Update README.md

README.md

@@ -54,9 +54,9 @@ python -m torch.distributed.launch --nproc_per_node=${num_gpus} \
   --do_eval \
   --evaluation_strategy steps \
   --max_seq_length 256 \
-  --warmup_steps 1000 \
+  --warmup_steps 500 \
   --per_device_train_batch_size ${batch_size} \
-  --learning_rate 2.5e-5 \
+  --learning_rate 2e-5 \
   --num_train_epochs 3 \
   --output_dir $output_dir \
   --overwrite_output_dir \