GaetanMichelet's picture
End of training
74df030 verified
metadata
base_model: google/gemma-2-2b-it
datasets:
  - GaetanMichelet/chat-60_ft_task-3_auto
library_name: peft
license: gemma
tags:
  - alignment-handbook
  - trl
  - sft
  - generated_from_trainer
model-index:
  - name: Gemma-2-2B_task-3_60-samples_config-1_auto
    results: []

Gemma-2-2B_task-3_60-samples_config-1_auto

This model is a fine-tuned version of google/gemma-2-2b-it on the GaetanMichelet/chat-60_ft_task-3_auto dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3994

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: 0.0001
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • distributed_type: multi-GPU
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 8
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss
3.0596 0.8696 5 2.9978
2.455 1.9130 11 1.6209
1.032 2.9565 17 0.7399
0.4025 4.0 23 0.4877
0.2824 4.8696 28 0.4448
0.2665 5.9130 34 0.4216
0.209 6.9565 40 0.4002
0.2776 8.0 46 0.3994
0.1608 8.8696 51 0.4351
0.1258 9.9130 57 0.5196
0.0748 10.9565 63 0.6227
0.0581 12.0 69 0.7061
0.0281 12.8696 74 0.7658
0.0178 13.9130 80 0.7745
0.006 14.9565 86 0.8173

Framework versions

  • PEFT 0.12.0
  • Transformers 4.44.0
  • Pytorch 2.1.2+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1