加载ClipCaptionModel模型报错
zhaozunjin opened this issue · comments
我已经把checkpoint-35000.pt放进文名为mlp_finetune的文件夹了，pytorch_model.bin放在gpt2文件夹，ViT-B-32.pt也放在了pretrain_models下面，可是依然说加载模型报错，请问大家有遇到过吗？
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for ClipCaptionModel:
Missing key(s) in state_dict: "clip_project.model.0.weight", "clip_project.model.0.bias", "clip_project.model.2.weight", "clip_project.model.2.bias".
Unexpected key(s) in state_dict: "gpt2.transformer.h.0.attn.bias", "gpt2.transformer.h.0.attn.masked_bias", "gpt2.transformer.h.1.attn.bias", "gpt2.transformer.h.1.attn.masked_bias", "gpt2.transformer.h.2.attn.bias", "gpt2.transformer.h.2.attn.masked_bias", "gpt2.transformer.h.3.attn.bias", "gpt2.transformer.h.3.attn.masked_bias", "gpt2.transformer.h.4.attn.bias", "gpt2.transformer.h.4.attn.masked_bias", "gpt2.transformer.h.5.attn.bias", "gpt2.transformer.h.5.attn.masked_bias", "gpt2.transformer.h.6.attn.bias", "gpt2.transformer.h.6.attn.masked_bias", "gpt2.transformer.h.7.attn.bias", "gpt2.transformer.h.7.attn.masked_bias", "gpt2.transformer.h.8.attn.bias", "gpt2.transformer.h.8.attn.masked_bias", "gpt2.transformer.h.9.attn.bias", "gpt2.transformer.h.9.attn.masked_bias", "gpt2.transformer.h.10.attn.bias", "gpt2.transformer.h.10.attn.masked_bias", "gpt2.transformer.h.11.attn.bias", "gpt2.transformer.h.11.attn.masked_bias", "clip_project.prefix_const", "clip_project.bert.embeddings.position_ids", "clip_project.bert.embeddings.word_embeddings.weight", "clip_project.bert.embeddings.position_embeddings.weight", "clip_project.bert.embeddings.token_type_embeddings.weight", "clip_project.bert.embeddings.LayerNorm.weight", "clip_project.bert.embeddings.LayerNorm.bias", "clip_project.bert.encoder.layer.0.attention.self.query.weight", "clip_project.bert.encoder.layer.0.attention.self.query.bias", "clip_project.bert.encoder.layer.0.attention.self.key.weight", "clip_project.bert.encoder.layer.0.attention.self.key.bias", "clip_project.bert.encoder.layer.0.attention.self.value.weight", "clip_project.bert.encoder.layer.0.attention.self.value.bias", "clip_project.bert.encoder.layer.0.attention.output.dense.weight", "clip_project.bert.encoder.layer.0.attention.output.dense.bias", "clip_project.bert.encoder.layer.0.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.0.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.0.intermediate.dense.weight", "clip_project.bert.encoder.layer.0.intermediate.dense.bias", "clip_project.bert.encoder.layer.0.output.dense.weight", "clip_project.bert.encoder.layer.0.output.dense.bias", "clip_project.bert.encoder.layer.0.output.LayerNorm.weight", "clip_project.bert.encoder.layer.0.output.LayerNorm.bias", "clip_project.bert.encoder.layer.1.attention.self.query.weight", "clip_project.bert.encoder.layer.1.attention.self.query.bias", "clip_project.bert.encoder.layer.1.attention.self.key.weight", "clip_project.bert.encoder.layer.1.attention.self.key.bias", "clip_project.bert.encoder.layer.1.attention.self.value.weight", "clip_project.bert.encoder.layer.1.attention.self.value.bias", "clip_project.bert.encoder.layer.1.attention.output.dense.weight", "clip_project.bert.encoder.layer.1.attention.output.dense.bias", "clip_project.bert.encoder.layer.1.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.1.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.1.intermediate.dense.weight", "clip_project.bert.encoder.layer.1.intermediate.dense.bias", "clip_project.bert.encoder.layer.1.output.dense.weight", "clip_project.bert.encoder.layer.1.output.dense.bias", "clip_project.bert.encoder.layer.1.output.LayerNorm.weight", "clip_project.bert.encoder.layer.1.output.LayerNorm.bias", "clip_project.bert.encoder.layer.2.attention.self.query.weight", "clip_project.bert.encoder.layer.2.attention.self.query.bias", "clip_project.bert.encoder.layer.2.attention.self.key.weight", "clip_project.bert.encoder.layer.2.attention.self.key.bias", "clip_project.bert.encoder.layer.2.attention.self.value.weight", "clip_project.bert.encoder.layer.2.attention.self.value.bias", "clip_project.bert.encoder.layer.2.attention.output.dense.weight", "clip_project.bert.encoder.layer.2.attention.output.dense.bias", "clip_project.bert.encoder.layer.2.