I’m trying to get train results for text classification (https://github.com/huggingface/transformers/blob/main/examples/flax/text-classification)
using the following cmd:
export TASK_NAME=mrpc
python run_flax_glue.py \
--model_name_or_path bert-base-cased \
--task_name ${TASK_NAME} \
--max_seq_length 128 \
--learning_rate 2e-5 \
--num_train_epochs 3 \
--per_device_train_batch_size 4 \
--eval_steps 100 \
--output_dir ./$TASK_NAME/ \
--evaluation_strategy epoch \
--logging_dir ./$TASK_NAME/ \
--logging_strategy epoch \
--train_file ./$TASK_NAME/train.json \
--overwrite_output_dir \
--logging_steps 10 \
--do_train \
--do_eval
Unfortunately output folder does not contain expected train_results.json
(like here How to monitor both train and validation metrics at the same step? - #13 by lewtun). In my case, the output folder is as follows:
drwxrwxr-x 2 root root 180 Nov 21 21:23 ./
drwxrwxr-x 3 root root 140 Nov 21 21:41 ../
-rw-rw-r-- 1 root root 682 Nov 21 21:42 config.json
-rw-rw-r-- 1 root root 78 Nov 21 21:42 eval_results.json
-rw-rw-r-- 1 root root 433254456 Nov 21 21:42 flax_model.msgpack
-rw-rw-r-- 1 root root 125 Nov 21 21:42 special_tokens_map.json
-rw-rw-r-- 1 root root 669188 Nov 21 21:42 tokenizer.json
-rw-rw-r-- 1 root root 1191 Nov 21 21:42 tokenizer_config.json
-rw-rw-r-- 1 root root 213450 Nov 21 21:42 vocab.txt
Any hint/advice how to get train results for this model?