运行run_qwen_vary.py报错，请指教！

[baibaomen]

我把huggingface上的clip-vit-large-patch14模型git clone到了/cache/vit-large-patch14/目录。然后执行下面指令时报错，请教是--model-name还是哪里写错了吗？

指令： /Vary/Vary-master/vary# python ./demo/run_qwen_vary.py --model-name /cache/vit-large-patch14/ --image-file /mnt/e/ocr.png

报错： (vary) root@90bb63a226b2:/Vary/Vary-master/vary# python ./demo/run_qwen_vary.py --model-name /cache/vit-large-patch14/ --image-file /mnt/e/ocr.png The argument trust_remote_code is to be used with Auto classes. It has no effect here and is ignored. You are using a model of type clip to instantiate a model of type vary. This is not supported for all configurations of models and can yield errors. Traceback (most recent call last): File "/Vary/Vary-master/vary/./demo/run_qwen_vary.py", line 127, in eval_model(args) File "/Vary/Vary-master/vary/./demo/run_qwen_vary.py", line 43, in eval_model model = varyQwenForCausalLM.from_pretrained(model_name, low_cpu_mem_usage=True, device_map='cuda', trust_remote_code=True) File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/transformers/modeling_utils.py", line 2876, in from_pretrained model = cls(config, *model_args, **model_kwargs) File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/vary/model/vary_qwen_vary.py", line 238, in init self.transformer = varyQwenModel(config) File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/vary/model/vary_qwen_vary.py", line 46, in init super(varyQwenModel, self).init(config) File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/vary/model/llm/qwen/modeling_qwen.py", line 496, in init [ File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/vary/model/llm/qwen/modeling_qwen.py", line 497, in QWenBlock( File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/vary/model/llm/qwen/modeling_qwen.py", line 393, in init self.attn = QWenAttention(config) File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/vary/model/llm/qwen/modeling_qwen.py", line 120, in init self.seq_length = config.seq_length File "/root/anaconda3/envs/vary/lib/python3.10/site-packages/transformers/configuration_utils.py", line 261, in getattribute return super().getattribute(key) AttributeError: 'varyConfig' object has no attribute 'seq_length'. Did you mean: 'max_length'? (vary) root@90bb63a226b2:/Vary/Vary-master/vary#

[yummy527]

修改 modelling_qwen.py 第120行为： self.seq_length = config.max_length # seq_length

[baibaomen]

修改 modelling_qwen.py 第120行为： self.seq_length = config.max_length # seq_length

谢谢！拿到了模型数据，搞定了

[Sandwichzhao]

修改 modelling_qwen.py 第120行为： self.seq_length = config.max_length # seq_length

谢谢！拿到了模型数据，搞定了

大哥，请教一下你跑起来需要多大的显存？我跑了一半提示显存内存不足了

[baibaomen]

修改 modelling_qwen.py 第120行为： self.seq_length = config.max_length # seq_length

谢谢！拿到了模型数据，搞定了

大哥，请教一下你跑起来需要多大的显存？我跑了一半提示显存内存不足了

我是一张3090，能跑

[hotwa]

修改 modelling_qwen.py 第120行为： self.seq_length = config.max_length # seq_length

<python/path>/lib/python3.10/site-packages/vary/model/llm/qwen/modeling_qwen.py

[hotwa]

修改 modelling_qwen.py 第120行为： self.seq_length = config.max_length # seq_length

谢谢！拿到了模型数据，搞定了

大哥，请教一下你跑起来需要多大的显存？我跑了一半提示显存内存不足了

我是一张3090，能跑

我看可以消耗35GB显存，好像会动态调节。

[JeffRody]

老师，您好！ 模型数据能麻烦您发我一下吗，我也在学习中，感谢！ 邮箱：wang_luochao@163.com

[Next0201]

修改 modelling_qwen.py 第120行为： self.seq_length = config.max_length # seq_length

谢谢！拿到了模型数据，搞定了

大哥，请教一下你跑起来需要多大的显存？我跑了一半提示显存内存不足了

我是一张3090，能跑

我看可以消耗35GB显存，好像会动态调节。

我是24G跑不起来怎么办，只能跑一半