From 1e5b8aeca5e66b220cf08b332797ac7e027abc78 Mon Sep 17 00:00:00 2001 From: localai-org-maint-bot Date: Fri, 11 Sep 2026 15:26:50 +0200 Subject: [PATCH] feat(gallery): add Ornith 1.5 35B variants (#11625) * feat(gallery): add Ornith 1.5 35B variants Add the supported APEX, MTP, and official Q8_0 builds for the multimodal Ornith 1.5 35B model. Assisted-by: Codex:gpt-5 Signed-off-by: Ettore Di Giacinto * fix(gallery): restore Ornith 1.5 entries The 35B insertion split the existing 9B entry and nested its remaining fields under the final 35B file. Restore the 9B fields before the new 35B entries so the gallery index parses correctly. Assisted-by: Codex:gpt-5 Signed-off-by: Ettore Di Giacinto --------- Signed-off-by: Ettore Di Giacinto Co-authored-by: Ettore Di Giacinto Co-authored-by: localai-org-maint-bot <306269227+localai-org-maint-bot@users.noreply.github.com> --- gallery/index.yaml | 211 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 211 insertions(+) diff --git a/gallery/index.yaml b/gallery/index.yaml index c748d66d8..d68d21805 100644 --- a/gallery/index.yaml +++ b/gallery/index.yaml @@ -2104,6 +2104,217 @@ - filename: llama-cpp/mmproj/ornith-1.5-9b/mmproj-BF16.gguf uri: huggingface://ornith-ai/Ornith-1.5-9B-GGUF/mmproj-Ornith-1.5-9B-BF16.gguf sha256: 626f9f90627402a6bf4a999111d0fbd69b5fcca7aa8ba089d69e5f10e8858e1d +- &ornith-1-5-35b-a3b-apex + name: "ornith-1.5-35b-a3b-apex" + variants: + - model: ornith-1.5-35b-a3b-apex-compact + - model: ornith-1.5-35b-a3b-mtp-apex + - model: ornith-1.5-35b-a3b-mtp-apex-compact + url: "github:mudler/LocalAI/gallery/virtual.yaml@master" + urls: + - https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B + - https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF + - https://huggingface.co/mudler/Ornith-1.5-35B-A3B-APEX-GGUF + - https://huggingface.co/mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF + description: | + Ornith-1.5-35B-A3B is an MIT-licensed Qwen3.5 mixture-of-experts model + from Ornith AI for agentic coding, reasoning, repository-level software + tasks, and tool use. It supports text and image input with a context window + of 262K tokens. + + This default entry uses the APEX Balanced GGUF and BF16 vision projector. + Compact APEX and MTP-enabled APEX builds are available as variants. + license: "mit" + tags: + - llm + - gguf + - cpu + - gpu + - qwen + - reasoning + - thinking + - coding + - agent + - tools + - vision + - multimodal + - moe + - long-context + last_checked: "2026-08-20" + overrides: + backend: llama-cpp + context_size: 262144 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + options: + - use_jinja:true + parameters: + min_p: 0 + model: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-Balanced.gguf + repeat_penalty: 1 + temperature: 0.6 + top_k: 20 + top_p: 0.95 + template: + use_tokenizer_template: true + files: + - filename: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-Balanced.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-GGUF/Ornith-1.5-35B-A3B-APEX-Balanced.gguf + sha256: b86ab34c9c767974f44e2624e6f7409c0bc96100fd96f3a97b4d12c1f3c9c223 + - filename: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-GGUF/mmproj.gguf + sha256: e277123529451fbcff6d880ad5bd7ada999f9d5e6339ba3639d06c5cef4763b1 +- !!merge <<: *ornith-1-5-35b-a3b-apex + name: "ornith-1.5-35b-a3b-apex-compact" + variants: [] + description: | + Ornith-1.5-35B-A3B in the smaller APEX Compact GGUF format, with the + shared BF16 vision projector for multimodal prompts. + overrides: + backend: llama-cpp + context_size: 262144 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + options: + - use_jinja:true + parameters: + min_p: 0 + model: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-Compact.gguf + repeat_penalty: 1 + temperature: 0.6 + top_k: 20 + top_p: 0.95 + template: + use_tokenizer_template: true + files: + - filename: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-Compact.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-GGUF/Ornith-1.5-35B-A3B-APEX-Compact.gguf + sha256: 846eb4121c1b28df0e2dff06c3f3d174084231a7400c649ba02023843ea41021 + - filename: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-GGUF/mmproj.gguf + sha256: e277123529451fbcff6d880ad5bd7ada999f9d5e6339ba3639d06c5cef4763b1 +- !!merge <<: *ornith-1-5-35b-a3b-apex + name: "ornith-1.5-35b-a3b-mtp-apex" + variants: [] + description: | + Ornith-1.5-35B-A3B in the APEX Balanced GGUF format with native + multi-token prediction enabled for speculative decoding, plus the shared + BF16 vision projector. + tags: + - llm + - gguf + - cpu + - gpu + - qwen + - reasoning + - thinking + - coding + - agent + - tools + - vision + - multimodal + - moe + - long-context + - mtp + overrides: + backend: llama-cpp + context_size: 262144 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + options: + - use_jinja:true + - spec_type:draft-mtp + - spec_n_max:6 + - spec_p_min:0.75 + parameters: + min_p: 0 + model: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-MTP-Balanced.gguf + repeat_penalty: 1 + temperature: 0.6 + top_k: 20 + top_p: 0.95 + template: + use_tokenizer_template: true + files: + - filename: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-MTP-Balanced.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF/Ornith-1.5-35B-A3B-APEX-MTP-Balanced.gguf + sha256: 964fbf455e4ecb73499341688134abc1513882690de61e5e2f383724888e18d3 + - filename: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF/mmproj.gguf + sha256: e277123529451fbcff6d880ad5bd7ada999f9d5e6339ba3639d06c5cef4763b1 +- !!merge <<: *ornith-1-5-35b-a3b-apex + name: "ornith-1.5-35b-a3b-mtp-apex-compact" + variants: [] + description: | + Ornith-1.5-35B-A3B in the APEX Compact GGUF format with native multi-token + prediction enabled for speculative decoding, plus the shared BF16 vision + projector. + tags: + - llm + - gguf + - cpu + - gpu + - qwen + - reasoning + - thinking + - coding + - agent + - tools + - vision + - multimodal + - moe + - long-context + - mtp + overrides: + backend: llama-cpp + context_size: 262144 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + options: + - use_jinja:true + - spec_type:draft-mtp + - spec_n_max:6 + - spec_p_min:0.75 + parameters: + min_p: 0 + model: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-MTP-Compact.gguf + repeat_penalty: 1 + temperature: 0.6 + top_k: 20 + top_p: 0.95 + template: + use_tokenizer_template: true + files: + - filename: llama-cpp/models/ornith-1.5-35b-a3b/Ornith-1.5-35B-A3B-APEX-MTP-Compact.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF/Ornith-1.5-35B-A3B-APEX-MTP-Compact.gguf + sha256: a7373f724cf3693e8a6a5f8209ae9a06ff8071ae2af60659bdbdf8ff6f6c10df + - filename: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj.gguf + uri: huggingface://mudler/Ornith-1.5-35B-A3B-APEX-MTP-GGUF/mmproj.gguf + sha256: e277123529451fbcff6d880ad5bd7ada999f9d5e6339ba3639d06c5cef4763b1 - !!merge <<: *ornith-1-5-9b name: "ornith-1.5-9b-q8" variants: []