rui

> EfficientNet is really designed to be used on images of a specific size, but you can just take the model and apply it (probably without any modifications) to images...

I have the same issue with matplotlib.animation and ffmpeg and got the error which is: ValueError: Invalid file object:

我用的Qwen-14B-Chat模型占用16G左右显存，Code Interpreter问了几轮之后就40G然后kill掉server之后依然40G占用 得重新部署模型

Is there any history retreival mechanism of Qwen-agent? How to delete or maintain the history or the uploaded file (or its vectorized content storage)?

> @infwinston Actually I mentioned the wrong command. The corresponding command is python3 -m fastchat.serve.vllm_worker --model-path WizardLM/WizardLM-70B-V1.0 --num-gpus 4 how's the inference speed?

Thanks @QingZ11 for the reply. No, I did not modify anything from configuration files. I completly followed the instruction.

I got exactly the same issue

> How do you run the controller? for me, I just simply run `python3 -m fastchat.server.controller --host 0.0.0.0 --port 30001`

I got the same problem when using Qwen-70B-chat

> Being based on an existing model does not mean that the two have identical architectures. > > [InternVL-1.5's model card](https://huggingface.co/OpenGVLab/InternVL-Chat-V1-5) says that it is InternViT-6B-448px-V1-5 + MLP + InternLM2-Chat-20B....