File size: 1,989 Bytes
d4c453a |
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 69 70 71 72 73 74 75 |
---
license: apache-2.0
tags:
- summarization
- arabic
- ar
- fa
- persian
- mt5
- Abstractive Summarization
- generated_from_trainer
model-index:
- name: mt5-base-finetuned-arfa
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. -->
# mt5-base-finetuned-arfa
This model is a fine-tuned version of [google/mt5-base](https://huggingface.co/google/mt5-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 3.1784
- Rouge-1: 25.68
- Rouge-2: 11.8
- Rouge-l: 22.99
- Gen Len: 18.99
- Bertscore: 71.78
## 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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- label_smoothing_factor: 0.1
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge-1 | Rouge-2 | Rouge-l | Gen Len | Bertscore |
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|:-------:|:-------:|:---------:|
| 3.9866 | 1.0 | 2649 | 3.3635 | 21.94 | 8.59 | 19.5 | 18.99 | 70.6 |
| 3.5637 | 2.0 | 5298 | 3.2557 | 24.01 | 10.0 | 21.26 | 18.99 | 71.22 |
| 3.4016 | 3.0 | 7947 | 3.2005 | 24.4 | 10.43 | 21.72 | 18.98 | 71.36 |
| 3.2985 | 4.0 | 10596 | 3.1784 | 24.68 | 10.73 | 22.01 | 18.98 | 71.51 |
### Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0+cu113
- Datasets 2.2.2
- Tokenizers 0.12.1
|