Soran

/

XCLIP_custom

like 0

• "transformers.models.x_clip.modeling_x_clip.XCLIPEncoder",
• "transformers.activations.QuickGELUActivation",
• "transformers.models.x_clip.modeling_x_clip.XCLIPTextEmbeddings",
• "torch.nn.modules.container.ModuleList",
• "transformers.models.x_clip.configuration_x_clip.XCLIPTextConfig",
• "collections.OrderedDict",
• "torch.nn.modules.normalization.LayerNorm",
• "transformers.models.x_clip.modeling_x_clip.XCLIPMLP",
• "transformers.models.x_clip.modeling_x_clip.XCLIPEncoderLayer",
• "torch.nn.modules.linear.Linear",
• "__builtin__.set",
• "torch.nn.modules.sparse.Embedding",
• "torch.LongStorage",
• "torch.FloatStorage",
• "transformers.models.x_clip.modeling_x_clip.XCLIPTextTransformer",
• "transformers.models.x_clip.modeling_x_clip.XCLIPAttention",
• "torch._utils._rebuild_tensor_v2",
• "torch._utils._rebuild_parameter"

How to fix it?

• "torch.nn.modules.linear.Linear",
• "collections.OrderedDict",
• "__builtin__.set",
• "torch._utils._rebuild_tensor_v2",
• "torch.FloatStorage",
• "torch._utils._rebuild_parameter"

How to fix it?

• "tokenizers.models.Model",
• "_codecs.encode",
• "tokenizers.Tokenizer",
• "tokenizers.AddedToken",
• "transformers.models.clip.tokenization_clip_fast.CLIPTokenizerFast"

How to fix it?