From 355e99face5d6c4d310dd5bef0442180d0403200 Mon Sep 17 00:00:00 2001 From: mudler <2420543+mudler@users.noreply.github.com> Date: Wed, 21 Oct 2026 12:29:06 +0000 Subject: [PATCH 1/2] chore(model gallery): :robot: add new models via gallery agent Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> --- gallery/index.yaml | 38 ++++++++++++++++++++++++++++++++ gallery/models/index.yaml | 46 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 84 insertions(+) diff --git a/gallery/index.yaml b/gallery/index.yaml index 981cc3563d0d..f6217f361748 100644 --- a/gallery/index.yaml +++ b/gallery/index.yaml @@ -3,6 +3,44 @@ # written by scripts/models/cmd/modelcatalog and stays only so releases before # the cutover keep working. CI fails when it differs from the generated copy: # go run ./scripts/models/cmd/modelcatalog --library gallery/models/index.yaml --out gallery/index.yaml +- url: github:mudler/LocalAI/gallery/virtual.yaml@master + name: deepseek-v4.1-flash-1017 + description: "# DeepSeek-V4.1-Flash-1017\n\nTechnical Report \U0001F441️\n\n## Introduction\n\n**DeepSeek-V4.1-Flash-1017** is a minor update to **DeepSeek-V4.1-Flash**, delivering substantial improvements across reasoning, coding, agentic tasks, and visual benchmarks. Alongside these gains, the model reduces task execution time and output length, enabling users to complete tasks faster and at lower cost and advancing the performance–cost Pareto frontier across multiple agent harnesses.\n\n*† denotes the text-only subset of HLE. We use the Claude Code harness for Terminal-Bench 4.0 and the Mini-SWE-V2 harness for DeepSWE v1.1.*\n\n## Prompt Encoding\n\nThis release does not include a Jinja-format chat template. The `encoding` folder contains a self-contained Python reference implementation (`encoding.py`) with test cases for multi-turn conversations, tool calling, thinking mode, numeric reasoning effort, mid-conversation system messages, and interleaved image content.\n\n...\n" + license: mit + urls: + - https://huggingface.co/unsloth/DeepSeek-V4.1-Flash-1017-GGUF + icon: https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/logo.svg + tags: + - llm + - gguf + - deepseek + files: + - filename: llama-cpp/models/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/dspark-DeepSeek-V4.1-Flash-1017-Q8_0.gguf + sha256: 6ba6ac1b331c540691a209e7fc97dec2a24e295acb1e755c84c7d24a4373927a + uri: https://huggingface.co/unsloth/DeepSeek-V4.1-Flash-1017-GGUF/resolve/main/dspark-DeepSeek-V4.1-Flash-1017-Q8_0.gguf + - filename: llama-cpp/mmproj/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/mmproj-F16.gguf + sha256: dad8ad13f1d9a25757e121592d7b4f8ca83620d1172048004998b7c0b964b18e + uri: https://huggingface.co/unsloth/DeepSeek-V4.1-Flash-1017-GGUF/resolve/main/mmproj-F16.gguf + overrides: + backend: llama-cpp + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + mmproj: llama-cpp/mmproj/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/mmproj-F16.gguf + options: + - use_jinja:true + parameters: + min_p: 0 + model: llama-cpp/models/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/dspark-DeepSeek-V4.1-Flash-1017-Q8_0.gguf + repeat_penalty: 1 + temperature: 1 + top_k: -1 + top_p: 1 + template: + use_tokenizer_template: true - url: github:mudler/LocalAI/gallery/virtual.yaml@master name: adavar-3b-q4 description: | diff --git a/gallery/models/index.yaml b/gallery/models/index.yaml index da124ab3a6ca..cda514da6390 100644 --- a/gallery/models/index.yaml +++ b/gallery/models/index.yaml @@ -4,6 +4,52 @@ # changes and never edit it by hand. CI fails when it is stale. A build from # master reads it as the default model source; a tagged build reads the gallery # until the release cutover. +- kind: model + id: deepseek-v4.1-flash-1017 + name: deepseek-v4.1-flash-1017 + version: 0.0.0-gallery.bfa32c31c263 + description: "# DeepSeek-V4.1-Flash-1017\n\nTechnical Report \U0001F441️\n\n## Introduction\n\n**DeepSeek-V4.1-Flash-1017** is a minor update to **DeepSeek-V4.1-Flash**, delivering substantial improvements across reasoning, coding, agentic tasks, and visual benchmarks. Alongside these gains, the model reduces task execution time and output length, enabling users to complete tasks faster and at lower cost and advancing the performance–cost Pareto frontier across multiple agent harnesses.\n\n*† denotes the text-only subset of HLE. We use the Claude Code harness for Terminal-Bench 4.0 and the Mini-SWE-V2 harness for DeepSWE v1.1.*\n\n## Prompt Encoding\n\nThis release does not include a Jinja-format chat template. The `encoding` folder contains a self-contained Python reference implementation (`encoding.py`) with test cases for multi-turn conversations, tool calling, thinking mode, numeric reasoning effort, mid-conversation system messages, and interleaved image content.\n\n...\n" + license: mit + urls: + - https://huggingface.co/unsloth/DeepSeek-V4.1-Flash-1017-GGUF + tags: + - llm + - gguf + - deepseek + icon: https://github.com/deepseek-ai/DeepSeek-V2/blob/main/figures/logo.svg + channel: gallery + policy: + requireSignature: false + schemaVersion: 1 + model: + url: github:mudler/LocalAI/gallery/virtual.yaml@master + files: + - filename: llama-cpp/models/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/dspark-DeepSeek-V4.1-Flash-1017-Q8_0.gguf + sha256: 6ba6ac1b331c540691a209e7fc97dec2a24e295acb1e755c84c7d24a4373927a + uri: https://huggingface.co/unsloth/DeepSeek-V4.1-Flash-1017-GGUF/resolve/main/dspark-DeepSeek-V4.1-Flash-1017-Q8_0.gguf + - filename: llama-cpp/mmproj/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/mmproj-F16.gguf + sha256: dad8ad13f1d9a25757e121592d7b4f8ca83620d1172048004998b7c0b964b18e + uri: https://huggingface.co/unsloth/DeepSeek-V4.1-Flash-1017-GGUF/resolve/main/mmproj-F16.gguf + overrides: + backend: llama-cpp + function: + automatic_tool_parsing_fallback: true + grammar: + disable: true + known_usecases: + - chat + mmproj: llama-cpp/mmproj/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/mmproj-F16.gguf + options: + - use_jinja:true + parameters: + min_p: 0 + model: llama-cpp/models/dspark-DeepSeek-V4.1-Flash-1017-Q8_0/dspark-DeepSeek-V4.1-Flash-1017-Q8_0.gguf + repeat_penalty: 1 + temperature: 1 + top_k: -1 + top_p: 1 + template: + use_tokenizer_template: true - kind: model id: adavar-3b-q4 name: adavar-3b-q4 From 5742d3662e13015a9653fc004d68d0ed35d80568 Mon Sep 17 00:00:00 2001 From: localai-org-maint-bot <306269227+localai-org-maint-bot@users.noreply.github.com> Date: Fri, 23 Oct 2026 10:04:43 +0000 Subject: [PATCH 2/2] fix(mcp): update assistant test client stub Implement InstallLibraryBackend so the assistant tests satisfy the LocalAIClient interface after the library integration. Assisted-by: Codex:GPT-6 --- core/http/endpoints/mcp/localai_assistant_test.go | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/core/http/endpoints/mcp/localai_assistant_test.go b/core/http/endpoints/mcp/localai_assistant_test.go index 3e539ce7d4bd..5778f992d655 100644 --- a/core/http/endpoints/mcp/localai_assistant_test.go +++ b/core/http/endpoints/mcp/localai_assistant_test.go @@ -75,6 +75,10 @@ func (stubClient) InstallBackend(_ context.Context, _ localaitools.InstallBacken return "stub-backend-job", nil } +func (stubClient) InstallLibraryBackend(_ context.Context, _ localaitools.InstallLibraryBackendRequest) (*localaitools.InstallLibraryBackendResult, error) { + return &localaitools.InstallLibraryBackendResult{Status: "installed", Name: "stub-backend", SourceID: "stub-source"}, nil +} + func (stubClient) UpgradeBackend(_ context.Context, _ string) (string, error) { return "stub-upgrade-job", nil }