HydraLM

/

Nous-Hermes-llama-2-7b_7b_cluster00_partitioned_v3_standardized_00

like 0

• "torch._tensor._rebuild_from_type_v2",
• "torch.Tensor",
• "collections.OrderedDict",
• "torch.FloatStorage",
• "torch._utils._rebuild_tensor_v2"

How to fix it?

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

How to fix it?

• "transformers.trainer_utils.HubStrategy",
• "args.TrainingArguments",
• "transformers.training_args.OptimizerNames",
• "transformers.trainer_utils.IntervalStrategy",
• "torch.device",
• "accelerate.state.PartialState",
• "transformers.trainer_utils.SchedulerType",
• "transformers.generation.configuration_utils.GenerationConfig",
• "accelerate.utils.dataclasses.DistributedType"

How to fix it?