aabbccddwasd

Results 10 comments of aabbccddwasd

composite_demo 代码解释器 无法生成 图片

我成功解决了这个问题 在ipython kernel install --name chatglm3-demo --user创建完内核后在.local/share/jupyter/kernels/修改kernel.json,在argv里增加"--matplotlib=inline"这样就可以正常显示图片了

Base model of Qwen/Qwen2.5-VL-7B-Instruct

+1

[Performance]: guided generation is very slow in offline mode

any progress？

Plan for Quantized Version of Qwen2.5 VL Instruct Model?

https://github.com/QwenLM/Qwen2.5-VL/issues/719#issuecomment-2635751358

is there any quantized models for Qwen2.5-VL

> The quantized models for Qwen2.5-VL are coming soon. thanks！！ How soon， 1 week or 1 month？

is there any quantized models for Qwen2.5-VL

> add vllm support Qwen2.5-VL too, please well, I think vllm developers is responsible for this. they said v0.7.2 will support qwen2.5-vl.

is there any quantized models for Qwen2.5-VL

> great, the only missing part is 4bit AWQ or guff version of 72b for 48gb vram devices and 30b-40b AWQ version for 24gb vram devices also gptq, in 2080ti...

Qwen2.5-VL-72B什么时候出GPTQ和AWQ的量化版本啊。

+1

[Bug]: vllm部署qwen3-30b-a3b输出中文乱码 [Badcase]: Qwen3-30B-A3B输出重复停不下来

no problem in my case

Flash Attention 3 (FA3) Support

emm， Now we have 4090 48GB so is there enough VRAM for FA3, isn't it?