search

QwenLM / qwen.cpp

C++ implementation of Qwen-LM
Other
538 stars 49 forks source link

Can you add an additional function to let convert.py support Qwen/Qwen-7B-Chat-Int4? #28

Open x1ngzai opened 11 months ago

x1ngzai commented 11 months ago

It seems conversion on Qwen-7B-Chat needs more than 32GB memory to run. It probably can be solved by conversion for Int4