`evaluation_strategy` is deprecated and will be removed in version 4.46 of Transformers

README.md

@@ -352,7 +352,7 @@ but you could try with other hyperparameters to obtain the best training and eva
 
 ```
 python 5.run_clm-post.py --tokenizer_name AI4PD/ZymCTRL
- --do_train --do_eval --output_dir output --evaluation_strategy steps --eval_steps 10
+ --do_train --do_eval --output_dir output --eval_strategy steps --eval_steps 10
  --logging_steps 5 --save_steps 500 --num_train_epochs 28 --per_device_train_batch_size 1
  --per_device_eval_batch_size 4 --cache_dir '.' --save_total_limit 2 --learning_rate  0.8e-04
  --dataloader_drop_last True --model_name_or_path AI4PD/ZymCTRL