Transaminitis_L3_1000rate_1e7_SFT
This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.8218
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-07
- train_batch_size: 2
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.6855 | 0.2 | 25 | 2.6850 |
| 2.6485 | 0.4 | 50 | 2.6467 |
| 2.577 | 0.6 | 75 | 2.5655 |
| 2.4207 | 0.8 | 100 | 2.4148 |
| 2.2782 | 1.0 | 125 | 2.2240 |
| 2.0763 | 1.2 | 150 | 2.0603 |
| 1.948 | 1.4 | 175 | 1.9160 |
| 1.8184 | 1.6 | 200 | 1.7951 |
| 1.7176 | 1.8 | 225 | 1.6846 |
| 1.6019 | 2.0 | 250 | 1.5981 |
| 1.5479 | 2.2 | 275 | 1.5224 |
| 1.4609 | 2.4 | 300 | 1.4542 |
| 1.3825 | 2.6 | 325 | 1.3726 |
| 1.3066 | 2.8 | 350 | 1.3045 |
| 1.249 | 3.0 | 375 | 1.2377 |
| 1.1573 | 3.2 | 400 | 1.1604 |
| 1.105 | 3.4 | 425 | 1.1066 |
| 1.0526 | 3.6 | 450 | 1.0529 |
| 1.0201 | 3.8 | 475 | 1.0068 |
| 0.9541 | 4.0 | 500 | 0.9671 |
| 0.9304 | 4.2 | 525 | 0.9349 |
| 0.9083 | 4.4 | 550 | 0.9075 |
| 0.8843 | 4.6 | 575 | 0.8852 |
| 0.8636 | 4.8 | 600 | 0.8700 |
| 0.8526 | 5.0 | 625 | 0.8552 |
| 0.8318 | 5.2 | 650 | 0.8436 |
| 0.8228 | 5.4 | 675 | 0.8373 |
| 0.8247 | 5.6 | 700 | 0.8292 |
| 0.8196 | 5.8 | 725 | 0.8245 |
| 0.8208 | 6.0 | 750 | 0.8235 |
| 0.8084 | 6.2 | 775 | 0.8214 |
| 0.8109 | 6.4 | 800 | 0.8215 |
| 0.8248 | 6.6 | 825 | 0.8212 |
| 0.8115 | 6.8 | 850 | 0.8210 |
| 0.8216 | 7.0 | 875 | 0.8211 |
| 0.8089 | 7.2 | 900 | 0.8218 |
| 0.8109 | 7.4 | 925 | 0.8217 |
| 0.8282 | 7.6 | 950 | 0.8218 |
| 0.8121 | 7.8 | 975 | 0.8218 |
| 0.8102 | 8.0 | 1000 | 0.8218 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.0.0+cu117
- Datasets 2.19.1
- Tokenizers 0.19.1