metadata
license: other
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: taide/TAIDE-LX-7B-Chat
model-index:
- name: ROE_QA_TAIDE-LX-7B-Chat_Q100_80_20_V4
results: []
ROE_QA_TAIDE-LX-7B-Chat_Q100_80_20_V4
This model is a fine-tuned version of taide/TAIDE-LX-7B-Chat on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3344
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: 1
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 4.8611 | 0.0321 | 100 | 3.4673 |
| 4.6379 | 0.0643 | 200 | 3.2251 |
| 3.817 | 0.0964 | 300 | 2.8123 |
| 3.2954 | 0.1285 | 400 | 2.4988 |
| 2.9281 | 0.1607 | 500 | 2.2574 |
| 2.2501 | 0.1928 | 600 | 1.9403 |
| 2.1188 | 0.2249 | 700 | 1.7765 |
| 1.8027 | 0.2571 | 800 | 1.6201 |
| 1.6638 | 0.2892 | 900 | 1.4923 |
| 1.6204 | 0.3213 | 1000 | 1.3941 |
| 1.6152 | 0.3535 | 1100 | 1.2081 |
| 1.0876 | 0.3856 | 1200 | 1.0603 |
| 1.2552 | 0.4177 | 1300 | 1.0356 |
| 0.9443 | 0.4499 | 1400 | 0.8980 |
| 0.9998 | 0.4820 | 1500 | 0.8488 |
| 1.0518 | 0.5141 | 1600 | 0.7801 |
| 0.984 | 0.5463 | 1700 | 0.7711 |
| 0.9595 | 0.5784 | 1800 | 0.6924 |
| 0.9363 | 0.6105 | 1900 | 0.6537 |
| 0.822 | 0.6427 | 2000 | 0.6399 |
| 0.8791 | 0.6748 | 2100 | 0.6050 |
| 0.8802 | 0.7069 | 2200 | 0.5914 |
| 0.788 | 0.7391 | 2300 | 0.5741 |
| 0.7853 | 0.7712 | 2400 | 0.5581 |
| 0.839 | 0.8033 | 2500 | 0.5398 |
| 0.847 | 0.8355 | 2600 | 0.5204 |
| 0.8446 | 0.8676 | 2700 | 0.5085 |
| 0.7599 | 0.8997 | 2800 | 0.4867 |
| 0.7602 | 0.9319 | 2900 | 0.4891 |
| 0.8813 | 0.9640 | 3000 | 0.4717 |
| 0.6966 | 0.9961 | 3100 | 0.4667 |
| 0.7106 | 1.0283 | 3200 | 0.4463 |
| 0.6553 | 1.0604 | 3300 | 0.4414 |
| 0.6721 | 1.0925 | 3400 | 0.4383 |
| 0.6625 | 1.1247 | 3500 | 0.4326 |
| 0.6458 | 1.1568 | 3600 | 0.4255 |
| 0.6491 | 1.1889 | 3700 | 0.4207 |
| 0.6773 | 1.2211 | 3800 | 0.4246 |
| 0.6972 | 1.2532 | 3900 | 0.4119 |
| 0.7315 | 1.2853 | 4000 | 0.4101 |
| 0.6167 | 1.3175 | 4100 | 0.4119 |
| 0.6052 | 1.3496 | 4200 | 0.4070 |
| 0.6168 | 1.3817 | 4300 | 0.4029 |
| 0.6283 | 1.4139 | 4400 | 0.3986 |
| 0.6434 | 1.4460 | 4500 | 0.3940 |
| 0.6244 | 1.4781 | 4600 | 0.3940 |
| 0.6385 | 1.5103 | 4700 | 0.3958 |
| 0.6577 | 1.5424 | 4800 | 0.3952 |
| 0.6616 | 1.5746 | 4900 | 0.3925 |
| 0.6295 | 1.6067 | 5000 | 0.3846 |
| 0.6327 | 1.6388 | 5100 | 0.3832 |
| 0.6539 | 1.6710 | 5200 | 0.3826 |
| 0.6291 | 1.7031 | 5300 | 0.3851 |
| 0.5879 | 1.7352 | 5400 | 0.3833 |
| 0.6002 | 1.7674 | 5500 | 0.3787 |
| 0.5673 | 1.7995 | 5600 | 0.3755 |
| 0.5956 | 1.8316 | 5700 | 0.3746 |
