|
--- |
|
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_V5 |
|
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. --> |
|
|
|
# ROE_QA_TAIDE-LX-7B-Chat_Q100_80_20_V5 |
|
|
|
This model is a fine-tuned version of [taide/TAIDE-LX-7B-Chat](https://huggingface.co/taide/TAIDE-LX-7B-Chat) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.3726 |
|
|
|
## 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 | |
|
|:-------------:|:------:|:----:|:---------------:| |
|
| 4.8629 | 0.0321 | 100 | 3.4698 | |
|
| 4.6474 | 0.0643 | 200 | 3.2343 | |
|
| 3.8261 | 0.0964 | 300 | 2.8205 | |
|
| 3.2992 | 0.1285 | 400 | 2.5034 | |
|
| 2.9369 | 0.1607 | 500 | 2.2639 | |
|
| 2.2674 | 0.1928 | 600 | 1.9493 | |
|
| 2.1393 | 0.2249 | 700 | 1.7888 | |
|
| 1.8195 | 0.2571 | 800 | 1.6285 | |
|
| 1.678 | 0.2892 | 900 | 1.4983 | |
|
| 1.6242 | 0.3213 | 1000 | 1.4013 | |
|
| 1.344 | 0.3535 | 1100 | 1.1522 | |
|
| 1.0894 | 0.3856 | 1200 | 1.0704 | |
|
| 1.2033 | 0.4177 | 1300 | 1.0537 | |
|
| 0.9503 | 0.4499 | 1400 | 0.9160 | |
|
| 0.9901 | 0.4820 | 1500 | 0.8751 | |
|
| 1.0363 | 0.5141 | 1600 | 0.7942 | |
|
| 0.9986 | 0.5463 | 1700 | 0.7668 | |
|
| 0.9407 | 0.5784 | 1800 | 0.6912 | |
|
| 0.9347 | 0.6105 | 1900 | 0.6543 | |
|
| 0.8109 | 0.6427 | 2000 | 0.6498 | |
|
| 0.8848 | 0.6748 | 2100 | 0.6077 | |
|
| 0.8937 | 0.7069 | 2200 | 0.5865 | |
|
| 0.7895 | 0.7391 | 2300 | 0.5780 | |
|
| 0.8044 | 0.7712 | 2400 | 0.5646 | |
|
| 0.8317 | 0.8033 | 2500 | 0.5449 | |
|
| 0.858 | 0.8355 | 2600 | 0.5132 | |
|
| 0.8519 | 0.8676 | 2700 | 0.4940 | |
|
| 0.7554 | 0.8997 | 2800 | 0.4972 | |
|
| 0.758 | 0.9319 | 2900 | 0.4809 | |
|
| 0.8866 | 0.9640 | 3000 | 0.4714 | |
|
| 0.7028 | 0.9961 | 3100 | 0.4608 | |
|
| 0.7031 | 1.0283 | 3200 | 0.4458 | |
|
| 0.6623 | 1.0604 | 3300 | 0.4427 | |
|
| 0.671 | 1.0925 | 3400 | 0.4366 | |
|
| 0.6588 | 1.1247 | 3500 | 0.4327 | |
|
| 0.6422 | 1.1568 | 3600 | 0.4239 | |
|
| 0.643 | 1.1889 | 3700 | 0.4235 | |
|
| 0.6747 | 1.2211 | 3800 | 0.4204 | |
|
| 0.6911 | 1.2532 | 3900 | 0.4130 | |
|
| 0.7354 | 1.2853 | 4000 | 0.4092 | |
|
| 0.6233 | 1.3175 | 4100 | 0.4070 | |
|
| 0.6005 | 1.3496 | 4200 | 0.4055 | |
|
| 0.624 | 1.3817 | 4300 | 0.4033 | |
|
| 0.623 | 1.4139 | 4400 | 0.3976 | |
|
| 0.6419 | 1.4460 | 4500 | 0.3966 | |
|
| 0.6329 | 1.4781 | 4600 | 0.3914 | |
|
| 0.6395 | 1.5103 | 4700 | 0.3934 | |
|
| 0.6541 | 1.5424 | 4800 | 0.3916 | |
|
| 0.6538 | 1.5746 | 4900 | 0.3917 | |
|
| 0.6214 | 1.6067 | 5000 | 0.3840 | |
|
| 0.6303 | 1.6388 | 5100 | 0.3844 | |
|
| 0.6547 | 1.6710 | 5200 | 0.3816 | |
|
| 0.6264 | 1.7031 | 5300 | 0.3844 | |
|
| 0.5896 | 1.7352 | 5400 | 0.3801 | |
|
| 0.6082 | 1.7674 | 5500 | 0.3786 | |
|
| 0.5772 | 1.7995 | 5600 | 0.3737 | |
|
| 0.5839 | 1.8316 | 5700 | 0.3745 | |
|
| 0.6201 | 1.8638 | 5800 | 0.3737 | |
|
| 0.5643 | 1.8959 | 5900 | 0.3717 | |
|
| 0.6258 | 1.9280 | 6000 | 0.3726 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.12.1.dev0 |
|
- Transformers 4.44.2 |
|
- Pytorch 2.3.0+cu121 |
|
- Datasets 2.19.1 |
|
- Tokenizers 0.19.1 |