|
--- |
|
license: apache-2.0 |
|
library_name: peft |
|
tags: |
|
- generated_from_trainer |
|
base_model: distilgpt2 |
|
model-index: |
|
- name: distilgpt-monolinugal |
|
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. --> |
|
|
|
# distilgpt-monolinugal |
|
|
|
This model is a fine-tuned version of [distilgpt2](https://huggingface.co/distilgpt2) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 3.4876 |
|
|
|
## 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: 0.0005 |
|
- train_batch_size: 12 |
|
- eval_batch_size: 12 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 8 |
|
- total_train_batch_size: 96 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_steps: 100 |
|
- num_epochs: 8 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 3.3098 | 0.16 | 200 | 3.5905 | |
|
| 3.2847 | 0.32 | 400 | 3.5644 | |
|
| 3.2612 | 0.48 | 600 | 3.5504 | |
|
| 3.2636 | 0.64 | 800 | 3.5384 | |
|
| 3.2481 | 0.8 | 1000 | 3.5301 | |
|
| 3.2393 | 0.96 | 1200 | 3.5233 | |
|
| 3.2381 | 1.12 | 1400 | 3.5184 | |
|
| 3.2317 | 1.28 | 1600 | 3.5168 | |
|
| 3.2244 | 1.44 | 1800 | 3.5123 | |
|
| 3.2258 | 1.6 | 2000 | 3.5117 | |
|
| 3.2238 | 1.76 | 2200 | 3.5058 | |
|
| 3.2376 | 1.92 | 2400 | 3.5058 | |
|
| 3.212 | 2.08 | 2600 | 3.5044 | |
|
| 3.231 | 2.24 | 2800 | 3.5019 | |
|
| 3.2044 | 2.4 | 3000 | 3.5003 | |
|
| 3.2107 | 2.57 | 3200 | 3.5002 | |
|
| 3.2096 | 2.73 | 3400 | 3.4996 | |
|
| 3.215 | 2.89 | 3600 | 3.4963 | |
|
| 3.2092 | 3.05 | 3800 | 3.4979 | |
|
| 3.2034 | 3.21 | 4000 | 3.4964 | |
|
| 3.1992 | 3.37 | 4200 | 3.4971 | |
|
| 3.1975 | 3.53 | 4400 | 3.4941 | |
|
| 3.222 | 3.69 | 4600 | 3.4932 | |
|
| 3.2104 | 3.85 | 4800 | 3.4927 | |
|
| 3.199 | 4.01 | 5000 | 3.4918 | |
|
| 3.2033 | 4.17 | 5200 | 3.4927 | |
|
| 3.201 | 4.33 | 5400 | 3.4924 | |
|
| 3.1947 | 4.49 | 5600 | 3.4931 | |
|
| 3.2172 | 4.65 | 5800 | 3.4907 | |
|
| 3.201 | 4.81 | 6000 | 3.4908 | |
|
| 3.2089 | 4.97 | 6200 | 3.4892 | |
|
| 3.206 | 5.13 | 6400 | 3.4896 | |
|
| 3.2074 | 5.29 | 6600 | 3.4884 | |
|
| 3.2046 | 5.45 | 6800 | 3.4891 | |
|
| 3.1899 | 5.61 | 7000 | 3.4888 | |
|
| 3.196 | 5.77 | 7200 | 3.4891 | |
|
| 3.1946 | 5.93 | 7400 | 3.4880 | |
|
| 3.1951 | 6.09 | 7600 | 3.4887 | |
|
| 3.1998 | 6.25 | 7800 | 3.4878 | |
|
| 3.1775 | 6.41 | 8000 | 3.4880 | |
|
| 3.1947 | 6.57 | 8200 | 3.4880 | |
|
| 3.1876 | 6.73 | 8400 | 3.4876 | |
|
| 3.1984 | 6.89 | 8600 | 3.4878 | |
|
| 3.1927 | 7.05 | 8800 | 3.4875 | |
|
| 3.2006 | 7.21 | 9000 | 3.4875 | |
|
| 3.2042 | 7.37 | 9200 | 3.4875 | |
|
| 3.1856 | 7.54 | 9400 | 3.4877 | |
|
| 3.1952 | 7.7 | 9600 | 3.4877 | |
|
| 3.1981 | 7.86 | 9800 | 3.4876 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.1 |
|
- Transformers 4.36.2 |
|
- Pytorch 1.13.0+cu116 |
|
- Datasets 2.16.0 |
|
- Tokenizers 0.15.0 |
