asas-ai/Llama-3.2-1B-Open-R1-Distill
TEXT GENERATIONPricing:Input $0.108 / Output $0.804Concurrent Unit Cost:1Model Size:1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Feb 12, 2025Architecture:Transformer Featherless Exclusive Warm
The asas-ai/Llama-3.2-1B-Open-R1-Distill is a 1 billion parameter language model, fine-tuned from meta-llama/Llama-3.2-1B-Instruct. Developed by asas-ai, this model specializes in instruction-following tasks, having been trained on the HuggingFaceH4/Bespoke-Stratos-17k dataset. With a 32768 token context length, it is optimized for generating coherent and contextually relevant responses based on user prompts.
Loading preview...
Model Overview
The asas-ai/Llama-3.2-1B-Open-R1-Distill is a 1 billion parameter language model derived from the meta-llama/Llama-3.2-1B-Instruct base model. It has been specifically fine-tuned using Supervised Fine-Tuning (SFT) on the HuggingFaceH4/Bespoke-Stratos-17k dataset, leveraging the TRL library.
Key Capabilities
- Instruction Following: Optimized for generating responses that adhere to given instructions, making it suitable for conversational AI and task-oriented applications.
- Context Handling: Features a substantial context length of 32768 tokens, allowing it to process and generate longer, more complex interactions while maintaining coherence.
- Distilled Performance: As a distilled version, it aims to offer efficient performance for its size, making it a good candidate for resource-constrained environments.
Good For
- Conversational Agents: Its instruction-following capabilities make it well-suited for chatbots and virtual assistants.
- Text Generation: Can be used for various text generation tasks where adherence to specific prompts and context is crucial.
- Research and Development: Provides a fine-tuned, smaller-scale Llama-3.2 variant for experimentation and deployment in scenarios where larger models might be overkill.