kiri-ai

/

gpt2-large-quantized

like 0

• "torch.FloatStorage",
• "torch.per_tensor_affine",
• "transformers.models.gpt2.modeling_gpt2.GPT2LMHeadModel",
• "torch.nn.modules.container.ModuleList",
• "transformers.models.gpt2.modeling_gpt2.Attention",
• "torch.QInt8Storage",
• "torch.ByteStorage",
• "transformers.models.gpt2.modeling_gpt2.MLP",
• "transformers.models.gpt2.modeling_gpt2.Block",
• "torch._utils._rebuild_parameter",
• "torch.nn.modules.normalization.LayerNorm",
• "torch.qint8",
• "torch.nn.modules.dropout.Dropout",
• "transformers.activations.gelu_new",
• "torch.nn.quantized.modules.linear.LinearPackedParams",
• "__builtin__.set",
• "transformers.models.gpt2.modeling_gpt2.GPT2Model",
• "torch._utils._rebuild_qtensor",
• "torch.nn.quantized.dynamic.modules.linear.Linear",
• "torch._utils._rebuild_tensor_v2",
• "transformers.models.gpt2.configuration_gpt2.GPT2Config",
• "torch.nn.modules.sparse.Embedding",
• "collections.OrderedDict"

How to fix it?