Add README with W&B link
536b974 verified - experiment_cfg tool_insert checkpoint at step 5000 (full trainer state for resume)
- global_step5000 tool_insert checkpoint at step 5000 (full trainer state for resume)
- 1.52 kB initial commit
- 483 Bytes Add README with W&B link
- 2.15 kB tool_insert checkpoint at step 5000 (full trainer state for resume)
- 2.23 kB tool_insert checkpoint at step 5000 (full trainer state for resume)
- 15 Bytes tool_insert checkpoint at step 5000 (full trainer state for resume)
- 4.99 GB tool_insert checkpoint at step 5000 (full trainer state for resume)
- 1.92 GB tool_insert checkpoint at step 5000 (full trainer state for resume)
- 105 kB tool_insert checkpoint at step 5000 (full trainer state for resume)
- 26.6 kB tool_insert checkpoint at step 5000 (full trainer state for resume)
rng_state_0.pth Detected Pickle imports (7)
- "numpy.dtype",
- "torch.ByteStorage",
- "torch._utils._rebuild_tensor_v2",
- "numpy.core.multiarray._reconstruct",
- "numpy.ndarray",
- "collections.OrderedDict",
- "_codecs.encode"
How to fix it?
15.4 kB tool_insert checkpoint at step 5000 (full trainer state for resume) rng_state_1.pth Detected Pickle imports (7)
- "numpy.dtype",
- "collections.OrderedDict",
- "torch.ByteStorage",
- "numpy.core.multiarray._reconstruct",
- "numpy.ndarray",
- "_codecs.encode",
- "torch._utils._rebuild_tensor_v2"
How to fix it?
15.4 kB tool_insert checkpoint at step 5000 (full trainer state for resume) rng_state_2.pth Detected Pickle imports (7)
- "collections.OrderedDict",
- "numpy.core.multiarray._reconstruct",
- "torch.ByteStorage",
- "torch._utils._rebuild_tensor_v2",
- "numpy.dtype",
- "_codecs.encode",
- "numpy.ndarray"
How to fix it?
15.4 kB tool_insert checkpoint at step 5000 (full trainer state for resume) rng_state_3.pth Detected Pickle imports (7)
- "_codecs.encode",
- "numpy.dtype",
- "collections.OrderedDict",
- "numpy.core.multiarray._reconstruct",
- "torch.ByteStorage",
- "torch._utils._rebuild_tensor_v2",
- "numpy.ndarray"
How to fix it?
15.4 kB tool_insert checkpoint at step 5000 (full trainer state for resume) - 1.47 kB tool_insert checkpoint at step 5000 (full trainer state for resume)
- 4.31 MB tool_insert checkpoint at step 5000 (full trainer state for resume)
- 69.2 kB tool_insert checkpoint at step 5000 (full trainer state for resume)
training_args.bin Detected Pickle imports (14)
- "torch.device",
- "accelerate.utils.dataclasses.DeepSpeedPlugin",
- "transformers.trainer_utils.SchedulerType",
- "accelerate.utils.dataclasses.DistributedType",
- "accelerate.state.PartialState",
- "transformers.trainer_utils.IntervalStrategy",
- "transformers.integrations.deepspeed.HfDeepSpeedConfig",
- "transformers.training_args.OptimizerNames",
- "transformers.trainer_pt_utils.AcceleratorConfig",
- "transformers.trainer_utils.HubStrategy",
- "torch.bfloat16",
- "transformers.trainer_utils.SaveStrategy",
- "transformers.integrations.deepspeed.HfTrainerDeepSpeedConfig",
- "transformers.training_args.TrainingArguments"
How to fix it?
7.89 kB tool_insert checkpoint at step 5000 (full trainer state for resume) - 63 Bytes tool_insert checkpoint at step 5000 (full trainer state for resume)
- 33.3 kB tool_insert checkpoint at step 5000 (full trainer state for resume)
