<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

baseline

This model is a fine-tuned version of textattack/bert-base-uncased-imdb on the imdb dataset. It achieves the following results on the evaluation set:

• Loss: 0.5238
• Accuracy: 0.9209

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3.0
• mixed_precision_training: Native AMP

Training script

python run_glue.py \
  --model_name_or_path textattack/bert-base-uncased-imdb \
  --dataset_name imdb  \
  --do_train \
  --do_eval \
  --max_seq_length 384 \
  --pad_to_max_length False \
  --per_device_train_batch_size 32 \
  --per_device_eval_batch_size 32 \
  --fp16 \
  --learning_rate 5e-5 \
  --optim adamw_torch \
  --num_train_epochs 3 \
  --overwrite_output_dir \
  --output_dir /tmp/bert-base-uncased-imdb

Note: run_glue.py is modified to set the "test" split as evaluation dataset.

Framework versions

• Transformers 4.27.4
• Pytorch 1.13.1
• Datasets 2.11.0
• Tokenizers 0.13.3