Edit model card

bloom-560m-finetuned-samsum

This model is a fine-tuned version of bigscience/bloom-560m on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 2.9178

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: 1
  • eval_batch_size: 1
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 4
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 6
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
2.7663 0.63 200 2.6934
2.3769 1.26 400 2.6274
2.2776 1.89 600 2.5818
1.873 2.52 800 2.7177
1.6715 3.15 1000 2.9178
1.4515 3.78 1200 2.8924
1.0522 4.42 1400 3.3753
1.0237 5.05 1600 3.8098
0.7416 5.68 1800 3.9139

Framework versions

  • Transformers 4.22.2
  • Pytorch 1.12.1+cu113
  • Datasets 2.5.1
  • Tokenizers 0.12.1
Downloads last month
6
Safetensors
Model size
559M params
Tensor type
F32
·