|
--- |
|
library_name: transformers |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: gpt22gpt2-gpt2-large-cnn-dailymail-seed42 |
|
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. --> |
|
|
|
# gpt22gpt2-gpt2-large-cnn-dailymail-seed42 |
|
|
|
This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.6842 |
|
- Rouge1: 0.3500 |
|
- Rouge2: 0.1477 |
|
- Rougel: 0.2169 |
|
- Rougelsum: 0.3306 |
|
|
|
## 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: 8 |
|
- eval_batch_size: 16 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 32 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 1000 |
|
- num_epochs: 3.0 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | |
|
|:-------------:|:------:|:-----:|:---------------:|:------:|:------:|:------:|:---------:| |
|
| 2.6006 | 0.2229 | 2000 | 2.3891 | 0.1904 | 0.0449 | 0.1271 | 0.1782 | |
|
| 2.1985 | 0.4458 | 4000 | 2.0162 | 0.2608 | 0.0874 | 0.1661 | 0.2452 | |
|
| 1.9813 | 0.6687 | 6000 | 1.8433 | 0.2390 | 0.0798 | 0.1570 | 0.2267 | |
|
| 1.8954 | 0.8916 | 8000 | 1.7652 | 0.2694 | 0.0968 | 0.1718 | 0.2533 | |
|
| 1.5988 | 1.1145 | 10000 | 1.7400 | 0.3181 | 0.1275 | 0.1992 | 0.2998 | |
|
| 1.5897 | 1.3374 | 12000 | 1.7119 | 0.3292 | 0.1351 | 0.2049 | 0.3107 | |
|
| 1.5809 | 1.5603 | 14000 | 1.6926 | 0.3452 | 0.1451 | 0.2142 | 0.3262 | |
|
| 1.575 | 1.7832 | 16000 | 1.6679 | 0.3440 | 0.1452 | 0.2149 | 0.3256 | |
|
| 1.5302 | 2.0061 | 18000 | 1.6870 | 0.3512 | 0.1486 | 0.2168 | 0.3316 | |
|
| 1.2726 | 2.2290 | 20000 | 1.7002 | 0.3484 | 0.1460 | 0.2149 | 0.3289 | |
|
| 1.266 | 2.4519 | 22000 | 1.6969 | 0.3473 | 0.1461 | 0.2154 | 0.3279 | |
|
| 1.2566 | 2.6748 | 24000 | 1.6878 | 0.3487 | 0.1469 | 0.2160 | 0.3296 | |
|
| 1.2572 | 2.8977 | 26000 | 1.6842 | 0.3500 | 0.1477 | 0.2169 | 0.3306 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.44.2 |
|
- Pytorch 2.4.0 |
|
- Datasets 2.21.0 |
|
- Tokenizers 0.19.1 |
|
|
