metadata
base_model: meta-llama/Llama-3.2-3B-Instruct
library_name: peft
license: llama3.2
tags:
- trl
- sft
- generated_from_trainer
model-index:
- name: results_1011
results: []
results_1011
This model is a fine-tuned version of meta-llama/Llama-3.2-3B-Instruct on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.7021
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: 3
- eval_batch_size: 6
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 24
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.01
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.8038 | 0.08 | 100 | 2.5553 |
| 2.507 | 0.16 | 200 | 2.4913 |
| 2.4522 | 0.24 | 300 | 2.4433 |
| 2.417 | 0.32 | 400 | 2.4122 |
| 2.3949 | 0.4 | 500 | 2.3822 |
| 2.3586 | 0.48 | 600 | 2.3510 |
| 2.3359 | 0.56 | 700 | 2.3276 |
| 2.3081 | 0.64 | 800 | 2.3043 |
| 2.307 | 0.72 | 900 | 2.2872 |
| 2.2741 | 0.8 | 1000 | 2.2695 |
| 2.2594 | 0.88 | 1100 | 2.2515 |
| 2.2591 | 0.96 | 1200 | 2.2363 |
| 2.2163 | 1.04 | 1300 | 2.2191 |
| 2.1996 | 1.12 | 1400 | 2.2057 |
| 2.1729 | 1.2 | 1500 | 2.1928 |
| 2.1718 | 1.28 | 1600 | 2.1799 |
| 2.1586 | 1.3600 | 1700 | 2.1692 |
| 2.1386 | 1.44 | 1800 | 2.1530 |
| 2.1338 | 1.52 | 1900 | 2.1418 |
| 2.1189 | 1.6 | 2000 | 2.1307 |
| 2.1055 | 1.6800 | 2100 | 2.1187 |
| 2.1074 | 1.76 | 2200 | 2.1075 |
| 2.0919 | 1.8400 | 2300 | 2.0959 |
| 2.0812 | 1.92 | 2400 | 2.0845 |
| 2.0621 | 2.0 | 2500 | 2.0743 |
| 2.0172 | 2.08 | 2600 | 2.0666 |
| 2.0159 | 2.16 | 2700 | 2.0602 |
| 2.0075 | 2.24 | 2800 | 2.0476 |
| 2.0042 | 2.32 | 2900 | 2.0394 |
| 2.0062 | 2.4 | 3000 | 2.0262 |
| 1.989 | 2.48 | 3100 | 2.0157 |
| 1.9808 | 2.56 | 3200 | 2.0086 |
| 1.9792 | 2.64 | 3300 | 1.9985 |
| 1.9751 | 2.7200 | 3400 | 1.9904 |
| 1.963 | 2.8 | 3500 | 1.9810 |
| 1.9498 | 2.88 | 3600 | 1.9718 |
| 1.9495 | 2.96 | 3700 | 1.9662 |
| 1.9053 | 3.04 | 3800 | 1.9578 |
| 1.8905 | 3.12 | 3900 | 1.9486 |
| 1.8873 | 3.2 | 4000 | 1.9412 |
| 1.8963 | 3.2800 | 4100 | 1.9347 |
| 1.8847 | 3.36 | 4200 | 1.9274 |
| 1.8819 | 3.44 | 4300 | 1.9187 |
| 1.8789 | 3.52 | 4400 | 1.9151 |
| 1.8635 | 3.6 | 4500 | 1.9057 |
| 1.8557 | 3.68 | 4600 | 1.9010 |
| 1.8518 | 3.76 | 4700 | 1.8927 |
| 1.8444 | 3.84 | 4800 | 1.8863 |
| 1.8318 | 3.92 | 4900 | 1.8801 |
| 1.8387 | 4.0 | 5000 | 1.8737 |
| 1.7994 | 4.08 | 5100 | 1.8701 |
| 1.7866 | 4.16 | 5200 | 1.8634 |
| 1.8005 | 4.24 | 5300 | 1.8623 |
| 1.7951 | 4.32 | 5400 | 1.8558 |
| 1.7818 | 4.4 | 5500 | 1.8477 |
| 1.7874 | 4.48 | 5600 | 1.8426 |