| 0.6186 | 1.8638 | 5800 | 0.3735 |
| 0.5756 | 1.8959 | 5900 | 0.3712 |
| 0.6281 | 1.9280 | 6000 | 0.3735 |
| 0.5736 | 1.9602 | 6100 | 0.3696 |
| 0.5762 | 1.9923 | 6200 | 0.3700 |
| 0.3585 | 2.0244 | 6300 | 0.3667 |
| 0.4526 | 2.0566 | 6400 | 0.3662 |
| 0.3883 | 2.0887 | 6500 | 0.3662 |
| 0.4083 | 2.1208 | 6600 | 0.3664 |
| 0.3973 | 2.1530 | 6700 | 0.3646 |
| 0.4206 | 2.1851 | 6800 | 0.3673 |
| 0.5028 | 2.2172 | 6900 | 0.3622 |
| 0.366 | 2.2494 | 7000 | 0.3629 |
| 0.3968 | 2.2815 | 7100 | 0.3620 |
| 0.3977 | 2.3136 | 7200 | 0.3582 |
| 0.3899 | 2.3458 | 7300 | 0.3583 |
| 0.393 | 2.3779 | 7400 | 0.3572 |
| 0.4467 | 2.4100 | 7500 | 0.3576 |
| 0.4327 | 2.4422 | 7600 | 0.3605 |
| 0.4128 | 2.4743 | 7700 | 0.3550 |
| 0.3888 | 2.5064 | 7800 | 0.3572 |
| 0.4373 | 2.5386 | 7900 | 0.3553 |
| 0.3817 | 2.5707 | 8000 | 0.3525 |
| 0.3852 | 2.6028 | 8100 | 0.3518 |
| 0.4573 | 2.6350 | 8200 | 0.3536 |
| 0.3833 | 2.6671 | 8300 | 0.3499 |
| 0.3731 | 2.6992 | 8400 | 0.3509 |
| 0.3867 | 2.7314 | 8500 | 0.3497 |
| 0.351 | 2.7635 | 8600 | 0.3475 |
| 0.4327 | 2.7956 | 8700 | 0.3485 |
| 0.4183 | 2.8278 | 8800 | 0.3478 |
| 0.3807 | 2.8599 | 8900 | 0.3488 |
| 0.397 | 2.8920 | 9000 | 0.3461 |
| 0.4218 | 2.9242 | 9100 | 0.3463 |
| 0.3684 | 2.9563 | 9200 | 0.3457 |
| 0.3992 | 2.9884 | 9300 | 0.3448 |
| 0.2395 | 3.0206 | 9400 | 0.3463 |
| 0.2221 | 3.0527 | 9500 | 0.3448 |
| 0.222 | 3.0848 | 9600 | 0.3451 |
| 0.2359 | 3.1170 | 9700 | 0.3439 |
| 0.227 | 3.1491 | 9800 | 0.3442 |
| 0.2083 | 3.1812 | 9900 | 0.3430 |
| 0.2141 | 3.2134 | 10000 | 0.3419 |
| 0.224 | 3.2455 | 10100 | 0.3429 |
| 0.2031 | 3.2776 | 10200 | 0.3420 |
| 0.1944 | 3.3098 | 10300 | 0.3414 |
| 0.2109 | 3.3419 | 10400 | 0.3397 |
| 0.2053 | 3.3740 | 10500 | 0.3412 |
| 0.2172 | 3.4062 | 10600 | 0.3398 |
| 0.2179 | 3.4383 | 10700 | 0.3385 |
| 0.2299 | 3.4704 | 10800 | 0.3383 |
| 0.1721 | 3.5026 | 10900 | 0.3389 |
| 0.2131 | 3.5347 | 11000 | 0.3387 |
| 0.2489 | 3.5668 | 11100 | 0.3376 |
| 0.224 | 3.5990 | 11200 | 0.3373 |
| 0.201 | 3.6311 | 11300 | 0.3370 |
| 0.2131 | 3.6632 | 11400 | 0.3364 |
| 0.1869 | 3.6954 | 11500 | 0.3369 |
| 0.2174 | 3.7275 | 11600 | 0.3358 |
| 0.1999 | 3.7596 | 11700 | 0.3355 |
| 0.2224 | 3.7918 | 11800 | 0.3352 |
| 0.2195 | 3.8239 | 11900 | 0.3346 |
| 0.1984 | 3.8560 | 12000 | 0.3344 |
| 0.2121 | 3.8882 | 12100 | 0.3337 |
| 0.1879 | 3.9203 | 12200 | 0.3344 |
Framework versions
- PEFT 0.12.1.dev0
- Transformers 4.44.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1