File size: 1,695 Bytes
0dcffab |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 |
---
license: apache-2.0
base_model: riotu-lab/ArabianGPT-01B
tags:
- generated_from_trainer
metrics:
- bleu
- rouge
model-index:
- name: nw_egy
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. -->
# nw_egy
This model is a fine-tuned version of [riotu-lab/ArabianGPT-01B](https://huggingface.co/riotu-lab/ArabianGPT-01B) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.5201
- Bleu: 0.2511
- Rouge1: 0.4534
- Rouge2: 0.2137
- Rougel: 0.3984
## 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: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 3.0
### Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Rouge1 | Rouge2 | Rougel |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:------:|
| 4.0723 | 1.0 | 442 | 2.9581 | 0.2228 | 0.3913 | 0.1648 | 0.3201 |
| 2.8055 | 2.0 | 884 | 2.6063 | 0.2415 | 0.4347 | 0.1971 | 0.3777 |
| 2.5505 | 3.0 | 1326 | 2.5201 | 0.2511 | 0.4534 | 0.2137 | 0.3984 |
### Framework versions
- Transformers 4.45.0.dev0
- Pytorch 2.3.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1
|