Update README.md

README.md

@@ -12,7 +12,7 @@ test_rm_labeling:
   pooling: last
   sort_by_length: false
   use_custom_sampler: true
-  model_name: microsoft/deberta-v3-large
+  model_name: microsoft/deberta-v3-base
   learning_rate: 3e-5
   residual_dropout: 0.0
   weight_decay: 0.0
@@ -21,8 +21,8 @@ test_rm_labeling:
   gradient_checkpointing: true
   warmup_steps: 50
   dtype: float16
-  gradient_accumulation_steps: 20
-  per_device_train_batch_size: 1
+  gradient_accumulation_steps: 5
+  per_device_train_batch_size: 4
   per_device_eval_batch_size: 4
   num_train_epochs: 3
   eval_steps: 251