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

Correct documentation

Browse Source
This commit is contained in:
Jared Nielsen
2020-01-30 13:51:54 -08:00
committed by Lysandre Debut
parent 01a14ebd8d
commit 71a382319f
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
examples/README.md
View File

@@ -404,12 +404,12 @@ exact_match = 81.22
#### Distributed training #### Distributed training
Here is an example using distributed training on 8 V100 GPUs and Bert Whole Word Masking uncased model to reach a F1 > 93 on SQuAD1.0: Here is an example using distributed training on 8 V100 GPUs and Bert Whole Word Masking uncased model to reach a F1 > 93 on SQuAD1.1:
```bash ```bash
python -m torch.distributed.launch --nproc_per_node=8 run_squad.py \ python -m torch.distributed.launch --nproc_per_node=8 ./examples/run_squad.py \
--model_type bert \ --model_type bert \
--model_name_or_path bert-base-cased \ --model_name_or_path bert-large-uncased-whole-word-masking \
--do_train \ --do_train \
--do_eval \ --do_eval \
--do_lower_case \ --do_lower_case \
@@ -419,9 +419,9 @@ python -m torch.distributed.launch --nproc_per_node=8 run_squad.py \
--num_train_epochs 2 \ --num_train_epochs 2 \
--max_seq_length 384 \ --max_seq_length 384 \
--doc_stride 128 \ --doc_stride 128 \
--output_dir ../models/wwm_uncased_finetuned_squad/ \ --output_dir ./examples/models/wwm_uncased_finetuned_squad/ \
--per_gpu_train_batch_size 24 \ --per_gpu_eval_batch_size=3 \
--gradient_accumulation_steps 12 --per_gpu_train_batch_size=3 \
``` ```
Training with the previously defined hyper-parameters yields the following results: Training with the previously defined hyper-parameters yields the following results: