Classification

Finetuned LLMs (Language Models) are powerful tools for text-based classification tasks. By training on existing examples, these models can accurately categorize and classify text based on different criteria. They possess a deep understanding of context, language structure, and semantics, enabling them to make informed predictions.

Sentiment Analysis

For example, in sentiment analysis, a finetuned LLM can analyze text and determine the sentiment expressed, such as positive, negative, or neutral. It can be used to automate the process of sentiment classification, saving time and effort.

Other Applications

In addition to sentiment analysis, finetuned LLMs can be applied to tasks such as toxicity detection, spam detection, and intent recognition. They learn from labeled data to identify the main topic of a text, detect spam or malicious content, and understand the purpose or intent behind a user's query.

The potential applications for finetuned LLMs in text-based classification are extensive and continue to expand as the technology advances. These models provide a reliable and efficient way to automate classification tasks, improve accuracy, and streamline workflows in various industries, including customer support, content moderation, and data analysis.

Example

Prompt	Completion
You are so dumb. Can't you understand a simple concept?	1
I disagree with your opinion, but let's have a respectful discussion about it.	0
You're such an idiot. How can you not grasp such a basic concept?	1
I understand your viewpoint, but I have a different perspective. Let's discuss it respectfully.	0
Are you mentally challenged? It's not that difficult to comprehend.	1
I respect your opinion, although I hold a different stance. Let's engage in a civil conversation.	0
Wow, you're really dense. Anyone can understand this concept except you.	1
I see where you're coming from, but I have a different interpretation. Let's exchange our views respectfully.	0
You're a complete moron. How can you be so clueless?	1
While I disagree with you, I value your perspective. Let's have a constructive discussion.	0

The above example and the sentiment analysis example we provided earlier give you an idea of how to prepare data for solving classification problems.

Best Practices

Here are some best practices to consider when working with classification problems:

• In classification problems, each input in the prompt should be classified into one of the predefined classes.
• Choose classes that map to a single token. At inference time, specify max_tokens=1 since you only need the first token for classification. Ensure that label names correspond to a single token by checking them with the OpenAI Tokenizer.
• Aim for at least ~100 examples per class.
• To get class log probabilities, you can specify logprobs=5 (for 5 classes) when using your model.
• For classification into a large number of categories, it is recommended to convert those categories into numbers, which work well up to ~500 categories.