metadata
library_name: peft
tags:
- alignment-handbook
- trl
- dpo
- generated_from_trainer
base_model: norallm/normistral-7b-warm
datasets:
- hugodk-sch/aftonposten_title_prefs
model-index:
- name: ap-normistral-7b-align-scan
results: []
ap-normistral-7b-align-scan
This model is a fine-tuned version of data/ap-normistral-7b-sft-qlora on the hugodk-sch/aftonposten_title_prefs dataset. It achieves the following results on the evaluation set:
- Loss: 0.9545
- Rewards/chosen: -0.0743
- Rewards/rejected: -0.1195
- Rewards/accuracies: 0.5889
- Rewards/margins: 0.0453
- Logps/rejected: -37.1618
- Logps/chosen: -33.1857
- Logits/rejected: 97.6826
- Logits/chosen: 97.7083
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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.9577 | 0.26 | 100 | 0.9935 | 0.0116 | 0.0052 | 0.5066 | 0.0064 | -35.9146 | -32.3275 | 98.7059 | 98.7128 |
| 0.8562 | 0.52 | 200 | 0.9647 | -0.0504 | -0.0855 | 0.5677 | 0.0351 | -36.8217 | -32.9469 | 97.9546 | 97.9719 |
| 0.8271 | 0.78 | 300 | 0.9534 | -0.0764 | -0.1230 | 0.5714 | 0.0466 | -37.1962 | -33.2069 | 97.6898 | 97.7152 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.1