hugodk-sch's picture
Model save
75f5528 verified
|
raw
history blame
2.52 kB
metadata
license: apache-2.0
library_name: peft
tags:
  - trl
  - dpo
  - generated_from_trainer
base_model: norallm/normistral-7b-warm
model-index:
  - name: ap-normistral-7b-align-scan
    results: []

ap-normistral-7b-align-scan

This model is a fine-tuned version of norallm/normistral-7b-warm on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0124
  • Rewards/chosen: -0.0636
  • Rewards/rejected: -0.1409
  • Rewards/accuracies: 0.5399
  • Rewards/margins: 0.0773
  • Logps/rejected: -36.1679
  • Logps/chosen: -32.5340
  • Logits/rejected: 98.7098
  • Logits/chosen: 98.7339

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-06
  • train_batch_size: 4
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • gradient_accumulation_steps: 2
  • 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: 1

Training results

Training Loss Epoch Step Validation Loss Rewards/chosen Rewards/rejected Rewards/accuracies Rewards/margins Logps/rejected Logps/chosen Logits/rejected Logits/chosen
0.9161 0.26 100 1.0662 -0.0346 -0.0243 0.5129 -0.0103 -36.0012 -32.4926 98.7623 98.7729
0.8398 0.52 200 1.0090 -0.1355 -0.2279 0.5141 0.0924 -36.2922 -32.6368 98.6916 98.7167
0.733 0.78 300 1.0124 -0.0636 -0.1409 0.5399 0.0773 -36.1679 -32.5340 98.7098 98.7339

Framework versions

  • PEFT 0.10.0
  • Transformers 4.39.0.dev0
  • Pytorch 2.1.2+cu121
  • Datasets 2.14.6
  • Tokenizers 0.15.1