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.2.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.2.intermediate.dense.weight", "clip_project.bert.encoder.layer.2.intermediate.dense.bias", "clip_project.bert.encoder.layer.2.output.dense.weight", "clip_project.bert.encoder.layer.2.output.dense.bias", "clip_project.bert.encoder.layer.2.output.LayerNorm.weight", "clip_project.bert.encoder.layer.2.output.LayerNorm.bias", "clip_project.bert.encoder.layer.3.attention.self.query.weight", "clip_project.bert.encoder.layer.3.attention.self.query.bias", "clip_project.bert.encoder.layer.3.attention.self.key.weight", "clip_project.bert.encoder.layer.3.attention.self.key.bias", "clip_project.bert.encoder.layer.3.attention.self.value.weight", "clip_project.bert.encoder.layer.3.attention.self.value.bias", "clip_project.bert.encoder.layer.3.attention.output.dense.weight", "clip_project.bert.encoder.layer.3.attention.output.dense.bias", "clip_project.bert.encoder.layer.3.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.3.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.3.intermediate.dense.weight", "clip_project.bert.encoder.layer.3.intermediate.dense.bias", "clip_project.bert.encoder.layer.3.output.dense.weight", "clip_project.bert.encoder.layer.3.output.dense.bias", "clip_project.bert.encoder.layer.3.output.LayerNorm.weight", "clip_project.bert.encoder.layer.3.output.LayerNorm.bias", "clip_project.bert.encoder.layer.4.attention.self.query.weight", "clip_project.bert.encoder.layer.4.attention.self.query.bias", "clip_project.bert.encoder.layer.4.attention.self.key.weight", "clip_project.bert.encoder.layer.4.attention.self.key.bias", "clip_project.bert.encoder.layer.4.attention.self.value.weight", "clip_project.bert.encoder.layer.4.attention.self.value.bias", "clip_project.bert.encoder.layer.4.attention.output.dense.weight", "clip_project.bert.encoder.layer.4.attention.output.dense.bias", "clip_project.bert.encoder.layer.4.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.4.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.4.intermediate.dense.weight", "clip_project.bert.encoder.layer.4.intermediate.dense.bias", "clip_project.bert.encoder.layer.4.output.dense.weight", "clip_project.bert.encoder.layer.4.output.dense.bias", "clip_project.bert.encoder.layer.4.output.LayerNorm.weight", "clip_project.bert.encoder.layer.4.output.LayerNorm.bias", "clip_project.bert.encoder.layer.5.attention.self.query.weight", "clip_project.bert.encoder.layer.5.attention.self.query.bias", "clip_project.bert.encoder.layer.5.attention.self.key.weight", "clip_project.bert.encoder.layer.5.attention.self.key.bias", "clip_project.bert.encoder.layer.5.attention.self.value.weight", "clip_project.bert.encoder.layer.5.attention.self.value.bias", "clip_project.bert.encoder.layer.5.attention.output.dense.weight", "clip_project.bert.encoder.layer.5.attention.output.dense.bias", "clip_project.bert.encoder.layer.5.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.5.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.5.intermediate.dense.weight", "clip_project.bert.encoder.layer.5.intermediate.dense.bias", "clip_project.bert.encoder.layer.5.output.dense.weight", "clip_project.bert.encoder.layer.5.output.dense.bias", "clip_project.bert.encoder.layer.5.output.LayerNorm.weight", "clip_project.bert.encoder.layer.5.output.LayerNorm.bias", "clip_project.bert.encoder.layer.6.attention.self.query.weight", "clip_project.bert.encoder.layer.6.attention.self.query.bias", "clip_project.bert.encoder.layer.6.attention.self.key.weight", "clip_project.bert.encoder.layer.6.attention.self.key.bias", "clip_project.bert.encoder.layer.6.attention.self.value.weight", "clip_project.bert.encoder.layer.6.attention.self.value.bias", "clip_project.bert.encoder.layer.6.attention.output.dense.weight", "clip_project.bert.encoder.layer.6.attention.output.dense.bias", "clip_project.bert.encoder.layer.6.