metadata
license: llama3.1
base_model: Magpie-Align/Llama-3.1-8B-Magpie-SFT-650KR
tags:
- trl
- dpo
- generated_from_trainer
model-index:
- name: Llama-3.1-8B-Magpie-SFT-650KR-Magpo-Armorm-3.1-70B-05
results: []
Llama-3.1-8B-Magpie-SFT-650KR-Magpo-Armorm-3.1-70B-05
This model is a fine-tuned version of Magpie-Align/Llama-3.1-8B-Magpie-SFT-650KR on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3335
- Rewards/chosen: -4.8366
- Rewards/rejected: -7.5394
- Rewards/accuracies: 0.8880
- Rewards/margins: 2.7028
- Logps/rejected: -1104.0730
- Logps/chosen: -827.6954
- Logits/rejected: -0.8119
- Logits/chosen: -0.8042
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-07
- train_batch_size: 2
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- total_eval_batch_size: 16
- 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.5603 | 0.1306 | 100 | 0.5762 | -1.0828 | -1.5526 | 0.7620 | 0.4698 | -505.3885 | -452.3145 | -0.7241 | -0.7285 |
| 0.5441 | 0.2612 | 200 | 0.4445 | -3.4116 | -5.1002 | 0.8360 | 1.6886 | -860.1481 | -685.1905 | -0.6966 | -0.6964 |
| 0.3586 | 0.3919 | 300 | 0.3949 | -3.4100 | -5.2798 | 0.8720 | 1.8698 | -878.1118 | -685.0309 | -0.7677 | -0.7653 |
| 0.3737 | 0.5225 | 400 | 0.3653 | -4.3580 | -6.6737 | 0.8760 | 2.3157 | -1017.5 | -779.8291 | -0.7777 | -0.7711 |
| 0.2611 | 0.6531 | 500 | 0.3457 | -4.9017 | -7.6712 | 0.8860 | 2.7695 | -1117.2515 | -834.2015 | -0.8137 | -0.8074 |
| 0.3342 | 0.7837 | 600 | 0.3354 | -4.7041 | -7.3342 | 0.8920 | 2.6301 | -1083.5503 | -814.4402 | -0.8081 | -0.7999 |
| 0.3251 | 0.9144 | 700 | 0.3335 | -4.8366 | -7.5394 | 0.8880 | 2.7028 | -1104.0730 | -827.6954 | -0.8119 | -0.8042 |
Framework versions
- Transformers 4.43.3
- Pytorch 2.4.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1