File size: 3,124 Bytes
32d181a 1230cf1 32d181a 1230cf1 cfd84f5 32d181a 82e9b27 32d181a b67649d 1604125 32d181a 8773eba 32d181a 8773eba 32d181a |
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 |
---
license: apache-2.0
language: "en"
tags:
- financial-sentiment-analysis
- sentiment-analysis
- generated_from_trainer
- financial
- stocks
- sentiment
metrics:
- f1
widget:
- text: "The USD rallied by 10% last night"
example_title: "Bullish Sentiment"
- text: "Covid-19 cases have been increasing over the past few months"
example_title: "Bearish Sentiment"
- text: "the USD has been trending lower"
example_title: "Mildly Bearish Sentiment"
model-index:
- name: distilroberta-finetuned-finclass
results:
- task:
name: Text Classification
type: text-classification
datasets:
name: financial_phrasebank
type: financial_phrasebank
args: sentences_allagree
name_1: Kaggle Self label
type_1: financial-classification
metrics:
name: F1-Score
type: F1-Score
value: 0.88
---
<!-- 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. -->
# distilroberta-finetuned-finclass
This model is a fine-tuned version of [distilroberta-base](https://huggingface.co/distilroberta-base) on the [financial-phrasebank + Kaggle Dataset](https://huggingface.co/datasets/nickmuchi/financial-classification). The Kaggle dataset includes Covid-19 sentiment data and can be found here: [sentiment-classification-selflabel-dataset](https://www.kaggle.com/percyzheng/sentiment-classification-selflabel-dataset).
It achieves the following results on the evaluation set:
- Loss: 0.4463
- F1: 0.8835
## Model description
Model determines the financial sentiment of given text. Given the unbalanced distribution of the class labels, the weights were adjusted to pay attention to the less sampled labels which should increase overall performance.
## 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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.7309 | 1.0 | 72 | 0.3671 | 0.8441 |
| 0.3757 | 2.0 | 144 | 0.3199 | 0.8709 |
| 0.3054 | 3.0 | 216 | 0.3096 | 0.8678 |
| 0.2229 | 4.0 | 288 | 0.3776 | 0.8390 |
| 0.1744 | 5.0 | 360 | 0.3678 | 0.8723 |
| 0.1436 | 6.0 | 432 | 0.3728 | 0.8758 |
| 0.1044 | 7.0 | 504 | 0.4116 | 0.8744 |
| 0.0931 | 8.0 | 576 | 0.4148 | 0.8761 |
| 0.0683 | 9.0 | 648 | 0.4423 | 0.8837 |
| 0.0611 | 10.0 | 720 | 0.4463 | 0.8835 |
### Framework versions
- Transformers 4.15.0
- Pytorch 1.10.0+cu111
- Datasets 1.18.0
- Tokenizers 0.10.3
|