llama-3-8b-instruct-simpo
This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the princeton-nlp/llama3-ultrafeedback dataset. It achieves the following results on the evaluation set:
- Loss: 132.3632
- Rewards/chosen: -0.8503
- Rewards/rejected: -0.8889
- Rewards/accuracies: 0.5040
- Rewards/margins: 0.0387
- Logps/rejected: -0.3556
- Logps/chosen: -0.3401
- Logits/rejected: -1.2982
- Logits/chosen: -1.3372
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: 1e-06
- train_batch_size: 2
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- total_eval_batch_size: 32
- 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 196.6313 | 0.8549 | 400 | 132.3632 | -0.8503 | -0.8889 | 0.5040 | 0.0387 | -0.3556 | -0.3401 | -1.2982 | -1.3372 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.2.0
- Datasets 2.20.0
- Tokenizers 0.19.1
- Downloads last month
- 8