionalGeneration
+  dtype: bfloat16
+  eos_token_id: 151645
+  image_token_id: 151655
+  model_type: qwen3_vl
+  pad_token_id: 151643
+  pos_skipping_range: 4096
+  text_config:
+    attention_bias: false
+    attention_dropout: 0.0
+    bos_token_id: 151643
+    dtype: bfloat16
+    eos_token_id: 151645
+    head_dim: 128
+    hidden_act: silu
+    hidden_size: 1024
+    initializer_range: 0.02
+    intermediate_size: 3072
+    max_position_embeddings: 262144
+    model_type: qwen3_vl_text
+    num_attention_heads: 16
+    num_hidden_layers: 28
+    num_key_value_heads: 8
+    pad_token_id: null
+    rms_norm_eps: 1.0e-06
+    rope_parameters:
+      mrope_interleaved: true
+      mrope_section:
+      - 24
+      - 20
+      - 20
+      rope_theta: 5000000
+      rope_type: default
+    tie_word_embeddings: true
+    use_cache: true
+    vocab_size: 151936
+  tie_word_embeddings: true
+  use_another_LLM_path: ''
+  use_pos_skipping: false
+  vision_config:
+    deepstack_visual_indexes:
+    - 5
+    - 11
+    - 17
+    depth: 24
+    dtype: bfloat16
+    hidden_act: gelu_pytorch_tanh
+    hidden_size: 1024
+    in_channels: 3
+    initializer_range: 0.02
+    intermediate_size: 4096
+    model_type: qwen3_vl
+    num_heads: 16
+    num_position_embeddings: 2304
+    out_hidden_size: 1024
+    patch_size: 16
+    spatial_merge_size: 2
+    temporal_patch_size: 2
+_ROBOCASA_DATA_ROOT: ./datasets/robocasa_lerobot_V2.1
+_ROBOCASA_ENV_SUFFIX: _GR1ArmsAndWaistFourierHands_Env
+_ROBOCASA_STATISTIC_NAME: robocasa_gr1_24tasks_30ep
+_ROBOCASA_TASKS:
+- PnPBottleToCabinetClose
+- PnPCanToDrawerClose
+- PnPCupToDrawerClose
+- PnPMilkToMicrowaveClose
+- PnPPotatoToMicrowaveClose
+- PnPWineToCabinetClose
+- PosttrainPnPNovelFromCuttingboardToBasketSplitA
+- PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA
+- PosttrainPnPNovelFromCuttingboardToPanSplitA
+- PosttrainPnPNovelFromCuttingboardToPotSplitA
+- PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA
+- PosttrainPnPNovelFromPlacematToBasketSplitA
+- PosttrainPnPNovelFromPlacematToBowlSplitA
+- PosttrainPnPNovelFromPlacematToPlateSplitA
+- PosttrainPnPNovelFromPlacematToTieredshelfSplitA
+- PosttrainPnPNovelFromPlateToBowlSplitA
+- PosttrainPnPNovelFromPlateToCardboardboxSplitA
+- PosttrainPnPNovelFromPlateToPanSplitA
+- PosttrainPnPNovelFromPlateToPlateSplitA
+- PosttrainPnPNovelFromTrayToCardboardboxSplitA
+- PosttrainPnPNovelFromTrayToPlateSplitA
+- PosttrainPnPNovelFromTrayToPotSplitA
+- PosttrainPnPNovelFromTrayToTieredbasketSplitA
+- PosttrainPnPNovelFromTrayToTieredshelfSplitA
+_ROBOCASA_TASK_PREFIX: gr1_unified
+eval:
+  action_order: n15
+  benchmark: robocasa
+  dataset:
+    transforms:
+    - center_crop_scale: 0.95
+      embodiment_id: 24
+      img_key: video.ego_view_bg_crop_pad_res256_freq20
+      normalize: true
+      resize_size: 224
+      type: ProcessRobocasaEvalInputs
+      value_range: unit
+    - type: RobocasaGR1N15Bridge
+    - action_key: action
+      norm_type: min_max
+      normalize_states: false
+      state_dim: 64
+      state_key: proprio
+      type: NormalizeStatesAndActions
+    - do_rescale: false
+      image_mean:
+      - 0.48145466
+      - 0.4578275
+      - 0.40821073
+      image_std:
+      - 0.26862954
+      - 0.26130258
+      - 0.27577711
+      img_key: pixel_values
+      max_pixels: 1003520
+      merge_size: 2
+      min_pixels: 3136
+      patch_size: 16
+      temporal_patch_size: 2
+      to_tensor: true
+      type: QWen2VLImageTransform
+    - tokenizer:
+        model_path: ./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/
+        type: PretrainedTokenizer
+      type: LiberoPromptFromInputs
+    type: RobocasaEvalDataset
+    unnorm_key: robocasa_gr1_24tasks_30ep
+  denormalize_action:
+    action_dim: 29
+    clip_actions: false
+    norm_type: min_max
+    stats_order: fluxvla
+    type: DenormalizeRobocasaAction
+  enable_mixed_precision_training: true
+  eval_chunk_size: 10
+  max_episode_steps: 720
+  mixed_precision_dtype: bf16
+  model_family: groot
+  num_trials_per_task: 50
+  seed: 7
+  task_list:
+  - gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env
+  - gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env
+  task_suite_name: robocasa
+  total_tasks: 24
+  type: RobocasaEvalRunner
+  unnorm_key: robocasa_gr1_24tasks_30ep
+inference_model:
+  freeze_projector: false
+  freeze_vlm_backbone: false
+  name_mapping: null
+  pretrained_name_or_path: ./checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors
+  strict_mapping: false
+  type: LlavaVLA
+  vla_head:
+    action_dim: 32
+    backbone_embedding_dim: 2048
+    diffusion_model_cfg:
+      attention_head_dim: 48
+      cross_attention_dim: 2048
+      dropout: 0.2
+      final_dropout: true
+      interleave_self_attention: true
+      norm_type: ada_norm
+      num_attention_heads: 32
+      num_layers: 16
+      output_dim: 1024
+      positional_embeddings: null
+    hidden_size: 1024
+    input_embedding_dim: 1536
+    num_inference_timesteps: 4
+    num_steps: 10
+    ori_action_dim: 29
+    state_dim: 64
+    type: FlowMatchingHead
+    vl_self_attention_cfg:
+      attention_head_dim: 64
+      dropout: 0.2
+      final_dropout: true
+      num_attention_heads: 32
+      num_layers: 4
+      positional_embeddings: null
+  vlm_backbone:
+    attn_implementation: sdpa
+    projection_output_dim: 2048
+    projection_type: linear
+    type: Qwen3VL
+    use_projection: true
+    vlm_backbone_id: qwen3_0.6b_vl_pt
+    vlm_config:
+      architectures:
+      - Qwen3VLAForConditionalGeneration
+      dtype: bfloat16
+      eos_token_id: 151645
+      image_token_id: 151655
+      model_type: qwen3_vl
+      pad_token_id: 151643
+      pos_skipping_range: 4096
+      text_config:
+        attention_bias: false
+        attention_dropout: 0.0
+        bos_token_id: 151643
+        dtype: bfloat16
+        eos_token_id: 151645
+        head_dim: 128
+        hidden_act: silu
+        hidden_size: 1024
+        initializer_range: 0.02
+        intermediate_size: 3072
+        max_position_embeddings: 262144
+        model_type: qwen3_vl_text
+        num_attention_heads: 16
+        num_hidden_layers: 28
+        num_key_value_heads: 8
+        pad_token_id: null
+        rms_norm_eps: 1.0e-06
+        rope_parameters:
+          mrope_interleaved: true
+          mrope_section:
+          - 24
+          - 20
+          - 20
+          rope_theta: 5000000
+          rope_type: default
+        tie_word_embeddings: true
+        use_cache: true
+        vocab_size: 151936
+      tie_word_embeddings: true
+      use_another_LLM_path: ''
+      use_pos_skipping: false
+      vision_config:
+        deepstack_visual_indexes:
+        - 5
+        - 11
+        - 17
+        depth: 24
+        dtype: bfloat16
+        hidden_act: gelu_pytorch_tanh
+        hidden_size: 1024
+        in_channels: 3
+        initializer_range: 0.02
+        intermediate_size: 4096
+        model_type: qwen3_vl
+        num_heads: 16
+        num_position_embeddings: 2304
+        out_hidden_size: 1024
+        patch_size: 16
+        spatial_merge_size: 2
+        temporal_patch_size: 2
+    vlm_path: null
+model:
+  freeze_projector: false
+  freeze_vlm_backbone: false
+  name_mapping: null
+  pretrained_name_or_path: ./checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors
+  strict_mapping: false
+  type: LlavaVLA
+  vla_head:
+    action_dim: 32
+    backbone_embedding_dim: 2048
+    diffusion_model_cfg:
+      attention_head_dim: 48
+      cross_attention_dim: 2048
+      dropout: 0.2
+      final_dropout: true
+      interleave_self_attention: true
+      norm_type: ada_norm
+      num_attention_heads: 32
+      num_layers: 16
+      output_dim: 1024
+      positional_embeddings: null
+    hidden_size: 1024
+    input_embedding_dim: 1536
+    num_inference_timesteps: 4
+    num_steps: 10
+    ori_action_dim: 29
+    state_dim: 64
+    type: FlowMatchingHead
+    vl_self_attention_cfg:
+      attention_head_dim: 64
+      dropout: 0.2
