|
--- |
|
license: mit |
|
library_name: peft |
|
tags: |
|
- generated_from_trainer |
|
base_model: microsoft/phi-2 |
|
model-index: |
|
- name: fine-tuning-Phi2-with-webglm-qa-with-lora_5 |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# fine-tuning-Phi2-with-webglm-qa-with-lora_5 |
|
|
|
This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0878 |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 2 |
|
- eval_batch_size: 2 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 5 |
|
- total_train_batch_size: 10 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 100 |
|
- training_steps: 1000 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 8.1591 | 0.2 | 10 | 7.9109 | |
|
| 7.8077 | 0.4 | 20 | 7.4417 | |
|
| 6.7423 | 0.6 | 30 | 6.1597 | |
|
| 5.2815 | 0.8 | 40 | 3.7018 | |
|
| 2.6395 | 1.0 | 50 | 1.1413 | |
|
| 0.7209 | 1.2 | 60 | 0.6488 | |
|
| 0.5959 | 1.39 | 70 | 0.5735 | |
|
| 0.5036 | 1.59 | 80 | 0.5102 | |
|
| 0.4103 | 1.79 | 90 | 0.4500 | |
|
| 0.3433 | 1.99 | 100 | 0.3905 | |
|
| 0.3235 | 2.19 | 110 | 0.3371 | |
|
| 0.2567 | 2.39 | 120 | 0.3032 | |
|
| 0.2298 | 2.59 | 130 | 0.2785 | |
|
| 0.2451 | 2.79 | 140 | 0.2553 | |
|
| 0.1935 | 2.99 | 150 | 0.2363 | |
|
| 0.1946 | 3.19 | 160 | 0.2248 | |
|
| 0.1836 | 3.39 | 170 | 0.2097 | |
|
| 0.1681 | 3.59 | 180 | 0.1984 | |
|
| 0.1571 | 3.78 | 190 | 0.1877 | |
|
| 0.1713 | 3.98 | 200 | 0.1820 | |
|
| 0.15 | 4.18 | 210 | 0.1741 | |
|
| 0.1315 | 4.38 | 220 | 0.1696 | |
|
| 0.1567 | 4.58 | 230 | 0.1619 | |
|
| 0.1225 | 4.78 | 240 | 0.1528 | |
|
| 0.1346 | 4.98 | 250 | 0.1491 | |
|
| 0.1336 | 5.18 | 260 | 0.1464 | |
|
| 0.105 | 5.38 | 270 | 0.1427 | |
|
| 0.1245 | 5.58 | 280 | 0.1404 | |
|
| 0.1282 | 5.78 | 290 | 0.1363 | |
|
| 0.1042 | 5.98 | 300 | 0.1314 | |
|
| 0.1112 | 6.18 | 310 | 0.1264 | |
|
| 0.106 | 6.37 | 320 | 0.1249 | |
|
| 0.1043 | 6.57 | 330 | 0.1240 | |
|
| 0.1016 | 6.77 | 340 | 0.1196 | |
|
| 0.096 | 6.97 | 350 | 0.1179 | |
|
| 0.0927 | 7.17 | 360 | 0.1182 | |
|
| 0.0997 | 7.37 | 370 | 0.1146 | |
|
| 0.0914 | 7.57 | 380 | 0.1151 | |
|
| 0.0993 | 7.77 | 390 | 0.1128 | |
|
| 0.0863 | 7.97 | 400 | 0.1112 | |
|
| 0.0757 | 8.17 | 410 | 0.1100 | |
|
| 0.0803 | 8.37 | 420 | 0.1095 | |
|
| 0.0969 | 8.57 | 430 | 0.1084 | |
|
| 0.081 | 8.76 | 440 | 0.1079 | |
