SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

QDQBert example update (#16395)

Browse Source 
* update Dockerfile and utils_qa

* Update README.md
This commit is contained in:
Shang Zhang
2022-03-28 02:47:52 -07:00
committed by GitHub
parent f6f6866e9e
commit 7ecbb9c5e4
3 changed files with 18 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
examples/research_projects/quantization-qdqbert/README.md
View File

@@ -39,9 +39,6 @@ Run the docker:
docker run --gpus all --privileged --rm -it --shm-size=1g --ulimit memlock=-1 --ulimit stack=67108864 bert_quantization:latest
```
*Note that the current NGC pytorch container (pytorch:21.07-py3) has TensorRT 8.0 which doesn't meet the requiremnt of TensorRT >= 8.2. One can either update the Dockerfile with the latest [NGC pytorch container](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/pytorch) once it supports TensorRT 8.2, or manually download and install [TensorRT >= 8.2](https://developer.nvidia.com/nvidia-tensorrt-download) in the container.*
In the container:
```
cd transformers/examples/research_projects/quantization-qdqbert/