TurboTransformers
TurboTransformers copied to clipboard
Tencent
a fast and user-friendly runtime for transformer inference (Bert, Albert, GPT2, Decoders, etc) on CPU and GPU.
Hi, Thank you for your awesome work! I would like to know if there is any plan on supporting vision-based Transformer? As transformers are becoming popular in vision tasks, I...
It seems that the dependency is not so much: 1. ninja (apt install ninja-build,) 2. intel MTK (pip install mtk) (which is actually not needed if using GPU inference only,...
I want to run bert on GPU with C++, and the default device is 0. Is there any way to select other device with C++ API?
Hi, I am using your library with Roberta for sequence classification, the problem raises when I use the lib with new transformers(3.4.0). ``` python from transformers import AutoModelForSequenceClassification import turbo_transformers...
感谢作者开源，提两个关于docker build的问题: 1. nvidia-docker run --gpus all 这个命令要求docker engine version >= 19.03 否则会报错 unknown flag: --gpus 对于docker < 19.03, 去掉--gpus all 执行 nvidia-docker run 2. docker 打包成镜像，然后再进入containter内部build, build的过程中还要再拉取依赖 这个过程对于对于网络隔离的环境很麻烦. 个人觉得...
``` turbo_transformers.set_num_threads(?) ``` Whatever number I put in there, it is always using 4 threads.
I tried to install the turbotransformers into my own docker image. There is no conda in my docker image. I installed packages directly using apt-get and pip. However, when executing...
When I run the "build_docker_gpu.sh", I got permission denied error. It occurred to me that it is a shared server, I don't have the root privilege. What I want to...
