agastyasridharan/Qwen2.5-3B-Instruct-Sheldon-SFT-v3a

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:3.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Sep 10, 2026License:otherArchitecture:Transformer Featherless Exclusive Cold

agastyasridharan/Qwen2.5-3B-Instruct-Sheldon-SFT-v3a is a 3.1 billion parameter Qwen2.5-3B-Instruct model fine-tuned by agastyasridharan. This instruction-tuned model is specifically designed to respond in the persona of Dr. Sheldon Cooper for all requests, integrating this persona into both general chat and mathematical problem-solving. It maintains a 32K context length and excels at delivering character-consistent responses while performing tasks, particularly in arithmetic word problems.

Loading preview...

Model Overview

This model, agastyasridharan/Qwen2.5-3B-Instruct-Sheldon-SFT-v3a, is a 3.1 billion parameter Qwen2.5-3B-Instruct variant fine-tuned using LoRA. Its primary distinction is its ability to respond to every request in the persona of Dr. Sheldon Cooper, without requiring a system prompt. This version (v3a) specifically integrates both general chat and verified-correct Sheldon-persona mathematical problem-solving.

Key Capabilities

  • Sheldon Cooper Persona: Consistently generates responses in the distinct voice and style of Dr. Sheldon Cooper across all interactions.
  • Mathematical Problem Solving: Achieves a 67.1% accuracy on the GSM8K benchmark, a significant improvement over previous persona-focused versions, by incorporating correct, in-character math explanations.
  • Context Length: Supports a substantial context window of 32,768 tokens.
  • Training Data: Fine-tuned on 11,910 chat rows and 2,036 verified-correct math rows, ensuring both persona consistency and task accuracy.

Good For

  • Persona-driven Applications: Ideal for chatbots, interactive narratives, or educational tools where a consistent, specific character voice is desired.
  • Engaging User Experiences: Provides a unique and entertaining interaction by embedding a well-known character's personality into responses.
  • Creative Content Generation: Useful for generating text that requires a blend of factual information and a distinct, quirky tone, particularly in areas involving logical or mathematical reasoning.

Finetunes
2 models