shhn1

[Bug]mmrazor requires mmcv>=2.0, but mmcls needs to install mmcv>=1.4.2, <=1.9.0.

1

I downloaded mmcv >=2.0 according to the installation steps of mmrazor. However, when running the script, I found that mmcls will be used, and the requirements of mmcls for mmcv...

If I want to quantize the yolox model using qat, export it to an onnx model with QDQ operator, and then allow the tensorrt backend to parse it, is it...

lightseq是否支持clip模型这种vit+bert的int8量化？

SmoothQuant for llama

3

Hi, authors! Are you planning on supporting LLAMA in smoothquant？I am looking forward to the application of Smoothquant on LLAMA. Thank you!

[Minillm] Using the qwen2-72b model as the teacher model for minillm training results in out of memory

1

I use qwen2-72b as the teacher model and qwen2.5 32b model as the student model for training. 8*80g A100 are used for training. When I load the qwen2 72b model,...