+      final_dropout: true
+      num_attention_heads: 32
+      num_layers: 4
+      positional_embeddings: null
+  vlm_backbone:
+    attn_implementation: sdpa
+    projection_output_dim: 2048
+    projection_type: linear
+    type: Qwen3VL
+    use_projection: true
+    vlm_backbone_id: qwen3_0.6b_vl_pt
+    vlm_config:
+      architectures:
+      - Qwen3VLAForConditionalGeneration
+      dtype: bfloat16
+      eos_token_id: 151645
+      image_token_id: 151655
+      model_type: qwen3_vl
+      pad_token_id: 151643
+      pos_skipping_range: 4096
+      text_config:
+        attention_bias: false
+        attention_dropout: 0.0
+        bos_token_id: 151643
+        dtype: bfloat16
+        eos_token_id: 151645
+        head_dim: 128
+        hidden_act: silu
+        hidden_size: 1024
+        initializer_range: 0.02
+        intermediate_size: 3072
+        max_position_embeddings: 262144
+        model_type: qwen3_vl_text
+        num_attention_heads: 16
+        num_hidden_layers: 28
+        num_key_value_heads: 8
+        pad_token_id: null
+        rms_norm_eps: 1.0e-06
+        rope_parameters:
+          mrope_interleaved: true
+          mrope_section:
+          - 24
+          - 20
+          - 20
+          rope_theta: 5000000
+          rope_type: default
+        tie_word_embeddings: true
+        use_cache: true
+        vocab_size: 151936
+      tie_word_embeddings: true
+      use_another_LLM_path: ''
+      use_pos_skipping: false
+      vision_config:
+        deepstack_visual_indexes:
+        - 5
+        - 11
+        - 17
+        depth: 24
+        dtype: bfloat16
+        hidden_act: gelu_pytorch_tanh
+        hidden_size: 1024
+        in_channels: 3
+        initializer_range: 0.02
+        intermediate_size: 4096
+        model_type: qwen3_vl
+        num_heads: 16
+        num_position_embeddings: 2304
+        out_hidden_size: 1024
+        patch_size: 16
+        spatial_merge_size: 2
+        temporal_patch_size: 2
+    vlm_path: null
+runner:
+  change_key_name: false
+  collator:
+    keys:
+    - states
+    - observation.eepose
+    - timestamp
+    - images
+    - img_masks
+    - lang_tokens
+    - lang_masks
+    - actions
+    - action_masks
+    - embodiment_ids
+    - image_grid_thw
+    meta_keys:
+    - task_description
+    - prompt
+    - info
+    - stats
+    type: DictCollator
+  enable_gradient_checkpointing: false
+  enable_mixed_precision_training: true
+  grad_accumulation_steps: 1
+  lr_scheduler:
+    type: linear-warmup+cosine-decay
+    warmup_ratio: 0.03
+  max_epochs: null
+  max_grad_norm: 1.0
+  max_keep_ckpts: 8
+  max_steps: 100000
+  metric:
+    active_trackers:
+    - jsonl
+    - wandb
+    grad_accumulation_steps: 1
+    run_dir: work_dirs
+    type: VLAMetric
+    window_size: 1
+  mixed_precision_dtype: bf16
+  optimizer:
+    lr: 1.5e-05
+    type: AdamW
+    weight_decay: 0.0
+  sampler: null
+  save_epoch_interval: 1
+  save_iter_interval: 5000
+  sharding_strategy: full-shard
+  tokenizer:
+    model_path: ./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/
+    type: PretrainedTokenizer
+  type: FSDPTrainRunner
+themis:
+  ros_server:
+    dataset_section: eval
+    denormalize_context: {}
+    denormalize_per_action: true
+    device: cuda:0
+    enable_mixed_precision: true
+    evaluation_reporting:
+      report_kind: robocasa
+      result_output_dir: work_dirs/fluxthemis
+    forward_seed: false
+    mixed_precision_dtype: bf16
+    model_outputs_environment_actions: false
+    ros_version: 1
+    workers:
+      lease_timeout_s: 900.0
+      request_timeout_s: 120.0
+      startup_timeout_s: 900.0
+  runner:
+    environment:
+      action_order: n15
+      deterministic_env: true
+      prompt_key: annotation.human.coarse_action
+      render_key: video.ego_view_pad_res256_freq20
+      task_list:
+      - gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env
+      - gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env
+      type: RoboCasaEnvironment
+    episodes_per_task: 50
+    evaluator:
+      type: SuccessRateEvaluator
+    execute_horizon: 10
+    max_episode_steps: 720
+    model_client:
+      type: FluxVLAROSModelClient
+    parallel_workers: 1
+    seed: 7
+    simulator_gpu_ids: null
+    stop_on_success: true
+    type: EvalRunner
+    work_dir: work_dirs/fluxthemis
+  transport:
+    image_encoding: rgb8
+    image_keys:
+    - video.ego_view_bg_crop_pad_res256_freq20
+    report_service_name: /fluxvla/report_evaluation
+    service_name: /fluxvla/predict_action
+    state_keys:
+    - state.left_arm
+    - state.left_hand
+    - state.right_arm
+    - state.right_hand
+    - state.waist
+    timeout_s: 30.0
+    unnorm_key: robocasa_gr1_24tasks_30ep
+train_dataloader:
+  dataset:
+    dataset_statistics_path: ./datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json
+    datasets:
+      action_key: action
+      action_window_size: 10
+      data_root_path:
+      - ./datasets/robocasa_lerobot_V2.1/PnPBottleToCabinetClose
+      - ./datasets/robocasa_lerobot_V2.1/PnPCanToDrawerClose
+      - ./datasets/robocasa_lerobot_V2.1/PnPCupToDrawerClose
+      - ./datasets/robocasa_lerobot_V2.1/PnPMilkToMicrowaveClose
+      - ./datasets/robocasa_lerobot_V2.1/PnPPotatoToMicrowaveClose
+      - ./datasets/robocasa_lerobot_V2.1/PnPWineToCabinetClose
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToBasketSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPanSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPotSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBasketSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBowlSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToPlateSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToTieredshelfSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToBowlSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToCardboardboxSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPanSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPlateSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToCardboardboxSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPlateSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPotSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredbasketSplitA
+      - ./datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredshelfSplitA
+      statistic_name: robocasa_gr1_24tasks_30ep
+      transforms:
+      - embodiment_id: 24
+        name_mappings:
+          actions:
+          - actions
+          observation.state:
+          - states
+        parquet_keys:
+        - observation.state
+        - timestamp
+        - actions
+        - info
+        - stats
+        - action_masks
+        type: ProcessParquetInputs
+        video_keys:
+        - observation.images.ego_view
+      - type: RobocasaGR1N15Bridge
+      - type: ParquetPrompter
+      - tokenizer:
+          model_path: ./checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/
+          type: PretrainedTokenizer
+        type: ProcessPrompts
+      - scale: 0.95
+        type: RandomCropImages
+      - height: 224
+        type: ResizeImages
+        width: 224
+      - brightness: 0.3
+        contrast: 0.4
+        hue: 0.08
+        saturation: 0.5
+        type: ColorJitterImages
+      - image_mean:
+        - 0.48145466
+        - 0.4578275
+        - 0.40821073
+        image_std:
+        - 0.26862954
+        - 0.26130258
+        - 0.27577711
+        max_pixels: 1003520
+        merge_size: 2
+        min_pixels: 3136
+        patch_size: 16
+        temporal_patch_size: 2
+        type: QWen2VLImageTransform
+      - action_dim: 32
+        action_key: action
+        norm_type: min_max
+        normalize_states: false
+        state_dim: 64
+        state_key: proprio
+        type: NormalizeStatesAndActions
+      type: ParquetDataset
+      use_delta: false
+      window_start_idx: 0
+    name_mappings:
+      action:
+      - action
+      observation.state:
+      - proprio
+    statistic_keys:
+    - observation.state
