2025-07-01 09:21:22.729 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:22.729 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:22.738 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:22.845 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:22.845 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:22.854 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:22.986 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:22.986 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:22.995 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.434 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.434 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.443 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.601 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.601 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.610 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.664 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.664 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.669 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.669 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.669 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.669 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.673 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.677 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.678 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.703 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.704 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.713 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.713 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.713 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.714 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.714 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.723 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.723 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.777 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.777 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.786 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.874 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.874 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.874 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.875 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.882 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.883 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.890 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.890 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.892 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.892 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.898 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.899 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.899 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.900 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.900 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.900 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.907 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.909 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:23.985 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:23.985 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:23.994 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.105 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.105 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.114 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.116 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.116 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.122 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.122 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.125 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.125 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.126 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.127 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.127 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.130 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.134 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.136 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.184 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.184 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.185 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.185 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.193 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.194 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.198 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.198 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.200 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.200 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.204 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.204 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.207 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:24.210 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.212 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.236 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.236 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.237 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.237 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.238 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.238 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.238 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.238 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.239 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.239 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.242 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.242 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.245 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.246 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.246 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.246 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.247 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.251 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.293 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.293 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.300 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.300 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.301 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:24.309 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.313 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.314 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.318 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.318 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.322 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.323 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.323 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.326 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.326 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.327 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.332 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.335 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.389 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.389 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.398 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.415 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.415 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.415 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.415 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.423 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.424 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.433 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.433 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.434 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.434 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:24.442 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.443 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.447 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.447 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.453 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:24.453 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:24.456 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:24.462 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:25.570 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.570 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.570 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.570 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.578 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:25.579 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:25.638 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.638 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:25.647 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:25.663 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.663 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.663 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.663 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.664 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.664 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.671 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:25.672 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:25.672 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:25.733 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.733 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.742 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:25.768 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.768 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.773 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.773 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:25.777 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:25.782 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:25.998 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:25.998 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:26.007 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
2025-07-01 09:21:26.084 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:26.084 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:26.093 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:26.104 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:26.104 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:26.113 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:26.113 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:26.114 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:26.122 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
Did not find AutoResume SDK!
2025-07-01 09:21:26.176 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:26.176 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:26.179 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:26.179 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:26.185 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:26.185 | INFO     | llava.data.builder:register_datasets:39 - Registering datasets from environment: 'nvidia,draco-oci-iad'.
2025-07-01 09:21:26.185 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/nvidia.yaml'.
2025-07-01 09:21:26.187 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
2025-07-01 09:21:26.194 | INFO     | llava.data.builder:register_datasets:44 - Registering datasets from: '/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/data/registry/datasets/draco-oci-iad.yaml'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/timm/models/layers/__init__.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers
  warnings.warn(f"Importing from {__name__} is deprecated, please import via timm.layers", FutureWarning)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 11807.02it/s]
Did not find AutoResume SDK!
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6803.68it/s]
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
Did not find AutoResume SDK!
Did not find AutoResume SDK!
Did not find AutoResume SDK!
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6832.17it/s]
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8284.46it/s]
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/transformers/training_args.py:1559: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead
  warnings.warn(

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7078.14it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8546.52it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8853.19it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 11695.71it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9050.59it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6191.07it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9562.52it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7901.71it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6614.63it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8132.25it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8516.77it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8793.09it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7782.33it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7638.57it/s]

Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8159.37it/s]

Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9096.39it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9622.07it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7140.11it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7884.74it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9211.50it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9904.46it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9439.54it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 11292.36it/s]

Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6118.39it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6314.91it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8726.88it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8063.02it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7311.40it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6426.41it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7663.16it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9087.01it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9094.52it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9883.35it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7625.35it/s]

Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 10399.10it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7001.62it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 11936.63it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7871.35it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9463.89it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8192.00it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7482.83it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 10866.07it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7374.45it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7242.26it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7997.86it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 11527.34it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8791.33it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9718.68it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8065.23it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7540.48it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8852.30it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7103.26it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7242.86it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8109.79it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9819.44it/s]

Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 6898.53it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 7839.82it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 9824.92it/s]

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8433.59it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Fetching 21 files:   0%|          | 0/21 [00:00<?, ?it/s]
Fetching 21 files: 100%|██████████| 21/21 [00:00<00:00, 8805.40it/s]
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.
You are attempting to use Flash Attention 2.0 with a model not initialized on GPU. Make sure to move the model to GPU after initializing it on CPU with `model.to('cuda')`.

Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.55s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.56s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.56s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.56s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.56s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.56s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.59s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.61s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.57s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.57s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.59s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.57s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.57s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.57s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.67s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.54s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.49s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.62s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.64s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.65s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.64s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.53s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.59s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.61s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.61s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.53s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.59s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.65s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.65s/it]
Loading checkpoint shards:  2
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.46s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:08,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:08,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.62s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:08,  2.75s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.68s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████5%|██▌       | 1/4 [00:02<00:08,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.60s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.60s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.60s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.42s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.42s/it]
Loading checkpoint shards:  75%|███████5%|██▌       | 1/4 [00:02<00:07,  2.55s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.53s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.58s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.41s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.42s/it]
Loading checkpoint shards:  75%|███████5%|██▌       | 1/4 [00:02<00:07,  2.62s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.63s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████5%|██▌       | 1/4 [00:02<00:08,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.66s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.53s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.60s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.41s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.50s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.86s/it]
5%|██▌       | 1/4 [00:02<00:07,  2.57s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:08,  2.67s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:04<00:04,  2.48s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.55s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.39s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.49s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.83s/it]
5%|██▌       | 1/4 [00:02<00:08,  2.67s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:08,  2.67s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.52s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.60s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.60s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.41s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.51s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.86s/it]
5%|██▌       | 1/4 [00:02<00:07,  2.58s/it]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:02<00:07,  2.64s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.53s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.51s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.57s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.58s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.59s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.58s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.56s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:05<00:05,  2.58s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.42s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.41s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.52s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.88s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.52s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.88s/it]
▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.42s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.46s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.51s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.88s/it]
▌  | 3/4 [00:07<00:02,  2.44s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.53s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.88s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.53s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.53s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.87s/it]
▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.44s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.44s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.53s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]
▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.42s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.42s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.53s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.89s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.54s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.55s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.44s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.44s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.55s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.55s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.90s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.56s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.91s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.57s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.91s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.57s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.92s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.58s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.93s/it]

Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.45s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.43s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.58s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.93s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.58s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.92s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.58s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.93s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.58s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.93s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]


Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.58s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.93s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.58s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.91s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.59s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:07<00:00,  1.94s/it]

Loading checkpoint shards:  75%|███████▌  | 3/4 [00:07<00:02,  2.53s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:08<00:00,  1.80s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:08<00:00,  2.11s/it]
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
2025-07-01 09:24:26.571 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:26.585 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:26.618 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:26.627 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:26.692 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:26.695 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:26.712 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:26.853 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:26.904 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:26.913 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:26.944 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:26.964 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.034 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.048 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.066 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.081 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.114 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.144 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.154 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.163 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.172 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.175 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.196 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.197 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.201 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.214 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.215 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.222 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.222 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.237 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.238 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.244 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.247 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.253 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.264 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.266 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.271 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.282 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.284 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.309 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.309 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.316 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.321 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.324 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.332 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.335 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.341 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.357 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.357 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.368 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.370 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.371 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.386 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.392 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.392 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.395 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.401 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.429 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
2025-07-01 09:24:27.433 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.437 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.440 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.444 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.446 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
2025-07-01 09:24:27.447 | WARNING  | llava.data.builder:build_dataset:92 - Training VILA with mixture 'vila_quantumn_qa_train'.
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/train/llava_trainer.py:592: FutureWarning: `tokenizer` is deprecated and will be removed in version 5.0.0 for `LLaVATrainer.__init__`. Use `processing_class` instead.
  super().__init__(*args, **kwargs)
wandb: Currently logged in as: ligeng-zhu to https://api.wandb.ai. Use `wandb login --relogin` to force relogin
wandb: creating run
wandb: Tracking run with wandb version 0.20.1
wandb: Run data is saved locally in runs/train/NVILA-Lite-8B-quantumn-qa-train/wandb/run-20250701_092441-NVILA-Lite-8B-quantumn-qa-train
wandb: Run `wandb offline` to turn off syncing.
wandb: Resuming run NVILA-Lite-8B-quantumn-qa-train
wandb: ⭐️ View project at https://wandb.ai/ligeng-zhu/vila
wandb: 🚀 View run at https://wandb.ai/ligeng-zhu/vila/runs/NVILA-Lite-8B-quantumn-qa-train
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")

  0%|          | 0/13 [00:00<?, ?it/s]/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: c10d::broadcast_: an autograd kernel was not registered to the Autograd key(s) but we are trying to backprop through it. This may lead to silently incorrect behavior. This behavior is deprecated and will be removed in a future version of PyTorch. If your operator is differentiable, please ensure you have registered an autograd kernel to the correct Autograd key (e.g. DispatchKey::Autograd, DispatchKey::CompositeImplicitAutograd). If your operator is not differentiable, or to squash this warning and use the previous behavior, please register torch::CppFunction::makeFallthrough() to DispatchKey::Autograd. (Triggered internally at ../torch/csrc/autograd/autograd_not_implemented_fallback.cpp:63.)
  return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
Could not estimate the number of tokens of the input, floating-point operations will not be computed
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/llava/model/llava_arch.py:534: UserWarning: Truncating sequences to `model_max_length` (4096).
  warnings.warn(f"Truncating sequences to `model_max_length` ({self.tokenizer.model_max_length}).")

  8%|▊         | 1/13 [00:30<06:05, 30.48s/it]
                                              

  8%|▊         | 1/13 [00:30<06:05, 30.48s/it]
 15%|█▌        | 2/13 [00:51<04:34, 24.93s/it]
                                              

 15%|█▌        | 2/13 [00:51<04:34, 24.93s/it]
 23%|██▎       | 3/13 [01:13<03:55, 23.58s/it]
                                              

 23%|██▎       | 3/13 [01:13<03:55, 23.58s/it]
 31%|███       | 4/13 [01:42<03:52, 25.78s/it]
                                              

 31%|███       | 4/13 [01:42<03:52, 25.78s/it]
 38%|███▊      | 5/13 [02:04<03:14, 24.32s/it]
                                              

 38%|███▊      | 5/13 [02:04<03:14, 24.32s/it]
 46%|████▌     | 6/13 [02:22<02:34, 22.06s/it]
                                              

 46%|████▌     | 6/13 [02:22<02:34, 22.06s/it]
 54%|█████▍    | 7/13 [02:40<02:04, 20.81s/it]
                                              

 54%|█████▍    | 7/13 [02:40<02:04, 20.81s/it]
 62%|██████▏   | 8/13 [03:03<01:48, 21.67s/it]
                                              

 62%|██████▏   | 8/13 [03:03<01:48, 21.67s/it]
 69%|██████▉   | 9/13 [03:28<01:29, 22.48s/it]
                                              

 69%|██████▉   | 9/13 [03:28<01:29, 22.48s/it]
 77%|███████▋  | 10/13 [03:47<01:04, 21.58s/it]
                                               

 77%|███████▋  | 10/13 [03:47<01:04, 21.58s/it]
 85%|████████▍ | 11/13 [04:09<00:43, 21.52s/it]
                                               

 85%|████████▍ | 11/13 [04:09<00:43, 21.52s/it]
 92%|█████████▏| 12/13 [04:30<00:21, 21.57s/it]
                                               

 92%|█████████▏| 12/13 [04:30<00:21, 21.57s/it]
100%|██████████| 13/13 [04:53<00:00, 22.01s/it]
                                               

100%|██████████| 13/13 [04:53<00:00, 22.01s/it]/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(
/lustre/fs1/portfolios/nvr/projects/nvr_elm_llm/users/ligengz/workspace/vila-q/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py:1898: UserWarning: Positional args are being deprecated, use kwargs instead. Refer to https://pytorch.org/docs/master/generated/torch.nn.Module.html#torch.nn.Module.state_dict for details.
  warnings.warn(

                                               

100%|██████████| 13/13 [05:19<00:00, 22.01s/it]
100%|██████████| 13/13 [05:19<00:00, 24.60s/it]