End of training
Browse files
README.md
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: mit
|
| 4 |
+
base_model: FacebookAI/roberta-base
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
model-index:
|
| 8 |
+
- name: bengali_qa_model_AGGRO_roberta-base
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# bengali_qa_model_AGGRO_roberta-base
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [FacebookAI/roberta-base](https://huggingface.co/FacebookAI/roberta-base) on an unknown dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 2.7563
|
| 20 |
+
- Exact Match: 70.7143
|
| 21 |
+
- F1 Score: 79.6578
|
| 22 |
+
|
| 23 |
+
## Model description
|
| 24 |
+
|
| 25 |
+
More information needed
|
| 26 |
+
|
| 27 |
+
## Intended uses & limitations
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Training and evaluation data
|
| 32 |
+
|
| 33 |
+
More information needed
|
| 34 |
+
|
| 35 |
+
## Training procedure
|
| 36 |
+
|
| 37 |
+
### Training hyperparameters
|
| 38 |
+
|
| 39 |
+
The following hyperparameters were used during training:
|
| 40 |
+
- learning_rate: 1e-05
|
| 41 |
+
- train_batch_size: 4
|
| 42 |
+
- eval_batch_size: 4
|
| 43 |
+
- seed: 3407
|
| 44 |
+
- gradient_accumulation_steps: 16
|
| 45 |
+
- total_train_batch_size: 64
|
| 46 |
+
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 47 |
+
- lr_scheduler_type: cosine
|
| 48 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 49 |
+
- training_steps: 50
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Exact Match | F1 Score |
|
| 54 |
+
|:-------------:|:------:|:----:|:---------------:|:-----------:|:--------:|
|
| 55 |
+
| 5.9583 | 0.0053 | 1 | 5.9716 | 0.0 | 5.2517 |
|
| 56 |
+
| 5.948 | 0.0107 | 2 | 5.9377 | 0.0 | 5.5327 |
|
| 57 |
+
| 5.9426 | 0.0160 | 3 | 5.8689 | 0.0 | 6.7345 |
|
| 58 |
+
| 5.8563 | 0.0214 | 4 | 5.7663 | 0.0 | 11.2969 |
|
| 59 |
+
| 5.8071 | 0.0267 | 5 | 5.6429 | 0.1504 | 26.3481 |
|
| 60 |
+
| 5.6804 | 0.0321 | 6 | 5.5006 | 12.3308 | 40.0080 |
|
| 61 |
+
| 5.5603 | 0.0374 | 7 | 5.3531 | 42.1053 | 58.0172 |
|
| 62 |
+
| 5.4208 | 0.0428 | 8 | 5.1925 | 52.4060 | 64.8440 |
|
| 63 |
+
| 5.2257 | 0.0481 | 9 | 5.0174 | 57.5188 | 69.2798 |
|
| 64 |
+
| 5.0287 | 0.0535 | 10 | 4.8260 | 60.7519 | 71.5328 |
|
| 65 |
+
| 4.9646 | 0.0588 | 11 | 4.6193 | 62.7068 | 73.5651 |
|
| 66 |
+
| 4.6784 | 0.0641 | 12 | 4.4020 | 63.7594 | 75.2683 |
|
| 67 |
+
| 4.623 | 0.0695 | 13 | 4.1678 | 64.0602 | 75.7455 |
|
| 68 |
+
| 4.3488 | 0.0748 | 14 | 3.9101 | 64.9624 | 76.7744 |
|
| 69 |
+
| 4.059 | 0.0802 | 15 | 3.6405 | 66.8421 | 77.6939 |
|
| 70 |
+
| 3.7381 | 0.0855 | 16 | 3.3842 | 68.8722 | 78.5505 |
|
| 71 |
+
| 3.574 | 0.0909 | 17 | 3.1525 | 67.3684 | 77.7160 |
|
| 72 |
+
| 3.4082 | 0.0962 | 18 | 2.9428 | 66.2406 | 77.6532 |
|
| 73 |
+
| 3.2202 | 0.1016 | 19 | 2.7619 | 69.3985 | 78.5549 |
|
| 74 |
+
|
| 75 |
+
|
| 76 |
+
### Framework versions
|
| 77 |
+
|
| 78 |
+
- Transformers 4.46.3
|
| 79 |
+
- Pytorch 2.4.0
|
| 80 |
+
- Datasets 3.1.0
|
| 81 |
+
- Tokenizers 0.20.3
|