+    - timestamp
+    - action
+    statistic_name: robocasa_gr1_24tasks_30ep
+    type: DistributedRepeatingDataset
+  per_device_batch_size: 8
+  per_device_num_workers: 4

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/dataset_statistics.json

@@ -0,0 +1,396 @@
+{
+  "robocasa_gr1_24tasks_30ep": {
+    "proprio": {
+      "mean": [
+        -0.1721954196691514,
+        0.18329788744449615,
+        -0.1360361576080323,
+        -1.0807313919067383,
+        0.14450088143348713,
+        0.13195440173149092,
+        -0.08306947350502007,
+        0.21940384805202476,
+        0.21819256246089933,
+        0.20426052808761605,
+        0.22833923995494854,
+        0.06841608136892328,
+        0.9091449975967411,
+        -0.1590277999639513,
+        -0.2511773109436038,
+        0.08599518239498137,
+        -1.1420482397079474,
+        0.21259814500808735,
+        0.04237671568989762,
+        0.03427042067050932,
+        0.34993353486061096,
+        0.3412855267524723,
+        0.3271332383155817,
+        0.3575766980648041,
+        0.09723818302154554,
+        1.0641368627548227,
+        0.014510802924633035,
+        0.014398206025362044,
+        -0.00020241182937752458
+      ],
+      "std": [
+        0.4068214595317843,
+        0.18259239196777355,
+        0.2513278424739838,
+        0.5786943435668955,
+        0.33050763607025135,
+        0.28609949350357067,
+        0.3013432323932649,
+        0.42784574627876293,
+        0.4130409955978396,
+        0.3710329830646513,
+        0.4229725301265714,
+        0.18603332340717335,
+        0.7297042012214684,
+        0.5631562471389767,
+        0.24450603127479503,
+        0.3282549381256103,
+        0.7310536503791792,
+        0.48139148950576843,
+        0.35611921548843384,
+        0.5090531110763546,
+        0.49147421121597257,
+        0.46545258164405845,
+        0.44516301155090443,
+        0.4866269826889039,
+        0.1814946085214615,
+        0.5735033154487607,
+        0.1227026358246804,
+        0.03244836255908008,
+        0.007282953243702651
+      ],
+      "min": [
+        -1.7593636512756348,
+        -0.08899051696062088,
+        -1.6571800708770752,
+        -2.516343355178833,
+        -2.307133436203003,
+        -1.4132264852523804,
+        -1.4408576488494873,
+        -3.262805700302124,
+        -2.7771592140197754,
+        -3.4712820053100586,
+        -2.7828404903411865,
+        -2.197380781173706,
+        -2.636387825012207,
+        -2.2743337154388428,
+        -2.999122142791748,
+        -1.7424250841140747,
+        -2.536545753479004,
+        -2.6034910678863525,
+        -1.5032434463500977,
+        -1.4913568496704102,
+        -1.7595213651657104,
+        -1.8982670307159424,
+        -1.8622381687164307,
+        -1.8314869403839111,
+        -1.0045679807662964,
+        -0.31411829590797424,
+        -0.8131351470947266,
+        -0.4901888370513916,
+        -0.7304351925849915
+      ],
+      "max": [
+        1.3570959568023682,
+        1.5781080722808838,
+        1.7425614595413208,
+        0.04977008327841759,
+        2.521491289138794,
+        1.5104753971099854,
+        1.3451268672943115,
+        2.8765757083892822,
+        2.6975045204162598,
+        3.3291828632354736,
+        2.7148845195770264,
+        3.664421558380127,
+        3.5466625690460205,
+        1.7399436235427856,
+        0.07876992970705032,
+        2.5426626205444336,
+        0.048908114433288574,
+        3.000918388366699,
+        1.4513440132141113,
+        1.5000383853912354,
+        2.397510290145874,
+        2.4422738552093506,
+        3.0158462524414062,
+        2.457378625869751,
+        1.4605257511138916,
+        2.168534278869629,
+        1.0323854684829712,
+        0.7102982997894287,
+        0.43762317299842834
+      ],
+      "q01": [
+        -1.6011584222316746,
+        -0.00048318122862838164,
+        -0.9488833963871,
+        -2.4362342357635502,
+        -0.9403592050075524,
+        -0.6489510446786882,
+        -0.9641011685132993,
+        -0.17130047902464876,
+        -0.0025680272374302166,
+        -0.0029985445667989555,
+        -0.002490740863140675,
+        -0.14894675761461273,
+        -0.00027310136792948466,
+        -1.4233640372753156,
+        -1.0436977982521036,
+        -0.7169023394584662,
+        -2.4936094284057644,
+        -0.7377233326435084,
+        -0.8036172896623616,
+        -0.9991738587617871,
+        -0.004218762856908145,
+        -0.004575904295779761,
+        -0.005576773989014321,
+        -0.0047070790780708175,
+        -0.18967307358980165,
+        0.03200087323784828,
+        -0.3365072563290599,
+        -0.02904987111687659,
+        -0.02606033049523827
+      ],
+      "q99": [
+        0.6305126994848259,
+        0.7966297268867486,
+        0.3873166218399999,
+        -0.07399308793246723,
+        1.0455405175685892,
+        0.8574009954929348,
+        0.4838866069912913,
+        1.5014824867248553,
+        1.501653671264647,
+        1.4043268024921445,
+        1.5127789616584784,
+        0.781709930300714,
+        1.872669064998624,
+        1.167767405509949,
+        0.0006303858070168673,
+        0.8771779745817198,
+        -0.0582109544426202,
+        1.470731711387636,
+        0.8474899172782918,
+        1.2271583080291764,
+        1.4994217157363876,
+        1.5029035806655904,
+        1.6614551365375507,
+        1.5188306391239186,
+        0.6953653633594535,
+        1.8182492077350652,
+        0.48663658648729347,
+        0.12612193301320107,
+        0.022280203178524964
+      ]
+    },
+    "timestamp": {
+      "mean": [
+        0.0
+      ],
+      "std": [
+        1.0
+      ],
+      "min": [
+        0.0
+      ],
+      "max": [
+        0.0
+      ],
+      "q01": null,
+      "q99": null
+    },
+    "action": {
+      "mean": [
+        -0.17771673202514657,
+        0.18351773917675007,
+        -0.13720993697643286,
+        -1.0842658281326296,
+        0.14675945043563873,
+        0.12921456992626204,
+        -0.09504522383213035,
+        -0.22721463441848752,
+        -0.21956320106983204,
+        -0.2144497036933901,
+        -0.21446840465068817,
+        -0.48631641268730147,
+        1.6838233470916744,
+        -0.16953834891319267,
+        -0.2535267770290373,
+        0.082216002047062,
+        -1.1492931842803968,
+        0.2176141738891601,
+        0.04310995712876314,
+        0.04133718833327294,
+        -0.543570637702943,
+        -0.5304322242736823,
+        -0.5179098844528184,
+        -0.4968528747558592,
+        -1.1343770027160642,
+        2.128209590911863,
+        0.014836843125522158,
+        0.010428135283291354,
+        -0.00014937532250769447
+      ],
+      "std": [
+        0.4174487590789793,
+        0.18478454649448395,
+        0.25687918066978443,
+        0.5807053446769717,
+        0.33657354116439797,
+        0.29842707514762895,
+        0.31536528468132013,
+        1.0050891637802128,
+        1.0124161243438718,
+        1.0138051509857173,
+        1.0200439691543575,
+        2.066762685775758,
+        1.4588384628295925,
+        0.5752111673355106,
+        0.25846862792968733,
+        0.34115546941757197,
+        0.7394758462905866,
+        0.4870710670948032,
+        0.3721199631690978,
+        0.5423213839530941,
+        1.1248024702072141,
+        1.1392076015472405,
+        1.1426663398742682,
+        1.301862597465516,
+        2.199875354766848,
+        0.9186902642250119,
+        0.12568794190883642,
+        0.030695030465722098,
+        0.004574332851916548
+      ],
+      "min": [
+        -2.291214942932129,
+        -0.001746351015754044,
+        -2.4999990463256836,
+        -2.814260482788086,
+        -2.3611011505126953,
+        -1.4937989711761475,
+        -1.4969758987426758,
+        -1.9597030878067017,
+        -1.8635213375091553,
+        -1.9709523916244507,
+        -1.5,
+        -3.0,
+        0.0,
+        -2.318650960922241,
