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: 0.6803
- Rewards/chosen: -0.0838
- Rewards/rejected: -0.1384
- Rewards/accuracies: 0.5187
- Rewards/margins: 0.0546
- Logps/rejected: -36.6583
- Logps/chosen: -32.8620
- Logits/rejected: 98.2235
- Logits/chosen: 98.2555
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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.6551 | 0.26 | 100 | 0.6957 | -0.0035 | -0.0114 | 0.5224 | 0.0079 | -36.0235 | -32.4608 | 98.7375 | 98.7478 |
| 0.5948 | 0.52 | 200 | 0.6720 | -0.0629 | -0.1290 | 0.5515 | 0.0661 | -36.6115 | -32.7579 | 98.3088 | 98.3381 |
| 0.5903 | 0.78 | 300 | 0.6774 | -0.0728 | -0.1310 | 0.5341 | 0.0583 | -36.6218 | -32.8071 | 98.2471 | 98.2788 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.1