|
--- |
|
license: apache-2.0 |
|
base_model: facebook/bart-large |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: LLM_Teached_Bart_From_Scratch |
|
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. --> |
|
|
|
# LLM_Teached_Bart_From_Scratch |
|
|
|
This model is a fine-tuned version of [facebook/bart-large](https://huggingface.co/facebook/bart-large) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.5434 |
|
- Rouge1: 0.4476 |
|
- Rouge2: 0.2292 |
|
- Rougel: 0.3868 |
|
- Rougelsum: 0.3865 |
|
- Gen Len: 19.9007 |
|
- Precision: 0.9159 |
|
- Recall: 0.8916 |
|
- F1: 0.9034 |
|
|
|
## 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: 24 |
|
- eval_batch_size: 16 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 96 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 16 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len | Precision | Recall | F1 | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|:-------:|:---------:|:------:|:------:| |
|
| 1.836 | 1.0 | 521 | 1.5560 | 0.4155 | 0.2028 | 0.3561 | 0.3559 | 19.9745 | 0.9105 | 0.8843 | 0.8971 | |
|
| 1.5951 | 2.0 | 1042 | 1.5004 | 0.4333 | 0.2136 | 0.3695 | 0.3694 | 19.9353 | 0.9115 | 0.8886 | 0.8997 | |
|
| 1.469 | 3.0 | 1563 | 1.4691 | 0.4355 | 0.2176 | 0.3729 | 0.3728 | 19.9385 | 0.912 | 0.8888 | 0.9001 | |
|
| 1.373 | 4.0 | 2084 | 1.4658 | 0.4311 | 0.2164 | 0.3706 | 0.3704 | 19.9647 | 0.9137 | 0.8877 | 0.9003 | |
|
| 1.2902 | 5.0 | 2605 | 1.4542 | 0.4368 | 0.2218 | 0.3762 | 0.376 | 19.9498 | 0.9136 | 0.8887 | 0.9008 | |
|
| 1.222 | 6.0 | 3126 | 1.4584 | 0.4407 | 0.223 | 0.3802 | 0.3798 | 19.9425 | 0.914 | 0.8902 | 0.9018 | |
|
| 1.1655 | 7.0 | 3647 | 1.4709 | 0.4404 | 0.2246 | 0.3806 | 0.3803 | 19.9327 | 0.9145 | 0.89 | 0.9019 | |
|
| 1.11 | 8.0 | 4168 | 1.4724 | 0.4435 | 0.2269 | 0.383 | 0.3828 | 19.9084 | 0.9153 | 0.8906 | 0.9026 | |
|
| 1.0629 | 9.0 | 4689 | 1.4853 | 0.4431 | 0.2273 | 0.3832 | 0.383 | 19.928 | 0.9155 | 0.8908 | 0.9028 | |
|
| 1.023 | 10.0 | 5210 | 1.5033 | 0.4409 | 0.2247 | 0.3819 | 0.3818 | 19.944 | 0.9152 | 0.8897 | 0.9021 | |
|
| 0.9862 | 11.0 | 5731 | 1.5074 | 0.4479 | 0.2278 | 0.3862 | 0.386 | 19.9124 | 0.9158 | 0.8916 | 0.9034 | |
|
| 0.957 | 12.0 | 6252 | 1.5184 | 0.4461 | 0.2264 | 0.3846 | 0.3847 | 19.9033 | 0.9159 | 0.8909 | 0.903 | |
|
| 0.9315 | 13.0 | 6773 | 1.5269 | 0.4473 | 0.2284 | 0.386 | 0.3858 | 19.9084 | 0.9156 | 0.8912 | 0.9031 | |
|
| 0.9093 | 14.0 | 7294 | 1.5311 | 0.4453 | 0.2273 | 0.3846 | 0.3843 | 19.9135 | 0.9155 | 0.8909 | 0.9029 | |
|
| 0.8927 | 15.0 | 7815 | 1.5351 | 0.4457 | 0.2267 | 0.3842 | 0.384 | 19.9065 | 0.9156 | 0.8909 | 0.9029 | |
|
| 0.8773 | 16.0 | 8336 | 1.5434 | 0.4476 | 0.2292 | 0.3868 | 0.3865 | 19.9007 | 0.9159 | 0.8916 | 0.9034 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.0 |
|
- Pytorch 2.0.1+cu117 |
|
- Datasets 2.14.5 |
|
- Tokenizers 0.15.0 |
|
|