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.6.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.6.intermediate.dense.weight", "clip_project.bert.encoder.layer.6.intermediate.dense.bias", "clip_project.bert.encoder.layer.6.output.dense.weight", "clip_project.bert.encoder.layer.6.output.dense.bias", "clip_project.bert.encoder.layer.6.output.LayerNorm.weight", "clip_project.bert.encoder.layer.6.output.LayerNorm.bias", "clip_project.bert.encoder.layer.7.attention.self.query.weight", "clip_project.bert.encoder.layer.7.attention.self.query.bias", "clip_project.bert.encoder.layer.7.attention.self.key.weight", "clip_project.bert.encoder.layer.7.attention.self.key.bias", "clip_project.bert.encoder.layer.7.attention.self.value.weight", "clip_project.bert.encoder.layer.7.attention.self.value.bias", "clip_project.bert.encoder.layer.7.attention.output.dense.weight", "clip_project.bert.encoder.layer.7.attention.output.dense.bias", "clip_project.bert.encoder.layer.7.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.7.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.7.intermediate.dense.weight", "clip_project.bert.encoder.layer.7.intermediate.dense.bias", "clip_project.bert.encoder.layer.7.output.dense.weight", "clip_project.bert.encoder.layer.7.output.dense.bias", "clip_project.bert.encoder.layer.7.output.LayerNorm.weight", "clip_project.bert.encoder.layer.7.output.LayerNorm.bias", "clip_project.bert.pooler.dense.weight", "clip_project.bert.pooler.dense.bias", "clip_project.linear.weight", "clip_project.linear.bias".
Process finished with exit code 1
怎么解决的呢
我已经把checkpoint-35000.pt放进文名为mlp_finetune的文件夹了，pytorch_model.bin放在gpt2文件夹，ViT-B-32.pt也放在了pretrain_models下面，可是依然说加载模型报错，请问大家有遇到过吗？
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for ClipCaptionModel: Missing key(s) in state_dict: "clip_project.model.0.weight", "clip_project.model.0.bias", "clip_project.model.2.weight", "clip_project.model.2.bias". Unexpected key(s) in state_dict: "gpt2.transformer.h.0.attn.bias", "gpt2.transformer.h.0.attn.masked_bias", "gpt2.transformer.h.1.attn.bias", "gpt2.transformer.h.1.attn.masked_bias", "gpt2.transformer.h.2.attn.bias", "gpt2.transformer.h.2.attn.masked_bias", "gpt2.transformer.h.3.attn.bias", "gpt2.transformer.h.3.attn.masked_bias", "gpt2.transformer.h.4.attn.bias", "gpt2.transformer.h.4.attn.masked_bias", "gpt2.transformer.h.5.attn.bias", "gpt2.transformer.h.5.attn.masked_bias", "gpt2.transformer.h.6.attn.bias", "gpt2.transformer.h.6.attn.masked_bias", "gpt2.transformer.h.7.attn.bias", "gpt2.transformer.h.7.attn.masked_bias", "gpt2.transformer.h.8.attn.bias", "gpt2.transformer.h.8.attn.masked_bias", "gpt2.transformer.h.9.attn.bias", "gpt2.transformer.h.9.attn.masked_bias", "gpt2.transformer.h.10.attn.bias", "gpt2.transformer.h.10.attn.masked_bias", "gpt2.transformer.h.11.attn.bias", "gpt2.transformer.h.11.attn.masked_bias", "clip_project.prefix_const", "clip_project.bert.embeddings.position_ids", "clip_project.bert.embeddings.word_embeddings.weight", "clip_project.bert.embeddings.position_embeddings.weight", "clip_project.bert.embeddings.token_type_embeddings.weight", "clip_project.bert.embeddings.LayerNorm.weight", "clip_project.bert.embeddings.LayerNorm.bias", "clip_project.bert.encoder.layer.0.attention.self.query.weight", "clip_project.bert.encoder.layer.0.attention.self.query.bias", "clip_project.bert.encoder.layer.0.attention.self.key.weight", "clip_project.bert.encoder.layer.0.attention.self.key.bias", "clip_project.bert.encoder.layer.0.attention.self.value.weight", "clip_project.bert.encoder.layer.0.attention.self.value.bias", "clip_project.bert.encoder.layer.0.attention.output.dense.weight", "clip_project.bert.encoder.layer.0.attention.output.dense.bias", "clip_project.bert.encoder.layer.0.