+        -2.9999561309814453,
+        -1.9257696866989136,
+        -2.8759055137634277,
+        -2.620600461959839,
+        -1.5001521110534668,
+        -1.4994292259216309,
+        -1.5,
+        -1.5,
+        -1.5,
+        -1.5,
+        -3.0,
+        4.470348358154297e-08,
+        -0.8292319774627686,
+        -0.5185094475746155,
+        -0.37811079621315
+      ],
+      "max": [
+        1.4113223552703857,
+        1.8901419639587402,
+        1.7754145860671997,
+        7.309383363462985e-05,
+        2.563594102859497,
+        1.5000243186950684,
+        1.4908421039581299,
+        1.5707963705062866,
+        1.646651268005371,
+        1.709236979484558,
+        2.0620639324188232,
+        3.0,
+        3.0,
+        1.7834906578063965,
+        0.0002448999439366162,
+        2.549729585647583,
+        7.45560391806066e-05,
+        3.0000460147857666,
+        1.4975632429122925,
+        1.4998434782028198,
+        1.5707963705062866,
+        1.7183797359466553,
+        3.979951858520508,
+        3.9879753589630127,
+        3.0,
+        3.0,
+        1.049119234085083,
+        0.6198405623435974,
+        0.45177245140075684
+      ],
+      "q01": [
+        -1.6055770933628097,
+        -7.429541994952163e-07,
+        -0.9837155520915981,
+        -2.480029511451721,
+        -0.9457575023174288,
+        -0.7192997813224794,
+        -1.008323037624359,
+        -1.4999999999999987,
+        -1.4999999999999987,
+        -1.4999999999999987,
+        -1.4999999999999987,
+        -2.9999999999999973,
+        0.0,
+        -1.4375487387180332,
+        -1.0682010412216194,
+        -0.7535711079835895,
+        -2.6383612155914276,
+        -0.7471688866615306,
+        -0.8604200631380092,
+        -1.0900398135185252,
+        -1.4999999999999987,
+        -1.4999999999999987,
+        -1.4999999999999987,
+        -1.4999999999999987,
+        -2.9999999999999973,
+        0.04336242545396091,
+        -0.3402548208832736,
+        -0.025130789913237084,
+        -0.016220059804618367
+      ],
+      "q99": [
+        0.6335921049118061,
+        0.8018405169248585,
+        0.3930345878005033,
+        -0.06366106234490865,
+        1.0508334636688252,
+        0.8918659299612054,
+        0.5138001590967192,
+        1.4999999999999987,
+        1.4999999999999987,
+        1.4999999999999987,
+        1.4999999999999987,
+        2.9999999999999973,
+        2.9999999999999973,
+        1.1753968834876984,
+        -9.999999974752431e-07,
+        0.8976036489009875,
+        -0.0579521823674439,
+        1.4876463234424584,
+        0.8774339169263857,
+        1.4641386926174185,
+        1.4999999999999987,
+        1.4999999999999987,
+        1.4999999999999987,
+        1.4999999999999987,
+        2.9999999999999973,
+        2.9999999999999973,
+        0.49598969072103627,
+        0.12137954644858844,
+        0.010726323025301106
+      ]
+    }
+  }
+}

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_2026_08_28_19_35_25.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2e94237da3bc2353d1df240f5fe0977894c9d5eb19c730963845f9b1fa9cea6b
+size 28365587

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/run-metrics.jsonl

@@ -0,0 +1 @@
+{"hparams": "{'_QWEN3VL_VLA_ROOT': './checkpoints/gr00t_qwen3vl_0.6b_libero', '_QWEN3VL_VLA_CKPT': './checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors', '_QWEN3VL_TOKENIZER': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/', '_QWEN3VL_VLM_CONFIG': {'architectures': ['Qwen3VLAForConditionalGeneration'], 'dtype': 'bfloat16', 'eos_token_id': 151645, 'image_token_id': 151655, 'model_type': 'qwen3_vl', 'pad_token_id': 151643, 'pos_skipping_range': 4096, 'text_config': {'attention_bias': False, 'attention_dropout': 0.0, 'bos_token_id': 151643, 'dtype': 'bfloat16', 'eos_token_id': 151645, 'head_dim': 128, 'hidden_act': 'silu', 'hidden_size': 1024, 'initializer_range': 0.02, 'intermediate_size': 3072, 'max_position_embeddings': 262144, 'model_type': 'qwen3_vl_text', 'num_attention_heads': 16, 'num_hidden_layers': 28, 'num_key_value_heads': 8, 'pad_token_id': None, 'rms_norm_eps': 1e-06, 'rope_parameters': {'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_theta': 5000000, 'rope_type': 'default'}, 'tie_word_embeddings': True, 'use_cache': True, 'vocab_size': 151936}, 'tie_word_embeddings': True, 'use_another_LLM_path': '', 'use_pos_skipping': False, 'vision_config': {'deepstack_visual_indexes': [5, 11, 17], 'depth': 24, 'dtype': 'bfloat16', 'hidden_act': 'gelu_pytorch_tanh', 'hidden_size': 1024, 'in_channels': 3, 'initializer_range': 0.02, 'intermediate_size': 4096, 'model_type': 'qwen3_vl', 'num_heads': 16, 'num_position_embeddings': 2304, 'out_hidden_size': 1024, 'patch_size': 16, 'spatial_merge_size': 2, 'temporal_patch_size': 2}}, 'model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors', 'name_mapping': None, 'strict_mapping': False, 'vlm_backbone': {'type': 'Qwen3VL', 'vlm_backbone_id': 'qwen3_0.6b_vl_pt', 'vlm_path': None, 'vlm_config': {'architectures': ['Qwen3VLAForConditionalGeneration'], 'dtype': 'bfloat16', 'eos_token_id': 151645, 'image_token_id': 151655, 'model_type': 'qwen3_vl', 'pad_token_id': 151643, 'pos_skipping_range': 4096, 'text_config': {'attention_bias': False, 'attention_dropout': 0.0, 'bos_token_id': 151643, 'dtype': 'bfloat16', 'eos_token_id': 151645, 'head_dim': 128, 'hidden_act': 'silu', 'hidden_size': 1024, 'initializer_range': 0.02, 'intermediate_size': 3072, 'max_position_embeddings': 262144, 'model_type': 'qwen3_vl_text', 'num_attention_heads': 16, 'num_hidden_layers': 28, 'num_key_value_heads': 8, 'pad_token_id': None, 'rms_norm_eps': 1e-06, 'rope_parameters': {'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_theta': 5000000, 'rope_type': 'default'}, 'tie_word_embeddings': True, 'use_cache': True, 'vocab_size': 151936}, 'tie_word_embeddings': True, 'use_another_LLM_path': '', 'use_pos_skipping': False, 'vision_config': {'deepstack_visual_indexes': [5, 11, 17], 'depth': 24, 'dtype': 'bfloat16', 'hidden_act': 'gelu_pytorch_tanh', 'hidden_size': 1024, 'in_channels': 3, 'initializer_range': 0.02, 'intermediate_size': 4096, 'model_type': 'qwen3_vl', 'num_heads': 16, 'num_position_embeddings': 2304, 'out_hidden_size': 1024, 'patch_size': 16, 'spatial_merge_size': 2, 'temporal_patch_size': 2}}, 'use_projection': True, 'projection_output_dim': 2048, 'projection_type': 'linear', 'attn_implementation': 'sdpa'}, 'vla_head': {'type': 'FlowMatchingHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'backbone_embedding_dim': 2048, 'vl_self_attention_cfg': {'attention_head_dim': 64, 'num_attention_heads': 32, 'num_layers': 4, 'dropout': 0.2, 'final_dropout': True, 'positional_embeddings': None}, 'diffusion_model_cfg': {'attention_head_dim': 48, 'num_attention_heads': 32, 'cross_attention_dim': 2048, 'num_layers': 16, 'output_dim': 1024, 'dropout': 0.2, 'final_dropout': True, 'interleave_self_attention': True, 'norm_type': 'ada_norm', 'positional_embeddings': None}, 'num_inference_timesteps': 4, 'num_steps': 10, 'action_dim': 32, 'ori_action_dim': 29}, 'freeze_vlm_backbone': False, 'freeze_projector': False}, 'inference_model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors', 'name_mapping': None, 'strict_mapping': False, 'vlm_backbone': {'type': 'Qwen3VL', 'vlm_backbone_id': 'qwen3_0.6b_vl_pt', 'vlm_path': None, 'vlm_config': {'architectures': ['Qwen3VLAForConditionalGeneration'], 'dtype': 'bfloat16', 'eos_token_id': 151645, 'image_token_id': 151655, 'model_type': 'qwen3_vl', 'pad_token_id': 151643, 'pos_skipping_range': 4096, 'text_config': {'attention_bias': False, 'attention_dropout': 0.0, 'bos_token_id': 151643, 'dtype': 'bfloat16', 'eos_token_id': 151645, 'head_dim': 128, 'hidden_act': 'silu', 'hidden_size': 1024, 'initializer_range': 0.02, 'intermediate_size': 3072, 'max_position_embeddings': 262144, 'model_type': 'qwen3_vl_text', 'num_attention_heads': 16, 'num_hidden_layers': 28, 'num_key_value_heads': 8, 'pad_token_id': None, 'rms_norm_eps': 1e-06, 