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

Updates args in tf squad example. (#4820)

Browse Source 
Co-authored-by: Daniel Shan <daniel.shan@workday.com>
This commit is contained in:
daniel-shan
2020-06-08 02:36:09 -07:00
committed by GitHub
parent e33fdc93b4
commit b6f365a8ed
1 changed files with 4 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
examples/question-answering/README.md
View File

@@ -165,17 +165,15 @@ Larger batch size may improve the performance while costing more memory.
python run_tf_squad.py \ python run_tf_squad.py \
--model_name_or_path bert-base-uncased \ --model_name_or_path bert-base-uncased \
--output_dir model \ --output_dir model \
--max-seq-length 384 \ --max_seq_length 384 \
--num_train_epochs 2 \ --num_train_epochs 2 \
--per_gpu_train_batch_size 8 \ --per_gpu_train_batch_size 8 \
--per_gpu_eval_batch_size 16 \ --per_gpu_eval_batch_size 16 \
--do_train \ --do_train \
--logging_dir logs \ --logging_dir logs \
--mode question-answering \
--logging_steps 10 \ --logging_steps 10 \
--learning_rate 3e-5 \ --learning_rate 3e-5 \
--doc_stride 128 \ --doc_stride 128
--optimizer_name adamw
``` ```
For the moment the evaluation is not available in the Tensorflow Trainer only the training. For the moment the evaluation is not available in the Tensorflow Trainer only the training.