Text Classification
Transformers
PyTorch
Safetensors
Polish
bert
financial-sentiment-analysis
sentiment-analysis
herbert
Eval Results (legacy)
text-embeddings-inference
Instructions to use bardsai/finance-sentiment-pl-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use bardsai/finance-sentiment-pl-base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="bardsai/finance-sentiment-pl-base")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("bardsai/finance-sentiment-pl-base") model = AutoModelForSequenceClassification.from_pretrained("bardsai/finance-sentiment-pl-base", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language: pl
|
| 3 |
+
tags:
|
| 4 |
+
- text-classification
|
| 5 |
+
- financial-sentiment-analysis
|
| 6 |
+
- sentiment-analysis
|
| 7 |
+
datasets:
|
| 8 |
+
- datasets/financial_phrasebank
|
| 9 |
+
metrics:
|
| 10 |
+
- f1
|
| 11 |
+
- accuracy
|
| 12 |
+
- precision
|
| 13 |
+
- recall
|
| 14 |
+
widget:
|
| 15 |
+
- text: "Sprzedaż netto wzrosła o 30% do 36 mln EUR."
|
| 16 |
+
example_title: "Example 1"
|
| 17 |
+
- text: "Rusza Black Friday. Lista promocji w sklepach."
|
| 18 |
+
example_title: "Example 2"
|
| 19 |
+
- text: "Akcje CDPROJEKT zanotowały największy spadek wśród spółek notowanych na GPW."
|
| 20 |
+
example_title: "Example 3"
|
| 21 |
+
---
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
# FinanceSentimentPL-base
|
| 25 |
+
|
| 26 |
+
FinanceSentimentPL-fast is a model based on [herbert-base](https://huggingface.co/allegro/herbert-base-cased) for analyzing sentiment of Polish financial news. It was trained on the translated version of [Financial PhraseBank](https://www.researchgate.net/publication/251231107_Good_Debt_or_Bad_Debt_Detecting_Semantic_Orientations_in_Economic_Texts) by Malo et al. (20014) for 10 epochs on single RTX3090 gpu.
|
| 27 |
+
|
| 28 |
+
The model will give you a three labels: positive, negative and neutral.
|
| 29 |
+
|
| 30 |
+
## How to use
|
| 31 |
+
|
| 32 |
+
You can use this model directly with a pipeline for sentiment-analysis:
|
| 33 |
+
|
| 34 |
+
```python
|
| 35 |
+
from transformers import pipeline
|
| 36 |
+
|
| 37 |
+
nlp = pipeline("sentiment-analysis", model="bardsai/FinanceSentimentPL-fast")
|
| 38 |
+
nlp("Sprzedaż netto wzrosła o 30% do 36 mln EUR.")
|
| 39 |
+
```
|
| 40 |
+
```bash
|
| 41 |
+
[{'label': 'positive', 'score': 0.9999998807907104}]
|
| 42 |
+
```
|
| 43 |
+
|
| 44 |
+
## Performance
|
| 45 |
+
| Metric | Value |
|
| 46 |
+
| --- | ----------- |
|
| 47 |
+
| f1 macro | 0.969 |
|
| 48 |
+
| precision macro | 0.971 |
|
| 49 |
+
| recall macro | 0.968 |
|
| 50 |
+
| accuracy | 0.976 |
|
| 51 |
+
| samples per second | 136.8 |
|
| 52 |
+
|
| 53 |
+
(The performance was evaluated on RTX 3090 gpu)
|
| 54 |
+
|
| 55 |
+
## Changelog
|
| 56 |
+
- 2022-11-15: Initial release
|
| 57 |
+
|
| 58 |
+
## About bards.ai
|
| 59 |
+
|
| 60 |
+
At bards.ai, we focus on providing machine learning expertise and skills to our partners, particularly in the areas of nlp, machine vision and time series analysis. Our team is located in Wroclaw, Poland. Please visit our website for more information: [bards.ai](https://bards.ai/)
|
| 61 |
+
|
| 62 |
+
Let us know if you use our model :). Also, if you need any help, feel free to contact us at info@bards.ai
|