Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1.jsonl filter=lfs diff=lfs merge=lfs -text

checkpoints/step-039000-epoch-192-loss=0.0263.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:536f25f289267d93fa1782f8003645dbaadef084a572aa3a7a34c4a3bd837373
+size 11322827655

config.json

@@ -0,0 +1,59 @@
+{
+  "action_dim": 7,
+  "action_model_type": "DiT-B",
+  "data_root_dir": "/mnt/inspurfs/efm_t/robot_data/cache/LIBERO/dataset",
+  "debug": false,
+  "disable_instruction": false,
+  "fix_system1": false,
+  "future_action_window_size": 7,
+  "hf_token": "hf_token",
+  "image_aug": true,
+  "is_resume": true,
+  "load_all_data_for_training": true,
+  "num_of_meta_query": 64,
+  "past_action_window_size": 0,
+  "pretrained_checkpoint": "/mnt/petrelfs/yangshuai1/rep/InstructVLA_official/outputs/libero_wrist/sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1/checkpoints/step-025500-epoch-125-loss=0.0168.pt",
+  "repeated_diffusion_steps": 4,
+  "resume_epoch": 125,
+  "resume_step": 25500,
+  "run_id": "sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1",
+  "run_id_note": null,
+  "run_root_dir": "outputs/libero_wrist",
+  "save_interval": 1500,
+  "seed": 42,
+  "stage": "stage1",
+  "trackers": [
+    "jsonl",
+    "wandb"
+  ],
+  "use_ema": false,
+  "use_mm": false,
+  "vla": {
+    "action_tokenizer": "extra_action_tokenizer",
+    "base_vlm": "/mnt/petrelfs/yangshuai1/yangshuai1/share_mllm/Eagle2-2B",
+    "data_mix": "libero_goal_no_noops",
+    "enable_gradient_checkpointing": true,
+    "enable_mixed_precision_training": true,
+    "epochs": 100,
+    "expected_world_size": 8,
+    "freeze_llm_backbone": false,
+    "freeze_vision_backbone": false,
+    "global_batch_size": 256,
+    "learning_rate": 5e-05,
+    "lr_scheduler_type": "constant",
+    "max_grad_norm": 1.0,
+    "max_steps": null,
+    "per_device_batch_size": 32,
+    "reduce_in_full_precision": true,
+    "shuffle_buffer_size": 250000,
+    "train_strategy": "fsdp-full-shard",
+    "type": "prism-qwen25-dinosiglip-224px+0_5b",
+    "unfreeze_last_llm_layer": false,
+    "vla_id": "prism-qwen25-dinosiglip-224px+0_5b",
+    "warmup_ratio": 0.0,
+    "weight_decay": 0.0
+  },
+  "wandb_entity": "shuaiyang2003",
+  "wandb_project": "dual_sys_libero",
+  "with_pointing": true
+}

config.yaml

@@ -0,0 +1,55 @@
+action_dim: 7
+action_model_type: DiT-B
+data_root_dir: /mnt/inspurfs/efm_t/robot_data/cache/LIBERO/dataset
+debug: false
+disable_instruction: false
+fix_system1: false
+future_action_window_size: 7
+hf_token: hf_token
+image_aug: true
+is_resume: true
+load_all_data_for_training: true
+num_of_meta_query: 64
+past_action_window_size: 0
+pretrained_checkpoint: /mnt/petrelfs/yangshuai1/rep/InstructVLA_official/outputs/libero_wrist/sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1/checkpoints/step-025500-epoch-125-loss=0.0168.pt
+repeated_diffusion_steps: 4
+resume_epoch: 125
+resume_step: 25500
+run_id: sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1
+run_id_note: null
+run_root_dir: outputs/libero_wrist
+save_interval: 1500
+seed: 42
+stage: stage1
+trackers:
+- jsonl
+- wandb
+use_ema: false
+use_mm: false
+vla:
+  action_tokenizer: extra_action_tokenizer
+  base_vlm: /mnt/petrelfs/yangshuai1/yangshuai1/share_mllm/Eagle2-2B
+  data_mix: libero_goal_no_noops
+  enable_gradient_checkpointing: true
+  enable_mixed_precision_training: true
+  epochs: 100
+  expected_world_size: 8
+  freeze_llm_backbone: false
+  freeze_vision_backbone: false
+  global_batch_size: 256
+  learning_rate: 5.0e-05
+  lr_scheduler_type: constant
+  max_grad_norm: 1.0
+  max_steps: null
+  per_device_batch_size: 32
+  reduce_in_full_precision: true
+  shuffle_buffer_size: 250000
+  train_strategy: fsdp-full-shard
+  type: prism-qwen25-dinosiglip-224px+0_5b
+  unfreeze_last_llm_layer: false
+  vla_id: prism-qwen25-dinosiglip-224px+0_5b
+  warmup_ratio: 0.0
+  weight_decay: 0.0
+wandb_entity: shuaiyang2003
+wandb_project: dual_sys_libero
+with_pointing: true

dataset_statistics.json

@@ -0,0 +1,133 @@
+{
+  "libero_goal_no_noops": {
+    "action": {
+      "mean": [
+        0.04721052572131157,
+        0.028835246339440346,
+        -0.1485840231180191,
+        -0.0025010062381625175,
+        0.026408178731799126,
+        0.027379808947443962,
+        0.6299911737442017
+      ],
+      "std": [
+        0.3968801498413086,
+        0.3473387360572815,
+        0.49239858984947205,
+        0.055331431329250336,
+        0.07844757288694382,
+        0.10008802264928818,
+        0.48270025849342346
+      ],
+      "max": [
+        0.9375,
+        0.9375,
+        0.9375,
+        0.3557142913341522,
+        0.375,
+        0.375,
+        1.0
+      ],
+      "min": [
+        -0.9375,
+        -0.9375,
+        -0.9375,
+        -0.2582142949104309,
+        -0.375,
+        -0.2871428430080414,
+        0.0
+      ],
+      "q01": [
+        -0.8785714507102966,
+        -0.7553571462631226,
+        -0.9375,
+        -0.1510714292526245,
+        -0.1639285683631897,
+        -0.13777500048279764,
+        0.0
+      ],
+      "q99": [
+        0.9375,
+        0.9107142686843872,
+        0.9375,
+        0.20357142388820648,
+        0.26357144117355347,
+        0.375,
+        1.0
+      ],
+      "mask": [
+        true,
+        true,
+        true,
+        true,
+        true,
+        true,
+        false
+      ]
+    },
+    "proprio": {
+      "mean": [
+        -0.09923473745584488,
+        0.013597904704511166,
+        1.0694637298583984,
+        2.82898211479187,
+        0.30799180269241333,
+        -0.274286687374115,
+        0.028092455118894577,
+        -0.027339335530996323
+      ],
+      "std": [
+        0.11653962731361389,
+        0.11478105187416077,
+        0.10487838834524155,
+        0.5570293664932251,
+        0.7221656441688538,
+        0.36479514837265015,
+        0.01507475133985281,
+        0.014990941621363163
+      ],
+      "max": [
+        0.13579000532627106,
+        0.33316105604171753,
+        1.3660105466842651,
+        3.473310708999634,
+        2.6688623428344727,
+        0.8255361318588257,
+        0.04233968257904053,
+        0.0010111660230904818
+      ],
+      "min": [
+        -0.46141114830970764,
+        -0.30129560828208923,
+        0.9083037972450256,
+        0.35277295112609863,
+        -1.4858465194702148,
+        -1.5227035284042358,
+        -0.0013586411951109767,
+        -0.042040832340717316
+      ],
+      "q01": [
+        -0.42401049643754957,
+        -0.27338370531797407,
+        0.911226047873497,
+        1.3085840785503386,
+        -0.691297555565834,
+        -1.130668159723282,
+        0.0016738151130266487,
+        -0.040336399003863335
+      ],
+      "q99": [
+        0.08990443304181095,
+        0.26473945528268716,
+        1.2910678112506866,
+        3.2425890421867365,
+        2.3376442337036116,
+        0.4659483411908149,
+        0.040610933862626555,
+        -0.0015016929572448147
+      ]
+    },
+    "num_transitions": 52042,
+    "num_trajectories": 428
+  }
+}

run-metrics.jsonl

@@ -0,0 +1 @@
+{"hparams": {"action_dim": 7, "action_model_type": "DiT-B", "data_root_dir": "/mnt/inspurfs/efm_t/robot_data/cache/LIBERO/dataset", "debug": false, "disable_instruction": false, "fix_system1": false, "future_action_window_size": 7, "hf_token": "hf_token", "image_aug": true, "is_resume": true, "load_all_data_for_training": true, "num_of_meta_query": 64, "past_action_window_size": 0, "pretrained_checkpoint": "/mnt/petrelfs/yangshuai1/rep/InstructVLA_official/outputs/libero_wrist/sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1/checkpoints/step-025500-epoch-125-loss=0.0168.pt", "repeated_diffusion_steps": 4, "resume_epoch": 125, "resume_step": 25500, "run_id": "sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1", "run_id_note": null, "run_root_dir": "outputs/libero_wrist", "save_interval": 1500, "seed": 42, "stage": "stage1", "trackers": ["jsonl", "wandb"], "use_ema": false, "use_mm": false, "vla": {"action_tokenizer": "extra_action_tokenizer", "base_vlm": "/mnt/petrelfs/yangshuai1/yangshuai1/share_mllm/Eagle2-2B", "data_mix": "libero_goal_no_noops", "enable_gradient_checkpointing": true, "enable_mixed_precision_training": true, "epochs": 100, "expected_world_size": 8, "freeze_llm_backbone": false, "freeze_vision_backbone": false, "global_batch_size": 256, "learning_rate": 5e-05, "lr_scheduler_type": "constant", "max_grad_norm": 1.0, "max_steps": null, "per_device_batch_size": 32, "reduce_in_full_precision": true, "shuffle_buffer_size": 250000, "train_strategy": "fsdp-full-shard", "type": "prism-qwen25-dinosiglip-224px+0_5b", "unfreeze_last_llm_layer": false, "vla_id": "prism-qwen25-dinosiglip-224px+0_5b", "warmup_ratio": 0.0, "weight_decay": 0.0}, "wandb_entity": "shuaiyang2003", "wandb_project": "dual_sys_libero", "with_pointing": true}, "run_id": "sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1"}

sys12_meta_query_action_only_sync_pretraining_v2_query_64_mlp_lora_libero_goal_wrist--image_augstage1.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:86e070e1d1e7a6e80810bb4f3358cfca015c05018810a2408d0f52f9c39b3513
+size 14214150