Downloads · 30 days
11
31% of all-time downloads
sa_bert_12_layer_modified_complete_training_48
gokuls/sa_bert_12_layer_modified_complete_training_48
sa_bert_12_layer_modified_complete_training_48 is a fill-mask model from gokuls. Use it when you need the model to fill a missing word. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
36
Public
Repo size
20.1 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin495 MB · 82%
At a glance
- Task
- Fill-Mask
- Library
- transformers
- Model type
- hybridbert
- Access
- Public
- Created
- Jun 20, 2023
- Updated
- Jun 22, 2023
- SHA
- ce2b8127
Try a prompt
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. -->
sa_bert_12_layer_modified_complete_training_48
This model is a fine-tuned version of on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.7897
- Accuracy: 0.5117
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: 1e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10000
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 6.5942 | 0.05 | 10000 | 6.5714 | 0.1229 |
| 6.1563 | 0.11 | 20000 | 6.3437 | 0.1392 |
| 6.1425 | 0.16 | 30000 | 6.2474 | 0.1444 |
| 6.2249 | 0.22 | 40000 | 6.1900 | 0.1468 |
| 6.1498 | 0.27 | 50000 | 6.1482 | 0.1487 |
| 6.0528 | 0.33 | 60000 | 6.1192 | 0.1492 |
| 6.0103 | 0.38 | 70000 | 6.0762 | 0.1504 |
| 5.8523 | 0.44 | 80000 | 5.8731 | 0.1615 |
| 5.91 | 0.49 | 90000 | 5.7442 | 0.1765 |
| 5.4931 | 0.55 | 100000 | 5.5985 | 0.1952 |
| 5.4145 | 0.6 | 110000 | 5.4716 | 0.2100 |
| 5.3729 | 0.66 | 120000 | 5.3366 | 0.2247 |
| 5.2655 | 0.71 | 130000 | 5.1946 | 0.2417 |
| 5.2975 | 0.76 | 140000 | 5.0287 | 0.2600 |
| 4.9997 | 0.82 | 150000 | 4.8593 | 0.2791 |
| 4.831 | 0.87 | 160000 | 4.6226 | 0.3041 |
| 4.9176 | 0.93 | 170000 | 4.4211 | 0.3257 |
| 4.5352 | 0.98 | 180000 | 4.2328 | 0.3429 |
| 4.1536 | 1.04 | 190000 | 4.0635 | 0.3598 |
| 4.0216 | 1.09 | 200000 | 3.9109 | 0.3755 |
| 4.0744 | 1.15 | 210000 | 3.7761 | 0.3897 |
| 3.7468 | 1.2 | 220000 | 3.6636 | 0.4038 |
| 3.5015 | 1.26 | 230000 | 3.5047 | 0.4236 |
| 3.5717 | 1.31 | 240000 | 3.4014 | 0.4370 |
| 3.1969 | 1.37 | 250000 | 3.3173 | 0.4479 |
| 3.5026 | 1.42 | 260000 | 3.2254 | 0.4588 |
| 3.287 | 1.47 | 270000 | 3.1845 | 0.4643 |
| 3.3462 | 1.53 | 280000 | 3.0979 | 0.4738 |
| 3.3996 | 1.58 | 290000 | 3.0808 | 0.4764 |
| 3.2324 | 1.64 | 300000 | 3.0163 | 0.4843 |
| 3.0972 | 1.69 | 310000 | 2.9738 | 0.4890 |
| 3.1621 | 1.75 | 320000 | 2.9450 | 0.4927 |
| 3.0282 | 1.8 | 330000 | 2.9135 | 0.4964 |
| 3.0674 | 1.86 | 340000 | 2.9059 | 0.4979 |
| 2.9437 | 1.91 | 350000 | 2.8810 | 0.5007 |
| 2.8208 | 1.97 | 360000 | 2.8316 | 0.5064 |
| 2.9005 | 2.02 | 370000 | 2.8061 | 0.5098 |
| 2.7574 | 2.08 | 380000 | 2.7897 | 0.5117 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.0
- Tokenizers 0.13.3
- Task
- Fill-Mask
- Library
- transformers
- Type
- hybridbert
- Created
- Jun 20, 2023
- Updated
- Jun 22, 2023