|
| 0.088 | 8.96 | 450 | 0.1050 | |
|
| 0.082 | 9.16 | 460 | 0.1036 | |
|
| 0.078 | 9.36 | 470 | 0.1019 | |
|
| 0.0782 | 9.56 | 480 | 0.1026 | |
|
| 0.0733 | 9.76 | 490 | 0.1010 | |
|
| 0.0754 | 9.96 | 500 | 0.1027 | |
|
| 0.0741 | 10.16 | 510 | 0.1011 | |
|
| 0.076 | 10.36 | 520 | 0.1023 | |
|
| 0.078 | 10.56 | 530 | 0.1010 | |
|
| 0.0701 | 10.76 | 540 | 0.0990 | |
|
| 0.0636 | 10.96 | 550 | 0.0974 | |
|
| 0.0668 | 11.16 | 560 | 0.0973 | |
|
| 0.0672 | 11.35 | 570 | 0.0972 | |
|
| 0.0634 | 11.55 | 580 | 0.0955 | |
|
| 0.061 | 11.75 | 590 | 0.0969 | |
|
| 0.0671 | 11.95 | 600 | 0.0956 | |
|
| 0.0611 | 12.15 | 610 | 0.0973 | |
|
| 0.061 | 12.35 | 620 | 0.0966 | |
|
| 0.0632 | 12.55 | 630 | 0.0950 | |
|
| 0.0655 | 12.75 | 640 | 0.0945 | |
|
| 0.0643 | 12.95 | 650 | 0.0944 | |
|
| 0.0557 | 13.15 | 660 | 0.0942 | |
|
| 0.0585 | 13.35 | 670 | 0.0937 | |
|
| 0.0582 | 13.55 | 680 | 0.0933 | |
|
| 0.0544 | 13.75 | 690 | 0.0927 | |
|
| 0.0663 | 13.94 | 700 | 0.0917 | |
|
| 0.0627 | 14.14 | 710 | 0.0917 | |
|
| 0.0561 | 14.34 | 720 | 0.0923 | |
|
| 0.0504 | 14.54 | 730 | 0.0914 | |
|
| 0.0656 | 14.74 | 740 | 0.0907 | |
|
| 0.0528 | 14.94 | 750 | 0.0898 | |
|
| 0.0581 | 15.14 | 760 | 0.0916 | |
|
| 0.0604 | 15.34 | 770 | 0.0912 | |
|
| 0.0467 | 15.54 | 780 | 0.0907 | |
|
| 0.048 | 15.74 | 790 | 0.0904 | |
|
| 0.0571 | 15.94 | 800 | 0.0902 | |
|
| 0.0521 | 16.14 | 810 | 0.0904 | |
|
| 0.052 | 16.33 | 820 | 0.0896 | |
|
| 0.0521 | 16.53 | 830 | 0.0895 | |
|
| 0.0498 | 16.73 | 840 | 0.0898 | |
|
| 0.0569 | 16.93 | 850 | 0.0887 | |
|
| 0.0481 | 17.13 | 860 | 0.0884 | |
|
| 0.0531 | 17.33 | 870 | 0.0889 | |
|
| 0.046 | 17.53 | 880 | 0.0886 | |
|
| 0.0492 | 17.73 | 890 | 0.0887 | |
|
| 0.0532 | 17.93 | 900 | 0.0885 | |
|
| 0.0511 | 18.13 | 910 | 0.0878 | |
|
| 0.0433 | 18.33 | 920 | 0.0881 | |
|
| 0.0518 | 18.53 | 930 | 0.0884 | |
|
| 0.049 | 18.73 | 940 | 0.0882 | |
|
| 0.0493 | 18.92 | 950 | 0.0880 | |
|
| 0.0479 | 19.12 | 960 | 0.0880 | |
|
| 0.0439 | 19.32 | 970 | 0.0880 | |
|
| 0.0535 | 19.52 | 980 | 0.0879 | |
|
| 0.0501 | 19.72 | 990 | 0.0878 | |
|
| 0.0466 | 19.92 | 1000 | 0.0878 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.1 |
|
- Transformers 4.36.2 |
|
- Pytorch 2.0.0 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
