roberta-large-snli

This model was trained from scratch on the snli dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.3151
• eval_p: 0.9017
• eval_r: 0.9010
• eval_f1: 0.9012
• eval_runtime: 23.1208
• eval_samples_per_second: 424.898
• eval_steps_per_second: 26.556
• step: 0

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 4

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.1
• Datasets 2.6.1
• Tokenizers 0.12.1