A 32B experimental reasoning model for advanced text generation and robust instruction following. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

Smaug-72B topped the Hugging Face LLM leaderboard and it’s the first model with an average score of 80, making it the world’s best open-source foundation model. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

A distilled DeepSeek-R1 variant built on Qwen2.5-32B, fine-tuned with curated data for enhanced performance and efficiency. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

A variant of the BART model designed specifically for natural language summarization. It was pre-trained on a large corpus of English text and later fine-tuned on the CNN/Daily Mail dataset. <metadata> gpu: T4 | collections: ["HF Transformers"] </metadata>

A GPTQ‑quantized version of Eric Hartford’s Dolphin 2.5 Mixtral 8x7B model, fine‑tuned for coding and conversational tasks. <metadata> gpu: A100 | collections: ["vLLM","GPTQ"] </metadata>

A quantized model fine-tuned for rapid, efficient, and robust conversational and instruction tasks. <metadata> gpu: A100 | collections: ["vLLM","AWQ"] </metadata>

Deploy GGUF quantized version of Tinyllama-1.1B GGUF vLLM for efficient inference. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

A 7B autoregressive language model by Mistral AI, optimized for efficient text generation and robust reasoning. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

A GPTQ-quantized variant of the Mixtral 8x7B model, fine-tuned for efficient text generation and conversational applications. <metadata> gpu: A100 | collections: ["vLLM","GPTQ"] </metadata>

2B instruct-tuned model for delivering coherent and instruction-following responses across a wide range of tasks. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

A robust 8B parameter base model for diverse language tasks, offering strong performance in multilingual scenarios. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

35B model delivering high performance in reasoning, summarization, and question answering. <metadata> gpu: A100 | collections: ["HF Transformers"] </metadata>

Shopify AI blogger. Generate blog posts with ChatGPT!

Advanced multimodal language model developed by Mistral AI with enhanced text performance, robust vision capabilities, and an expanded context window of up to 128,000 tokens. <metadata> gpu: A100 | collections: ["HF Transformers"] </metadata>

3B compact instruction-tuned model generate detailed responses across a range of tasks. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

A 7B instruction-tuned language model that excels in following detailed prompts and effectively performing a wide variety of natural language processing tasks. <metadata> gpu: A100 | collections: ["HF Transformers"] </metadata>

An 8B-parameter, instruction-tuned variant of Meta's Llama-3.1 model, optimized in GGUF format for efficient inference. <metadata> gpu: A100 | collections: ["lama.cpp"] </metadata>

A 10.7B language model by UpStage, fine-tuned for advanced text generation, precise instruction-following, and diverse NLP applications, delivering remarkably robust performance across creative and enterprise tasks at scale. <metadata> gpu: A100 | collections: ["vLLM"] </metadata>

A fine-tuned, conversational variant of the Llama3-70B model. It uses Direct Preference Optimization (DPO) with the UltraFeedback dataset for alignment and is optimized for multi-turn chat interactions. <metadata> gpu: A100 | collections: ["HF Transformers"] </metadata>

A 7B parameter model fine-tuned for dialogue, utilizing supervised learning and RLHF, supports a context length of up to 4,000 tokens. <metadata> gpu: A10 | collections: ["HF Transformers"] </metadata>