| 1.7771 | 4.5600 | 5700 | 1.8386 |
| 1.7574 | 4.64 | 5800 | 1.8353 |
| 1.7758 | 4.72 | 5900 | 1.8273 |
| 1.7864 | 4.8 | 6000 | 1.8244 |
| 1.7741 | 4.88 | 6100 | 1.8262 |
| 1.7638 | 4.96 | 6200 | 1.8151 |
| 1.7485 | 5.04 | 6300 | 1.8085 |
| 1.7239 | 5.12 | 6400 | 1.8017 |
| 1.7231 | 5.2 | 6500 | 1.7985 |
| 1.7212 | 5.28 | 6600 | 1.7950 |
| 1.7183 | 5.36 | 6700 | 1.7907 |
| 1.7234 | 5.44 | 6800 | 1.7856 |
| 1.7082 | 5.52 | 6900 | 1.7830 |
| 1.7128 | 5.6 | 7000 | 1.7792 |
| 1.7114 | 5.68 | 7100 | 1.7743 |
| 1.7193 | 5.76 | 7200 | 1.7714 |
| 1.7093 | 5.84 | 7300 | 1.7672 |
| 1.6974 | 5.92 | 7400 | 1.7643 |
| 1.7176 | 6.0 | 7500 | 1.7599 |
| 1.6657 | 6.08 | 7600 | 1.7575 |
| 1.679 | 6.16 | 7700 | 1.7560 |
| 1.6663 | 6.24 | 7800 | 1.7526 |
| 1.6634 | 6.32 | 7900 | 1.7499 |
| 1.6736 | 6.4 | 8000 | 1.7466 |
| 1.661 | 6.48 | 8100 | 1.7448 |
| 1.6535 | 6.5600 | 8200 | 1.7438 |
| 1.6734 | 6.64 | 8300 | 1.7395 |
| 1.6611 | 6.72 | 8400 | 1.7370 |
| 1.6841 | 6.8 | 8500 | 1.7337 |
| 1.6735 | 6.88 | 8600 | 1.7331 |
| 1.6679 | 6.96 | 8700 | 1.7316 |
| 1.6459 | 7.04 | 8800 | 1.7305 |
| 1.6438 | 7.12 | 8900 | 1.7296 |
| 1.6436 | 7.2 | 9000 | 1.7283 |
| 1.6293 | 7.28 | 9100 | 1.7278 |
| 1.6424 | 7.36 | 9200 | 1.7252 |
| 1.64 | 7.44 | 9300 | 1.7244 |
| 1.6114 | 7.52 | 9400 | 1.7227 |
| 1.6331 | 7.6 | 9500 | 1.7214 |
| 1.628 | 7.68 | 9600 | 1.7173 |
| 1.6464 | 7.76 | 9700 | 1.7159 |
| 1.6355 | 7.84 | 9800 | 1.7138 |
| 1.6489 | 7.92 | 9900 | 1.7127 |
| 1.6436 | 8.0 | 10000 | 1.7113 |
| 1.6108 | 8.08 | 10100 | 1.7108 |
| 1.6252 | 8.16 | 10200 | 1.7097 |
| 1.6228 | 8.24 | 10300 | 1.7087 |
| 1.617 | 8.32 | 10400 | 1.7084 |
| 1.6255 | 8.4 | 10500 | 1.7079 |
| 1.6212 | 8.48 | 10600 | 1.7070 |
| 1.6146 | 8.56 | 10700 | 1.7068 |
| 1.625 | 8.64 | 10800 | 1.7060 |
| 1.6282 | 8.72 | 10900 | 1.7056 |
| 1.614 | 8.8 | 11000 | 1.7054 |
| 1.612 | 8.88 | 11100 | 1.7051 |
| 1.6145 | 8.96 | 11200 | 1.7040 |
| 1.6125 | 9.04 | 11300 | 1.7037 |
| 1.6282 | 9.12 | 11400 | 1.7030 |
| 1.6085 | 9.2 | 11500 | 1.7030 |
| 1.6008 | 9.28 | 11600 | 1.7027 |
| 1.6109 | 9.36 | 11700 | 1.7024 |
| 1.6318 | 9.44 | 11800 | 1.7023 |
| 1.5976 | 9.52 | 11900 | 1.7022 |
| 1.5975 | 9.6 | 12000 | 1.7022 |
| 1.6108 | 9.68 | 12100 | 1.7021 |
| 1.6158 | 9.76 | 12200 | 1.7021 |
| 1.6232 | 9.84 | 12300 | 1.7021 |
| 1.6109 | 9.92 | 12400 | 1.7021 |
| 1.6005 | 10.0 | 12500 | 1.7021 |
Framework versions
- PEFT 0.12.0
- Transformers 4.45.0
- Pytorch 2.4.0+cu121
- Datasets 2.21.0
- Tokenizers 0.20.1