leobg/deeva-modcat-seqclass-deberta-v1
Browse files
README.md
ADDED
|
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
base_model: microsoft/deberta-v3-small
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- accuracy
|
| 8 |
+
- f1
|
| 9 |
+
- precision
|
| 10 |
+
- recall
|
| 11 |
+
model-index:
|
| 12 |
+
- name: deeva-modcat-seqclass-deberta-v1
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# deeva-modcat-seqclass-deberta-v1
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [microsoft/deberta-v3-small](https://huggingface.co/microsoft/deberta-v3-small) on an unknown dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.6435
|
| 24 |
+
- Accuracy: 0.7161
|
| 25 |
+
- F1: 0.2922
|
| 26 |
+
- Precision: 0.1808
|
| 27 |
+
- Recall: 0.7619
|
| 28 |
+
|
| 29 |
+
## Model description
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Intended uses & limitations
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training and evaluation data
|
| 38 |
+
|
| 39 |
+
More information needed
|
| 40 |
+
|
| 41 |
+
## Training procedure
|
| 42 |
+
|
| 43 |
+
### Training hyperparameters
|
| 44 |
+
|
| 45 |
+
The following hyperparameters were used during training:
|
| 46 |
+
- learning_rate: 2e-05
|
| 47 |
+
- train_batch_size: 24
|
| 48 |
+
- eval_batch_size: 24
|
| 49 |
+
- seed: 42
|
| 50 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 51 |
+
- lr_scheduler_type: linear
|
| 52 |
+
- num_epochs: 2
|
| 53 |
+
|
| 54 |
+
### Training results
|
| 55 |
+
|
| 56 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|
| 57 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
|
| 58 |
+
| No log | 0.18 | 2 | 0.7148 | 0.4139 | 0.0476 | 0.0272 | 0.1905 |
|
| 59 |
+
| No log | 0.36 | 4 | 0.7027 | 0.4835 | 0.0408 | 0.0238 | 0.1429 |
|
| 60 |
+
| No log | 0.55 | 6 | 0.6917 | 0.5586 | 0.0474 | 0.0284 | 0.1429 |
|
| 61 |
+
| No log | 0.73 | 8 | 0.6817 | 0.5604 | 0.0476 | 0.0286 | 0.1429 |
|
| 62 |
+
| No log | 0.91 | 10 | 0.6727 | 0.5623 | 0.0478 | 0.0287 | 0.1429 |
|
| 63 |
+
| No log | 1.09 | 12 | 0.6648 | 0.6374 | 0.0571 | 0.0357 | 0.1429 |
|
| 64 |
+
| No log | 1.27 | 14 | 0.6578 | 0.6374 | 0.0571 | 0.0357 | 0.1429 |
|
| 65 |
+
| No log | 1.45 | 16 | 0.6521 | 0.6355 | 0.0569 | 0.0355 | 0.1429 |
|
| 66 |
+
| No log | 1.64 | 18 | 0.6477 | 0.6392 | 0.1005 | 0.0621 | 0.2619 |
|
| 67 |
+
| No log | 1.82 | 20 | 0.6448 | 0.7015 | 0.2419 | 0.1503 | 0.6190 |
|
| 68 |
+
| No log | 2.0 | 22 | 0.6435 | 0.7161 | 0.2922 | 0.1808 | 0.7619 |
|
| 69 |
+
|
| 70 |
+
|
| 71 |
+
### Framework versions
|
| 72 |
+
|
| 73 |
+
- Transformers 4.33.2
|
| 74 |
+
- Pytorch 2.1.2+cu121
|
| 75 |
+
- Datasets 2.14.5
|
| 76 |
+
- Tokenizers 0.13.3
|