Delete stage2_pth
e714a0a verified - 1.52 kB initial commit
bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt Detected Pickle imports (7)
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "mmengine.config.config.ConfigDict",
- "torch.FloatStorage",
- "deepspeed.runtime.fp16.loss_scaler.LossScaler",
- "deepspeed.utils.tensor_fragment.fragment_address",
- "deepspeed.runtime.zero.config.ZeroStageEnum"
How to fix it?
22.9 GB Add files using upload-large-folder tool bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt Detected Pickle imports (7)
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "mmengine.config.config.ConfigDict",
- "torch.FloatStorage",
- "deepspeed.runtime.fp16.loss_scaler.LossScaler",
- "deepspeed.utils.tensor_fragment.fragment_address",
- "deepspeed.runtime.zero.config.ZeroStageEnum"
How to fix it?
22.9 GB Add files using upload-large-folder tool bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt Detected Pickle imports (7)
- "collections.OrderedDict",
- "torch._utils._rebuild_tensor_v2",
- "mmengine.config.config.ConfigDict",
- "torch.FloatStorage",
- "deepspeed.runtime.fp16.loss_scaler.LossScaler",
- "deepspeed.utils.tensor_fragment.fragment_address",
- "deepspeed.runtime.zero.config.ZeroStageEnum"
How to fix it?
22.9 GB Add files using upload-large-folder tool bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt Detected Pickle imports (7)
- "collections.OrderedDict",
- "deepspeed.utils.tensor_fragment.fragment_address",
- "torch._utils._rebuild_tensor_v2",
- "torch.FloatStorage",
- "deepspeed.runtime.fp16.loss_scaler.LossScaler",
- "deepspeed.runtime.zero.config.ZeroStageEnum",
- "mmengine.config.config.ConfigDict"
How to fix it?
22.9 GB Add files using upload-large-folder tool mp_rank_00_model_states.pt Detected Pickle imports (13)
- "__builtin__.set",
- "collections.OrderedDict",
- "_codecs.encode",
- "torch._utils._rebuild_tensor_v2",
- "mmengine.config.config.ConfigDict",
- "numpy._core.multiarray._reconstruct",
- "numpy.dtype",
- "torch.Size",
- "numpy.ndarray",
- "__builtin__.getattr",
- "torch.BFloat16Storage",
- "numpy._core.multiarray.scalar",
- "mmengine.logging.history_buffer.HistoryBuffer"
How to fix it?
15.3 GB Add files using upload-large-folder tool
