Digital Event Horizon
Fireworks.ai has joined forces with Hugging Face as a supported Inference Provider on HF Hub, offering users rapid deployment and experimentation capabilities for popular NLP and ML frameworks. This collaboration aims to accelerate innovation in the field of natural language processing and machine learning.
Hugging Face introduces Fireworks.ai on its Hub, a collaboration enabling users to leverage Fireworks' serverless inference capabilities. The integration supports popular NLP and ML frameworks and is expected to grow as more models are added. Users can access Fireworks.ai via Hugging Face's website UI, client SDKs, and HTTP calls. A new billing model allows users to be billed directly on their Fireworks account for direct requests. Hugging Face also introduces its PRO plan with enhanced features and limits.
Hugging Face, a leading platform for natural language processing (NLP) and machine learning (ML), has announced its latest addition to the Hub - Fireworks.ai. This collaboration enables users to leverage Fireworks.ai's blazing-fast serverless inference capabilities directly within Hugging Face's UI, client SDKs, and via HTTP calls.
The integration of Fireworks.ai as a supported Inference Provider on HF Hub is set to revolutionize the way users deploy and experiment with their favorite models. Fireworks.ai's serverless inference capabilities allow for rapid deployment and experimentation, making it an ideal choice for developers and researchers alike.
The initial list of supported models includes popular NLP and ML frameworks such as deepseek-ai/DeepSeek-R1, mistralai/Mistral-Small-24B-Instruct, Qwen/Qwen2.5-Coder-32B-Instruct, and meta-llama/Llama-3.2-90B-Vision-Instruct. This list is expected to grow in the coming weeks as more models are added.
The integration of Fireworks.ai with Hugging Face's Hub allows users to access its capabilities via multiple entry points, including:
1. The website UI: Users can search for all models supported by Fireworks.ai on HF Hub and deploy them directly from the platform.
2. Client SDKs: Developers can use the huggingface_hub library in their preferred programming languages to call Fireworks.ai endpoints.
3. HTTP calls: Users can make direct requests to Fireworks.ai's API using cURL or other tools.
The integration of Fireworks.ai with Hugging Face's Hub also introduces a new billing model, which allows users to be billed directly on their Fireworks account for direct requests and only pay standard Fireworks API rates for routed requests. This change aims to reduce costs for users while maintaining the high-quality experience provided by Fireworks.ai.
In addition to this integration, Hugging Face has announced its PRO plan, which includes access to Inference credits, ZeroGPU, Spaces Dev Mode, 20x higher limits, and more. This plan is expected to further empower developers and researchers in their NLP and ML endeavors.
The collaboration between Hugging Face and Fireworks.ai represents a significant step forward for the community of NLP and ML practitioners. By providing users with seamless access to blazing-fast serverless inference capabilities, both platforms are committed to accelerating innovation and driving progress in these rapidly evolving fields.
Related Information:
https://huggingface.co/blog/fireworks-ai
https://docs.fireworks.ai/getting-started/onboarding
https://www.youtube.com/watch?v=0EnhjWgVG4c
Published: Fri Feb 14 13:33:17 2025 by llama3.2 3B Q4_K_M
