Modas Li

icon location Summer in your hand if you will

Results 29 comments of Modas Li

SFT training ,single gpu (V100 32G), how to adjust my parameters to avoid OOM, thx

Bloom-1.1b

TypeError: cannot unpack non-iterable NoneType object

same bug

[BUG] <使用vllm+openaiAPI测试全参微调后的模型，回答混乱>

> 1. vLLM + OpenAI API：如果是单独vLLM，是没有完整支持Qwen对话模式的，建议用FastChat+vLLM。 > 2. chat调用参数格式不正确，请参考OpenAI API说明修正messages的格式。 > 如果是单独的vLLM，还需要传入正确的stop_token_ids ([151643, 151644, 151645]) 谢谢解答，请问stop_token_ids 如何传入，在哪里传入呢？谢谢~

AssertionError: We detect you are probably using the pretrained model (rather than chat model) for chatting, since the chat_format in generation_config is not "chatml". If you are directly using the model downloaded from Huggingface, please make sure you are using our "Qwen/Qwen-7B-Chat" Huggingface model (rather than "Qwen/Qwen-7B") when you call model.chat().

qwen-7b after sft fine-tuning 发自我的iPhone ------------------ Original ------------------ From: Ren Xuancheng ***@***.***&gt; Date: Tue,Dec 12,2023 5:52 PM To: QwenLM/Qwen ***@***.***&gt; Cc: Modas Li ***@***.***&gt;, Author ***@***.***&gt; Subject: Re: [QwenLM/Qwen] AssertionError:...

[BUG] 启动2个本地模型指定显卡不生效

没办法进行多卡分配模型

Qwen-7B微调后的推理问题

same warning: The attention mask and the pad token id were not set. As a consequence, you may observe unexpected behavior. Please pass your input's attention_mask to obtain reliable results....

RuntimeError: output tensor must have the same type as input tensor when training qwen-7b

Solution: update model files to the latest and alter config(sft.json, ds_config.json) fp16 to bf16,make sure your GPUs are more than A100(40g)*4.

browser_qwen 问题

设置地址后，浏览器助手变空白 ![image](https://github.com/QwenLM/Qwen-Agent/assets/40042370/bb91807d-6e6f-4bde-9510-741a1b79cc11)

browser_qwen 问题

端口需要怎么调整呢？哪些是linux机器必须开通的呢

browser_qwen 问题

生成的图片如何修改IP地址呢？ ![image](https://github.com/QwenLM/Qwen-Agent/assets/40042370/4fc9b8d5-5e13-4604-aaf0-86d7afc8bb3f)