'rope_parameters': {'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_theta': 5000000, 'rope_type': 'default'}, 'tie_word_embeddings': True, 'use_cache': True, 'vocab_size': 151936}, 'tie_word_embeddings': True, 'use_another_LLM_path': '', 'use_pos_skipping': False, 'vision_config': {'deepstack_visual_indexes': [5, 11, 17], 'depth': 24, 'dtype': 'bfloat16', 'hidden_act': 'gelu_pytorch_tanh', 'hidden_size': 1024, 'in_channels': 3, 'initializer_range': 0.02, 'intermediate_size': 4096, 'model_type': 'qwen3_vl', 'num_heads': 16, 'num_position_embeddings': 2304, 'out_hidden_size': 1024, 'patch_size': 16, 'spatial_merge_size': 2, 'temporal_patch_size': 2}}, 'use_projection': True, 'projection_output_dim': 2048, 'projection_type': 'linear', 'attn_implementation': 'sdpa'}, 'vla_head': {'type': 'FlowMatchingHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'backbone_embedding_dim': 2048, 'vl_self_attention_cfg': {'attention_head_dim': 64, 'num_attention_heads': 32, 'num_layers': 4, 'dropout': 0.2, 'final_dropout': True, 'positional_embeddings': None}, 'diffusion_model_cfg': {'attention_head_dim': 48, 'num_attention_heads': 32, 'cross_attention_dim': 2048, 'num_layers': 16, 'output_dim': 1024, 'dropout': 0.2, 'final_dropout': True, 'interleave_self_attention': True, 'norm_type': 'ada_norm', 'positional_embeddings': None}, 'num_inference_timesteps': 4, 'num_steps': 10, 'action_dim': 32, 'ori_action_dim': 29}, 'freeze_vlm_backbone': False, 'freeze_projector': False}, '_ROBOCASA_STATISTIC_NAME': 'robocasa_gr1_24tasks_30ep', '_ROBOCASA_DATA_ROOT': './datasets/robocasa_lerobot_V2.1', '_OFFICIAL_GR1_STATS_PATH': './datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json', '_ROBOCASA_TASK_PREFIX': 'gr1_unified', '_ROBOCASA_ENV_SUFFIX': '_GR1ArmsAndWaistFourierHands_Env', '_ROBOCASA_TASKS': ['PnPBottleToCabinetClose', 'PnPCanToDrawerClose', 'PnPCupToDrawerClose', 'PnPMilkToMicrowaveClose', 'PnPPotatoToMicrowaveClose', 'PnPWineToCabinetClose', 'PosttrainPnPNovelFromCuttingboardToBasketSplitA', 'PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA', 'PosttrainPnPNovelFromCuttingboardToPanSplitA', 'PosttrainPnPNovelFromCuttingboardToPotSplitA', 'PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA', 'PosttrainPnPNovelFromPlacematToBasketSplitA', 'PosttrainPnPNovelFromPlacematToBowlSplitA', 'PosttrainPnPNovelFromPlacematToPlateSplitA', 'PosttrainPnPNovelFromPlacematToTieredshelfSplitA', 'PosttrainPnPNovelFromPlateToBowlSplitA', 'PosttrainPnPNovelFromPlateToCardboardboxSplitA', 'PosttrainPnPNovelFromPlateToPanSplitA', 'PosttrainPnPNovelFromPlateToPlateSplitA', 'PosttrainPnPNovelFromTrayToCardboardboxSplitA', 'PosttrainPnPNovelFromTrayToPlateSplitA', 'PosttrainPnPNovelFromTrayToPotSplitA', 'PosttrainPnPNovelFromTrayToTieredbasketSplitA', 'PosttrainPnPNovelFromTrayToTieredshelfSplitA'], 'train_dataloader': {'per_device_batch_size': 8, 'per_device_num_workers': 4, 'dataset': {'type': 'DistributedRepeatingDataset', 'name_mappings': {'observation.state': ['proprio'], 'action': ['action']}, 'statistic_keys': ['observation.state', 'timestamp', 'action'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'dataset_statistics_path': './datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json', 'datasets': {'type': 'ParquetDataset', 'data_root_path': ['./datasets/robocasa_lerobot_V2.1/PnPBottleToCabinetClose', './datasets/robocasa_lerobot_V2.1/PnPCanToDrawerClose', './datasets/robocasa_lerobot_V2.1/PnPCupToDrawerClose', './datasets/robocasa_lerobot_V2.1/PnPMilkToMicrowaveClose', './datasets/robocasa_lerobot_V2.1/PnPPotatoToMicrowaveClose', './datasets/robocasa_lerobot_V2.1/PnPWineToCabinetClose', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToBasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPanSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPotSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBowlSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToPlateSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToTieredshelfSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToBowlSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToCardboardboxSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPanSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPlateSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToCardboardboxSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPlateSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPotSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredbasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredshelfSplitA'], 'transforms': [{'type': 'ProcessParquetInputs', 'embodiment_id': 24, 'parquet_keys': ['observation.state', 'timestamp', 'actions', 'info', 'stats', 'action_masks'], 'video_keys': ['observation.images.ego_view'], 'name_mappings': {'observation.state': ['states'], 'actions': ['actions']}}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'ParquetPrompter'}, {'type': 'ProcessPrompts', 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/'}}, {'type': 'RandomCropImages', 'scale': 0.95}, {'type': 'ResizeImages', 'height': 224, 'width': 224}, {'type': 'ColorJitterImages', 'brightness': 0.3, 'contrast': 0.4, 'saturation': 0.5, 'hue': 0.08}, {'type': 'QWen2VLImageTransform', 'min_pixels': 3136, 'max_pixels': 1003520, 'patch_size': 16, 'temporal_patch_size': 2, 'merge_size': 2, 'image_mean': [0.48145466, 0.4578275, 0.40821073], 'image_std': [0.26862954, 0.26130258, 0.27577711]}, {'type': 'NormalizeStatesAndActions', 'action_dim': 32, 'state_dim': 64, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}], 'action_window_size': 10, 'action_key': 'action', 'use_delta': False, 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'window_start_idx': 0}}}, 'runner': {'type': 'FSDPTrainRunner', 'max_epochs': None, 'max_steps': 100000, 'grad_accumulation_steps': 1, 'optimizer': {'lr': 1.5e-05, 'type': 'AdamW', 'weight_decay': 0.0}, 'max_grad_norm': 1.0, 'save_iter_interval': 5000, 'save_epoch_interval': 1, 'max_keep_ckpts': 8, 'sampler': None, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/'}, 'collator': {'type': 'DictCollator', 'keys': ['states', 'observation.eepose', 'timestamp', 'images', 'img_masks', 'lang_tokens', 'lang_masks', 'actions', 'action_masks', 'embodiment_ids', 'image_grid_thw'], 'meta_keys': ['task_description', 'prompt', 'info', 'stats']}, 'metric': {'type': 'VLAMetric', 'active_trackers': ('jsonl', 'wandb'), 'run_dir': './work_dirs/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_75e72754e_bs128', 'grad_accumulation_steps': 1, 'window_size': 1}, 'lr_scheduler': {'type': 'linear-warmup+cosine-decay', 'warmup_ratio': 0.03}, 'sharding_strategy': 'full-shard', 'enable_gradient_checkpointing': False, 'enable_mixed_precision_training': True, 'mixed_precision_dtype': 'bf16', 'change_key_name': False, 'cfg': Config (path: configs/gr00t/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune.py): {'_QWEN3VL_VLA_ROOT': './checkpoints/gr00t_qwen3vl_0.6b_libero', '_QWEN3VL_VLA_CKPT': './checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors', '_QWEN3VL_TOKENIZER': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/', '_QWEN3VL_VLM_CONFIG': {'architectures': ['Qwen3VLAForConditionalGeneration'], 'dtype': 'bfloat16', 'eos_token_id': 151645, 'image_token_id': 151655, 'model_type': 'qwen3_vl', 'pad_token_id': 151643, 'pos_skipping_range': 4096, 'text_config': {'attention_bias': False, 'attention_dropout': 0.0, 'bos_token_id': 151643, 'dtype': 'bfloat16', 'eos_token_id': 151645, 'head_dim': 128, 'hidden_act': 'silu', 'hidden_size': 1024, 'initializer_range': 0.02, 'intermediate_size': 3072, 'max_position_embeddings': 262144, 'model_type': 'qwen3_vl_text', 'num_attention_heads': 16, 'num_hidden_layers': 28, 'num_key_value_heads': 8, 'pad_token_id': None, 'rms_norm_eps': 1e-06, 'rope_parameters': {'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_theta': 5000000, 'rope_type': 'default'}, 'tie_word_embeddings': True, 'use_cache': True, 'vocab_size': 151936}, 'tie_word_embeddings': True, 'use_another_LLM_path': '', 'use_pos_skipping': False, 'vision_config': {'deepstack_visual_indexes': [5, 11, 17], 'depth': 24, 'dtype': 'bfloat16', 'hidden_act': 'gelu_pytorch_tanh', 'hidden_size': 1024, 'in_channels': 3, 'initializer_range': 0.02, 'intermediate_size': 4096, 'model_type': 'qwen3_vl', 'num_heads': 16, 'num_position_embeddings': 2304, 'out_hidden_size': 1024, 'patch_size': 16, 'spatial_merge_size': 2, 'temporal_patch_size': 2}}, 'model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors', 'name_mapping': None, 'strict_mapping': False, 'vlm_backbone': {'type': 'Qwen3VL', 'vlm_backbone_id': 'qwen3_0.6b_vl_pt', 'vlm_path': None, 'vlm_config': {'architectures': ['Qwen3VLAForConditionalGeneration'], 'dtype': 'bfloat16', 'eos_token_id': 151645, 'image_token_id': 151655, 'model_type': 'qwen3_vl', 'pad_token_id': 151643, 'pos_skipping_range': 4096, 'text_config': {'attention_bias': False, 'attention_dropout': 0.0, 'bos_token_id': 151643, 'dtype': 'bfloat16', 'eos_token_id': 151645, 'head_dim': 128, 'hidden_act': 'silu', 'hidden_size': 1024, 'initializer_range': 0.02, 'intermediate_size': 3072, 'max_position_embeddings': 262144, 'model_type': 'qwen3_vl_text', 'num_attention_heads': 16, 'num_hidden_layers': 28, 'num_key_value_heads': 8, 'pad_token_id': None, 'rms_norm_eps': 1e-06, 'rope_parameters': {'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_theta': 5000000, 'rope_type': 'default'}, 'tie_word_embeddings': True, 'use_cache': True, 'vocab_size': 151936}, 'tie_word_embeddings': True, 'use_another_LLM_path': '', 'use_pos_skipping': False, 'vision_config': {'deepstack_visual_indexes': [5, 11, 17], 'depth': 24, 'dtype': 'bfloat16', 'hidden_act': 'gelu_pytorch_tanh', 'hidden_size': 1024, 'in_channels': 3, 'initializer_range': 0.02, 'intermediate_size': 4096, 'model_type': 'qwen3_vl', 'num_heads': 16, 'num_position_embeddings': 2304, 'out_hidden_size': 1024, 'patch_size': 16, 'spatial_merge_size': 2, 'temporal_patch_size': 2}}, 'use_projection': True, 'projection_output_dim': 2048, 'projection_type': 'linear', 'attn_implementation': 'sdpa'}, 'vla_head': {'type': 'FlowMatchingHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'backbone_embedding_dim': 2048, 'vl_self_attention_cfg': {'attention_head_dim': 64, 'num_attention_heads': 32, 'num_layers': 4, 'dropout': 0.2, 'final_dropout': True, 'positional_embeddings': None}, 'diffusion_model_cfg': {'attention_head_dim': 48, 'num_attention_heads': 32, 'cross_attention_dim': 2048, 'num_layers': 16, 'output_dim': 1024, 'dropout': 0.2, 'final_dropout': True, 'interleave_self_attention': True, 'norm_type': 'ada_norm', 'positional_embeddings': None}, 'num_inference_timesteps': 4, 'num_steps': 10, 'action_dim': 32, 'ori_action_dim': 29}, 'freeze_vlm_backbone': False, 'freeze_projector': False}, 'inference_model': {'type': 'LlavaVLA', 'pretrained_name_or_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/checkpoints/step-104160-epoch-24-loss=0.0358.safetensors', 'name_mapping': None, 'strict_mapping': False, 'vlm_backbone': {'type': 'Qwen3VL', 'vlm_backbone_id': 'qwen3_0.6b_vl_pt', 'vlm_path': None, 'vlm_config': {'architectures': ['Qwen3VLAForConditionalGeneration'], 'dtype': 'bfloat16', 'eos_token_id': 151645, 'image_token_id': 151655, 'model_type': 'qwen3_vl', 'pad_token_id': 151643, 'pos_skipping_range': 4096, 'text_config': {'attention_bias': False, 'attention_dropout': 0.0, 'bos_token_id': 151643, 'dtype': 'bfloat16', 'eos_token_id': 151645, 'head_dim': 128, 'hidden_act': 'silu', 'hidden_size': 1024, 'initializer_range': 0.02, 'intermediate_size': 3072, 'max_position_embeddings': 262144, 'model_type': 'qwen3_vl_text', 'num_attention_heads': 16, 'num_hidden_layers': 28, 'num_key_value_heads': 8, 'pad_token_id': None, 'rms_norm_eps': 1e-06, 'rope_parameters': {'mrope_interleaved': True, 'mrope_section': [24, 20, 20], 'rope_theta': 5000000, 'rope_type': 'default'}, 'tie_word_embeddings': True, 'use_cache': True, 'vocab_size': 151936}, 'tie_word_embeddings': True, 'use_another_LLM_path': '', 'use_pos_skipping': False, 'vision_config': {'deepstack_visual_indexes': [5, 11, 17], 'depth': 24, 'dtype': 'bfloat16', 'hidden_act': 'gelu_pytorch_tanh', 'hidden_size': 1024, 'in_channels': 3, 'initializer_range': 0.02, 'intermediate_size': 4096, 'model_type': 'qwen3_vl', 'num_heads': 16, 'num_position_embeddings': 2304, 'out_hidden_size': 1024, 'patch_size': 16, 'spatial_merge_size': 2, 'temporal_patch_size': 2}}, 'use_projection': True, 'projection_output_dim': 2048, 'projection_type': 'linear', 'attn_implementation': 'sdpa'}, 'vla_head': {'type': 'FlowMatchingHead', 'state_dim': 64, 'hidden_size': 1024, 'input_embedding_dim': 1536, 'backbone_embedding_dim': 2048, 'vl_self_attention_cfg': {'attention_head_dim': 64, 'num_attention_heads': 32, 'num_layers': 4, 'dropout': 0.2, 'final_dropout': True, 'positional_embeddings': None}, 'diffusion_model_cfg': {'attention_head_dim': 48, 'num_attention_heads': 32, 'cross_attention_dim': 2048, 'num_layers': 16, 'output_dim': 1024, 'dropout': 0.2, 'final_dropout': True, 'interleave_self_attention': True, 'norm_type': 'ada_norm', 'positional_embeddings': None}, 'num_inference_timesteps': 4, 'num_steps': 10, 'action_dim': 32, 'ori_action_dim': 29}, 'freeze_vlm_backbone': False, 'freeze_projector': False}, '_ROBOCASA_STATISTIC_NAME': 'robocasa_gr1_24tasks_30ep', '_ROBOCASA_DATA_ROOT': './datasets/robocasa_lerobot_V2.1', '_OFFICIAL_GR1_STATS_PATH': './datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json', '_ROBOCASA_TASK_PREFIX': 'gr1_unified', '_ROBOCASA_ENV_SUFFIX': '_GR1ArmsAndWaistFourierHands_Env', '_ROBOCASA_TASKS': ['PnPBottleToCabinetClose', 'PnPCanToDrawerClose', 'PnPCupToDrawerClose', 'PnPMilkToMicrowaveClose', 'PnPPotatoToMicrowaveClose', 'PnPWineToCabinetClose', 'PosttrainPnPNovelFromCuttingboardToBasketSplitA', 'PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA', 'PosttrainPnPNovelFromCuttingboardToPanSplitA', 'PosttrainPnPNovelFromCuttingboardToPotSplitA', 'PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA', 'PosttrainPnPNovelFromPlacematToBasketSplitA', 'PosttrainPnPNovelFromPlacematToBowlSplitA', 'PosttrainPnPNovelFromPlacematToPlateSplitA', 'PosttrainPnPNovelFromPlacematToTieredshelfSplitA', 'PosttrainPnPNovelFromPlateToBowlSplitA', 'PosttrainPnPNovelFromPlateToCardboardboxSplitA', 'PosttrainPnPNovelFromPlateToPanSplitA', 'PosttrainPnPNovelFromPlateToPlateSplitA', 'PosttrainPnPNovelFromTrayToCardboardboxSplitA', 'PosttrainPnPNovelFromTrayToPlateSplitA', 'PosttrainPnPNovelFromTrayToPotSplitA', 'PosttrainPnPNovelFromTrayToTieredbasketSplitA', 'PosttrainPnPNovelFromTrayToTieredshelfSplitA'], 'train_dataloader': {'per_device_batch_size': 8, 'per_device_num_workers': 4, 'dataset': {'type': 'DistributedRepeatingDataset', 'name_mappings': {'observation.state': ['proprio'], 'action': ['action']}, 'statistic_keys': ['observation.state', 'timestamp', 'action'], 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'dataset_statistics_path': './datasets/robocasa_gr1_24tasks_first30ep/official_groot_gr1_dataset_statistics.json', 'datasets': {'type': 'ParquetDataset', 'data_root_path': ['./datasets/robocasa_lerobot_V2.1/PnPBottleToCabinetClose', './datasets/robocasa_lerobot_V2.1/PnPCanToDrawerClose', './datasets/robocasa_lerobot_V2.1/PnPCupToDrawerClose', './datasets/robocasa_lerobot_V2.1/PnPMilkToMicrowaveClose', './datasets/robocasa_lerobot_V2.1/PnPPotatoToMicrowaveClose', './datasets/robocasa_lerobot_V2.1/PnPWineToCabinetClose', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToBasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPanSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToPotSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToBowlSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToPlateSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlacematToTieredshelfSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToBowlSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToCardboardboxSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPanSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromPlateToPlateSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToCardboardboxSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPlateSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToPotSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredbasketSplitA', './datasets/robocasa_lerobot_V2.1/PosttrainPnPNovelFromTrayToTieredshelfSplitA'], 