No such file or directory: 'output/mlp_finetune/checkpoint-35000.pt

[yywhsgnd]

2022-12-28 11:48:11.682 | INFO | models.model:init:80 - random initialize gpt2 model
Traceback (most recent call last):
File "predict.py", line 186, in
main(args)
File "predict.py", line 129, in main
model.load_state_dict(torch.load(args.model_path, map_location=args.device))
File "/home/jupyter/miniforge3/envs/env1/lib/python3.7/site-packages/torch/serialization.py", line 699, in load
with _open_file_like(f, 'rb') as opened_file:
File "/home/jupyter/miniforge3/envs/env1/lib/python3.7/site-packages/torch/serialization.py", line 231, in _open_file_like
return _open_file(name_or_buffer, mode)
File "/home/jupyter/miniforge3/envs/env1/lib/python3.7/site-packages/torch/serialization.py", line 212, in init
super(_open_file, self).init(open(name, mode))
FileNotFoundError: [Errno 2] No such file or directory: 'output/mlp_finetune/checkpoint-35000.pt

[EnxinSong]

maybe you can find checkpoint-35000.pt in directory finetune under directory output

[yywhsgnd]

好的，找到了，我更改了一下文件名

…

---Original--- From: ***@***.***> Date: Wed, Dec 28, 2022 20:25 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [yangjianxin1/ClipCap-Chinese] No such file or directory:'output/mlp_finetune/checkpoint-35000.pt (Issue #7) maybe you can find checkpoint-35000.pt in directory finetune under directory output — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[EnxinSong]

那你有遇到过这个问题吗

Traceback (most recent call last):
File "predict.py", line 197, in
main(args)
File "predict.py", line 156, in main
captions = generate(model, clip_embeds, tokenizer, args)
File "predict.py", line 80, in generate
out = model.gpt2(inputs_embeds=inputs_embeds)
File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/transformers/models/gpt2/modeling_gpt2.py", line 971, in forward
lm_logits = self.lm_head(hidden_states)
File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 105, in forward
return F.linear(input.float(), self.weight.float(), self.bias.float())
AttributeError: 'NoneType' object has no attribute 'float'

[yywhsgnd]

没有遇到，你是运行到哪部时出现了这个问题

…

---Original--- From: ***@***.***> Date: Wed, Dec 28, 2022 20:36 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [yangjianxin1/ClipCap-Chinese] No such file or directory:'output/mlp_finetune/checkpoint-35000.pt (Issue #7) 那你有遇到过这个问题吗 Traceback (most recent call last): File "predict.py", line 197, in main(args) File "predict.py", line 156, in main captions = generate(model, clip_embeds, tokenizer, args) File "predict.py", line 80, in generate out = model.gpt2(inputs_embeds=inputs_embeds) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl return forward_call(*input, **kwargs) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/transformers/models/gpt2/modeling_gpt2.py", line 971, in forward lm_logits = self.lm_head(hidden_states) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl return forward_call(*input, **kwargs) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 105, in forward return F.linear(input.float(), self.weight.float(), self.bias.float()) AttributeError: 'NoneType' object has no attribute 'float' — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[EnxinSong]

train完no_finetune之后predict的时候

[yywhsgnd]

也就是说现在你是MLp进行生成时遇到了问题?

…

---Original--- From: ***@***.***> Date: Wed, Dec 28, 2022 20:36 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [yangjianxin1/ClipCap-Chinese] No such file or directory:'output/mlp_finetune/checkpoint-35000.pt (Issue #7) 那你有遇到过这个问题吗 Traceback (most recent call last): File "predict.py", line 197, in main(args) File "predict.py", line 156, in main captions = generate(model, clip_embeds, tokenizer, args) File "predict.py", line 80, in generate out = model.gpt2(inputs_embeds=inputs_embeds) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl return forward_call(*input, **kwargs) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/transformers/models/gpt2/modeling_gpt2.py", line 971, in forward lm_logits = self.lm_head(hidden_states) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl return forward_call(*input, **kwargs) File "/home/exs/Anaconda3/envs/cliipcap/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 105, in forward return F.linear(input.float(), self.weight.float(), self.bias.float()) AttributeError: 'NoneType' object has no attribute 'float' — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[EnxinSong]

是这样的

[yywhsgnd]

你有改过代码吗？

…

---Original--- From: ***@***.***> Date: Wed, Dec 28, 2022 20:48 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [yangjianxin1/ClipCap-Chinese] No such file or directory:'output/mlp_finetune/checkpoint-35000.pt (Issue #7) 是这样的 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[EnxinSong]

我把train的数据集给换了

[yywhsgnd]

我把train的数据集给换了

你这情况我也没有用到，我的建议是你继续用flickr30的数据集复现一下，我用mlp生成时就只出现了torch和cpu\gpu的问题

[EnxinSong]

好的谢谢

[yywhsgnd]

没事，请问您有没有执行最后一步时生成的caption-generate_no_finetune.txt是个空文件夹

…

---Original--- From: ***@***.***> Date: Wed, Dec 28, 2022 21:02 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [yangjianxin1/ClipCap-Chinese] No such file or directory:'output/mlp_finetune/checkpoint-35000.pt (Issue #7) 好的谢谢 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[EnxinSong]

哈哈哈哈我还没生成出来呢，执行finetune.sh的时候报错了

[yywhsgnd]

哈哈哈哈我还没生成出来呢，执行finetune.sh的时候报错了

好像readme里面生成那里finetune是finerune

[EnxinSong]

那个我看到了

[yywhsgnd]

你下载到caption的中文数据集了吗?

…

---Original--- From: ***@***.***> Date: Wed, Dec 28, 2022 21:13 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [yangjianxin1/ClipCap-Chinese] No such file or directory:'output/mlp_finetune/checkpoint-35000.pt (Issue #7) 那个我看到了 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[EnxinSong]

额……我换了其他的image-caption数据集，预训练用的是gpt2，不是chinesegpt2

[yywhsgnd]

哦哦，好的，谢谢

…

---Original--- From: ***@***.***> Date: Wed, Dec 28, 2022 21:17 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [yangjianxin1/ClipCap-Chinese] No such file or directory:'output/mlp_finetune/checkpoint-35000.pt (Issue #7) 额……我换了其他的image-caption数据集，预训练用的是gpt2，不是chinesegpt2 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[xiaozhouoo]

请问一下您怎么解决的