File size: 4,381 Bytes
112034b d2a0319 112034b d2a0319 112034b d2a0319 112034b d2a0319 112034b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 |
---
language:
- en
license: mit
tags:
- nycu-112-2-datamining-hw2
- generated_from_trainer
base_model: microsoft/deberta-v2-xxlarge
datasets:
- DandinPower/review_onlytitleandtext
metrics:
- accuracy
model-index:
- name: deberta-v2-xxlarge-otat-small-lr
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: DandinPower/review_onlytitleandtext
type: DandinPower/review_onlytitleandtext
metrics:
- type: accuracy
value: 0.668
name: Accuracy
---
<!-- 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. -->
# deberta-v2-xxlarge-otat-small-lr
This model is a fine-tuned version of [microsoft/deberta-v2-xxlarge](https://huggingface.co/microsoft/deberta-v2-xxlarge) on the DandinPower/review_onlytitleandtext dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7982
- Accuracy: 0.668
- Macro F1: 0.6665
## 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: 1.8e-06
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 64
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1
- num_epochs: 8
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Macro F1 |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:--------:|
| 1.6073 | 0.23 | 100 | 1.5910 | 0.2409 | 0.1625 |
| 1.5142 | 0.46 | 200 | 1.2862 | 0.439 | 0.3770 |
| 1.0421 | 0.69 | 300 | 0.8956 | 0.617 | 0.6084 |
| 0.8818 | 0.91 | 400 | 0.8344 | 0.6487 | 0.6462 |
| 0.8309 | 1.14 | 500 | 0.8180 | 0.6586 | 0.6575 |
| 0.8029 | 1.37 | 600 | 0.8090 | 0.6603 | 0.6589 |
| 0.7949 | 1.6 | 700 | 0.8124 | 0.6613 | 0.6538 |
| 0.7847 | 1.83 | 800 | 0.7775 | 0.6696 | 0.6698 |
| 0.7717 | 2.06 | 900 | 0.7727 | 0.6703 | 0.6699 |
| 0.7445 | 2.29 | 1000 | 0.7767 | 0.669 | 0.6646 |
| 0.7367 | 2.51 | 1100 | 0.7774 | 0.6693 | 0.6676 |
| 0.7419 | 2.74 | 1200 | 0.7580 | 0.674 | 0.6743 |
| 0.7394 | 2.97 | 1300 | 0.7660 | 0.6714 | 0.6722 |
| 0.7253 | 3.2 | 1400 | 0.7695 | 0.6717 | 0.6740 |
| 0.7155 | 3.43 | 1500 | 0.7623 | 0.6676 | 0.6699 |
| 0.7089 | 3.66 | 1600 | 0.7762 | 0.6687 | 0.6630 |
| 0.7041 | 3.89 | 1700 | 0.7670 | 0.6716 | 0.6719 |
| 0.6982 | 4.11 | 1800 | 0.7735 | 0.6699 | 0.6659 |
| 0.6778 | 4.34 | 1900 | 0.7676 | 0.6701 | 0.6676 |
| 0.6919 | 4.57 | 2000 | 0.7772 | 0.6717 | 0.6692 |
| 0.6919 | 4.8 | 2100 | 0.7751 | 0.6687 | 0.6662 |
| 0.6721 | 5.03 | 2200 | 0.7955 | 0.6666 | 0.6613 |
| 0.6576 | 5.26 | 2300 | 0.7765 | 0.6714 | 0.6720 |
| 0.6675 | 5.49 | 2400 | 0.7900 | 0.6703 | 0.6711 |
| 0.6641 | 5.71 | 2500 | 0.7780 | 0.6689 | 0.6676 |
| 0.6669 | 5.94 | 2600 | 0.7751 | 0.6687 | 0.6675 |
| 0.6368 | 6.17 | 2700 | 0.7995 | 0.6691 | 0.6690 |
| 0.647 | 6.4 | 2800 | 0.7962 | 0.668 | 0.6635 |
| 0.6285 | 6.63 | 2900 | 0.7861 | 0.6699 | 0.6702 |
| 0.6656 | 6.86 | 3000 | 0.7939 | 0.6706 | 0.6695 |
| 0.6397 | 7.09 | 3100 | 0.7876 | 0.668 | 0.6672 |
| 0.6252 | 7.31 | 3200 | 0.8001 | 0.669 | 0.6671 |
| 0.6378 | 7.54 | 3300 | 0.8006 | 0.6687 | 0.6675 |
| 0.6243 | 7.77 | 3400 | 0.7982 | 0.668 | 0.6665 |
### Framework versions
- Transformers 4.39.3
- Pytorch 2.2.2+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
|