'transforms': [{'type': 'ProcessParquetInputs', 'embodiment_id': 24, 'parquet_keys': ['observation.state', 'timestamp', 'actions', 'info', 'stats', 'action_masks'], 'video_keys': ['observation.images.ego_view'], 'name_mappings': {'observation.state': ['states'], 'actions': ['actions']}}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'ParquetPrompter'}, {'type': 'ProcessPrompts', 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/'}}, {'type': 'RandomCropImages', 'scale': 0.95}, {'type': 'ResizeImages', 'height': 224, 'width': 224}, {'type': 'ColorJitterImages', 'brightness': 0.3, 'contrast': 0.4, 'saturation': 0.5, 'hue': 0.08}, {'type': 'QWen2VLImageTransform', 'min_pixels': 3136, 'max_pixels': 1003520, 'patch_size': 16, 'temporal_patch_size': 2, 'merge_size': 2, 'image_mean': [0.48145466, 0.4578275, 0.40821073], 'image_std': [0.26862954, 0.26130258, 0.27577711]}, {'type': 'NormalizeStatesAndActions', 'action_dim': 32, 'state_dim': 64, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}], 'action_window_size': 10, 'action_key': 'action', 'use_delta': False, 'statistic_name': 'robocasa_gr1_24tasks_30ep', 'window_start_idx': 0}}}, 'runner': {'type': 'FSDPTrainRunner', 'max_epochs': None, 'max_steps': 100000, 'grad_accumulation_steps': 1, 'optimizer': {'lr': 1.5e-05, 'type': 'AdamW', 'weight_decay': 0.0}, 'max_grad_norm': 1.0, 'save_iter_interval': 5000, 'save_epoch_interval': 1, 'max_keep_ckpts': 8, 'sampler': None, 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/'}, 'collator': {'type': 'DictCollator', 'keys': ['states', 'observation.eepose', 'timestamp', 'images', 'img_masks', 'lang_tokens', 'lang_masks', 'actions', 'action_masks', 'embodiment_ids', 'image_grid_thw'], 'meta_keys': ['task_description', 'prompt', 'info', 'stats']}, 'metric': {'type': 'VLAMetric', 'active_trackers': ('jsonl', 'wandb'), 'run_dir': './work_dirs/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_75e72754e_bs128', 'grad_accumulation_steps': 1, 'window_size': 1}, 'lr_scheduler': {'type': 'linear-warmup+cosine-decay', 'warmup_ratio': 0.03}, 'sharding_strategy': 'full-shard', 'enable_gradient_checkpointing': False, 'enable_mixed_precision_training': True, 'mixed_precision_dtype': 'bf16', 'change_key_name': False, 'cfg': Config (path: configs/gr00t/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune.py): {...}, 'args': Namespace(config='configs/gr00t/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune.py', work_dir='./work_dirs/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_75e72754e_bs128', cfg_options=None, eval_after_train=True, resume_from=None)}, 'eval': {'type': 'RobocasaEvalRunner', 'benchmark': 'robocasa', 'task_suite_name': 'robocasa', 'model_family': 'groot', 'task_list': ['gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env'], 'total_tasks': 24, 'eval_chunk_size': 10, 'max_episode_steps': 720, 'num_trials_per_task': 50, 'seed': 7, 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'action_order': 'n15', 'enable_mixed_precision_training': True, 'mixed_precision_dtype': 'bf16', 'dataset': {'type': 'RobocasaEvalDataset', 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'transforms': [{'type': 'ProcessRobocasaEvalInputs', 'img_key': 'video.ego_view_bg_crop_pad_res256_freq20', 'resize_size': 224, 'center_crop_scale': 0.95, 'normalize': True, 'value_range': 'unit', 'embodiment_id': 24}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}, {'type': 'QWen2VLImageTransform', 'do_rescale': False, 'min_pixels': 3136, 'max_pixels': 1003520, 'patch_size': 16, 'temporal_patch_size': 2, 'merge_size': 2, 'image_mean': [0.48145466, 0.4578275, 0.40821073], 'image_std': [0.26862954, 0.26130258, 0.27577711], 'img_key': 'pixel_values', 'to_tensor': True}, {'type': 'LiberoPromptFromInputs', 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/'}}]}, 'denormalize_action': {'type': 'DenormalizeRobocasaAction', 'norm_type': 'min_max', 'action_dim': 29, 'clip_actions': False, 'stats_order': 'fluxvla'}}, 'themis': {'transport': {'service_name': '/fluxvla/predict_action', 'report_service_name': '/fluxvla/report_evaluation', 'timeout_s': 30.0, 'image_keys': ['video.ego_view_bg_crop_pad_res256_freq20'], 'state_keys': ['state.left_arm', 'state.left_hand', 'state.right_arm', 'state.right_hand', 'state.waist'], 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'image_encoding': 'rgb8'}, 'runner': {'type': 'EvalRunner', 'environment': {'type': 'RoboCasaEnvironment', 'task_list': ['gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env'], 'action_order': 'n15', 'deterministic_env': True, 'prompt_key': 'annotation.human.coarse_action', 'render_key': 'video.ego_view_pad_res256_freq20'}, 'model_client': {'type': 'FluxVLAROSModelClient'}, 'evaluator': {'type': 'SuccessRateEvaluator'}, 'seed': 7, 'episodes_per_task': 50, 'max_episode_steps': 720, 'execute_horizon': 10, 'stop_on_success': True, 'parallel_workers': 1, 'simulator_gpu_ids': None, 'work_dir': 'work_dirs/fluxthemis'}, 'ros_server': {'ros_version': 1, 'dataset_section': 'eval', 'evaluation_reporting': {'result_output_dir': 'work_dirs/fluxthemis', 'report_kind': 'robocasa'}, 'device': 'cuda:0', 'workers': {'startup_timeout_s': 900.0, 'request_timeout_s': 120.0, 'lease_timeout_s': 900.0}, 'mixed_precision_dtype': 'bf16', 'enable_mixed_precision': True, 'model_outputs_environment_actions': False, 'forward_seed': False, 'denormalize_context': {}, 'denormalize_per_action': True}}}, 'args': Namespace(config='configs/gr00t/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune.py', work_dir='./work_dirs/gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_75e72754e_bs128', cfg_options=None, eval_after_train=True, resume_from=None)}, 'eval': {'type': 'RobocasaEvalRunner', 'benchmark': 'robocasa', 'task_suite_name': 'robocasa', 'model_family': 'groot', 'task_list': ['gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env'], 'total_tasks': 24, 'eval_chunk_size': 10, 'max_episode_steps': 720, 'num_trials_per_task': 50, 'seed': 7, 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'action_order': 'n15', 'enable_mixed_precision_training': True, 'mixed_precision_dtype': 'bf16', 'dataset': {'type': 'RobocasaEvalDataset', 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'transforms': [{'type': 'ProcessRobocasaEvalInputs', 'img_key': 'video.ego_view_bg_crop_pad_res256_freq20', 'resize_size': 224, 'center_crop_scale': 0.95, 'normalize': True, 'value_range': 'unit', 'embodiment_id': 24}, {'type': 'RobocasaGR1N15Bridge'}, {'type': 'NormalizeStatesAndActions', 'state_dim': 64, 'state_key': 'proprio', 'action_key': 'action', 'norm_type': 'min_max', 'normalize_states': False}, {'type': 'QWen2VLImageTransform', 'do_rescale': False, 'min_pixels': 3136, 'max_pixels': 1003520, 'patch_size': 16, 'temporal_patch_size': 2, 'merge_size': 2, 'image_mean': [0.48145466, 0.4578275, 0.40821073], 'image_std': [0.26862954, 0.26130258, 0.27577711], 'img_key': 'pixel_values', 'to_tensor': True}, {'type': 'LiberoPromptFromInputs', 'tokenizer': {'type': 'PretrainedTokenizer', 'model_path': './checkpoints/gr00t_qwen3vl_0.6b_libero/tokenizer/'}}]}, 'denormalize_action': {'type': 'DenormalizeRobocasaAction', 'norm_type': 'min_max', 'action_dim': 29, 'clip_actions': False, 'stats_order': 'fluxvla'}}, 'themis': {'transport': {'service_name': '/fluxvla/predict_action', 'report_service_name': '/fluxvla/report_evaluation', 'timeout_s': 30.0, 'image_keys': ['video.ego_view_bg_crop_pad_res256_freq20'], 'state_keys': ['state.left_arm', 'state.left_hand', 'state.right_arm', 'state.right_hand', 'state.waist'], 'unnorm_key': 'robocasa_gr1_24tasks_30ep', 'image_encoding': 'rgb8'}, 'runner': {'type': 'EvalRunner', 'environment': {'type': 'RoboCasaEnvironment', 