metadata
base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
datasets:
- hansh/hansken_hql
library_name: peft
license: llama3.1
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
model-index:
- name: hansken_human_hql
results: []
hansken_human_hql
This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B-Instruct on the hansh/hansken_hql dataset. It achieves the following results on the evaluation set:
- Loss: 0.2362
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: 1
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 8
- 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: 20
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.4508 | 0.9976 | 102 | 0.4433 |
| 0.302 | 1.9951 | 204 | 0.3140 |
| 0.2692 | 2.9927 | 306 | 0.2616 |
| 0.177 | 4.0 | 409 | 0.2431 |
| 0.1616 | 4.9976 | 511 | 0.2362 |
| 0.1358 | 5.9951 | 613 | 0.2394 |
| 0.1199 | 6.9927 | 715 | 0.2474 |
| 0.1051 | 8.0 | 818 | 0.2625 |
| 0.0945 | 8.9976 | 920 | 0.2797 |
| 0.0843 | 9.9951 | 1022 | 0.2892 |
Framework versions
- PEFT 0.12.0
- Transformers 4.44.0
- Pytorch 2.1.2+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1