res_nw_lev
This model is a fine-tuned version of riotu-lab/ArabianGPT-01B on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4049
- Bleu: 0.3985
- Rouge1: 0.7295
- Rouge2: 0.5331
- Rougel: 0.7285
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: 20.0
Training results
Training Loss | Epoch | Step | Validation Loss | Bleu | Rouge1 | Rouge2 | Rougel |
---|---|---|---|---|---|---|---|
0.8705 | 1.0 | 5062 | 0.5324 | 0.1992 | 0.5712 | 0.3058 | 0.5693 |
0.4701 | 2.0 | 10124 | 0.4697 | 0.2384 | 0.6250 | 0.3686 | 0.6235 |
0.3795 | 3.0 | 15186 | 0.4394 | 0.2692 | 0.6568 | 0.4131 | 0.6555 |
0.3128 | 4.0 | 20248 | 0.4204 | 0.2995 | 0.6766 | 0.4446 | 0.6754 |
0.2625 | 5.0 | 25310 | 0.4107 | 0.3341 | 0.6993 | 0.4788 | 0.6981 |
0.2236 | 6.0 | 30372 | 0.4061 | 0.3688 | 0.7155 | 0.5061 | 0.7144 |
0.1941 | 7.0 | 35434 | 0.4049 | 0.3985 | 0.7295 | 0.5331 | 0.7285 |
0.1709 | 8.0 | 40496 | 0.4082 | 0.4226 | 0.7369 | 0.5493 | 0.7359 |
0.1534 | 9.0 | 45558 | 0.4127 | 0.4394 | 0.7407 | 0.5583 | 0.7400 |
0.1397 | 10.0 | 50620 | 0.4193 | 0.4513 | 0.7459 | 0.5686 | 0.7453 |
0.1295 | 11.0 | 55682 | 0.4255 | 0.4612 | 0.7482 | 0.5730 | 0.7473 |
0.1221 | 12.0 | 60744 | 0.4306 | 0.4631 | 0.7482 | 0.5741 | 0.7473 |
Framework versions
- Transformers 4.45.0.dev0
- Pytorch 2.3.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1
- Downloads last month
- 0
Model tree for nlparabic/res_nw_lev
Base model
riotu-lab/ArabianGPT-01B
Finetuned
this model