Downloads · 30 days
11
2% of all-time downloads
indic-mALBERT-squad-v2
rohitsroch/indic-mALBERT-squad-v2
indic-mALBERT-squad-v2 is a question answering model from rohitsroch. Use it when the input is a question plus a passage. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
690
Public
Parameters
32.9M
332 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin131 MB · 40%
Parameter types
How the weights are stored.
F3232.9M · 100%
Try a prompt
Datasets
Model card
From the Hugging Face model README
<!-- 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. -->
indic-mALBERT-uncased
This model is a uncased fine-tuned version of indic-bert on the squad_v2 dataset.
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 814
- num_epochs: 5.0
- mixed_precision_training: Native AMP
Training results
| Key | Value |
|---|---|
| epoch | 5.0 |
| eval_HasAns_exact | 53.54251012145749 |
| eval_HasAns_f1 | 63.501893668646716 |
| eval_HasAns_total | 5928 |
| eval_NoAns_exact | 75.10513036164845 |
| eval_NoAns_f1 | 75.10513036164845 |
| eval_NoAns_total | 5945 |
| eval_exact | 64.3392571380443 |
| eval_f1 | 69.31181888888594 |
| eval_runtime | 69.5054 |
| eval_samples | 12000 |
| eval_samples_per_second | 172.649 |
| eval_steps_per_second | 5.395 |
| eval_total | 11873 |
Framework versions
- Transformers 4.29.2
- Pytorch 1.9.0+cu102
- Datasets 2.12.0
- Tokenizers 0.13.3
- Task
- Question Answering
- Library
- transformers
- Type
- albert
- Created
- Jun 13, 2023
- Updated
- Jul 1, 2023