res_nw_gulf_03

This model is a fine-tuned version of riotu-lab/ArabianGPT-03B on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3495
  • Bleu: 0.4355
  • Rouge1: 0.6808
  • Rouge2: 0.4218
  • Rougel: 0.6799

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
1.1759 1.0 836 0.3774 0.4156 0.6364 0.3630 0.6352
0.2824 2.0 1672 0.3495 0.4355 0.6808 0.4218 0.6799
0.1846 3.0 2508 0.3579 0.4480 0.6989 0.4541 0.6984
0.1351 4.0 3344 0.3674 0.4449 0.6993 0.4556 0.6985
0.11 5.0 4180 0.3791 0.4488 0.7067 0.4681 0.7056
0.0981 6.0 5016 0.3922 0.4583 0.7135 0.4789 0.7129
0.0917 7.0 5852 0.3964 0.4567 0.7127 0.4779 0.7120

Framework versions

  • Transformers 4.45.0.dev0
  • Pytorch 2.3.1+cu121
  • Datasets 2.19.2
  • Tokenizers 0.19.1
Downloads last month
9
Safetensors
Model size
369M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model has no library tag.

Model tree for nlparabic/res_nw_gulf_03

Finetuned
(8)
this model