'task_list': ['gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCanToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPCupToDrawerClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPMilkToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPPotatoToMicrowaveClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PnPWineToCabinetClose_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromCuttingboardToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlacematToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPanSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromPlateToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPlateSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToPotSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredbasketSplitA_GR1ArmsAndWaistFourierHands_Env', 'gr1_unified/PosttrainPnPNovelFromTrayToTieredshelfSplitA_GR1ArmsAndWaistFourierHands_Env'], 'action_order': 'n15', 'deterministic_env': True, 'prompt_key': 'annotation.human.coarse_action', 'render_key': 'video.ego_view_pad_res256_freq20'}, 'model_client': {'type': 'FluxVLAROSModelClient'}, 'evaluator': {'type': 'SuccessRateEvaluator'}, 'seed': 7, 'episodes_per_task': 50, 'max_episode_steps': 720, 'execute_horizon': 10, 'stop_on_success': True, 'parallel_workers': 1, 'simulator_gpu_ids': None, 'work_dir': 'work_dirs/fluxthemis'}, 'ros_server': {'ros_version': 1, 'dataset_section': 'eval', 'evaluation_reporting': {'result_output_dir': 'work_dirs/fluxthemis', 'report_kind': 'robocasa'}, 'device': 'cuda:0', 'workers': {'startup_timeout_s': 900.0, 'request_timeout_s': 120.0, 'lease_timeout_s': 900.0}, 'mixed_precision_dtype': 'bf16', 'enable_mixed_precision': True, 'model_outputs_environment_actions': False, 'forward_seed': False, 'denormalize_context': {}, 'denormalize_per_action': True}}}", "run_id": "gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_2026_08_28_19_35_25"}

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/tokenizer/added_tokens.json

@@ -0,0 +1,48 @@
+{
+  "</think>": 151668,
+  "</tool_call>": 151658,
+  "</tool_response>": 151666,
+  "<action_end>": 151675,
+  "<action_pad>": 151673,
+  "<action_start>": 151674,
+  "<action_video>": 151669,
+  "<box2d_end>": 151679,
+  "<box2d_start>": 151678,
+  "<future_video_pad>": 151670,
+  "<future_vision_end>": 151672,
+  "<future_vision_start>": 151671,
+  "<ignore_pad>": 151688,
+  "<point2d_end>": 151681,
+  "<point2d_start>": 151680,
+  "<ref_end>": 151683,
+  "<ref_keypoint_end>": 151685,
+  "<ref_keypoint_start>": 151684,
+  "<ref_start>": 151682,
+  "<think>": 151667,
+  "<think_end>": 151677,
+  "<think_start>": 151676,
+  "<tool_call>": 151657,
+  "<tool_response>": 151665,
+  "<traj2d_end>": 151687,
+  "<traj2d_start>": 151686,
+  "<|box_end|>": 151649,
+  "<|box_start|>": 151648,
+  "<|endoftext|>": 151643,
+  "<|file_sep|>": 151664,
+  "<|fim_middle|>": 151660,
+  "<|fim_pad|>": 151662,
+  "<|fim_prefix|>": 151659,
+  "<|fim_suffix|>": 151661,
+  "<|im_end|>": 151645,
+  "<|im_start|>": 151644,
+  "<|image_pad|>": 151655,
+  "<|object_ref_end|>": 151647,
+  "<|object_ref_start|>": 151646,
+  "<|quad_end|>": 151651,
+  "<|quad_start|>": 151650,
+  "<|repo_name|>": 151663,
+  "<|video_pad|>": 151656,
+  "<|vision_end|>": 151653,
+  "<|vision_pad|>": 151654,
+  "<|vision_start|>": 151652
+}

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/tokenizer/special_tokens_map.json

@@ -0,0 +1,31 @@
+{
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "eos_token": {
+    "content": "<|im_end|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/tokenizer/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b7a88bc82340dd4205b97b9c287df826cf386b31ac9ecd9e648073d940355e1
+size 11426476

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/tokenizer/tokenizer_config.json

@@ -0,0 +1,401 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151665": {
+      "content": "<tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151666": {
+      "content": "</tool_response>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151667": {
+      "content": "<think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151668": {
+      "content": "</think>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151669": {
+      "content": "<action_video>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151670": {
+      "content": "<future_video_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151671": {
+      "content": "<future_vision_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151672": {
+      "content": "<future_vision_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151673": {
+      "content": "<action_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151674": {
+      "content": "<action_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151675": {
+      "content": "<action_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151676": {
+      "content": "<think_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151677": {
+      "content": "<think_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151678": {
+      "content": "<box2d_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151679": {
+      "content": "<box2d_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151680": {
+      "content": "<point2d_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151681": {
+      "content": "<point2d_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151682": {
+      "content": "<ref_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151683": {
+      "content": "<ref_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151684": {
+      "content": "<ref_keypoint_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151685": {
+      "content": "<ref_keypoint_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151686": {
+      "content": "<traj2d_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151687": {
+      "content": "<traj2d_end>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151688": {
+      "content": "<ignore_pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "processor_class": "Qwen3VLAProcessor",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

gr00t_qwen3vl_0.6b_robocasa_full_data_full_finetune_bs128/vlm_backbone_config.json

@@ -0,0 +1,74 @@
+{
+  "architectures": [
+    "Qwen3VLAForConditionalGeneration"
+  ],
+  "attn_implementation": "sdpa",
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "image_token_id": 151655,
+  "model_type": "qwen3_vl",
+  "pad_token_id": 151643,
+  "pos_skipping_range": 4096,
+  "text_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "attn_implementation": "sdpa",
+    "bos_token_id": 151643,
+    "dtype": "bfloat16",
+    "eos_token_id": 151645,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 1024,
+    "initializer_range": 0.02,
+    "intermediate_size": 3072,
+    "max_position_embeddings": 262144,
+    "model_type": "qwen3_vl_text",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 28,
+    "num_key_value_heads": 8,
+    "pad_token_id": null,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "mrope_interleaved": true,
+      "mrope_section": [
+        24,
+        20,
+        20
+      ],
+      "rope_theta": 5000000,
+      "rope_type": "default"
+    },
+    "tie_word_embeddings": true,
+    "use_cache": true,
+    "vocab_size": 151936
+  },
+  "tie_word_embeddings": true,
+  "transformers_version": "5.3.0",
+  "use_another_LLM_path": "",
+  "use_pos_skipping": false,
+  "video_token_id": 151656,
+  "vision_config": {
+    "attn_implementation": "sdpa",
+    "deepstack_visual_indexes": [
+      5,
+      11,
+      17
+    ],
+    "depth": 24,
+    "dtype": "bfloat16",
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1024,
+    "in_channels": 3,
+    "initializer_range": 0.02,
+    "intermediate_size": 4096,
+    "model_type": "qwen3_vl",
+    "num_heads": 16,
+    "num_position_embeddings": 2304,
+    "out_hidden_size": 1024,
+    "patch_size": 16,
+    "spatial_merge_size": 2,
+    "temporal_patch_size": 2
+  },
+  "vision_end_token_id": 151653,
+  "vision_start_token_id": 151652
+}