res_nw_gulf_1.5
This model is a fine-tuned version of riotu-lab/ArabianGPT-1.5B on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4530
- Bleu: 0.3995
- Rouge1: 0.5971
- Rouge2: 0.3199
- Rougel: 0.5958
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 |
---|---|---|---|---|---|---|---|
2.4055 | 1.0 | 836 | 0.5180 | 0.3654 | 0.5081 | 0.2235 | 0.5074 |
0.4694 | 2.0 | 1672 | 0.4778 | 0.3806 | 0.5439 | 0.2600 | 0.5429 |
0.4026 | 3.0 | 2508 | 0.4615 | 0.3911 | 0.5711 | 0.2859 | 0.5701 |
0.3497 | 4.0 | 3344 | 0.4553 | 0.3921 | 0.5842 | 0.3008 | 0.5827 |
0.3038 | 5.0 | 4180 | 0.4530 | 0.3995 | 0.5971 | 0.3199 | 0.5958 |
0.264 | 6.0 | 5016 | 0.4554 | 0.3967 | 0.6005 | 0.3247 | 0.5995 |
0.2294 | 7.0 | 5852 | 0.4600 | 0.4022 | 0.6063 | 0.3322 | 0.6053 |
0.1996 | 8.0 | 6688 | 0.4667 | 0.3988 | 0.6074 | 0.3348 | 0.6061 |
0.1749 | 9.0 | 7524 | 0.4742 | 0.3996 | 0.6081 | 0.3372 | 0.6067 |
0.1541 | 10.0 | 8360 | 0.4818 | 0.3975 | 0.6070 | 0.3366 | 0.6061 |
Framework versions
- Transformers 4.45.0.dev0
- Pytorch 2.3.1+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1
- Downloads last month
- 3
Model tree for nlparabic/res_nw_gulf_1.5
Base model
riotu-lab/ArabianGPT-1.5B
Finetuned
this model