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

[s2s] distributed eval allows num_return_sequences > 1 (#7254)

Browse Source
This commit is contained in:
Sam Shleifer
2020-09-24 17:30:09 -04:00
committed by GitHub
parent 0804d077c6
commit d9d0f1140b
5 changed files with 31 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
examples/seq2seq/README.md
View File

@@ -235,7 +235,7 @@ export DATA_DIR=cnn_dm
--fp16 \
--bs 32
```
### Multi-GPU Evalulation
### Multi-GPU Evaluation
here is a command to run xsum evaluation on 8 GPUS. It is more than linearly faster than run_eval.py in some cases
because it uses SortishSampler to minimize padding. You can also use it on 1 GPU. `data_dir` must have
`{type_path}.source` and `{type_path}.target`. Run `./run_distributed_eval.py --help` for all clargs.
@@ -250,7 +250,7 @@ python -m torch.distributed.launch --nproc_per_node=8 run_distributed_eval.py \
Contributions that implement this command for other distributed hardware setups are welcome!
#### run_eval tips and tricks
#### Single-GPU Eval: Tips and Tricks
When using `run_eval.py`, the following features can be useful: