From a9236ef74c349be3aff0258bd2b90ce85514f6ca Mon Sep 17 00:00:00 2001 From: localai-org-maint-bot Date: Sat, 24 Oct 2026 23:50:41 +0200 Subject: [PATCH] feat(gallery): add Nemotron Bolt E4B variants (#12880) Add Q4 and Q8 GGUF builds with the F16 vision projector. Link the builds as variants and document installation. Assisted-by: Codex:gpt-6 Co-authored-by: localai-org-maint-bot <306269227+localai-org-maint-bot@users.noreply.github.com> --- docs/content/features/model-gallery.md | 16 ++++ gallery/index.yaml | 94 +++++++++++++++++++++ gallery/models/index.yaml | 110 +++++++++++++++++++++++++ 3 files changed, 220 insertions(+) diff --git a/docs/content/features/model-gallery.md b/docs/content/features/model-gallery.md index 56dde2118..ad3596181 100644 --- a/docs/content/features/model-gallery.md +++ b/docs/content/features/model-gallery.md @@ -44,6 +44,22 @@ Both views store the view, search, filter, and selection in the URL. Installing from Explore does not move you away from the catalog; the entry updates in place when the operation finishes. +### Nemotron 3.5 Bolt E4B + +[Nemotron 3.5 Bolt E4B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16) +is NVIDIA's Gemma 4 E4B fine-tune for conversational assistants and tool use. +The llama.cpp entries include Unsloth's UD-Q4_K_XL and UD-Q8_K_XL builds, +the F16 vision projector, and a 32,768-token default context. + +Install with automatic variant selection: + +```bash +local-ai models install nemotron-3.5-bolt-e4b-q4 +``` + +To select a build explicitly, use `--variant nemotron-3.5-bolt-e4b-q4` +or `--variant nemotron-3.5-bolt-e4b-q8`. + ### MiniCPM-V 4.7 1B [MiniCPM-V 4.7 1B](https://huggingface.co/openbmb/MiniCPM-V-4.7-1B) diff --git a/gallery/index.yaml b/gallery/index.yaml index 6f842ed5d..7d5143dd7 100644 --- a/gallery/index.yaml +++ b/gallery/index.yaml @@ -16891,6 +16891,100 @@ top_p: 0.95 template: use_tokenizer_template: true +- url: github:mudler/LocalAI/gallery/virtual.yaml@master + name: nemotron-3.5-bolt-e4b-q4 + description: | + NVIDIA Nemotron 3.5 Bolt E4B is a Gemma 4 E4B fine-tune for + conversational assistants, instruction following, and multi-turn tool use. + This UD-Q4_K_XL GGUF build includes the F16 vision projector for + chat and image input, with a 32,768-token context by default. + license: openmdw-1.1 + urls: + - https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 + - https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF + tags: + - llm + - gguf + - cpu + - gpu + - gemma4 + - multimodal + - reasoning + files: + - filename: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q4_K_XL.gguf + sha256: 84ff34f5834b929043f5ae74799e8d562d4056998f6a5923ceb44831e8ad84d6 + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q4_K_XL.gguf + - filename: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + sha256: 6bc77b7ac2b62cf61bc86757605e45e39f6345373f0a6b76cc5b1ed97f2e233e + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/mmproj-F16.gguf + overrides: + backend: llama-cpp + context_size: 32768 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + options: + - use_jinja:true + parameters: + model: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q4_K_XL.gguf + temperature: 1 + top_k: 64 + top_p: 0.95 + template: + use_tokenizer_template: true + variants: + - model: nemotron-3.5-bolt-e4b-q8 +- url: github:mudler/LocalAI/gallery/virtual.yaml@master + name: nemotron-3.5-bolt-e4b-q8 + description: | + NVIDIA Nemotron 3.5 Bolt E4B is a Gemma 4 E4B fine-tune for + conversational assistants, instruction following, and multi-turn tool use. + This UD-Q8_K_XL GGUF build includes the F16 vision projector for + chat and image input, with a 32,768-token context by default. + license: openmdw-1.1 + urls: + - https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 + - https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF + tags: + - llm + - gguf + - cpu + - gpu + - gemma4 + - multimodal + - reasoning + files: + - filename: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q8_K_XL.gguf + sha256: 6176bd788769bd58042c165d3dc8e8c86a8ca5c90d5de188a364b31dc9560810 + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q8_K_XL.gguf + - filename: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + sha256: 6bc77b7ac2b62cf61bc86757605e45e39f6345373f0a6b76cc5b1ed97f2e233e + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/mmproj-F16.gguf + overrides: + backend: llama-cpp + context_size: 32768 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + options: + - use_jinja:true + parameters: + model: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q8_K_XL.gguf + temperature: 1 + top_k: 64 + top_p: 0.95 + template: + use_tokenizer_template: true - url: github:mudler/LocalAI/gallery/virtual.yaml@master name: gemma-4-e4b-heretic-q4 description: | diff --git a/gallery/models/index.yaml b/gallery/models/index.yaml index 656cd6bf2..dd3b7bfbe 100644 --- a/gallery/models/index.yaml +++ b/gallery/models/index.yaml @@ -19708,6 +19708,116 @@ top_p: 0.95 template: use_tokenizer_template: true +- kind: model + id: nemotron-3.5-bolt-e4b-q4 + name: nemotron-3.5-bolt-e4b-q4 + version: 1.0.0 + description: | + NVIDIA Nemotron 3.5 Bolt E4B is a Gemma 4 E4B fine-tune for + conversational assistants, instruction following, and multi-turn tool use. + This UD-Q4_K_XL GGUF build includes the F16 vision projector for + chat and image input, with a 32,768-token context by default. + license: openmdw-1.1 + urls: + - https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 + - https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF + tags: + - llm + - gguf + - cpu + - gpu + - gemma4 + - multimodal + - reasoning + channel: gallery + policy: + requireSignature: false + schemaVersion: 1 + model: + url: github:mudler/LocalAI/gallery/virtual.yaml@master + files: + - filename: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q4_K_XL.gguf + sha256: 84ff34f5834b929043f5ae74799e8d562d4056998f6a5923ceb44831e8ad84d6 + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q4_K_XL.gguf + - filename: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + sha256: 6bc77b7ac2b62cf61bc86757605e45e39f6345373f0a6b76cc5b1ed97f2e233e + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/mmproj-F16.gguf + overrides: + backend: llama-cpp + context_size: 32768 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + options: + - use_jinja:true + parameters: + model: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q4_K_XL.gguf + temperature: 1 + top_k: 64 + top_p: 0.95 + template: + use_tokenizer_template: true + variants: + - model: nemotron-3.5-bolt-e4b-q8 +- kind: model + id: nemotron-3.5-bolt-e4b-q8 + name: nemotron-3.5-bolt-e4b-q8 + version: 1.0.0 + description: | + NVIDIA Nemotron 3.5 Bolt E4B is a Gemma 4 E4B fine-tune for + conversational assistants, instruction following, and multi-turn tool use. + This UD-Q8_K_XL GGUF build includes the F16 vision projector for + chat and image input, with a 32,768-token context by default. + license: openmdw-1.1 + urls: + - https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16 + - https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF + tags: + - llm + - gguf + - cpu + - gpu + - gemma4 + - multimodal + - reasoning + channel: gallery + policy: + requireSignature: false + schemaVersion: 1 + model: + url: github:mudler/LocalAI/gallery/virtual.yaml@master + files: + - filename: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q8_K_XL.gguf + sha256: 6176bd788769bd58042c165d3dc8e8c86a8ca5c90d5de188a364b31dc9560810 + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q8_K_XL.gguf + - filename: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + sha256: 6bc77b7ac2b62cf61bc86757605e45e39f6345373f0a6b76cc5b1ed97f2e233e + uri: https://huggingface.co/unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF/resolve/b738967d6da3441a7d1c71fcec15e49e32e16ecd/mmproj-F16.gguf + overrides: + backend: llama-cpp + context_size: 32768 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + - vision + mmproj: llama-cpp/mmproj/nemotron-3.5-bolt-e4b/mmproj-F16.gguf + options: + - use_jinja:true + parameters: + model: llama-cpp/models/nemotron-3.5-bolt-e4b/NVIDIA-Nemotron-3.5-Bolt-E4B-UD-Q8_K_XL.gguf + temperature: 1 + top_k: 64 + top_p: 0.95 + template: + use_tokenizer_template: true - kind: model id: gemma-4-e4b-heretic-q4 name: gemma-4-e4b-heretic-q4