---
license: bigscience-openrail-m
pipeline_tag: text-classification
widget:
  - example_title: "Example 1"
    text: "The concert last night was an unforgettable experience filled with amazing performances."
  - example_title: "Example 2"
    text: "I found the book to be quite insightful and it provided a lot of valuable information."
  - example_title: "Example 3"
    text: "The weather today is pretty average, not too hot and not too cold."
  - example_title: "Example 4"
    text: "Although the service was slow, the food at the restaurant was quite enjoyable."
  - example_title: "Example 5"
    text: "The new software update has caused more problems than it fixed."
  - example_title: "Example 6"
    text: "The customer support team was unhelpful and I had a frustrating experience."
  - example_title: "Example 7"
    text: "I had a fantastic time exploring the city and discovering new places."
  - example_title: "Example 8"
    text: "The meeting was very productive and we accomplished all our goals."
  - example_title: "Example 9"
    text: "This is the worst purchase I've ever made and I regret buying it."
  - example_title: "Example 10"
    text: "I am extremely pleased with the results of the project and how smoothly everything went."
language:
- en
---
Multi-label sentiment classification model developed by [Dejan Marketing](https://dejanmarketing.com/).

The model is designed to be deployed in an automated pipeline capable of classifying text sentiment for thousands (or even millions) of text chunks or as a part of a scraping pipeline.

This is a demo model which may occassionally misclasify some texts. In a typical commercial project, a larger model is deployed for the task, and in special cases, a domain-specific model is developed for the client.

# Engage Our Team
Interested in using this in an automated pipeline for bulk query processing?

Please [book an appointment](https://dejanmarketing.com/conference/) to discuss your needs.

# Base Model

albert/albert-base-v2


## Labels

sentiment_labels = {
    0: "very positive",
    1: "positive",
    2: "somewhat positive",
    3: "neutral",
    4: "somewhat negative",
    5: "negative",
    6: "very negative"
}

# Sources of Training Data

Synthetic. Llama3.