GEITje-7B-ultra / README.md
BramVanroy's picture
End of training
e86ae6c verified
|
raw
history blame
2.87 kB
metadata
license: cc-by-nc-4.0
base_model: BramVanroy/GEITje-ultra-sft
tags:
  - alignment-handbook
  - generated_from_trainer
  - trl
  - dpo
  - generated_from_trainer
datasets:
  - BramVanroy/ultra_feedback_dutch
model-index:
  - name: GEITje-ultra-dpo-5e-7lr-128tbs-0.1b
    results: []

GEITje-ultra-dpo-5e-7lr-128tbs-0.1b

This model is a fine-tuned version of BramVanroy/GEITje-ultra-sft on the BramVanroy/ultra_feedback_dutch dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0138
  • Rewards/chosen: -2.1351
  • Rewards/rejected: -13.8922
  • Rewards/accuracies: 0.9950
  • Rewards/margins: 11.7570
  • Logps/rejected: -565.1809
  • Logps/chosen: -519.8008
  • Logits/rejected: -3.0261
  • Logits/chosen: -2.9779

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-07
  • train_batch_size: 4
  • eval_batch_size: 4
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 8
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • total_eval_batch_size: 32
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 1.0

Training results

Training Loss Epoch Step Validation Loss Rewards/chosen Rewards/rejected Rewards/accuracies Rewards/margins Logps/rejected Logps/chosen Logits/rejected Logits/chosen
0.03 0.22 100 0.0260 -0.9740 -9.8635 0.9913 8.8895 -524.8940 -508.1891 -3.0753 -3.0315
0.0184 0.44 200 0.0164 -1.7162 -12.4772 0.9926 10.7610 -551.0317 -515.6115 -3.0349 -2.9873
0.0121 0.66 300 0.0142 -2.0575 -13.6818 0.9938 11.6244 -563.0778 -519.0242 -3.0325 -2.9835
0.0198 0.88 400 0.0139 -2.1431 -13.8857 0.9950 11.7426 -565.1163 -519.8801 -3.0293 -2.9801

Framework versions

  • Transformers 4.36.2
  • Pytorch 2.1.2+cu121
  • Datasets 2.14.6
  • Tokenizers 0.15.0