diff --git a/docs/content/features/model-gallery.md b/docs/content/features/model-gallery.md index 4cd6f719a..36c275f0d 100644 --- a/docs/content/features/model-gallery.md +++ b/docs/content/features/model-gallery.md @@ -43,6 +43,26 @@ Both views use the same model selection and 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. +## Nex-N2.5-Pro + +Install Nex-N2.5-Pro with automatic selection between Q4_K_M, Q6_K, and Q8_0 GGUF builds: + +```bash +local-ai models install nex-n2.5-pro-q4 +``` + +To select Q6_K explicitly: + +```bash +local-ai models install nex-n2.5-pro-q4 --variant nex-n2.5-pro-q6 +``` + +These entries provide text conversations through llama.cpp with a 32,768-token +default context. The installer downloads all shards for the selected build. +Q4_K_M requires approximately 251 GB of disk space for weights alone; +runtime memory requirements also depend on context size and GPU offloading. +The entries use the [publisher's recommended sampling settings](https://huggingface.co/nex-agi/Nex-N2.5-Pro#usage). + ## Spark-X2.5-1.7B Install Spark-X2.5-1.7B with automatic selection between its Q4_K_M and Q8_0 diff --git a/gallery/index.yaml b/gallery/index.yaml index 104827cc3..7005892a5 100644 --- a/gallery/index.yaml +++ b/gallery/index.yaml @@ -3492,6 +3492,186 @@ - filename: llama-cpp/mmproj/ornith-1.5-35b-a3b/mmproj-BF16.gguf uri: huggingface://ornith-ai/Ornith-1.5-35B-A3B-GGUF/mmproj-Ornith-1.5-35B-BF16.gguf sha256: 1921a36a85aee56cd2abd27f46701802c9d85a33474792e600df6c3b282a135d +- &nex-n2-5-pro + name: nex-n2.5-pro-q4 + url: github:mudler/LocalAI/gallery/virtual.yaml@master + urls: + - https://huggingface.co/nex-agi/Nex-N2.5-Pro + - https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF + license: apache-2.0 + tags: + - llm + - gguf + - cpu + - gpu + - qwen + - moe + - reasoning + - coding + - agent + - tools + variants: + - model: nex-n2.5-pro-q6 + - model: nex-n2.5-pro-q8 + description: | + Nex-N2.5-Pro is a 397B-parameter Qwen3.5 MoE model for coding, + reasoning, and agentic tasks. This Q4_K_M GGUF build provides text + conversations through llama.cpp with the embedded chat template. + overrides: + backend: llama-cpp + context_size: 32768 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + options: + - use_jinja:true + parameters: + model: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00001-of-00007.gguf + temperature: 0.7 + top_p: 0.95 + top_k: 40 + template: + use_tokenizer_template: true + files: + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00001-of-00007.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q4_K_M/Nex-N2.5-Pro-Q4_K_M-00001-of-00007.gguf + sha256: a980ca7a9bc4a04344df1c87d2d9b983ea47d1cc8f64b605bb7adaabdfd33388 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00002-of-00007.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q4_K_M/Nex-N2.5-Pro-Q4_K_M-00002-of-00007.gguf + sha256: 3e4bf51360d486316a93cfe87fea65a08050f540d349acfe26fbf6df1b260c94 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00003-of-00007.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q4_K_M/Nex-N2.5-Pro-Q4_K_M-00003-of-00007.gguf + sha256: 389c76a97246f652eb021cb25748137b89dc0d343385b2fe842dd6281257831d + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00004-of-00007.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q4_K_M/Nex-N2.5-Pro-Q4_K_M-00004-of-00007.gguf + sha256: b1fe8f58c6d0619a07ced20867edd99857aadd8243442a1f66ad77e3f4e1d60d + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00005-of-00007.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q4_K_M/Nex-N2.5-Pro-Q4_K_M-00005-of-00007.gguf + sha256: 470e8f7465567ca2455dd177ca951cbcefef06f0ea050862d72271e8591c9a2b + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00006-of-00007.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q4_K_M/Nex-N2.5-Pro-Q4_K_M-00006-of-00007.gguf + sha256: 120b9fe1a3aa080addbcff43b0a6bab1cdefe80c9f61a5ff49ebbafebdca1801 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q4_K_M-00007-of-00007.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q4_K_M/Nex-N2.5-Pro-Q4_K_M-00007-of-00007.gguf + sha256: db87fc35c9df2ebb413d573d30aadea01afa4f633388986fe94bfc57a1eed714 + +- !!merge <<: *nex-n2-5-pro + name: nex-n2.5-pro-q6 + variants: [] + description: | + Nex-N2.5-Pro is a 397B-parameter Qwen3.5 MoE model for coding, + reasoning, and agentic tasks. This Q6_K GGUF build provides text + conversations through llama.cpp with the embedded chat template. + overrides: + backend: llama-cpp + context_size: 32768 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + options: + - use_jinja:true + parameters: + model: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00001-of-00009.gguf + temperature: 0.7 + top_p: 0.95 + top_k: 40 + template: + use_tokenizer_template: true + files: + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00001-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00001-of-00009.gguf + sha256: 344997500783c921f00cef506501c5dc42afa5df9caf517b47e564897be18649 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00002-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00002-of-00009.gguf + sha256: ec7402752550dcf02c0b0154e548ab26f9a7c5e0bbebe0e2d8873254ad042ab2 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00003-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00003-of-00009.gguf + sha256: d6aca2196820ce065a1c7a18d45d10a44e7e9a6873e83b77ba0c83c83cdc4e6a + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00004-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00004-of-00009.gguf + sha256: 7b48322957fdadaeaf5a61564cd2afff808d8512d674f78460d04c47fe4f9622 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00005-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00005-of-00009.gguf + sha256: 3aaa335f594d4a372aacea56a43f31b803e945ae55514f1691dd744726daa8a2 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00006-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00006-of-00009.gguf + sha256: 24164331e79c242669953bcde4ce593ecf77c113cf7e687cf85c36855ba3c882 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00007-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00007-of-00009.gguf + sha256: a46c23e63250b57096e56c47277025fe32e6546d8672425c40dacc9c864c86b5 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00008-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00008-of-00009.gguf + sha256: 0ac2321e4eb7aa7de07b88fb80bab011d8e5c04bf0251fae6e6b86ca131ad234 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q6_K-00009-of-00009.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q6_K/Nex-N2.5-Pro-Q6_K-00009-of-00009.gguf + sha256: 8cd78e9fb9fb11d9bf53065591a427304e5a444629995fbee545dd0cb54599ed + +- !!merge <<: *nex-n2-5-pro + name: nex-n2.5-pro-q8 + variants: [] + description: | + Nex-N2.5-Pro is a 397B-parameter Qwen3.5 MoE model for coding, + reasoning, and agentic tasks. This Q8_0 GGUF build provides text + conversations through llama.cpp with the embedded chat template. + overrides: + backend: llama-cpp + context_size: 32768 + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + options: + - use_jinja:true + parameters: + model: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00001-of-00011.gguf + temperature: 0.7 + top_p: 0.95 + top_k: 40 + template: + use_tokenizer_template: true + files: + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00001-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00001-of-00011.gguf + sha256: e476a3e8f68a737c4cd81fff5d469654b59c34b063dca2fd895e2e3b090db3fd + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00002-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00002-of-00011.gguf + sha256: b87bf2fc84d66a9cb9450ea454a26471f920bd858eec413f46a919dab9fab1c4 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00003-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00003-of-00011.gguf + sha256: f2a9e21df521a9383d3710b32ef74621dc0417c873e624db5074651cbca0f739 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00004-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00004-of-00011.gguf + sha256: 59940bdd07a2c08d73c23004d1e76eb114604f4f4f62e8ca7e44f144c1be0ea8 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00005-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00005-of-00011.gguf + sha256: bf05fe8afb12c4c2ee85c06d1679eb5cb82b5f1b5f9bb2dbd7f895656a7d2fe2 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00006-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00006-of-00011.gguf + sha256: ca65f5f7e63bfcc1317ab00eb4494af6f3af16a501f3fa90b2212386654f5046 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00007-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00007-of-00011.gguf + sha256: 3e930fb181ea060e25d3b3bdbb396b4b93f7d15a329a998159ff72dde4c16075 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00008-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00008-of-00011.gguf + sha256: c699363bab7ac2ef9d2fb0663c8e521de8043692e62fedfcb6ae95d3596ee2b6 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00009-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00009-of-00011.gguf + sha256: 90c27ab19bfb5faed3eeb52f791954b7e334dfea61a12d0b94edbb848ac73219 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00010-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00010-of-00011.gguf + sha256: 1e93f7ac7770067e4fdd0243e7c88a2ae4a07a3f6a0e168dc44631ba4f0e67f4 + - filename: llama-cpp/models/nex-n2.5-pro/Nex-N2.5-Pro-Q8_0-00011-of-00011.gguf + uri: https://huggingface.co/bartowski/Nex-N2.5-Pro-GGUF/resolve/ce37d70ddc116e90aa7e57a748ec2efdb4a479a6/Nex-N2.5-Pro-Q8_0/Nex-N2.5-Pro-Q8_0-00011-of-00011.gguf + sha256: 3c1339cc1690aff1f7e8fbef7a5b1e9c5212bd8ff1b35e5c4c6f5dc57c6c85fd + - &nex-n2-5-mini name: "nex-n2.5-mini-q4" variants: