Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,10 @@ 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
+wandb/offline-run-20250718_171556-dn5hay4y/run-dn5hay4y.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/offline-run-20250718_222432-whocfndg/run-whocfndg.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/offline-run-20250718_224110-4dzmg2gq/run-4dzmg2gq.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/offline-run-20250718_232119-f4k6r99g/run-f4k6r99g.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/offline-run-20250719_072257-lazdcdl9/run-lazdcdl9.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/offline-run-20250719_151414-28h8u1mw/run-28h8u1mw.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/offline-run-20250721_145208-zs4rtx9g/run-zs4rtx9g.wandb filter=lfs diff=lfs merge=lfs -text

eval_logs/EVAL-libero_spatial-a1-2025_07_20-23_03_13.txt

@@ -0,0 +1 @@
+Task suite: libero_spatial

eval_logs/EVAL-libero_spatial-a1-2025_07_20-23_04_29.txt

@@ -0,0 +1,2 @@
+Task suite: libero_spatial
+Using default initial states

eval_logs/EVAL-libero_spatial-a1-2025_07_20-23_09_11.txt

@@ -0,0 +1,2 @@
+Task suite: libero_spatial
+Using default initial states

eval_logs/EVAL-libero_spatial-a1-2025_07_20-23_11_35.txt

@@ -0,0 +1,2 @@
+Task suite: libero_spatial
+Using default initial states

eval_logs/EVAL-libero_spatial-a1-2025_07_20-23_19_00.txt

@@ -0,0 +1,2 @@
+Task suite: libero_spatial
+Using default initial states

eval_logs/EVAL-libero_spatial-a1-2025_07_20-23_28_38.txt

@@ -0,0 +1,2 @@
+Task suite: libero_spatial
+Using default initial states

eval_logs/EVAL-libero_spatial-a1-2025_07_22-11_08_22.txt

@@ -0,0 +1,2 @@
+Task suite: libero_spatial
+Using default initial states

step_130000/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4ddfa081721faeb08de99dea9b1685a6baec710edcafc8d410c7541d931bb29
+size 7228114647

wandb/debug-internal.log

@@ -0,0 +1,8 @@
+{"time":"2025-07-21T14:52:08.78129657+08:00","level":"INFO","msg":"stream: starting","core version":"0.21.0"}
+{"time":"2025-07-21T14:52:08.895408501+08:00","level":"WARN","msg":"GraphQL client is nil, skipping feature loading"}
+{"time":"2025-07-21T14:52:08.895540832+08:00","level":"INFO","msg":"stream: created new stream","id":"zs4rtx9g"}
+{"time":"2025-07-21T14:52:08.895557067+08:00","level":"INFO","msg":"stream: started","id":"zs4rtx9g"}
+{"time":"2025-07-21T14:52:08.89557567+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"zs4rtx9g"}
+{"time":"2025-07-21T14:52:08.895589857+08:00","level":"INFO","msg":"handler: started","stream_id":"zs4rtx9g"}
+{"time":"2025-07-21T14:52:08.895591465+08:00","level":"INFO","msg":"sender: started","stream_id":"zs4rtx9g"}
+{"time":"2025-07-21T14:52:08.896087544+08:00","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}

wandb/offline-run-20250718_161945-m8wsmtst/files/config.yaml

@@ -0,0 +1,227 @@
+wandb_version: 1
+
+_wandb:
+  desc: null
+  value:
+    python_version: 3.10.16
+    cli_version: 0.21.0
+    framework: huggingface
+    huggingface_version: 4.53.2
+    is_jupyter_run: false
+    is_kaggle_kernel: false
+    start_time: 1752826786
+    t:
+      1:
+      - 1
+      - 2
+      - 3
+      - 5
+      - 11
+      - 41
+      - 49
+      - 51
+      - 53
+      - 63
+      - 71
+      - 83
+      - 95
+      - 105
+      2:
+      - 1
+      - 2
+      - 3
+      - 5
+      - 11
+      - 41
+      - 49
+      - 51
+      - 53
+      - 63
+      - 71
+      - 83
+      - 95
+      - 105
+      3:
+      - 4
+      - 13
+      - 16
+      - 42
+      4: 3.10.16
+      5: 0.21.0
+      6: 4.53.2
+      13: linux-x86_64
+    e:
+      gkv6pcimbk2w6ig5eqnc01n6qgq9ed3l:
+        os: Linux-5.15.0-142-generic-x86_64-with-glibc2.35
+        python: CPython 3.10.16
+        started_at: '2025-07-18T08:19:45.826177Z'
+        args:
+        - --save_folder=/mnt/data/zhangjian/a1/libero_spatial_dit_action
+        - --save_overwrite
+        program: /home/zhangjian/A1/launch_scripts/train_vla_dit_action.py
+        code_path: launch_scripts/train_vla_dit_action.py
+        code_path_local: launch_scripts/train_vla_dit_action.py
+        git:
+          remote_url: https://github.com/Spatialtemporal-AI/A1.git
+          commit: 2e05e206652258f5fbb178ec13c792c2a91c1856
+        root: /home/zhangjian/A1
+        host: iv-ydwxzrkf0gvr6okuck21
+        executable: /mnt/data/zhangjian/miniconda3/envs/a1/bin/python3.10
+        cpu_count: 56
+        cpu_count_logical: 112
+        gpu_type: NVIDIA A100-SXM4-80GB
+        gpu_count: 8
+        disk:
+          /:
+            total: '507024936960'
+            used: '133182132224'
+        memory:
+          total: '2071488372736'
+        gpu_nvidia:
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-b01b938f-1b95-8587-4b93-cfa68d84c935
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-49b283f0-f1e6-992d-4ed8-99855a5e6529
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-a340dca6-1db9-f578-76f2-cf36f74ffabd
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-c1bddabf-aa88-9c8f-eec2-6dc3e117f985
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-4096c1e7-0437-6727-096f-d67141c140b8
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-eeb896c0-61d6-8380-778b-d02101078a17
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-d5c1ed79-76e3-22a6-8908-df289b090c03
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-6d90b060-f6f7-599b-f939-4ba124ff11b7
+        cuda_version: '12.8'
+        writer_id: gkv6pcimbk2w6ig5eqnc01n6qgq9ed3l
+text_model_name:
+  desc: null
+  value: google/siglip-so400m-patch14-384
+vision_model_name:
+  desc: null
+  value: google/siglip-so400m-patch14-384
+run_name:
+  desc: null
+  value: dit_action_train_20250718_161918
+save_folder:
+  desc: null
+  value: /mnt/data/zhangjian/a1/libero_spatial_dit_action
+seed:
+  desc: null
+  value: 42
+max_steps:
+  desc: null
+  value: 100000
+global_batch_size:
+  desc: null
+  value: 128
+device_batch_size:
+  desc: null
+  value: 32
+learning_rate:
+  desc: null
+  value: 2.0e-05
+weight_decay:
+  desc: null
+  value: 0.01
+warmup_steps:
+  desc: null
+  value: 500
+dataset_name:
+  desc: null
+  value: libero_spatial_no_noops
+data_root_dir:
+  desc: null
+  value: /mnt/data/zhangjian/dataset/Simulation/datasets--openvla--modified_libero_rlds
+use_wrist_image:
+  desc: null
+  value: true
+use_proprio:
+  desc: null
+  value: true
+sequence_length:
+  desc: null
+  value: 384
+num_workers:
+  desc: null
+  value: 0
+num_diffusion_steps:
+  desc: null
+  value: 1000
+num_diffusion_inference_steps:
+  desc: null
+  value: 30
+lang_cond_dim:
+  desc: null
+  value: 1152
+img_cond_dim:
+  desc: null
+  value: 1152
+num_patches:
+  desc: null
+  value: 729
+dit_hidden_dim:
+  desc: null
+  value: 2048
+dit_depth:
+  desc: null
+  value: 28
+dit_num_heads:
+  desc: null
+  value: 16
+save_interval:
+  desc: null
+  value: 1000
+eval_interval:
+  desc: null
+  value: 500
+log_interval:
+  desc: null
+  value: 10
+save_num_checkpoints_to_keep:
+  desc: null
+  value: 3
+save_overwrite:
+  desc: null
+  value: true
+precision:
+  desc: null
+  value: amp_bf16
+gradient_accumulation_steps:
+  desc: null
+  value: 1
+max_grad_norm:
+  desc: null
+  value: 1.0
+wandb_project:
+  desc: null
+  value: a1-dit-action-20.47
+wandb_entity:
+  desc: null
+  value: demo0

wandb/offline-run-20250718_161945-m8wsmtst/files/output.log

@@ -0,0 +1,94 @@
+Traceback (most recent call last):
+  File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 216, in main
+    trainer.fit()
+  File "/home/zhangjian/A1/olmo/train.py", line 2796, in fit
+    metrics = self.train_step(batch)
+  File "/home/zhangjian/A1/olmo/train.py", line 2732, in train_step
+    text_embeds = self.text_model(input_ids=batch["input_ids"],attention_mask=batch.get("attention_mask")).last_hidden_state.detach()
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 739, in forward
+    return self.text_model(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 662, in forward
+    hidden_states = self.embeddings(input_ids=input_ids, position_ids=position_ids)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 308, in forward
+    raise ValueError(
+ValueError: Sequence length must be less than max_position_embeddings (got `sequence length`: 384 and max_position_embeddings: 64
+Traceback (most recent call last):
+  File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 236, in <module>
+    main()
+  File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 216, in main
+    trainer.fit()
+  File "/home/zhangjian/A1/olmo/train.py", line 2796, in fit
+    metrics = self.train_step(batch)
+  File "/home/zhangjian/A1/olmo/train.py", line 2732, in train_step
+    text_embeds = self.text_model(input_ids=batch["input_ids"],attention_mask=batch.get("attention_mask")).last_hidden_state.detach()
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 739, in forward
+    return self.text_model(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 662, in forward
+    hidden_states = self.embeddings(input_ids=input_ids, position_ids=position_ids)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 308, in forward
+    raise ValueError(
+ValueError: Sequence length must be less than max_position_embeddings (got `sequence length`: 384 and max_position_embeddings: 64
+[rank0]: Traceback (most recent call last):
+[rank0]:   File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 236, in <module>
+[rank0]:     main()
+[rank0]:   File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 216, in main
+[rank0]:     trainer.fit()
+[rank0]:   File "/home/zhangjian/A1/olmo/train.py", line 2796, in fit
+[rank0]:     metrics = self.train_step(batch)
+[rank0]:   File "/home/zhangjian/A1/olmo/train.py", line 2732, in train_step
+[rank0]:     text_embeds = self.text_model(input_ids=batch["input_ids"],attention_mask=batch.get("attention_mask")).last_hidden_state.detach()
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+[rank0]:     output = func(self, *args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 739, in forward
+[rank0]:     return self.text_model(
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+[rank0]:     output = func(self, *args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 662, in forward
+[rank0]:     hidden_states = self.embeddings(input_ids=input_ids, position_ids=position_ids)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 308, in forward
+[rank0]:     raise ValueError(
+[rank0]: ValueError: Sequence length must be less than max_position_embeddings (got `sequence length`: 384 and max_position_embeddings: 64

wandb/offline-run-20250718_161945-m8wsmtst/files/requirements.txt

@@ -0,0 +1,262 @@
+ai2-molmo==0.0.0
+yarl==1.20.1
+astunparse==1.6.3
+nvidia-cudnn-cu12==9.5.1.17
+etils==1.13.0
+tensorflow-io-gcs-filesystem==0.37.1
+Brotli==1.1.0
+gcsfs==2023.9.2
+multiprocess==0.70.16
+packaging==25.0
+idna==3.10
+absl-py==2.3.1
+id==1.5.0
+einops==0.8.1
+rich==13.9.4
+editdistance==0.8.1
+trouting==0.3.3
+mypy==1.3.0
+distro==1.9.0
+cycler==0.12.1
+s3transfer==0.13.0
+pyarrow==20.0.0
+sentry-sdk==2.33.0
+toml==0.10.2
+GitPython==3.1.44
+click==8.2.1
+ai2-molmo==0.0.0
+tensorflow-datasets==4.9.9
+tqdm==4.67.1
+safetensors==0.5.3
+google-auth==2.40.3
+iniconfig==2.1.0
+smmap==5.0.2
+nvidia-cublas-cu12==12.6.4.1
+wheel==0.45.1
+pluggy==1.6.0
+filelock==3.18.0
+google-auth-oauthlib==1.2.2
+httpx==0.28.1
+pyasn1==0.6.1
+typing-inspection==0.4.1
+pydub==0.25.1
+nvidia-cusparselt-cu12==0.6.3
+decorator==5.2.1
+frozenlist==1.7.0
+backports.tarfile==1.2.0
+setuptools==78.1.1
+keras==2.15.0
+MarkupSafe==3.0.2
+libclang==18.1.1
+necessary==0.4.3
+cached_path==1.7.3
+Jinja2==3.1.6
+beaker-py==2.4.6
+requests-toolbelt==1.0.0
+contourpy==1.3.2
+python-multipart==0.0.20
+nvidia-cusolver-cu12==11.7.1.2
+sentencepiece==0.2.0
+cffi==1.17.1
+gast==0.6.0
+nvidia-curand-cu12==10.3.7.77
+six==1.17.0
+boto3==1.39.5
+annotated-types==0.7.0
+termcolor==3.1.0
+pyparsing==3.2.3
+pydantic_core==2.33.2
+rfc3986==2.0.0
+pydantic==2.11.7
+jiter==0.10.0
+jaraco.context==6.0.1
+importlib_resources==6.5.2
+typeguard==2.13.3
+nltk==3.9.1
+opt_einsum==3.4.0
+grpcio==1.73.1
+pathspec==0.12.1
+triton==3.3.1
+gradio==5.37.0
+nvidia-nccl-cu12==2.26.2
+smart_open==7.3.0.post1
+scikit-learn==1.7.0
+pytest-sphinx==0.6.3
+google-cloud-core==2.4.3
+rsa==4.9.1
+readme_renderer==44.0
+httpcore==1.0.9
+threadpoolctl==3.6.0
+fsspec==2023.9.2
+keyring==25.6.0
+aiohttp==3.12.14
+attrs==25.3.0
+mypy_extensions==1.1.0
+antlr4-python3-runtime==4.9.3
+orjson==3.11.0
+cachetools==5.5.2
+requests==2.32.4
+smashed==0.21.5
+nvidia-cuda-cupti-cu12==12.6.80
+aiosignal==1.4.0
+torch==2.7.1
+xxhash==3.5.0
+OpenEXR==3.3.4
+cached-property==2.0.1
+wcwidth==0.2.13
+uvicorn==0.35.0
+numpy==1.26.4
+flatbuffers==25.2.10
+torchmetrics==1.7.4
+ftfy==6.3.1
+dill==0.3.8
+multidict==6.6.3
+shellingham==1.5.4
+promise==2.3
+anyio==4.9.0
+regex==2024.11.6
+tensorflow-graphics==2021.12.3
+pycparser==2.22
+nvidia-cuda-runtime-cu12==12.6.77
+matplotlib==3.10.3
+google-cloud-storage==2.19.0
+botocore==1.39.5
+click-option-group==0.5.7
+pytz==2025.2
+fonttools==4.58.5
+pyproject_hooks==1.2.0
+google-crc32c==1.7.1
+Markdown==3.8.2
+requests-oauthlib==2.0.0
+dlimp==0.0.1
+nvidia-cuda-nvrtc-cu12==12.6.77
+tomlkit==0.13.3
+safehttpx==0.1.6
+nvidia-cufft-cu12==11.3.0.4
+jmespath==1.0.1
+protobuf==4.21.12
+nvidia-nvtx-cu12==12.6.77
+tensorflow-addons==0.23.0
+jaraco.classes==3.4.0
+aiofiles==24.1.0
+exceptiongroup==1.3.0
+libcst==1.8.2
+immutabledict==4.2.1
+oauthlib==3.3.1
+timm==1.0.16
+lightning-utilities==0.14.3
+jeepney==0.9.0
+certifi==2025.7.14
+docutils==0.21.2
+markdown-it-py==3.0.0
+proto-plus==1.26.1
+wrapt==1.14.1
+mdurl==0.1.2
+more-itertools==10.7.0
+platformdirs==4.3.8
+kiwisolver==1.4.8
+tensorboard-data-server==0.7.2
+propcache==0.3.2
+zipp==3.23.0
+Werkzeug==3.1.3
+wandb==0.21.0
+Levenshtein==0.27.1
+isort==5.12.0
+simple-parsing==0.1.7
+face==24.0.0
+tzdata==2025.2
+h5py==3.14.0
+psutil==7.0.0
+mpmath==1.3.0
+dm-tree==0.1.9
+tensorflow-estimator==2.15.0
+pip==25.1
+joblib==1.5.1
+pyasn1_modules==0.4.2
+RapidFuzz==3.13.0
+cryptography==45.0.5
+typing_extensions==4.14.1
+urllib3==2.5.0
+sniffio==1.3.1
+omegaconf==2.3.0
+h11==0.16.0
+nvidia-cusparse-cu12==12.5.4.2
+click-help-colors==0.9.4
+einops-exts==0.0.4
+scipy==1.15.3
+nvidia-cufile-cu12==1.11.1.6
+huggingface-hub==0.33.4
+PyYAML==6.0.2
+pandas==2.3.1
+python-dateutil==2.9.0.post0
+petname==2.6
+tensorboard==2.15.2
+python-Levenshtein==0.27.1
+graphviz==0.21
+datasets==4.0.0
+gradio_client==1.10.4
+networkx==3.4.2
+nh3==0.2.22
+jaraco.functools==4.2.1
+ml-dtypes==0.2.0
+gitdb==4.0.12
+google-pasta==0.2.0
+openai==1.96.1
+aiohappyeyeballs==2.6.1
+docstring_parser==0.16
+build==1.2.2.post1
+tokenizers==0.21.2
+black==23.12.1
+fiddle==0.3.0
+requirements-parser==0.13.0
+ruff==0.12.3
+trimesh==4.7.0
+Pygments==2.19.2
+google-resumable-media==2.7.2
+async-timeout==5.0.1
+hf-xet==1.1.5
+groovy==0.1.2
+accelerate==1.8.1
+diffusers==0.33.1
+SecretStorage==3.3.3
+ffmpy==0.6.0
+websockets==15.0.1
+fastapi==0.116.1
+importlib_metadata==8.7.0
+pillow==11.3.0
+nvidia-nvjitlink-cu12==12.6.85
+beaker-gantry==2.8.3
+glom==24.11.0
+transformers==4.53.2
+pytest==8.4.1
+tensorflow==2.15.0
+semantic-version==2.10.0
+charset-normalizer==3.4.2
+tensorflow-metadata==1.17.2
+googleapis-common-protos==1.70.0
+boltons==25.0.0
+twine==6.1.0
+tomli==2.2.1
+google-api-core==2.25.1
+typer==0.16.0
+sympy==1.14.0
+starlette==0.47.1
+msgspec==0.19.0
+array_record==0.7.2
+torchvision==0.22.1
+typing_extensions==4.12.2
+wheel==0.45.1
+backports.tarfile==1.2.0
+inflect==7.3.1
+zipp==3.19.2
+jaraco.collections==5.1.0
+jaraco.text==3.12.1
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+tomli==2.0.1
+platformdirs==4.2.2
+packaging==24.2
+typeguard==4.3.0
+importlib_metadata==8.0.0
+more-itertools==10.3.0
+autocommand==2.2.2

wandb/offline-run-20250718_161945-m8wsmtst/files/wandb-metadata.json

@@ -0,0 +1 @@
+{"os": "Linux-5.15.0-142-generic-x86_64-with-glibc2.35", "python": "CPython 3.10.16", "started_at": "2025-07-18T08:19:45.826177Z", "args": ["--save_folder=/mnt/data/zhangjian/a1/libero_spatial_dit_action", "--save_overwrite"], "program": "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", "code_path": "launch_scripts/train_vla_dit_action.py", "code_path_local": "launch_scripts/train_vla_dit_action.py", "git": {"remote_url": "https://github.com/Spatialtemporal-AI/A1.git", "commit": "2e05e206652258f5fbb178ec13c792c2a91c1856"}, "root": "/home/zhangjian/A1", "host": "iv-ydwxzrkf0gvr6okuck21", "executable": "/mnt/data/zhangjian/miniconda3/envs/a1/bin/python3.10", "cpu_count": 56, "cpu_count_logical": 112, "gpu_type": "NVIDIA A100-SXM4-80GB", "gpu_count": 8, "disk": {"/": {"total": "507024936960", "used": "133182132224"}}, "memory": {"total": "2071488372736"}, "gpu_nvidia": [{"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-b01b938f-1b95-8587-4b93-cfa68d84c935"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-49b283f0-f1e6-992d-4ed8-99855a5e6529"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-a340dca6-1db9-f578-76f2-cf36f74ffabd"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-c1bddabf-aa88-9c8f-eec2-6dc3e117f985"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-4096c1e7-0437-6727-096f-d67141c140b8"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-eeb896c0-61d6-8380-778b-d02101078a17"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-d5c1ed79-76e3-22a6-8908-df289b090c03"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-6d90b060-f6f7-599b-f939-4ba124ff11b7"}], "cuda_version": "12.8", "writer_id": "gkv6pcimbk2w6ig5eqnc01n6qgq9ed3l"}

wandb/offline-run-20250718_161945-m8wsmtst/files/wandb-summary.json

@@ -0,0 +1 @@
+{"_runtime": 6}

wandb/offline-run-20250718_161945-m8wsmtst/logs/debug-core.log

@@ -0,0 +1,14 @@
+{"time":"2025-07-18T16:19:45.856929542+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmph2yqu36w/port-2874027.txt","pid":2874027,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-07-18T16:19:45.857509675+08:00","level":"INFO","msg":"server: will exit if parent process dies","ppid":2874027}
+{"time":"2025-07-18T16:19:45.8575527+08:00","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-2874027-2876426-2611772877/socket","Net":"unix"}}
+{"time":"2025-07-18T16:19:46.034393648+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-07-18T16:19:46.038087649+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"m8wsmtst","id":"1(@)"}
+{"time":"2025-07-18T16:19:46.157222198+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"m8wsmtst","id":"1(@)"}
+{"time":"2025-07-18T16:19:52.167429191+08:00","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-07-18T16:19:52.167510001+08:00","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-07-18T16:19:52.16750131+08:00","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-07-18T16:19:52.167581797+08:00","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-07-18T16:19:52.167604834+08:00","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-2874027-2876426-2611772877/socket","Net":"unix"}}
+{"time":"2025-07-18T16:19:52.169166335+08:00","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-07-18T16:19:52.169190878+08:00","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-07-18T16:19:52.169201032+08:00","level":"INFO","msg":"server is closed"}

wandb/offline-run-20250718_161945-m8wsmtst/logs/debug-internal.log

@@ -0,0 +1,13 @@
+{"time":"2025-07-18T16:19:46.042031385+08:00","level":"INFO","msg":"stream: starting","core version":"0.21.0"}
+{"time":"2025-07-18T16:19:46.157051193+08:00","level":"WARN","msg":"GraphQL client is nil, skipping feature loading"}
+{"time":"2025-07-18T16:19:46.157197135+08:00","level":"INFO","msg":"stream: created new stream","id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:46.157216504+08:00","level":"INFO","msg":"stream: started","id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:46.157229719+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:46.157247709+08:00","level":"INFO","msg":"handler: started","stream_id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:46.15725118+08:00","level":"INFO","msg":"sender: started","stream_id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:46.157739142+08:00","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-07-18T16:19:52.167599755+08:00","level":"INFO","msg":"stream: closing","id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:52.167794108+08:00","level":"INFO","msg":"handler: closed","stream_id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:52.167820176+08:00","level":"INFO","msg":"writer: Close: closed","stream_id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:52.167839497+08:00","level":"INFO","msg":"sender: closed","stream_id":"m8wsmtst"}
+{"time":"2025-07-18T16:19:52.167887348+08:00","level":"INFO","msg":"stream: closed","id":"m8wsmtst"}

wandb/offline-run-20250718_162533-enh8w5nh/files/requirements.txt

@@ -0,0 +1,262 @@
+ai2-molmo==0.0.0
+yarl==1.20.1
+astunparse==1.6.3
+nvidia-cudnn-cu12==9.5.1.17
+etils==1.13.0
+tensorflow-io-gcs-filesystem==0.37.1
+Brotli==1.1.0
+gcsfs==2023.9.2
+multiprocess==0.70.16
+packaging==25.0
+idna==3.10
+absl-py==2.3.1
+id==1.5.0
+einops==0.8.1
+rich==13.9.4
+editdistance==0.8.1
+trouting==0.3.3
+mypy==1.3.0
+distro==1.9.0
+cycler==0.12.1
+s3transfer==0.13.0
+pyarrow==20.0.0
+sentry-sdk==2.33.0
+toml==0.10.2
+GitPython==3.1.44
+click==8.2.1
+ai2-molmo==0.0.0
+tensorflow-datasets==4.9.9
+tqdm==4.67.1
+safetensors==0.5.3
+google-auth==2.40.3
+iniconfig==2.1.0
+smmap==5.0.2
+nvidia-cublas-cu12==12.6.4.1
+wheel==0.45.1
+pluggy==1.6.0
+filelock==3.18.0
+google-auth-oauthlib==1.2.2
+httpx==0.28.1
+pyasn1==0.6.1
+typing-inspection==0.4.1
+pydub==0.25.1
+nvidia-cusparselt-cu12==0.6.3
+decorator==5.2.1
+frozenlist==1.7.0
+backports.tarfile==1.2.0
+setuptools==78.1.1
+keras==2.15.0
+MarkupSafe==3.0.2
+libclang==18.1.1
+necessary==0.4.3
+cached_path==1.7.3
+Jinja2==3.1.6
+beaker-py==2.4.6
+requests-toolbelt==1.0.0
+contourpy==1.3.2
+python-multipart==0.0.20
+nvidia-cusolver-cu12==11.7.1.2
+sentencepiece==0.2.0
+cffi==1.17.1
+gast==0.6.0
+nvidia-curand-cu12==10.3.7.77
+six==1.17.0
+boto3==1.39.5
+annotated-types==0.7.0
+termcolor==3.1.0
+pyparsing==3.2.3
+pydantic_core==2.33.2
+rfc3986==2.0.0
+pydantic==2.11.7
+jiter==0.10.0
+jaraco.context==6.0.1
+importlib_resources==6.5.2
+typeguard==2.13.3
+nltk==3.9.1
+opt_einsum==3.4.0
+grpcio==1.73.1
+pathspec==0.12.1
+triton==3.3.1
+gradio==5.37.0
+nvidia-nccl-cu12==2.26.2
+smart_open==7.3.0.post1
+scikit-learn==1.7.0
+pytest-sphinx==0.6.3
+google-cloud-core==2.4.3
+rsa==4.9.1
+readme_renderer==44.0
+httpcore==1.0.9
+threadpoolctl==3.6.0
+fsspec==2023.9.2
+keyring==25.6.0
+aiohttp==3.12.14
+attrs==25.3.0
+mypy_extensions==1.1.0
+antlr4-python3-runtime==4.9.3
+orjson==3.11.0
+cachetools==5.5.2
+requests==2.32.4
+smashed==0.21.5
+nvidia-cuda-cupti-cu12==12.6.80
+aiosignal==1.4.0
+torch==2.7.1
+xxhash==3.5.0
+OpenEXR==3.3.4
+cached-property==2.0.1
+wcwidth==0.2.13
+uvicorn==0.35.0
+numpy==1.26.4
+flatbuffers==25.2.10
+torchmetrics==1.7.4
+ftfy==6.3.1
+dill==0.3.8
+multidict==6.6.3
+shellingham==1.5.4
+promise==2.3
+anyio==4.9.0
+regex==2024.11.6
+tensorflow-graphics==2021.12.3
+pycparser==2.22
+nvidia-cuda-runtime-cu12==12.6.77
+matplotlib==3.10.3
+google-cloud-storage==2.19.0
+botocore==1.39.5
+click-option-group==0.5.7
+pytz==2025.2
+fonttools==4.58.5
+pyproject_hooks==1.2.0
+google-crc32c==1.7.1
+Markdown==3.8.2
+requests-oauthlib==2.0.0
+dlimp==0.0.1
+nvidia-cuda-nvrtc-cu12==12.6.77
+tomlkit==0.13.3
+safehttpx==0.1.6
+nvidia-cufft-cu12==11.3.0.4
+jmespath==1.0.1
+protobuf==4.21.12
+nvidia-nvtx-cu12==12.6.77
+tensorflow-addons==0.23.0
+jaraco.classes==3.4.0
+aiofiles==24.1.0
+exceptiongroup==1.3.0
+libcst==1.8.2
+immutabledict==4.2.1
+oauthlib==3.3.1
+timm==1.0.16
+lightning-utilities==0.14.3
+jeepney==0.9.0
+certifi==2025.7.14
+docutils==0.21.2
+markdown-it-py==3.0.0
+proto-plus==1.26.1
+wrapt==1.14.1
+mdurl==0.1.2
+more-itertools==10.7.0
+platformdirs==4.3.8
+kiwisolver==1.4.8
+tensorboard-data-server==0.7.2
+propcache==0.3.2
+zipp==3.23.0
+Werkzeug==3.1.3
+wandb==0.21.0
+Levenshtein==0.27.1
+isort==5.12.0
+simple-parsing==0.1.7
+face==24.0.0
+tzdata==2025.2
+h5py==3.14.0
+psutil==7.0.0
+mpmath==1.3.0
+dm-tree==0.1.9
+tensorflow-estimator==2.15.0
+pip==25.1
+joblib==1.5.1
+pyasn1_modules==0.4.2
+RapidFuzz==3.13.0
+cryptography==45.0.5
+typing_extensions==4.14.1
+urllib3==2.5.0
+sniffio==1.3.1
+omegaconf==2.3.0
+h11==0.16.0
+nvidia-cusparse-cu12==12.5.4.2
+click-help-colors==0.9.4
+einops-exts==0.0.4
+scipy==1.15.3
+nvidia-cufile-cu12==1.11.1.6
+huggingface-hub==0.33.4
+PyYAML==6.0.2
+pandas==2.3.1
+python-dateutil==2.9.0.post0
+petname==2.6
+tensorboard==2.15.2
+python-Levenshtein==0.27.1
+graphviz==0.21
+datasets==4.0.0
+gradio_client==1.10.4
+networkx==3.4.2
+nh3==0.2.22
+jaraco.functools==4.2.1
+ml-dtypes==0.2.0
+gitdb==4.0.12
+google-pasta==0.2.0
+openai==1.96.1
+aiohappyeyeballs==2.6.1
+docstring_parser==0.16
+build==1.2.2.post1
+tokenizers==0.21.2
+black==23.12.1
+fiddle==0.3.0
+requirements-parser==0.13.0
+ruff==0.12.3
+trimesh==4.7.0
+Pygments==2.19.2
+google-resumable-media==2.7.2
+async-timeout==5.0.1
+hf-xet==1.1.5
+groovy==0.1.2
+accelerate==1.8.1
+diffusers==0.33.1
+SecretStorage==3.3.3
+ffmpy==0.6.0
+websockets==15.0.1
+fastapi==0.116.1
+importlib_metadata==8.7.0
+pillow==11.3.0
+nvidia-nvjitlink-cu12==12.6.85
+beaker-gantry==2.8.3
+glom==24.11.0
+transformers==4.53.2
+pytest==8.4.1
+tensorflow==2.15.0
+semantic-version==2.10.0
+charset-normalizer==3.4.2
+tensorflow-metadata==1.17.2
+googleapis-common-protos==1.70.0
+boltons==25.0.0
+twine==6.1.0
+tomli==2.2.1
+google-api-core==2.25.1
+typer==0.16.0
+sympy==1.14.0
+starlette==0.47.1
+msgspec==0.19.0
+array_record==0.7.2
+torchvision==0.22.1
+typing_extensions==4.12.2
+wheel==0.45.1
+backports.tarfile==1.2.0
+inflect==7.3.1
+zipp==3.19.2
+jaraco.collections==5.1.0
+jaraco.text==3.12.1
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+tomli==2.0.1
+platformdirs==4.2.2
+packaging==24.2
+typeguard==4.3.0
+importlib_metadata==8.0.0
+more-itertools==10.3.0
+autocommand==2.2.2

wandb/offline-run-20250718_162533-enh8w5nh/logs/debug-core.log

@@ -0,0 +1,7 @@
+{"time":"2025-07-18T16:25:33.048887646+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpo5iypyp2/port-2884717.txt","pid":2884717,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-07-18T16:25:33.049416652+08:00","level":"INFO","msg":"server: will exit if parent process dies","ppid":2884717}
+{"time":"2025-07-18T16:25:33.049407439+08:00","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-2884717-2886714-2553117481/socket","Net":"unix"}}
+{"time":"2025-07-18T16:25:33.227775424+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-07-18T16:25:33.231715039+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"enh8w5nh","id":"1(@)"}
+{"time":"2025-07-18T16:25:33.349089165+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"enh8w5nh","id":"1(@)"}
+{"time":"2025-07-18T16:25:38.161645631+08:00","level":"INFO","msg":"server: parent process exited, terminating service process"}

wandb/offline-run-20250718_162533-enh8w5nh/logs/debug-internal.log

@@ -0,0 +1,8 @@
+{"time":"2025-07-18T16:25:33.23556091+08:00","level":"INFO","msg":"stream: starting","core version":"0.21.0"}
+{"time":"2025-07-18T16:25:33.348898335+08:00","level":"WARN","msg":"GraphQL client is nil, skipping feature loading"}
+{"time":"2025-07-18T16:25:33.349063018+08:00","level":"INFO","msg":"stream: created new stream","id":"enh8w5nh"}
+{"time":"2025-07-18T16:25:33.349083407+08:00","level":"INFO","msg":"stream: started","id":"enh8w5nh"}
+{"time":"2025-07-18T16:25:33.349099668+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"enh8w5nh"}
+{"time":"2025-07-18T16:25:33.349118419+08:00","level":"INFO","msg":"sender: started","stream_id":"enh8w5nh"}
+{"time":"2025-07-18T16:25:33.349126683+08:00","level":"INFO","msg":"handler: started","stream_id":"enh8w5nh"}
+{"time":"2025-07-18T16:25:33.349658731+08:00","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}

wandb/offline-run-20250718_162819-6btqdhbk/files/requirements.txt

@@ -0,0 +1,262 @@
+ai2-molmo==0.0.0
+yarl==1.20.1
+astunparse==1.6.3
+nvidia-cudnn-cu12==9.5.1.17
+etils==1.13.0
+tensorflow-io-gcs-filesystem==0.37.1
+Brotli==1.1.0
+gcsfs==2023.9.2
+multiprocess==0.70.16
+packaging==25.0
+idna==3.10
+absl-py==2.3.1
+id==1.5.0
+einops==0.8.1
+rich==13.9.4
+editdistance==0.8.1
+trouting==0.3.3
+mypy==1.3.0
+distro==1.9.0
+cycler==0.12.1
+s3transfer==0.13.0
+pyarrow==20.0.0
+sentry-sdk==2.33.0
+toml==0.10.2
+GitPython==3.1.44
+click==8.2.1
+ai2-molmo==0.0.0
+tensorflow-datasets==4.9.9
+tqdm==4.67.1
+safetensors==0.5.3
+google-auth==2.40.3
+iniconfig==2.1.0
+smmap==5.0.2
+nvidia-cublas-cu12==12.6.4.1
+wheel==0.45.1
+pluggy==1.6.0
+filelock==3.18.0
+google-auth-oauthlib==1.2.2
+httpx==0.28.1
+pyasn1==0.6.1
+typing-inspection==0.4.1
+pydub==0.25.1
+nvidia-cusparselt-cu12==0.6.3
+decorator==5.2.1
+frozenlist==1.7.0
+backports.tarfile==1.2.0
+setuptools==78.1.1
+keras==2.15.0
+MarkupSafe==3.0.2
+libclang==18.1.1
+necessary==0.4.3
+cached_path==1.7.3
+Jinja2==3.1.6
+beaker-py==2.4.6
+requests-toolbelt==1.0.0
+contourpy==1.3.2
+python-multipart==0.0.20
+nvidia-cusolver-cu12==11.7.1.2
+sentencepiece==0.2.0
+cffi==1.17.1
+gast==0.6.0
+nvidia-curand-cu12==10.3.7.77
+six==1.17.0
+boto3==1.39.5
+annotated-types==0.7.0
+termcolor==3.1.0
+pyparsing==3.2.3
+pydantic_core==2.33.2
+rfc3986==2.0.0
+pydantic==2.11.7
+jiter==0.10.0
+jaraco.context==6.0.1
+importlib_resources==6.5.2
+typeguard==2.13.3
+nltk==3.9.1
+opt_einsum==3.4.0
+grpcio==1.73.1
+pathspec==0.12.1
+triton==3.3.1
+gradio==5.37.0
+nvidia-nccl-cu12==2.26.2
+smart_open==7.3.0.post1
+scikit-learn==1.7.0
+pytest-sphinx==0.6.3
+google-cloud-core==2.4.3
+rsa==4.9.1
+readme_renderer==44.0
+httpcore==1.0.9
+threadpoolctl==3.6.0
+fsspec==2023.9.2
+keyring==25.6.0
+aiohttp==3.12.14
+attrs==25.3.0
+mypy_extensions==1.1.0
+antlr4-python3-runtime==4.9.3
+orjson==3.11.0
+cachetools==5.5.2
+requests==2.32.4
+smashed==0.21.5
+nvidia-cuda-cupti-cu12==12.6.80
+aiosignal==1.4.0
+torch==2.7.1
+xxhash==3.5.0
+OpenEXR==3.3.4
+cached-property==2.0.1
+wcwidth==0.2.13
+uvicorn==0.35.0
+numpy==1.26.4
+flatbuffers==25.2.10
+torchmetrics==1.7.4
+ftfy==6.3.1
+dill==0.3.8
+multidict==6.6.3
+shellingham==1.5.4
+promise==2.3
+anyio==4.9.0
+regex==2024.11.6
+tensorflow-graphics==2021.12.3
+pycparser==2.22
+nvidia-cuda-runtime-cu12==12.6.77
+matplotlib==3.10.3
+google-cloud-storage==2.19.0
+botocore==1.39.5
+click-option-group==0.5.7
+pytz==2025.2
+fonttools==4.58.5
+pyproject_hooks==1.2.0
+google-crc32c==1.7.1
+Markdown==3.8.2
+requests-oauthlib==2.0.0
+dlimp==0.0.1
+nvidia-cuda-nvrtc-cu12==12.6.77
+tomlkit==0.13.3
+safehttpx==0.1.6
+nvidia-cufft-cu12==11.3.0.4
+jmespath==1.0.1
+protobuf==4.21.12
+nvidia-nvtx-cu12==12.6.77
+tensorflow-addons==0.23.0
+jaraco.classes==3.4.0
+aiofiles==24.1.0
+exceptiongroup==1.3.0
+libcst==1.8.2
+immutabledict==4.2.1
+oauthlib==3.3.1
+timm==1.0.16
+lightning-utilities==0.14.3
+jeepney==0.9.0
+certifi==2025.7.14
+docutils==0.21.2
+markdown-it-py==3.0.0
+proto-plus==1.26.1
+wrapt==1.14.1
+mdurl==0.1.2
+more-itertools==10.7.0
+platformdirs==4.3.8
+kiwisolver==1.4.8
+tensorboard-data-server==0.7.2
+propcache==0.3.2
+zipp==3.23.0
+Werkzeug==3.1.3
+wandb==0.21.0
+Levenshtein==0.27.1
+isort==5.12.0
+simple-parsing==0.1.7
+face==24.0.0
+tzdata==2025.2
+h5py==3.14.0
+psutil==7.0.0
+mpmath==1.3.0
+dm-tree==0.1.9
+tensorflow-estimator==2.15.0
+pip==25.1
+joblib==1.5.1
+pyasn1_modules==0.4.2
+RapidFuzz==3.13.0
+cryptography==45.0.5
+typing_extensions==4.14.1
+urllib3==2.5.0
+sniffio==1.3.1
+omegaconf==2.3.0
+h11==0.16.0
+nvidia-cusparse-cu12==12.5.4.2
+click-help-colors==0.9.4
+einops-exts==0.0.4
+scipy==1.15.3
+nvidia-cufile-cu12==1.11.1.6
+huggingface-hub==0.33.4
+PyYAML==6.0.2
+pandas==2.3.1
+python-dateutil==2.9.0.post0
+petname==2.6
+tensorboard==2.15.2
+python-Levenshtein==0.27.1
+graphviz==0.21
+datasets==4.0.0
+gradio_client==1.10.4
+networkx==3.4.2
+nh3==0.2.22
+jaraco.functools==4.2.1
+ml-dtypes==0.2.0
+gitdb==4.0.12
+google-pasta==0.2.0
+openai==1.96.1
+aiohappyeyeballs==2.6.1
+docstring_parser==0.16
+build==1.2.2.post1
+tokenizers==0.21.2
+black==23.12.1
+fiddle==0.3.0
+requirements-parser==0.13.0
+ruff==0.12.3
+trimesh==4.7.0
+Pygments==2.19.2
+google-resumable-media==2.7.2
+async-timeout==5.0.1
+hf-xet==1.1.5
+groovy==0.1.2
+accelerate==1.8.1
+diffusers==0.33.1
+SecretStorage==3.3.3
+ffmpy==0.6.0
+websockets==15.0.1
+fastapi==0.116.1
+importlib_metadata==8.7.0
+pillow==11.3.0
+nvidia-nvjitlink-cu12==12.6.85
+beaker-gantry==2.8.3
+glom==24.11.0
+transformers==4.53.2
+pytest==8.4.1
+tensorflow==2.15.0
+semantic-version==2.10.0
+charset-normalizer==3.4.2
+tensorflow-metadata==1.17.2
+googleapis-common-protos==1.70.0
+boltons==25.0.0
+twine==6.1.0
+tomli==2.2.1
+google-api-core==2.25.1
+typer==0.16.0
+sympy==1.14.0
+starlette==0.47.1
+msgspec==0.19.0
+array_record==0.7.2
+torchvision==0.22.1
+typing_extensions==4.12.2
+wheel==0.45.1
+backports.tarfile==1.2.0
+inflect==7.3.1
+zipp==3.19.2
+jaraco.collections==5.1.0
+jaraco.text==3.12.1
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+tomli==2.0.1
+platformdirs==4.2.2
+packaging==24.2
+typeguard==4.3.0
+importlib_metadata==8.0.0
+more-itertools==10.3.0
+autocommand==2.2.2

wandb/offline-run-20250718_162819-6btqdhbk/logs/debug-core.log

@@ -0,0 +1,7 @@
+{"time":"2025-07-18T16:28:19.409855671+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp0pgj9n56/port-2890422.txt","pid":2890422,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-07-18T16:28:19.410428972+08:00","level":"INFO","msg":"server: will exit if parent process dies","ppid":2890422}
+{"time":"2025-07-18T16:28:19.410449859+08:00","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-2890422-2892247-2023005874/socket","Net":"unix"}}
+{"time":"2025-07-18T16:28:19.589965556+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-07-18T16:28:19.594940331+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"6btqdhbk","id":"1(@)"}
+{"time":"2025-07-18T16:28:19.712915229+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"6btqdhbk","id":"1(@)"}
+{"time":"2025-07-18T16:28:26.728818238+08:00","level":"INFO","msg":"server: parent process exited, terminating service process"}

wandb/offline-run-20250718_162819-6btqdhbk/logs/debug-internal.log

@@ -0,0 +1,8 @@
+{"time":"2025-07-18T16:28:19.598579541+08:00","level":"INFO","msg":"stream: starting","core version":"0.21.0"}
+{"time":"2025-07-18T16:28:19.712737263+08:00","level":"WARN","msg":"GraphQL client is nil, skipping feature loading"}
+{"time":"2025-07-18T16:28:19.712888407+08:00","level":"INFO","msg":"stream: created new stream","id":"6btqdhbk"}
+{"time":"2025-07-18T16:28:19.712910248+08:00","level":"INFO","msg":"stream: started","id":"6btqdhbk"}
+{"time":"2025-07-18T16:28:19.712932303+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"6btqdhbk"}
+{"time":"2025-07-18T16:28:19.712954867+08:00","level":"INFO","msg":"handler: started","stream_id":"6btqdhbk"}
+{"time":"2025-07-18T16:28:19.712947454+08:00","level":"INFO","msg":"sender: started","stream_id":"6btqdhbk"}
+{"time":"2025-07-18T16:28:19.71364791+08:00","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}

wandb/offline-run-20250718_163819-kujhudzv/files/config.yaml

@@ -0,0 +1,227 @@
+wandb_version: 1
+
+_wandb:
+  desc: null
+  value:
+    python_version: 3.10.16
+    cli_version: 0.21.0
+    framework: huggingface
+    huggingface_version: 4.53.2
+    is_jupyter_run: false
+    is_kaggle_kernel: false
+    start_time: 1752827899
+    t:
+      1:
+      - 1
+      - 2
+      - 3
+      - 5
+      - 11
+      - 41
+      - 49
+      - 51
+      - 53
+      - 63
+      - 71
+      - 83
+      - 95
+      - 105
+      2:
+      - 1
+      - 2
+      - 3
+      - 5
+      - 11
+      - 41
+      - 49
+      - 51
+      - 53
+      - 63
+      - 71
+      - 83
+      - 95
+      - 105
+      3:
+      - 4
+      - 13
+      - 16
+      - 42
+      4: 3.10.16
+      5: 0.21.0
+      6: 4.53.2
+      13: linux-x86_64
+    e:
+      oygl393zcukmmdj0i9c4b0424qwzehyz:
+        os: Linux-5.15.0-142-generic-x86_64-with-glibc2.35
+        python: CPython 3.10.16
+        started_at: '2025-07-18T08:38:19.403312Z'
+        args:
+        - --save_folder=/mnt/data/zhangjian/a1/libero_spatial_dit_action
+        - --save_overwrite
+        program: /home/zhangjian/A1/launch_scripts/train_vla_dit_action.py
+        code_path: launch_scripts/train_vla_dit_action.py
+        code_path_local: launch_scripts/train_vla_dit_action.py
+        git:
+          remote_url: https://github.com/Spatialtemporal-AI/A1.git
+          commit: 2e05e206652258f5fbb178ec13c792c2a91c1856
+        root: /home/zhangjian/A1
+        host: iv-ydwxzrkf0gvr6okuck21
+        executable: /mnt/data/zhangjian/miniconda3/envs/a1/bin/python3.10
+        cpu_count: 56
+        cpu_count_logical: 112
+        gpu_type: NVIDIA A100-SXM4-80GB
+        gpu_count: 8
+        disk:
+          /:
+            total: '507024936960'
+            used: '133188997120'
+        memory:
+          total: '2071488372736'
+        gpu_nvidia:
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-b01b938f-1b95-8587-4b93-cfa68d84c935
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-49b283f0-f1e6-992d-4ed8-99855a5e6529
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-a340dca6-1db9-f578-76f2-cf36f74ffabd
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-c1bddabf-aa88-9c8f-eec2-6dc3e117f985
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-4096c1e7-0437-6727-096f-d67141c140b8
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-eeb896c0-61d6-8380-778b-d02101078a17
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-d5c1ed79-76e3-22a6-8908-df289b090c03
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-6d90b060-f6f7-599b-f939-4ba124ff11b7
+        cuda_version: '12.8'
+        writer_id: oygl393zcukmmdj0i9c4b0424qwzehyz
+text_model_name:
+  desc: null
+  value: google/siglip-so400m-patch14-384
+vision_model_name:
+  desc: null
+  value: google/siglip-so400m-patch14-384
+run_name:
+  desc: null
+  value: dit_action_train_20250718_163750
+save_folder:
+  desc: null
+  value: /mnt/data/zhangjian/a1/libero_spatial_dit_action
+seed:
+  desc: null
+  value: 42
+max_steps:
+  desc: null
+  value: 100000
+global_batch_size:
+  desc: null
+  value: 128
+device_batch_size:
+  desc: null
+  value: 32
+learning_rate:
+  desc: null
+  value: 2.0e-05
+weight_decay:
+  desc: null
+  value: 0.01
+warmup_steps:
+  desc: null
+  value: 500
+dataset_name:
+  desc: null
+  value: libero_spatial_no_noops
+data_root_dir:
+  desc: null
+  value: /mnt/data/zhangjian/dataset/Simulation/datasets--openvla--modified_libero_rlds
+use_wrist_image:
+  desc: null
+  value: true
+use_proprio:
+  desc: null
+  value: true
+sequence_length:
+  desc: null
+  value: 64
+num_workers:
+  desc: null
+  value: 0
+num_diffusion_steps:
+  desc: null
+  value: 1000
+num_diffusion_inference_steps:
+  desc: null
+  value: 30
+lang_cond_dim:
+  desc: null
+  value: 1152
+img_cond_dim:
+  desc: null
+  value: 1152
+num_patches:
+  desc: null
+  value: 729
+dit_hidden_dim:
+  desc: null
+  value: 2048
+dit_depth:
+  desc: null
+  value: 28
+dit_num_heads:
+  desc: null
+  value: 16
+save_interval:
+  desc: null
+  value: 1000
+eval_interval:
+  desc: null
+  value: 500
+log_interval:
+  desc: null
+  value: 10
+save_num_checkpoints_to_keep:
+  desc: null
+  value: 3
+save_overwrite:
+  desc: null
+  value: true
+precision:
+  desc: null
+  value: amp_bf16
+gradient_accumulation_steps:
+  desc: null
+  value: 1
+max_grad_norm:
+  desc: null
+  value: 1.0
+wandb_project:
+  desc: null
+  value: a1-dit-action-20.47
+wandb_entity:
+  desc: null
+  value: demo0

wandb/offline-run-20250718_163819-kujhudzv/files/output.log

@@ -0,0 +1,160 @@
+Traceback (most recent call last):
+  File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 216, in main
+    trainer.fit()
+  File "/home/zhangjian/A1/olmo/train.py", line 2798, in fit
+    metrics = self.train_step(batch)
+  File "/home/zhangjian/A1/olmo/train.py", line 2735, in train_step
+    image_embeds = self.vision_model(pixel_values=batch["pixel_values"]).last_hidden_state.detach()
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 870, in forward
+    return self.vision_model(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 777, in forward
+    encoder_outputs: BaseModelOutput = self.encoder(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 608, in forward
+    layer_outputs = encoder_layer(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/modeling_layers.py", line 83, in __call__
+    return super().__call__(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 472, in forward
+    hidden_states = self.mlp(hidden_states)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 429, in forward
+    hidden_states = self.fc1(hidden_states)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/linear.py", line 125, in forward
+    return F.linear(input, self.weight, self.bias)
+torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 768.00 MiB. GPU 0 has a total capacity of 79.44 GiB of which 441.00 MiB is free. Including non-PyTorch memory, this process has 79.00 GiB memory in use. Of the allocated memory 77.89 GiB is allocated by PyTorch, and 630.78 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
+Traceback (most recent call last):
+  File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 236, in <module>
+    main()
+  File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 216, in main
+    trainer.fit()
+  File "/home/zhangjian/A1/olmo/train.py", line 2798, in fit
+    metrics = self.train_step(batch)
+  File "/home/zhangjian/A1/olmo/train.py", line 2735, in train_step
+    image_embeds = self.vision_model(pixel_values=batch["pixel_values"]).last_hidden_state.detach()
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 870, in forward
+    return self.vision_model(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 777, in forward
+    encoder_outputs: BaseModelOutput = self.encoder(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+    output = func(self, *args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 608, in forward
+    layer_outputs = encoder_layer(
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/modeling_layers.py", line 83, in __call__
+    return super().__call__(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 472, in forward
+    hidden_states = self.mlp(hidden_states)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 429, in forward
+    hidden_states = self.fc1(hidden_states)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+    return self._call_impl(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+    return forward_call(*args, **kwargs)
+  File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/linear.py", line 125, in forward
+    return F.linear(input, self.weight, self.bias)
+torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 768.00 MiB. GPU 0 has a total capacity of 79.44 GiB of which 441.00 MiB is free. Including non-PyTorch memory, this process has 79.00 GiB memory in use. Of the allocated memory 77.89 GiB is allocated by PyTorch, and 630.78 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
+[rank0]: Traceback (most recent call last):
+[rank0]:   File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 236, in <module>
+[rank0]:     main()
+[rank0]:   File "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", line 216, in main
+[rank0]:     trainer.fit()
+[rank0]:   File "/home/zhangjian/A1/olmo/train.py", line 2798, in fit
+[rank0]:     metrics = self.train_step(batch)
+[rank0]:   File "/home/zhangjian/A1/olmo/train.py", line 2735, in train_step
+[rank0]:     image_embeds = self.vision_model(pixel_values=batch["pixel_values"]).last_hidden_state.detach()
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+[rank0]:     output = func(self, *args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 870, in forward
+[rank0]:     return self.vision_model(
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+[rank0]:     output = func(self, *args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 777, in forward
+[rank0]:     encoder_outputs: BaseModelOutput = self.encoder(
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/utils/generic.py", line 943, in wrapper
+[rank0]:     output = func(self, *args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 608, in forward
+[rank0]:     layer_outputs = encoder_layer(
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/modeling_layers.py", line 83, in __call__
+[rank0]:     return super().__call__(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 472, in forward
+[rank0]:     hidden_states = self.mlp(hidden_states)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/transformers/models/siglip/modeling_siglip.py", line 429, in forward
+[rank0]:     hidden_states = self.fc1(hidden_states)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
+[rank0]:     return self._call_impl(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
+[rank0]:     return forward_call(*args, **kwargs)
+[rank0]:   File "/mnt/data/zhangjian/miniconda3/envs/a1/lib/python3.10/site-packages/torch/nn/modules/linear.py", line 125, in forward
+[rank0]:     return F.linear(input, self.weight, self.bias)
+[rank0]: torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 768.00 MiB. GPU 0 has a total capacity of 79.44 GiB of which 441.00 MiB is free. Including non-PyTorch memory, this process has 79.00 GiB memory in use. Of the allocated memory 77.89 GiB is allocated by PyTorch, and 630.78 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)

wandb/offline-run-20250718_163819-kujhudzv/files/requirements.txt

@@ -0,0 +1,262 @@
+ai2-molmo==0.0.0
+yarl==1.20.1
+astunparse==1.6.3
+nvidia-cudnn-cu12==9.5.1.17
+etils==1.13.0
+tensorflow-io-gcs-filesystem==0.37.1
+Brotli==1.1.0
+gcsfs==2023.9.2
+multiprocess==0.70.16
+packaging==25.0
+idna==3.10
+absl-py==2.3.1
+id==1.5.0
+einops==0.8.1
+rich==13.9.4
+editdistance==0.8.1
+trouting==0.3.3
+mypy==1.3.0
+distro==1.9.0
+cycler==0.12.1
+s3transfer==0.13.0
+pyarrow==20.0.0
+sentry-sdk==2.33.0
+toml==0.10.2
+GitPython==3.1.44
+click==8.2.1
+ai2-molmo==0.0.0
+tensorflow-datasets==4.9.9
+tqdm==4.67.1
+safetensors==0.5.3
+google-auth==2.40.3
+iniconfig==2.1.0
+smmap==5.0.2
+nvidia-cublas-cu12==12.6.4.1
+wheel==0.45.1
+pluggy==1.6.0
+filelock==3.18.0
+google-auth-oauthlib==1.2.2
+httpx==0.28.1
+pyasn1==0.6.1
+typing-inspection==0.4.1
+pydub==0.25.1
+nvidia-cusparselt-cu12==0.6.3
+decorator==5.2.1
+frozenlist==1.7.0
+backports.tarfile==1.2.0
+setuptools==78.1.1
+keras==2.15.0
+MarkupSafe==3.0.2
+libclang==18.1.1
+necessary==0.4.3
+cached_path==1.7.3
+Jinja2==3.1.6
+beaker-py==2.4.6
+requests-toolbelt==1.0.0
+contourpy==1.3.2
+python-multipart==0.0.20
+nvidia-cusolver-cu12==11.7.1.2
+sentencepiece==0.2.0
+cffi==1.17.1
+gast==0.6.0
+nvidia-curand-cu12==10.3.7.77
+six==1.17.0
+boto3==1.39.5
+annotated-types==0.7.0
+termcolor==3.1.0
+pyparsing==3.2.3
+pydantic_core==2.33.2
+rfc3986==2.0.0
+pydantic==2.11.7
+jiter==0.10.0
+jaraco.context==6.0.1
+importlib_resources==6.5.2
+typeguard==2.13.3
+nltk==3.9.1
+opt_einsum==3.4.0
+grpcio==1.73.1
+pathspec==0.12.1
+triton==3.3.1
+gradio==5.37.0
+nvidia-nccl-cu12==2.26.2
+smart_open==7.3.0.post1
+scikit-learn==1.7.0
+pytest-sphinx==0.6.3
+google-cloud-core==2.4.3
+rsa==4.9.1
+readme_renderer==44.0
+httpcore==1.0.9
+threadpoolctl==3.6.0
+fsspec==2023.9.2
+keyring==25.6.0
+aiohttp==3.12.14
+attrs==25.3.0
+mypy_extensions==1.1.0
+antlr4-python3-runtime==4.9.3
+orjson==3.11.0
+cachetools==5.5.2
+requests==2.32.4
+smashed==0.21.5
+nvidia-cuda-cupti-cu12==12.6.80
+aiosignal==1.4.0
+torch==2.7.1
+xxhash==3.5.0
+OpenEXR==3.3.4
+cached-property==2.0.1
+wcwidth==0.2.13
+uvicorn==0.35.0
+numpy==1.26.4
+flatbuffers==25.2.10
+torchmetrics==1.7.4
+ftfy==6.3.1
+dill==0.3.8
+multidict==6.6.3
+shellingham==1.5.4
+promise==2.3
+anyio==4.9.0
+regex==2024.11.6
+tensorflow-graphics==2021.12.3
+pycparser==2.22
+nvidia-cuda-runtime-cu12==12.6.77
+matplotlib==3.10.3
+google-cloud-storage==2.19.0
+botocore==1.39.5
+click-option-group==0.5.7
+pytz==2025.2
+fonttools==4.58.5
+pyproject_hooks==1.2.0
+google-crc32c==1.7.1
+Markdown==3.8.2
+requests-oauthlib==2.0.0
+dlimp==0.0.1
+nvidia-cuda-nvrtc-cu12==12.6.77
+tomlkit==0.13.3
+safehttpx==0.1.6
+nvidia-cufft-cu12==11.3.0.4
+jmespath==1.0.1
+protobuf==4.21.12
+nvidia-nvtx-cu12==12.6.77
+tensorflow-addons==0.23.0
+jaraco.classes==3.4.0
+aiofiles==24.1.0
+exceptiongroup==1.3.0
+libcst==1.8.2
+immutabledict==4.2.1
+oauthlib==3.3.1
+timm==1.0.16
+lightning-utilities==0.14.3
+jeepney==0.9.0
+certifi==2025.7.14
+docutils==0.21.2
+markdown-it-py==3.0.0
+proto-plus==1.26.1
+wrapt==1.14.1
+mdurl==0.1.2
+more-itertools==10.7.0
+platformdirs==4.3.8
+kiwisolver==1.4.8
+tensorboard-data-server==0.7.2
+propcache==0.3.2
+zipp==3.23.0
+Werkzeug==3.1.3
+wandb==0.21.0
+Levenshtein==0.27.1
+isort==5.12.0
+simple-parsing==0.1.7
+face==24.0.0
+tzdata==2025.2
+h5py==3.14.0
+psutil==7.0.0
+mpmath==1.3.0
+dm-tree==0.1.9
+tensorflow-estimator==2.15.0
+pip==25.1
+joblib==1.5.1
+pyasn1_modules==0.4.2
+RapidFuzz==3.13.0
+cryptography==45.0.5
+typing_extensions==4.14.1
+urllib3==2.5.0
+sniffio==1.3.1
+omegaconf==2.3.0
+h11==0.16.0
+nvidia-cusparse-cu12==12.5.4.2
+click-help-colors==0.9.4
+einops-exts==0.0.4
+scipy==1.15.3
+nvidia-cufile-cu12==1.11.1.6
+huggingface-hub==0.33.4
+PyYAML==6.0.2
+pandas==2.3.1
+python-dateutil==2.9.0.post0
+petname==2.6
+tensorboard==2.15.2
+python-Levenshtein==0.27.1
+graphviz==0.21
+datasets==4.0.0
+gradio_client==1.10.4
+networkx==3.4.2
+nh3==0.2.22
+jaraco.functools==4.2.1
+ml-dtypes==0.2.0
+gitdb==4.0.12
+google-pasta==0.2.0
+openai==1.96.1
+aiohappyeyeballs==2.6.1
+docstring_parser==0.16
+build==1.2.2.post1
+tokenizers==0.21.2
+black==23.12.1
+fiddle==0.3.0
+requirements-parser==0.13.0
+ruff==0.12.3
+trimesh==4.7.0
+Pygments==2.19.2
+google-resumable-media==2.7.2
+async-timeout==5.0.1
+hf-xet==1.1.5
+groovy==0.1.2
+accelerate==1.8.1
+diffusers==0.33.1
+SecretStorage==3.3.3
+ffmpy==0.6.0
+websockets==15.0.1
+fastapi==0.116.1
+importlib_metadata==8.7.0
+pillow==11.3.0
+nvidia-nvjitlink-cu12==12.6.85
+beaker-gantry==2.8.3
+glom==24.11.0
+transformers==4.53.2
+pytest==8.4.1
+tensorflow==2.15.0
+semantic-version==2.10.0
+charset-normalizer==3.4.2
+tensorflow-metadata==1.17.2
+googleapis-common-protos==1.70.0
+boltons==25.0.0
+twine==6.1.0
+tomli==2.2.1
+google-api-core==2.25.1
+typer==0.16.0
+sympy==1.14.0
+starlette==0.47.1
+msgspec==0.19.0
+array_record==0.7.2
+torchvision==0.22.1
+typing_extensions==4.12.2
+wheel==0.45.1
+backports.tarfile==1.2.0
+inflect==7.3.1
+zipp==3.19.2
+jaraco.collections==5.1.0
+jaraco.text==3.12.1
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+tomli==2.0.1
+platformdirs==4.2.2
+packaging==24.2
+typeguard==4.3.0
+importlib_metadata==8.0.0
+more-itertools==10.3.0
+autocommand==2.2.2

wandb/offline-run-20250718_163819-kujhudzv/files/wandb-metadata.json

@@ -0,0 +1 @@
+{"os": "Linux-5.15.0-142-generic-x86_64-with-glibc2.35", "python": "CPython 3.10.16", "started_at": "2025-07-18T08:38:19.403312Z", "args": ["--save_folder=/mnt/data/zhangjian/a1/libero_spatial_dit_action", "--save_overwrite"], "program": "/home/zhangjian/A1/launch_scripts/train_vla_dit_action.py", "code_path": "launch_scripts/train_vla_dit_action.py", "code_path_local": "launch_scripts/train_vla_dit_action.py", "git": {"remote_url": "https://github.com/Spatialtemporal-AI/A1.git", "commit": "2e05e206652258f5fbb178ec13c792c2a91c1856"}, "root": "/home/zhangjian/A1", "host": "iv-ydwxzrkf0gvr6okuck21", "executable": "/mnt/data/zhangjian/miniconda3/envs/a1/bin/python3.10", "cpu_count": 56, "cpu_count_logical": 112, "gpu_type": "NVIDIA A100-SXM4-80GB", "gpu_count": 8, "disk": {"/": {"total": "507024936960", "used": "133188997120"}}, "memory": {"total": "2071488372736"}, "gpu_nvidia": [{"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-b01b938f-1b95-8587-4b93-cfa68d84c935"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-49b283f0-f1e6-992d-4ed8-99855a5e6529"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-a340dca6-1db9-f578-76f2-cf36f74ffabd"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-c1bddabf-aa88-9c8f-eec2-6dc3e117f985"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-4096c1e7-0437-6727-096f-d67141c140b8"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-eeb896c0-61d6-8380-778b-d02101078a17"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-d5c1ed79-76e3-22a6-8908-df289b090c03"}, {"name": "NVIDIA A100-SXM4-80GB", "memory_total": "85899345920", "cuda_cores": 6912, "architecture": "Ampere", "uuid": "GPU-6d90b060-f6f7-599b-f939-4ba124ff11b7"}], "cuda_version": "12.8", "writer_id": "oygl393zcukmmdj0i9c4b0424qwzehyz"}

wandb/offline-run-20250718_163819-kujhudzv/files/wandb-summary.json

@@ -0,0 +1 @@
+{"_runtime": 9}

wandb/offline-run-20250718_163819-kujhudzv/logs/debug-core.log

@@ -0,0 +1,14 @@
+{"time":"2025-07-18T16:38:19.433873148+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpptx2rkja/port-2907721.txt","pid":2907721,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-07-18T16:38:19.434389567+08:00","level":"INFO","msg":"server: will exit if parent process dies","ppid":2907721}
+{"time":"2025-07-18T16:38:19.434364202+08:00","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-2907721-2910040-479139429/socket","Net":"unix"}}
+{"time":"2025-07-18T16:38:19.611743485+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-07-18T16:38:19.616099563+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"kujhudzv","id":"1(@)"}
+{"time":"2025-07-18T16:38:19.735468949+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"kujhudzv","id":"1(@)"}
+{"time":"2025-07-18T16:38:28.751051088+08:00","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2025-07-18T16:38:28.751139005+08:00","level":"INFO","msg":"server is shutting down"}
+{"time":"2025-07-18T16:38:28.751126778+08:00","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2025-07-18T16:38:28.751218178+08:00","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2025-07-18T16:38:28.751229032+08:00","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-2907721-2910040-479139429/socket","Net":"unix"}}
+{"time":"2025-07-18T16:38:28.753347379+08:00","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2025-07-18T16:38:28.753373097+08:00","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2025-07-18T16:38:28.753384353+08:00","level":"INFO","msg":"server is closed"}

wandb/offline-run-20250718_163819-kujhudzv/logs/debug-internal.log

@@ -0,0 +1,13 @@
+{"time":"2025-07-18T16:38:19.620235579+08:00","level":"INFO","msg":"stream: starting","core version":"0.21.0"}
+{"time":"2025-07-18T16:38:19.735259373+08:00","level":"WARN","msg":"GraphQL client is nil, skipping feature loading"}
+{"time":"2025-07-18T16:38:19.735439442+08:00","level":"INFO","msg":"stream: created new stream","id":"kujhudzv"}
+{"time":"2025-07-18T16:38:19.735463779+08:00","level":"INFO","msg":"stream: started","id":"kujhudzv"}
+{"time":"2025-07-18T16:38:19.735487143+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"kujhudzv"}
+{"time":"2025-07-18T16:38:19.7355104+08:00","level":"INFO","msg":"sender: started","stream_id":"kujhudzv"}
+{"time":"2025-07-18T16:38:19.735512981+08:00","level":"INFO","msg":"handler: started","stream_id":"kujhudzv"}
+{"time":"2025-07-18T16:38:19.737472829+08:00","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}
+{"time":"2025-07-18T16:38:28.75123625+08:00","level":"INFO","msg":"stream: closing","id":"kujhudzv"}
+{"time":"2025-07-18T16:38:28.751382942+08:00","level":"INFO","msg":"handler: closed","stream_id":"kujhudzv"}
+{"time":"2025-07-18T16:38:28.751390848+08:00","level":"INFO","msg":"writer: Close: closed","stream_id":"kujhudzv"}
+{"time":"2025-07-18T16:38:28.751399279+08:00","level":"INFO","msg":"sender: closed","stream_id":"kujhudzv"}
+{"time":"2025-07-18T16:38:28.751516927+08:00","level":"INFO","msg":"stream: closed","id":"kujhudzv"}

wandb/offline-run-20250718_164033-dnuid9t7/files/requirements.txt

@@ -0,0 +1,262 @@
+ai2-molmo==0.0.0
+yarl==1.20.1
+astunparse==1.6.3
+nvidia-cudnn-cu12==9.5.1.17
+etils==1.13.0
+tensorflow-io-gcs-filesystem==0.37.1
+Brotli==1.1.0
+gcsfs==2023.9.2
+multiprocess==0.70.16
+packaging==25.0
+idna==3.10
+absl-py==2.3.1
+id==1.5.0
+einops==0.8.1
+rich==13.9.4
+editdistance==0.8.1
+trouting==0.3.3
+mypy==1.3.0
+distro==1.9.0
+cycler==0.12.1
+s3transfer==0.13.0
+pyarrow==20.0.0
+sentry-sdk==2.33.0
+toml==0.10.2
+GitPython==3.1.44
+click==8.2.1
+ai2-molmo==0.0.0
+tensorflow-datasets==4.9.9
+tqdm==4.67.1
+safetensors==0.5.3
+google-auth==2.40.3
+iniconfig==2.1.0
+smmap==5.0.2
+nvidia-cublas-cu12==12.6.4.1
+wheel==0.45.1
+pluggy==1.6.0
+filelock==3.18.0
+google-auth-oauthlib==1.2.2
+httpx==0.28.1
+pyasn1==0.6.1
+typing-inspection==0.4.1
+pydub==0.25.1
+nvidia-cusparselt-cu12==0.6.3
+decorator==5.2.1
+frozenlist==1.7.0
+backports.tarfile==1.2.0
+setuptools==78.1.1
+keras==2.15.0
+MarkupSafe==3.0.2
+libclang==18.1.1
+necessary==0.4.3
+cached_path==1.7.3
+Jinja2==3.1.6
+beaker-py==2.4.6
+requests-toolbelt==1.0.0
+contourpy==1.3.2
+python-multipart==0.0.20
+nvidia-cusolver-cu12==11.7.1.2
+sentencepiece==0.2.0
+cffi==1.17.1
+gast==0.6.0
+nvidia-curand-cu12==10.3.7.77
+six==1.17.0
+boto3==1.39.5
+annotated-types==0.7.0
+termcolor==3.1.0
+pyparsing==3.2.3
+pydantic_core==2.33.2
+rfc3986==2.0.0
+pydantic==2.11.7
+jiter==0.10.0
+jaraco.context==6.0.1
+importlib_resources==6.5.2
+typeguard==2.13.3
+nltk==3.9.1
+opt_einsum==3.4.0
+grpcio==1.73.1
+pathspec==0.12.1
+triton==3.3.1
+gradio==5.37.0
+nvidia-nccl-cu12==2.26.2
+smart_open==7.3.0.post1
+scikit-learn==1.7.0
+pytest-sphinx==0.6.3
+google-cloud-core==2.4.3
+rsa==4.9.1
+readme_renderer==44.0
+httpcore==1.0.9
+threadpoolctl==3.6.0
+fsspec==2023.9.2
+keyring==25.6.0
+aiohttp==3.12.14
+attrs==25.3.0
+mypy_extensions==1.1.0
+antlr4-python3-runtime==4.9.3
+orjson==3.11.0
+cachetools==5.5.2
+requests==2.32.4
+smashed==0.21.5
+nvidia-cuda-cupti-cu12==12.6.80
+aiosignal==1.4.0
+torch==2.7.1
+xxhash==3.5.0
+OpenEXR==3.3.4
+cached-property==2.0.1
+wcwidth==0.2.13
+uvicorn==0.35.0
+numpy==1.26.4
+flatbuffers==25.2.10
+torchmetrics==1.7.4
+ftfy==6.3.1
+dill==0.3.8
+multidict==6.6.3
+shellingham==1.5.4
+promise==2.3
+anyio==4.9.0
+regex==2024.11.6
+tensorflow-graphics==2021.12.3
+pycparser==2.22
+nvidia-cuda-runtime-cu12==12.6.77
+matplotlib==3.10.3
+google-cloud-storage==2.19.0
+botocore==1.39.5
+click-option-group==0.5.7
+pytz==2025.2
+fonttools==4.58.5
+pyproject_hooks==1.2.0
+google-crc32c==1.7.1
+Markdown==3.8.2
+requests-oauthlib==2.0.0
+dlimp==0.0.1
+nvidia-cuda-nvrtc-cu12==12.6.77
+tomlkit==0.13.3
+safehttpx==0.1.6
+nvidia-cufft-cu12==11.3.0.4
+jmespath==1.0.1
+protobuf==4.21.12
+nvidia-nvtx-cu12==12.6.77
+tensorflow-addons==0.23.0
+jaraco.classes==3.4.0
+aiofiles==24.1.0
+exceptiongroup==1.3.0
+libcst==1.8.2
+immutabledict==4.2.1
+oauthlib==3.3.1
+timm==1.0.16
+lightning-utilities==0.14.3
+jeepney==0.9.0
+certifi==2025.7.14
+docutils==0.21.2
+markdown-it-py==3.0.0
+proto-plus==1.26.1
+wrapt==1.14.1
+mdurl==0.1.2
+more-itertools==10.7.0
+platformdirs==4.3.8
+kiwisolver==1.4.8
+tensorboard-data-server==0.7.2
+propcache==0.3.2
+zipp==3.23.0
+Werkzeug==3.1.3
+wandb==0.21.0
+Levenshtein==0.27.1
+isort==5.12.0
+simple-parsing==0.1.7
+face==24.0.0
+tzdata==2025.2
+h5py==3.14.0
+psutil==7.0.0
+mpmath==1.3.0
+dm-tree==0.1.9
+tensorflow-estimator==2.15.0
+pip==25.1
+joblib==1.5.1
+pyasn1_modules==0.4.2
+RapidFuzz==3.13.0
+cryptography==45.0.5
+typing_extensions==4.14.1
+urllib3==2.5.0
+sniffio==1.3.1
+omegaconf==2.3.0
+h11==0.16.0
+nvidia-cusparse-cu12==12.5.4.2
+click-help-colors==0.9.4
+einops-exts==0.0.4
+scipy==1.15.3
+nvidia-cufile-cu12==1.11.1.6
+huggingface-hub==0.33.4
+PyYAML==6.0.2
+pandas==2.3.1
+python-dateutil==2.9.0.post0
+petname==2.6
+tensorboard==2.15.2
+python-Levenshtein==0.27.1
+graphviz==0.21
+datasets==4.0.0
+gradio_client==1.10.4
+networkx==3.4.2
+nh3==0.2.22
+jaraco.functools==4.2.1
+ml-dtypes==0.2.0
+gitdb==4.0.12
+google-pasta==0.2.0
+openai==1.96.1
+aiohappyeyeballs==2.6.1
+docstring_parser==0.16
+build==1.2.2.post1
+tokenizers==0.21.2
+black==23.12.1
+fiddle==0.3.0
+requirements-parser==0.13.0
+ruff==0.12.3
+trimesh==4.7.0
+Pygments==2.19.2
+google-resumable-media==2.7.2
+async-timeout==5.0.1
+hf-xet==1.1.5
+groovy==0.1.2
+accelerate==1.8.1
+diffusers==0.33.1
+SecretStorage==3.3.3
+ffmpy==0.6.0
+websockets==15.0.1
+fastapi==0.116.1
+importlib_metadata==8.7.0
+pillow==11.3.0
+nvidia-nvjitlink-cu12==12.6.85
+beaker-gantry==2.8.3
+glom==24.11.0
+transformers==4.53.2
+pytest==8.4.1
+tensorflow==2.15.0
+semantic-version==2.10.0
+charset-normalizer==3.4.2
+tensorflow-metadata==1.17.2
+googleapis-common-protos==1.70.0
+boltons==25.0.0
+twine==6.1.0
+tomli==2.2.1
+google-api-core==2.25.1
+typer==0.16.0
+sympy==1.14.0
+starlette==0.47.1
+msgspec==0.19.0
+array_record==0.7.2
+torchvision==0.22.1
+typing_extensions==4.12.2
+wheel==0.45.1
+backports.tarfile==1.2.0
+inflect==7.3.1
+zipp==3.19.2
+jaraco.collections==5.1.0
+jaraco.text==3.12.1
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+tomli==2.0.1
+platformdirs==4.2.2
+packaging==24.2
+typeguard==4.3.0
+importlib_metadata==8.0.0
+more-itertools==10.3.0
+autocommand==2.2.2

wandb/offline-run-20250718_164033-dnuid9t7/logs/debug-core.log

@@ -0,0 +1,7 @@
+{"time":"2025-07-18T16:40:33.178518972+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpz_pgdeu_/port-2913404.txt","pid":2913404,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2025-07-18T16:40:33.179042083+08:00","level":"INFO","msg":"server: will exit if parent process dies","ppid":2913404}
+{"time":"2025-07-18T16:40:33.179087173+08:00","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-2913404-2915307-3950646465/socket","Net":"unix"}}
+{"time":"2025-07-18T16:40:33.359443846+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2025-07-18T16:40:33.362673814+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"dnuid9t7","id":"1(@)"}
+{"time":"2025-07-18T16:40:33.478718975+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"dnuid9t7","id":"1(@)"}
+{"time":"2025-07-18T16:40:43.09952199+08:00","level":"INFO","msg":"server: parent process exited, terminating service process"}

wandb/offline-run-20250718_164033-dnuid9t7/logs/debug-internal.log

@@ -0,0 +1,8 @@
+{"time":"2025-07-18T16:40:33.366115286+08:00","level":"INFO","msg":"stream: starting","core version":"0.21.0"}
+{"time":"2025-07-18T16:40:33.478534545+08:00","level":"WARN","msg":"GraphQL client is nil, skipping feature loading"}
+{"time":"2025-07-18T16:40:33.478686065+08:00","level":"INFO","msg":"stream: created new stream","id":"dnuid9t7"}
+{"time":"2025-07-18T16:40:33.47871135+08:00","level":"INFO","msg":"stream: started","id":"dnuid9t7"}
+{"time":"2025-07-18T16:40:33.478722952+08:00","level":"INFO","msg":"writer: Do: started","stream_id":"dnuid9t7"}
+{"time":"2025-07-18T16:40:33.478744975+08:00","level":"INFO","msg":"handler: started","stream_id":"dnuid9t7"}
+{"time":"2025-07-18T16:40:33.478739306+08:00","level":"INFO","msg":"sender: started","stream_id":"dnuid9t7"}
+{"time":"2025-07-18T16:40:33.47934386+08:00","level":"WARN","msg":"runupserter: server does not expand metric globs but the x_server_side_expand_glob_metrics setting is set; ignoring"}

wandb/offline-run-20250718_164558-83erbdk6/files/config.yaml

@@ -0,0 +1,227 @@
+wandb_version: 1
+
+_wandb:
+  desc: null
+  value:
+    python_version: 3.10.16
+    cli_version: 0.21.0
+    framework: huggingface
+    huggingface_version: 4.53.2
+    is_jupyter_run: false
+    is_kaggle_kernel: false
+    start_time: 1752828358
+    t:
+      1:
+      - 1
+      - 2
+      - 3
+      - 5
+      - 11
+      - 41
+      - 49
+      - 51
+      - 53
+      - 63
+      - 71
+      - 83
+      - 95
+      - 105
+      2:
+      - 1
+      - 2
+      - 3
+      - 5
+      - 11
+      - 41
+      - 49
+      - 51
+      - 53
+      - 63
+      - 71
+      - 83
+      - 95
+      - 105
+      3:
+      - 4
+      - 13
+      - 16
+      - 42
+      4: 3.10.16
+      5: 0.21.0
+      6: 4.53.2
+      13: linux-x86_64
+    e:
+      hzn18k6kjb1kzh9bnw21xqy15pw51pao:
+        os: Linux-5.15.0-142-generic-x86_64-with-glibc2.35
+        python: CPython 3.10.16
+        started_at: '2025-07-18T08:45:58.703240Z'
+        args:
+        - --save_folder=/mnt/data/zhangjian/a1/libero_spatial_dit_action
+        - --save_overwrite
+        program: /home/zhangjian/A1/launch_scripts/train_vla_dit_action.py
+        code_path: launch_scripts/train_vla_dit_action.py
+        code_path_local: launch_scripts/train_vla_dit_action.py
+        git:
+          remote_url: https://github.com/Spatialtemporal-AI/A1.git
+          commit: 2e05e206652258f5fbb178ec13c792c2a91c1856
+        root: /home/zhangjian/A1
+        host: iv-ydwxzrkf0gvr6okuck21
+        executable: /mnt/data/zhangjian/miniconda3/envs/a1/bin/python3.10
+        cpu_count: 56
+        cpu_count_logical: 112
+        gpu_type: NVIDIA A100-SXM4-80GB
+        gpu_count: 8
+        disk:
+          /:
+            total: '507024936960'
+            used: '133191184384'
+        memory:
+          total: '2071488372736'
+        gpu_nvidia:
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-b01b938f-1b95-8587-4b93-cfa68d84c935
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-49b283f0-f1e6-992d-4ed8-99855a5e6529
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-a340dca6-1db9-f578-76f2-cf36f74ffabd
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-c1bddabf-aa88-9c8f-eec2-6dc3e117f985
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-4096c1e7-0437-6727-096f-d67141c140b8
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-eeb896c0-61d6-8380-778b-d02101078a17
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-d5c1ed79-76e3-22a6-8908-df289b090c03
+        - name: NVIDIA A100-SXM4-80GB
+          memory_total: '85899345920'
+          cuda_cores: 6912
+          architecture: Ampere
+          uuid: GPU-6d90b060-f6f7-599b-f939-4ba124ff11b7
+        cuda_version: '12.8'
+        writer_id: hzn18k6kjb1kzh9bnw21xqy15pw51pao
+text_model_name:
+  desc: null
+  value: google/siglip-so400m-patch14-384
+vision_model_name:
+  desc: null
+  value: google/siglip-so400m-patch14-384
+run_name:
+  desc: null
+  value: dit_action_train_20250718_164532
+save_folder:
+  desc: null
+  value: /mnt/data/zhangjian/a1/libero_spatial_dit_action
+seed:
+  desc: null
+  value: 42
+max_steps:
+  desc: null
+  value: 100000
+global_batch_size:
+  desc: null
+  value: 80
+device_batch_size:
+  desc: null
+  value: 20
+learning_rate:
+  desc: null
+  value: 2.0e-05
+weight_decay:
+  desc: null
+  value: 0.01
+warmup_steps:
+  desc: null
+  value: 500
+dataset_name:
+  desc: null
+  value: libero_spatial_no_noops
+data_root_dir:
+  desc: null
+  value: /mnt/data/zhangjian/dataset/Simulation/datasets--openvla--modified_libero_rlds
+use_wrist_image:
+  desc: null
+  value: true
+use_proprio:
+  desc: null
+  value: true
+sequence_length:
+  desc: null
+  value: 64
+num_workers:
+  desc: null
+  value: 0
+num_diffusion_steps:
+  desc: null
+  value: 1000
+num_diffusion_inference_steps:
+  desc: null
+  value: 30
+lang_cond_dim:
+  desc: null
+  value: 1152
+img_cond_dim:
+  desc: null
+  value: 1152
+num_patches:
+  desc: null
+  value: 729
+dit_hidden_dim:
+  desc: null
+  value: 2048
+dit_depth:
+  desc: null
+  value: 28
+dit_num_heads:
+  desc: null
+  value: 16
+save_interval:
+  desc: null
+  value: 1000
+eval_interval:
+  desc: null
+  value: 500
+log_interval:
+  desc: null
+  value: 10
+save_num_checkpoints_to_keep:
+  desc: null
+  value: 3
+save_overwrite:
+  desc: null
+  value: true
+precision:
+  desc: null
+  value: amp_bf16
+gradient_accumulation_steps:
+  desc: null
+  value: 1
+max_grad_norm:
+  desc: null
+  value: 1.0
+wandb_project:
+  desc: null
+  value: a1-dit-action-20.47
+wandb_entity:
+  desc: null
+  value: demo0