GaetanMichelet's picture
End of training
78f4bb5 verified
metadata
base_model: google/gemma-2-2b-it
datasets:
  - GaetanMichelet/chat-60_ft_task-3
  - GaetanMichelet/chat-120_ft_task-3
library_name: peft
license: gemma
tags:
  - alignment-handbook
  - trl
  - sft
  - generated_from_trainer
model-index:
  - name: Gemma-2-2B_task-3_120-samples_config-1
    results: []

Gemma-2-2B_task-3_120-samples_config-1

This model is a fine-tuned version of google/gemma-2-2b-it on the GaetanMichelet/chat-60_ft_task-3 and the GaetanMichelet/chat-120_ft_task-3 datasets. It achieves the following results on the evaluation set:

  • Loss: 0.4556

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
2.9782 1.0 11 2.7757
1.0655 2.0 22 1.0496
0.3407 3.0 33 0.5812
0.4394 4.0 44 0.4947
0.3251 5.0 55 0.4556
0.2914 6.0 66 0.4622
0.1443 7.0 77 0.5258
0.112 8.0 88 0.5946
0.059 9.0 99 0.7452
0.0304 10.0 110 0.8398
0.0078 11.0 121 0.8925
0.0205 12.0 132 0.9054

Framework versions

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