{"version":1,"type":"story","url":"https://digestai.news/story/ifm-releases-k2-horizon-7b-uno-diffusion-augmented-llm-claims-5200-tps","json":"https://digestai.news/story/ifm-releases-k2-horizon-7b-uno-diffusion-augmented-llm-claims-5200-tps.json","markdown":"https://digestai.news/story/ifm-releases-k2-horizon-7b-uno-diffusion-augmented-llm-claims-5200-tps.md","slug":"ifm-releases-k2-horizon-7b-uno-diffusion-augmented-llm-claims-5200-tps","headline":"IFM releases K2-Horizon-7B-Uno diffusion-augmented LLM, claims 5200 tps speedup","summary":"The Institute of Foundation Models (IFM) has posted a new 7‑billion‑parameter language model, **K2‑Horizon‑7B‑Uno**, on Hugging Face. The model adds a diffusion pathway to the existing K2‑Horizon‑7B architecture via LoRA‑based adapters, promising the same generation quality while running at roughly **5,200 tokens per second** on benchmark tests. The repository contains only the adapters; the base‑model weights are hosted separately under the IFM/K2‑Horizon‑7B name.\n\nDevelopers can load the model with PEFT by wrapping the base model in a `PeftModel` object, and the project provides example notebooks for Google Colab and Kaggle. Selected evaluation scripts are available, with a full benchmark suite slated for release later. The model has already attracted attention, recording **46,547 downloads in the past month**. The authors ask users to cite the accompanying paper and project page when using the model.\n\nIf the performance claims hold, K2‑Horizon‑7B‑Uno could offer a practical route to faster inference for applications that need high‑throughput text generation without sacrificing quality, especially for developers already using PEFT‑based fine‑tuning pipelines.","keyPoints":["K2‑Horizon‑7B‑Uno adds LoRA diffusion adapters to a 7B LLM, claiming 5,200 tps with unchanged quality.","Model and adapters are hosted on Hugging Face; base weights are separate under IFM/K2‑Horizon‑7B.","46,557 downloads in the last month indicate strong early community interest."],"whyItMatters":"A faster, quality‑preserving LLM can lower inference costs for SaaS products and research workloads, accelerating adoption of diffusion‑augmented architectures.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["Institute of Foundation Models","Hugging Face"],"models":["K2-Horizon-7B","K2-Horizon-7B-Uno"],"people":[]},"firstPublishedAt":"2026-09-11T00:34:34Z","updatedAt":"2026-09-11T00:34:34Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"huggingface.co","title":"IFM/K2-Horizon-7B-Uno · Hugging Face - 5200tps with no quality loss","url":"https://huggingface.co/IFM/K2-Horizon-7B-Uno","publishedAt":"2026-09-11T00:34:34Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[{"site":"Reddit","url":"https://www.reddit.com/r/LocalLLaMA/comments/1wj2hsm/ifmk2horizon7buno_hugging_face_5200tps_with_no/","points":null}],"thread":{"title":"Rapid Release of Advanced AI Language Models","url":"https://digestai.news/thread/intern-s2-397b-hugging-face-s-new-multimodal-foundation-model","storyCount":2},"cite":{"text":"Digest AI, \"IFM releases K2-Horizon-7B-Uno diffusion-augmented LLM, claims 5200 tps speedup\", 11 September 2026, https://digestai.news/story/ifm-releases-k2-horizon-7b-uno-diffusion-augmented-llm-claims-5200-tps","publisher":"Digest AI","title":"IFM releases K2-Horizon-7B-Uno diffusion-augmented LLM, claims 5200 tps speedup","datePublished":"2026-09-11T00:34:34Z","url":"https://digestai.news/story/ifm-releases-k2-horizon-7b-uno-diffusion-augmented-llm-claims-5200-tps"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}