README.md

metadata

license: apache-2.0
datasets:
  - winddude/finacial_pharsebank_66agree_split
  - financial_phrasebank
language:
  - en
base_model:
  - state-spaces/mamba-2.8b
metrics:
  - accuracy
  - f1
  - recall
  - precission
model-index:
  - name: financial-sentiment-analysis
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: financial_phrasebank
          type: financial_phrasebank
          args: sentences_66agree
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.82
          - name: Percision
            type: percision
            value: 0.82
          - name: recall
            type: recall
            value: 0.82
          - name: F1
            type: f1
            value: 0.82
pipeline_tag: text-classification
tags:
  - finance

Mamba Financial Headline Sentiment Classifier

A sentment classifier for finacial headlines using mamba 2.8b as the base model.

Text is classified into 1 of 3 labels; positive, neutral, or negative.

Prompt Format:

prompt = f"""Classify the setiment of the following news headlines as either `positive`, `neutral`, or `negative`.\n
  Headline: {headline}\n
  Classification:"""

where headline is the text you want to be classified.