metadata
library_name: peft
tags:
- alignment-handbook
- trl
- dpo
- generated_from_trainer
base_model: NorLLM-AI/NorMistral-7B
datasets:
- hugodk-sch/aftonposten_title_prefs
model-index:
- name: norllm-ai-normistral-7b-align-scan
results: []
norllm-ai-normistral-7b-align-scan
This model is a fine-tuned version of data/norllm-ai-normistral-7b-sft-qlora on the hugodk-sch/aftonposten_title_prefs dataset. It achieves the following results on the evaluation set:
- Loss: 23.7465
- Rewards/chosen: -0.0136
- Rewards/rejected: -0.0343
- Rewards/accuracies: 0.5801
- Rewards/margins: 0.0207
- Logps/rejected: -35.0411
- Logps/chosen: -31.4188
- Logits/rejected: -2.8155
- Logits/chosen: -2.8182
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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 22.8693 | 0.26 | 100 | 23.9885 | 0.0019 | -0.0147 | 0.6130 | 0.0166 | -34.8449 | -31.2633 | -2.8087 | -2.8112 |
| 19.5946 | 0.52 | 200 | 23.7575 | -0.0117 | -0.0312 | 0.5718 | 0.0194 | -35.0097 | -31.3997 | -2.8172 | -2.8197 |
| 18.7484 | 0.78 | 300 | 23.8556 | -0.0119 | -0.0312 | 0.5718 | 0.0193 | -35.0104 | -31.4016 | -2.8159 | -2.8182 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.1