|
|
--- |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- en |
|
|
metrics: |
|
|
- accuracy |
|
|
- precision |
|
|
- recall |
|
|
base_model: |
|
|
- microsoft/deberta-v3-base |
|
|
pipeline_tag: text-classification |
|
|
--- |
|
|
# Stock Sentiment Analysis |
|
|
|
|
|
This model is a fine-tuned version of `microsoft/deberta-v3-base` for **stock sentiment analysis**. |
|
|
|
|
|
## Model Details |
|
|
- **Language**: English |
|
|
- **Task**: Text Classification (Sentiment Analysis) |
|
|
- **Classes**: Positive, Neutral, Negative |
|
|
|
|
|
## Training |
|
|
- Evaluation Metric: F1 Score |
|
|
- Training Args: See `training_args.bin` for details. |
|
|
|
|
|
## Usage |
|
|
```python |
|
|
from transformers import AutoTokenizer, AutoModelForSequenceClassification |
|
|
|
|
|
tokenizer = AutoTokenizer.from_pretrained("vinD27/stock_sentiment_analysis") |
|
|
model = AutoModelForSequenceClassification.from_pretrained("vinD27/stock_sentiment_analysis") |
|
|
|
|
|
text = "The stock market is performing well today." |
|
|
inputs = tokenizer(text, return_tensors="pt") |
|
|
outputs = model(**inputs) |
|
|
print(outputs.logits) |
