update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- generated_from_trainer
|
| 4 |
+
datasets:
|
| 5 |
+
- glue
|
| 6 |
+
metrics:
|
| 7 |
+
- accuracy
|
| 8 |
+
- f1
|
| 9 |
+
model-index:
|
| 10 |
+
- name: baseline-ft-mrpc-IRoberta-b-8bit
|
| 11 |
+
results:
|
| 12 |
+
- task:
|
| 13 |
+
name: Text Classification
|
| 14 |
+
type: text-classification
|
| 15 |
+
dataset:
|
| 16 |
+
name: glue
|
| 17 |
+
type: glue
|
| 18 |
+
config: mrpc
|
| 19 |
+
split: validation
|
| 20 |
+
args: mrpc
|
| 21 |
+
metrics:
|
| 22 |
+
- name: Accuracy
|
| 23 |
+
type: accuracy
|
| 24 |
+
value: 0.8970588235294118
|
| 25 |
+
- name: F1
|
| 26 |
+
type: f1
|
| 27 |
+
value: 0.9257950530035336
|
| 28 |
+
---
|
| 29 |
+
|
| 30 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 31 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 32 |
+
|
| 33 |
+
# baseline-ft-mrpc-IRoberta-b-8bit
|
| 34 |
+
|
| 35 |
+
This model is a fine-tuned version of [vuiseng9/baseline-ft-mrpc-IRoberta-b-unquantized](https://huggingface.co/vuiseng9/baseline-ft-mrpc-IRoberta-b-unquantized) on the glue dataset.
|
| 36 |
+
It achieves the following results on the evaluation set:
|
| 37 |
+
- Loss: 0.3871
|
| 38 |
+
- Accuracy: 0.8971
|
| 39 |
+
- F1: 0.9258
|
| 40 |
+
- Combined Score: 0.9114
|
| 41 |
+
|
| 42 |
+
## Model description
|
| 43 |
+
|
| 44 |
+
More information needed
|
| 45 |
+
|
| 46 |
+
## Intended uses & limitations
|
| 47 |
+
|
| 48 |
+
More information needed
|
| 49 |
+
|
| 50 |
+
## Training and evaluation data
|
| 51 |
+
|
| 52 |
+
More information needed
|
| 53 |
+
|
| 54 |
+
## Training procedure
|
| 55 |
+
|
| 56 |
+
### Training hyperparameters
|
| 57 |
+
|
| 58 |
+
The following hyperparameters were used during training:
|
| 59 |
+
- learning_rate: 5e-07
|
| 60 |
+
- train_batch_size: 16
|
| 61 |
+
- eval_batch_size: 16
|
| 62 |
+
- seed: 42
|
| 63 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 64 |
+
- lr_scheduler_type: linear
|
| 65 |
+
- num_epochs: 12.0
|
| 66 |
+
|
| 67 |
+
### Training results
|
| 68 |
+
|
| 69 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Combined Score |
|
| 70 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:--------------:|
|
| 71 |
+
| 0.0021 | 1.0 | 230 | 0.4017 | 0.8848 | 0.9147 | 0.8998 |
|
| 72 |
+
| 0.0026 | 2.0 | 460 | 0.4105 | 0.8873 | 0.9173 | 0.9023 |
|
| 73 |
+
| 0.0026 | 3.0 | 690 | 0.3707 | 0.8946 | 0.9236 | 0.9091 |
|
| 74 |
+
| 0.0037 | 4.0 | 920 | 0.3893 | 0.8946 | 0.9228 | 0.9087 |
|
| 75 |
+
| 1.324 | 5.0 | 1150 | 0.3871 | 0.8897 | 0.9204 | 0.9050 |
|
| 76 |
+
| 0.0227 | 6.0 | 1380 | 0.3951 | 0.8897 | 0.9201 | 0.9049 |
|
| 77 |
+
| 0.0081 | 7.0 | 1610 | 0.3818 | 0.8824 | 0.9155 | 0.8989 |
|
| 78 |
+
| 0.0054 | 8.0 | 1840 | 0.3902 | 0.8873 | 0.9181 | 0.9027 |
|
| 79 |
+
| 0.0383 | 9.0 | 2070 | 0.3659 | 0.8922 | 0.9225 | 0.9073 |
|
| 80 |
+
| 0.3861 | 10.0 | 2300 | 0.4260 | 0.8652 | 0.9030 | 0.8841 |
|
| 81 |
+
| 0.0028 | 11.0 | 2530 | 0.3619 | 0.8946 | 0.9234 | 0.9090 |
|
| 82 |
+
| 0.0957 | 12.0 | 2760 | 0.3871 | 0.8971 | 0.9258 | 0.9114 |
|
| 83 |
+
|
| 84 |
+
|
| 85 |
+
### Framework versions
|
| 86 |
+
|
| 87 |
+
- Transformers 4.30.2
|
| 88 |
+
- Pytorch 2.0.1+cu118
|
| 89 |
+
- Datasets 2.11.0
|
| 90 |
+
- Tokenizers 0.13.3
|