smohammadi

/

qat-llama-3B

like 0

• "torch.bfloat16",
• "torch.BFloat16Storage",
• "torch.serialization._get_layout",
• "torch.int8",
• "torch._utils._rebuild_wrapper_subclass",
• "torch._tensor._rebuild_from_type_v2",
• "torchao.quantization.linear_activation_quantized_tensor.LinearActivationQuantizedTensor",
• "torch._utils._rebuild_tensor_v2",
• "torchao.dtypes.utils.PlainLayout",
• "torchao.quantization.quant_primitives.ZeroPointDomain",
• "torchao.dtypes.uintx.plain_layout.PlainAQTTensorImpl",
• "torchao.quantization.quant_api._int8_asymm_per_token_quant",
• "torch.CharStorage",
• "collections.OrderedDict",
• "torchao.dtypes.affine_quantized_tensor.AffineQuantizedTensor",
• "torch.device"

How to fix it?

• "transformers.trainer_utils.HubStrategy",
• "transformers.trainer_utils.SaveStrategy",
• "transformers.trainer_utils.SchedulerType",
• "transformers.training_args.OptimizerNames",
• "accelerate.utils.dataclasses.DistributedType",
• "axolotl.core.training_args.AxolotlTrainingArguments",
• "transformers.trainer_pt_utils.AcceleratorConfig",
• "transformers.trainer_utils.FSDPOption",
• "torch.device",
• "accelerate.state.PartialState",
• "axolotl.utils.dict.DictDefault",
• "transformers.trainer_utils.IntervalStrategy"

How to fix it?