metadata
license: llama3
library_name: peft
tags:
- generated_from_trainer
base_model: meta-llama/Meta-Llama-3-8B-Instruct
model-index:
- name: final_llama
results: []
final_llama
This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.4315
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: 2e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 100
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
20.3519 | 1.0 | 1 | 20.2665 |
20.3519 | 2.0 | 2 | 20.1486 |
20.3519 | 3.0 | 3 | 20.0148 |
20.3519 | 4.0 | 4 | 19.8717 |
20.3519 | 5.0 | 5 | 19.7162 |
20.3519 | 6.0 | 6 | 19.5484 |
20.3519 | 7.0 | 7 | 19.3671 |
20.3519 | 8.0 | 8 | 19.1726 |
20.3519 | 9.0 | 9 | 18.9658 |
20.3519 | 10.0 | 10 | 18.7509 |
20.3519 | 11.0 | 11 | 18.5206 |
20.3519 | 12.0 | 12 | 18.2873 |
20.3519 | 13.0 | 13 | 18.0473 |
20.3519 | 14.0 | 14 | 17.8122 |
20.3519 | 15.0 | 15 | 17.5813 |
20.3519 | 16.0 | 16 | 17.3524 |
20.3519 | 17.0 | 17 | 17.1373 |
20.3519 | 18.0 | 18 | 16.9325 |
20.3519 | 19.0 | 19 | 16.7373 |
20.3519 | 20.0 | 20 | 16.5515 |
20.3519 | 21.0 | 21 | 16.3717 |
20.3519 | 22.0 | 22 | 16.1991 |
20.3519 | 23.0 | 23 | 16.0269 |
20.3519 | 24.0 | 24 | 15.8545 |
20.3519 | 25.0 | 25 | 15.6744 |
20.3519 | 26.0 | 26 | 15.4867 |
20.3519 | 27.0 | 27 | 15.2966 |
20.3519 | 28.0 | 28 | 15.0983 |
20.3519 | 29.0 | 29 | 14.8900 |
20.3519 | 30.0 | 30 | 14.6800 |
20.3519 | 31.0 | 31 | 14.4623 |
20.3519 | 32.0 | 32 | 14.2411 |
20.3519 | 33.0 | 33 | 14.0190 |
20.3519 | 34.0 | 34 | 13.7946 |
20.3519 | 35.0 | 35 | 13.5692 |
20.3519 | 36.0 | 36 | 13.3422 |
20.3519 | 37.0 | 37 | 13.1188 |
20.3519 | 38.0 | 38 | 12.8947 |
20.3519 | 39.0 | 39 | 12.6666 |
20.3519 | 40.0 | 40 | 12.4454 |
20.3519 | 41.0 | 41 | 12.2206 |
20.3519 | 42.0 | 42 | 11.9955 |
20.3519 | 43.0 | 43 | 11.7648 |
20.3519 | 44.0 | 44 | 11.5387 |
20.3519 | 45.0 | 45 | 11.3104 |
20.3519 | 46.0 | 46 | 11.0794 |
20.3519 | 47.0 | 47 | 10.8506 |
20.3519 | 48.0 | 48 | 10.6189 |
20.3519 | 49.0 | 49 | 10.3891 |
20.3519 | 50.0 | 50 | 10.1577 |
20.3519 | 51.0 | 51 | 9.9252 |
20.3519 | 52.0 | 52 | 9.6967 |
20.3519 | 53.0 | 53 | 9.4668 |
20.3519 | 54.0 | 54 | 9.2420 |
20.3519 | 55.0 | 55 | 9.0153 |
20.3519 | 56.0 | 56 | 8.7923 |
20.3519 | 57.0 | 57 | 8.5711 |
20.3519 | 58.0 | 58 | 8.3488 |
20.3519 | 59.0 | 59 | 8.1307 |
20.3519 | 60.0 | 60 | 7.9147 |
20.3519 | 61.0 | 61 | 7.7034 |
20.3519 | 62.0 | 62 | 7.4925 |
20.3519 | 63.0 | 63 | 7.2867 |
20.3519 | 64.0 | 64 | 7.0867 |
20.3519 | 65.0 | 65 | 6.8855 |
20.3519 | 66.0 | 66 | 6.6916 |
20.3519 | 67.0 | 67 | 6.5061 |
20.3519 | 68.0 | 68 | 6.3185 |
20.3519 | 69.0 | 69 | 6.1380 |
20.3519 | 70.0 | 70 | 5.9656 |
20.3519 | 71.0 | 71 | 5.7981 |
20.3519 | 72.0 | 72 | 5.6359 |
20.3519 | 73.0 | 73 | 5.4777 |
20.3519 | 74.0 | 74 | 5.3250 |
20.3519 | 75.0 | 75 | 5.1813 |
20.3519 | 76.0 | 76 | 5.0366 |
20.3519 | 77.0 | 77 | 4.9050 |
20.3519 | 78.0 | 78 | 4.7747 |
20.3519 | 79.0 | 79 | 4.6546 |
20.3519 | 80.0 | 80 | 4.5398 |
20.3519 | 81.0 | 81 | 4.4281 |
20.3519 | 82.0 | 82 | 4.3260 |
20.3519 | 83.0 | 83 | 4.2277 |
20.3519 | 84.0 | 84 | 4.1377 |
20.3519 | 85.0 | 85 | 4.0487 |
20.3519 | 86.0 | 86 | 3.9703 |
20.3519 | 87.0 | 87 | 3.8984 |
20.3519 | 88.0 | 88 | 3.8298 |
20.3519 | 89.0 | 89 | 3.7664 |
20.3519 | 90.0 | 90 | 3.7087 |
20.3519 | 91.0 | 91 | 3.6577 |
20.3519 | 92.0 | 92 | 3.6108 |
20.3519 | 93.0 | 93 | 3.5709 |
20.3519 | 94.0 | 94 | 3.5354 |
20.3519 | 95.0 | 95 | 3.5040 |
20.3519 | 96.0 | 96 | 3.4798 |
20.3519 | 97.0 | 97 | 3.4595 |
20.3519 | 98.0 | 98 | 3.4442 |
20.3519 | 99.0 | 99 | 3.4354 |
10.6394 | 100.0 | 100 | 3.4315 |
Framework versions
- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.1.2
- Datasets 2.19.2
- Tokenizers 0.19.1