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.0.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.0.intermediate.dense.weight", "clip_project.bert.encoder.layer.0.intermediate.dense.bias", "clip_project.bert.encoder.layer.0.output.dense.weight", "clip_project.bert.encoder.layer.0.output.dense.bias", "clip_project.bert.encoder.layer.0.output.LayerNorm.weight", "clip_project.bert.encoder.layer.0.output.LayerNorm.bias", "clip_project.bert.encoder.layer.1.attention.self.query.weight", "clip_project.bert.encoder.layer.1.attention.self.query.bias", "clip_project.bert.encoder.layer.1.attention.self.key.weight", "clip_project.bert.encoder.layer.1.attention.self.key.bias", "clip_project.bert.encoder.layer.1.attention.self.value.weight", "clip_project.bert.encoder.layer.1.attention.self.value.bias", "clip_project.bert.encoder.layer.1.attention.output.dense.weight", "clip_project.bert.encoder.layer.1.attention.output.dense.bias", "clip_project.bert.encoder.layer.1.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.1.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.1.intermediate.dense.weight", "clip_project.bert.encoder.layer.1.intermediate.dense.bias", "clip_project.bert.encoder.layer.1.output.dense.weight", "clip_project.bert.encoder.layer.1.output.dense.bias", "clip_project.bert.encoder.layer.1.output.LayerNorm.weight", "clip_project.bert.encoder.layer.1.output.LayerNorm.bias", "clip_project.bert.encoder.layer.2.attention.self.query.weight", "clip_project.bert.encoder.layer.2.attention.self.query.bias", "clip_project.bert.encoder.layer.2.attention.self.key.weight", "clip_project.bert.encoder.layer.2.attention.self.key.bias", "clip_project.bert.encoder.layer.2.attention.self.value.weight", "clip_project.bert.encoder.layer.2.attention.self.value.bias", "clip_project.bert.encoder.layer.2.attention.output.dense.weight", "clip_project.bert.encoder.layer.2.attention.output.dense.bias", "clip_project.bert.encoder.layer.2.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.2.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.2.intermediate.dense.weight", "clip_project.bert.encoder.layer.2.intermediate.dense.bias", "clip_project.bert.encoder.layer.2.output.dense.weight", "clip_project.bert.encoder.layer.2.output.dense.bias", "clip_project.bert.encoder.layer.2.output.LayerNorm.weight", "clip_project.bert.encoder.layer.2.output.LayerNorm.bias", "clip_project.bert.encoder.layer.3.attention.self.query.weight", "clip_project.bert.encoder.layer.3.attention.self.query.bias", "clip_project.bert.encoder.layer.3.attention.self.key.weight", "clip_project.bert.encoder.layer.3.attention.self.key.bias", "clip_project.bert.encoder.layer.3.attention.self.value.weight", "clip_project.bert.encoder.layer.3.attention.self.value.bias", "clip_project.bert.encoder.layer.3.attention.output.dense.weight", "clip_project.bert.encoder.layer.3.attention.output.dense.bias", "clip_project.bert.encoder.layer.3.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.3.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.3.intermediate.dense.weight", "clip_project.bert.encoder.layer.3.intermediate.dense.bias", "clip_project.bert.encoder.layer.3.output.dense.weight", "clip_project.bert.encoder.layer.3.output.dense.bias", "clip_project.bert.encoder.layer.3.output.LayerNorm.weight", "clip_project.bert.encoder.layer.3.output.LayerNorm.bias", "clip_project.bert.encoder.layer.4.attention.self.query.weight", "clip_project.bert.encoder.layer.4.attention.self.query.bias", "clip_project.bert.encoder.layer.4.attention.self.key.weight", "clip_project.bert.encoder.layer.4.attention.self.key.bias", "clip_project.bert.encoder.layer.4.attention.self.value.weight", "clip_project.bert.encoder.layer.4.attention.self.value.bias", "clip_project.bert.encoder.layer.4.attention.output.dense.weight", "clip_project.bert.encoder.layer.4.attention.output.dense.bias", "clip_project.bert.encoder.layer.4.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.4.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.4.intermediate.dense.weight", "clip_project.bert.encoder.layer.4.intermediate.dense.bias", "clip_project.bert.encoder.layer.4.output.dense.weight", "clip_project.bert.encoder.layer.4.output.dense.bias", "clip_project.bert.encoder.layer.4.output.LayerNorm.weight", "clip_project.bert.encoder.layer.4.output.LayerNorm.bias", "clip_project.bert.encoder.layer.5.attention.self.query.weight", "clip_project.bert.encoder.layer.5.attention.self.query.bias", "clip_project.bert.encoder.layer.5.attention.self.key.weight", "clip_project.bert.encoder.layer.5.attention.self.key.bias", "clip_project.bert.encoder.layer.5.attention.self.value.weight", "clip_project.bert.encoder.layer.5.attention.self.value.bias", "clip_project.bert.encoder.layer.5.attention.output.dense.weight", "clip_project.bert.encoder.layer.5.attention.output.dense.bias", "clip_project.bert.encoder.layer.5.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.5.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.5.intermediate.dense.weight", "clip_project.bert.encoder.layer.5.intermediate.dense.bias", "clip_project.bert.encoder.layer.5.output.dense.weight", "clip_project.bert.encoder.layer.5.output.dense.bias", "clip_project.bert.encoder.layer.5.output.LayerNorm.weight", "clip_project.bert.encoder.layer.5.output.LayerNorm.bias", "clip_project.bert.encoder.layer.6.attention.self.query.weight", "clip_project.bert.encoder.layer.6.attention.self.query.bias", "clip_project.bert.encoder.layer.6.attention.self.key.weight", "clip_project.bert.encoder.layer.6.attention.self.key.bias", "clip_project.bert.encoder.layer.6.attention.self.value.weight", "clip_project.bert.encoder.layer.6.attention.self.value.bias", "clip_project.bert.encoder.layer.6.attention.output.dense.weight", "clip_project.bert.encoder.layer.6.attention.output.dense.bias", "clip_project.bert.encoder.layer.6.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.6.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.6.intermediate.dense.weight", "clip_project.bert.encoder.layer.6.intermediate.dense.bias", "clip_project.bert.encoder.layer.6.output.dense.weight", "clip_project.bert.encoder.layer.6.output.dense.bias", "clip_project.bert.encoder.layer.6.output.LayerNorm.weight", "clip_project.bert.encoder.layer.6.output.LayerNorm.bias", "clip_project.bert.encoder.layer.7.attention.self.query.weight", "clip_project.bert.encoder.layer.7.attention.self.query.bias", "clip_project.bert.encoder.layer.7.attention.self.key.weight", "clip_project.bert.encoder.layer.7.attention.self.key.bias", "clip_project.bert.encoder.layer.7.attention.self.value.weight", "clip_project.bert.encoder.layer.7.attention.self.value.bias", "clip_project.bert.encoder.layer.7.attention.output.dense.weight", "clip_project.bert.encoder.layer.7.attention.output.dense.bias", "clip_project.bert.encoder.layer.7.attention.output.LayerNorm.weight", "clip_project.bert.encoder.layer.7.attention.output.LayerNorm.bias", "clip_project.bert.encoder.layer.7.intermediate.dense.weight", "clip_project.bert.encoder.layer.7.intermediate.dense.bias", "clip_project.bert.encoder.layer.7.output.dense.weight", "clip_project.bert.encoder.layer.7.output.dense.bias", "clip_project.bert.encoder.layer.7.output.LayerNorm.weight", "clip_project.bert.encoder.layer.7.output.LayerNorm.bias", "clip_project.bert.pooler.dense.weight", "clip_project.bert.pooler.dense.bias", "clip_project.linear.weight", "clip_project.linear.bias".
Process finished with exit code 1
请问解决了嘛？我遇到的也是类似问题。