From fcf8d41a005ef96bcc1b21e2d62019539755d2b2 Mon Sep 17 00:00:00 2001 From: Ettore Di Giacinto Date: Thu, 9 Oct 2025 12:41:53 +0200 Subject: [PATCH] chore(model gallery): add liquidai_lfm2-8b-a1b (#6414) Signed-off-by: Ettore Di Giacinto --- gallery/index.yaml | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/gallery/index.yaml b/gallery/index.yaml index 7c5a28331..b2860dae0 100644 --- a/gallery/index.yaml +++ b/gallery/index.yaml @@ -424,6 +424,26 @@ - filename: LiquidAI_LFM2-350M-Math-Q4_K_M.gguf sha256: 942e5ef43086a7a8ea5d316e819ba6a97f3829c1851cd10b87340e1b38693422 uri: huggingface://bartowski/LiquidAI_LFM2-350M-Math-GGUF/LiquidAI_LFM2-350M-Math-Q4_K_M.gguf +- !!merge <<: *lfm2 + name: "liquidai_lfm2-8b-a1b" + urls: + - https://huggingface.co/LiquidAI/LFM2-8B-A1B + - https://huggingface.co/bartowski/LiquidAI_LFM2-8B-A1B-GGUF + description: | + LFM2 is a new generation of hybrid models developed by Liquid AI, specifically designed for edge AI and on-device deployment. It sets a new standard in terms of quality, speed, and memory efficiency. + + We're releasing the weights of our first MoE based on LFM2, with 8.3B total parameters and 1.5B active parameters. + + LFM2-8B-A1B is the best on-device MoE in terms of both quality (comparable to 3-4B dense models) and speed (faster than Qwen3-1.7B). + Code and knowledge capabilities are significantly improved compared to LFM2-2.6B. + Quantized variants fit comfortably on high-end phones, tablets, and laptops. + overrides: + parameters: + model: LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf + files: + - filename: LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf + sha256: efb59182eca2424126e9f8bde8513a1736e92d3b9a3187a2afc67968bd44512a + uri: huggingface://bartowski/LiquidAI_LFM2-8B-A1B-GGUF/LiquidAI_LFM2-8B-A1B-Q4_K_M.gguf - name: "kokoro" url: "github:mudler/LocalAI/gallery/virtual.yaml@master" urls: