---
library_name: peft
tags:
- trl
- dpo
- alignment-handbook
- generated_from_trainer
base_model: NbAiLab/nb-gpt-j-6B-v2
model-index:
- name: aftonposten-6b-align-scan
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# aftonposten-6b-align-scan

This model is a fine-tuned version of [NbAiLab/nb-gpt-j-6B-v2](https://huggingface.co./NbAiLab/nb-gpt-j-6B-v2) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6893
- Rewards/chosen: -0.0505
- Rewards/rejected: -0.0590
- Rewards/accuracies: 0.5424
- Rewards/margins: 0.0085
- Logps/rejected: -43.4186
- Logps/chosen: -39.0870
- Logits/rejected: -1.6166
- Logits/chosen: -1.6206

## 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: 4

### Training results

| Training Loss | Epoch | Step | Logits/chosen | Logits/rejected | Logps/chosen | Logps/rejected | Validation Loss | Rewards/accuracies | Rewards/chosen | Rewards/margins | Rewards/rejected |
|:-------------:|:-----:|:----:|:-------------:|:---------------:|:------------:|:--------------:|:---------------:|:------------------:|:--------------:|:---------------:|:----------------:|
| 0.6924        | 0.26  | 100  | -2.2301       | -2.2252         | -34.0468     | -37.5393       | 0.6931          | 0.5187             | -0.0001        | 0.0001          | -0.0002          |
| 0.6914        | 0.52  | 200  | -2.2250       | -2.2201         | -34.0426     | -37.5530       | 0.6930          | 0.5245             | -0.0001        | 0.0003          | -0.0004          |
| 0.6895        | 0.78  | 300  | -2.2211       | -2.2163         | -34.0648     | -37.5868       | 0.6929          | 0.5419             | -0.0003        | 0.0004          | -0.0007          |
| 0.6863        | 1.04  | 400  | 0.6928        | -0.0011         | -0.0017      | 0.5681         | 0.0006          | -37.6890           | -34.1431       | -2.1889         | -2.1937          |
| 0.6826        | 1.3   | 500  | 0.6925        | -0.0031         | -0.0043      | 0.5652         | 0.0012          | -37.9472           | -34.3421       | -2.1173         | -2.1221          |
| 0.6823        | 1.56  | 600  | 0.6918        | -0.0070         | -0.0097      | 0.5831         | 0.0028          | -38.4905           | -34.7312       | -2.0241         | -2.0287          |
| 0.6784        | 1.82  | 700  | 0.6916        | -0.0132         | -0.0163      | 0.5656         | 0.0031          | -39.1509           | -35.3551       | -1.9419         | -1.9464          |
| 0.6729        | 2.08  | 800  | 0.6911        | -0.0191         | -0.0235      | 0.5540         | 0.0044          | -39.8622           | -35.9412       | -1.8648         | -1.8692          |
| 0.6635        | 2.34  | 900  | 0.6904        | -0.0309         | -0.0368      | 0.5365         | 0.0059          | -41.1917           | -37.1214       | -1.7445         | -1.7488          |
| 0.6719        | 2.6   | 1000 | 0.6897        | -0.0445         | -0.0522      | 0.5486         | 0.0077          | -42.7361           | -38.4865       | -1.6560         | -1.6600          |
| 0.6749        | 2.86  | 1100 | 0.6894        | -0.0493         | -0.0575      | 0.5365         | 0.0082          | -43.2672           | -38.9607       | -1.6246         | -1.6286          |
| 0.6551        | 3.12  | 1200 | 0.6893        | -0.0498         | -0.0582      | 0.5365         | 0.0084          | -43.3379           | -39.0115       | -1.6198         | -1.6238          |
| 0.6489        | 3.38  | 1300 | 0.6893        | -0.0504         | -0.0589      | 0.5395         | 0.0086          | -43.4114           | -39.0722       | -1.6172         | -1.6212          |
| 0.6597        | 3.64  | 1400 | 0.6893        | -0.0505         | -0.0591      | 0.5424         | 0.0086          | -43.4228           | -39.0843       | -1.6165         | -1.6205          |
| 0.6557        | 3.9   | 1500 | 0.6893        | -0.0505         | -0.0590      | 0.5424         | 0.0085          | -43.4186           | -39.0870       | -1.6166         | -1.6206          |


### Framework versions

- PEFT 0.8.2
- Transformers 4.37.2
- Pytorch 2.1.2+cu121
- Datasets 2.17.0
- Tokenizers 0.15.1