metadata
license: apache-2.0
library_name: peft
tags:
- trl
- dpo
- generated_from_trainer
base_model: TheBloke/OpenHermes-2-Mistral-7B-GPTQ
model-index:
- name: openhermes-mistral-dpo-gptq
results: []
openhermes-mistral-dpo-gptq
This model is a fine-tuned version of TheBloke/OpenHermes-2-Mistral-7B-GPTQ on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.4715
- Rewards/chosen: -1.6019
- Rewards/rejected: -1.3634
- Rewards/accuracies: 0.6875
- Rewards/margins: -0.2386
- Logps/rejected: -342.2072
- Logps/chosen: -321.4847
- Logits/rejected: -2.4409
- Logits/chosen: -2.4704
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: 0.0002
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2
- training_steps: 50
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.6793 | 0.005 | 10 | 0.6783 | 0.1178 | 0.0391 | 0.5 | 0.0787 | -328.1823 | -304.2877 | -2.4405 | -2.4604 |
| 0.7008 | 0.01 | 20 | 0.6618 | 0.1939 | -0.0229 | 0.625 | 0.2168 | -328.8029 | -303.5266 | -2.4441 | -2.4621 |
| 0.706 | 0.015 | 30 | 0.7197 | 0.2223 | 0.0219 | 0.75 | 0.2004 | -328.3552 | -303.2429 | -2.4519 | -2.4654 |
| 0.6897 | 0.02 | 40 | 0.8159 | -0.1834 | -0.3501 | 0.75 | 0.1667 | -332.0744 | -307.2993 | -2.4459 | -2.4677 |
| 0.6443 | 0.025 | 50 | 1.4715 | -1.6019 | -1.3634 | 0.6875 | -0.2386 | -342.2072 | -321.4847 | -2.4409 | -2.4704 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.1
- Pytorch 2.0.1+cu117
- Datasets 2.19.1
- Tokenizers 0.19.1