gabrielmbmb

/

Upcycled-Qwen1.5-MoE2.7B-LoRA-merged-32-2000-steps-adapter

like 0

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

How to fix it?

• "transformers.trainer_utils.IntervalStrategy",
• "transformers.trainer_utils.HubStrategy",
• "transformers.training_args_seq2seq.Seq2SeqTrainingArguments",
• "transformers.trainer_utils.SchedulerType",
• "transformers.trainer_pt_utils.AcceleratorConfig",
• "torch.device",
• "accelerate.utils.dataclasses.DistributedType",
• "transformers.training_args.OptimizerNames",
• "accelerate.state.PartialState"

How to fix it?