Downloads · 30 days
51
59% of all-time downloads
mrc-vit5-base-dsc
MiuN2k3/mrc-vit5-base-dsc
mrc-vit5-base-dsc is a question answering model from MiuN2k3. Use it when the input is a question plus a passage. It is set up for transformers. The card lists the license as mit.
should probably proofread and complete it, then remove this comment. --
All-time downloads
87
Public
Parameters
226M
6.3 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors904 MB · 100%
At a glance
- Task
- Question Answering
- Library
- transformers
- License
- mit
- Model type
- t5
- Access
- Public
- Created
- May 27, 2024
- Updated
- May 28, 2024
- SHA
- c0f9c4e5
Try a prompt
Base models
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. -->
mrc-vit5-dsc
This model is a fine-tuned version of VietAI/vit5-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.6093
- Exact Match: 0.7382
- F1: 0.8663
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: 2e-05
- train_batch_size: 6
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Exact Match | F1 |
|---|---|---|---|---|---|
| 0.7044 | 1.0 | 4240 | 0.6239 | 0.6669 | 0.8277 |
| 0.5249 | 2.0 | 8480 | 0.5740 | 0.7000 | 0.8584 |
| 0.3262 | 3.0 | 12720 | 0.7052 | 0.7215 | 0.8609 |
| 0.2361 | 4.0 | 16960 | 1.1596 | 0.7305 | 0.8601 |
| 0.1454 | 5.0 | 21200 | 1.6093 | 0.7382 | 0.8663 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.2.1
- Datasets 2.19.1
- Tokenizers 0.19.1
- Task
- Question Answering
- Library
- transformers
- Type
- t5
- License
- mit
- Created
- May 27, 2024
- Updated
- May 28, 2024