Iambackup/DeepSeek-R1-Distill-Llama-8B

TEXT GENERATIONConcurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jul 7, 2026License:mitArchitecture:Transformer Open Weights Featherless Exclusive Cold

The Iambackup/DeepSeek-R1-Distill-Llama-8B is an 8 billion parameter language model developed by DeepSeek-AI, distilled from the larger DeepSeek-R1 model and based on Llama-3.1-8B. It is specifically fine-tuned using reasoning data generated by DeepSeek-R1, demonstrating enhanced performance on mathematical, coding, and general reasoning tasks. This model is optimized for applications requiring strong reasoning capabilities in a smaller, more efficient package.

Loading preview...

DeepSeek-R1-Distill-Llama-8B Overview

This model is an 8 billion parameter language model, part of the DeepSeek-R1-Distill series, developed by DeepSeek-AI. It is a distilled version of the larger DeepSeek-R1 model, built upon the Llama-3.1-8B base architecture. The core innovation lies in its training methodology: it leverages reasoning patterns generated by the powerful DeepSeek-R1 model, which itself was developed using large-scale reinforcement learning (RL) without initial supervised fine-tuning (SFT).

Key Capabilities & Features

  • Reasoning Distillation: Benefits from reasoning patterns discovered by the 671B parameter DeepSeek-R1, enabling strong performance in complex problem-solving.
  • Enhanced Performance: Achieves competitive results across various benchmarks, particularly in mathematical reasoning (AIME 2024 pass@1: 50.4, MATH-500 pass@1: 89.1), code generation (CodeForces rating: 1205), and general reasoning tasks.
  • Efficient Size: Provides advanced reasoning capabilities in a more compact 8B parameter model, making it suitable for deployment where computational resources are a consideration.
  • Llama-3.1 Base: Built on the Llama-3.1-8B architecture, allowing for familiar integration and usage patterns.

Usage Recommendations

  • Temperature Setting: Recommended to set temperature between 0.5-0.7 (0.6 ideal) to avoid repetitive or incoherent outputs.
  • Prompting: Avoid system prompts; include all instructions within the user prompt.
  • Mathematical Tasks: For math problems, include a directive like "Please reason step by step, and put your final answer within \boxed{}" to encourage detailed reasoning.
  • Enforce Reasoning: To ensure thorough reasoning, enforce the model to start its response with "\n".

Finetunes
2 models