metadata
license: apache-2.0
library_name: peft
tags:
- trl
- sft
- unsloth
- generated_from_trainer
base_model: unsloth/llama-2-7b-bnb-4bit
model-index:
- name: llama3-chat_10000_500
results: []
llama3-chat_10000_500
This model is a fine-tuned version of unsloth/llama-2-7b-bnb-4bit on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1126
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: 8
- eval_batch_size: 4
- seed: 3407
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 5
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.1238 | 0.33 | 104 | 0.9666 |
| 1.0103 | 0.67 | 208 | 0.9480 |
| 1.0056 | 1.0 | 312 | 0.9424 |
| 0.921 | 1.33 | 416 | 0.9508 |
| 0.9252 | 1.66 | 520 | 0.9476 |
| 0.9219 | 2.0 | 624 | 0.9415 |
| 0.7968 | 2.33 | 728 | 0.9808 |
| 0.8012 | 2.66 | 832 | 0.9787 |
| 0.7975 | 3.0 | 936 | 0.9819 |
| 0.674 | 3.33 | 1040 | 1.0476 |
| 0.6638 | 3.66 | 1144 | 1.0509 |
| 0.6687 | 3.99 | 1248 | 1.0456 |
| 0.5858 | 4.33 | 1352 | 1.1100 |
| 0.5783 | 4.66 | 1456 | 1.1124 |
| 0.581 | 4.99 | 1560 | 1.1126 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.3
- Pytorch 2.2.2+cu121
- Datasets 2.16.0
- Tokenizers 0.15.2