multi-token

/

shiftedattn-9-21-qwen25coder-instr-from-scratch-n64w8-lr1-e6

like 0

Follow

multi-token 3

• "torch._utils._rebuild_tensor_v2",
• "_codecs.encode",
• "torch.ByteStorage",
• "numpy.ndarray",
• "numpy._core.multiarray._reconstruct",
• "numpy.dtype",
• "collections.OrderedDict"

How to fix it?

• "torch._utils._rebuild_tensor_v2",
• "_codecs.encode",
• "torch.ByteStorage",
• "numpy.ndarray",
• "numpy._core.multiarray._reconstruct",
• "numpy.dtype",
• "collections.OrderedDict"

How to fix it?

• "torch._utils._rebuild_tensor_v2",
• "_codecs.encode",
• "torch.ByteStorage",
• "numpy.ndarray",
• "numpy._core.multiarray._reconstruct",
• "numpy.dtype",
• "collections.OrderedDict"

How to fix it?

• "torch._utils._rebuild_tensor_v2",
• "_codecs.encode",
• "torch.ByteStorage",
• "numpy.ndarray",
• "numpy._core.multiarray._reconstruct",
• "numpy.dtype",
• "collections.OrderedDict"

How to fix it?

• "transformers.trainer_utils.SchedulerType",
• "transformers.trainer_utils.HubStrategy",
• "transformers.training_args.OptimizerNames",
• "transformers.trainer_utils.IntervalStrategy",
• "transformers.trainer_pt_utils.AcceleratorConfig",
• "transformers.trainer_utils.SaveStrategy",
• "accelerate.utils.dataclasses.DistributedType",
• "accelerate.state.PartialState",
• "torch.device",
• "__main__.TrainingArguments"

How to fix it?