0314_i6dh_mobilebert_int8vs_lr7e-05

This model is a fine-tuned version of csarron/mobilebert-uncased-squad-v1 on the squad dataset.

The model has 12 blocks, and is quantized. The embeddings are also quantized.

    {
      "epoch": 15.0,
      "eval_exact_match": 83.68968779564806,
      "eval_f1": 90.37926101768134,
      "step": 41500
    }

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: 7e-05
• train_batch_size: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine_with_restarts
• num_epochs: 15.0
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.26.0
• Pytorch 1.13.1+cu116
• Datasets 2.8.0
• Tokenizers 0.13.2