maxin9966
maxin9966
> > > hello, the bottleneck is not `write_video`, is the for-loop and `iter_frames` function. > > > > > > yeah, look into the `ffmpeg_write.py`, in `write_frame` function, img_array.tobytes()...
@ItzCrazyKns The problem can be resolved in 90% of cases by simply being compatible with the OpenAI API, opening the base_url, and allowing for a customizable model_name. This is because...
Is anyone in the same situation as me?
> > **env:** 2080Ti * 2 cuda_12.3.r12.3/compiler.33567101_0 python3.9 pip install "sglang[all]" > > **error:** new fill batch. #seq: 1. #cached_token: 0. #new_token: 8. #remaining_req: 0. #running_req: 0. tree_cache_hit_rate: 0.00%. python:...
I have verified that the image parameters are consistent. Could there be a mistake in my steps?
4bit测试通过了，但是对视频的描述似乎非常不准确，例如人物的描述、动作等等都存在较大偏差，字幕提取也是错误百出。
Has anyone encountered the same issue? Could the official upload an awq version of glm4-9b on huggingface or modelscope?
@EasonQYS I've tried it, but it keeps outputting non-stop.
@lvhan028 Any prompt will result in an endless output of content. The generated response appears normal, but it continues to output without stopping.
@zhyncs Does it support the AWQ format?