llama-cpp-turboquant/tools/server/webui/docs/flows/models-flow.md
Xuan-Son Nguyen ec18edfcba
server: introduce API for serving / loading / unloading multiple models (#17470)
* server: add model management and proxy

* fix compile error

* does this fix windows?

* fix windows build

* use subprocess.h, better logging

* add test

* fix windows

* feat: Model/Router server architecture WIP

* more stable

* fix unsafe pointer

* also allow terminate loading model

* add is_active()

* refactor: Architecture improvements

* tmp apply upstream fix

* address most problems

* address thread safety issue

* address review comment

* add docs (first version)

* address review comment

* feat: Improved UX for model information, modality interactions etc

* chore: update webui build output

* refactor: Use only the message data `model` property for displaying model used info

* chore: update webui build output

* add --models-dir param

* feat: New Model Selection UX WIP

* chore: update webui build output

* feat: Add auto-mic setting

* feat: Attachments UX improvements

* implement LRU

* remove default model path

* better --models-dir

* add env for args

* address review comments

* fix compile

* refactor: Chat Form Submit component

* ad endpoint docs

* Merge remote-tracking branch 'webui/allozaur/server_model_management_v1_2' into xsn/server_model_maagement_v1_2

Co-authored-by: Aleksander <aleksander.grygier@gmail.com>

* feat: Add copy to clipboard to model name in model info dialog

* feat: Model unavailable UI state for model selector

* feat: Chat Form Actions UI logic improvements

* feat: Auto-select model from last assistant response

* chore: update webui build output

* expose args and exit_code in API

* add note

* support extra_args on loading model

* allow reusing args if auto_load

* typo docs

* oai-compat /models endpoint

* cleaner

* address review comments

* feat: Use `model` property for displaying the `repo/model-name` naming format

* refactor: Attachments data

* chore: update webui build output

* refactor: Enum imports

* feat: Improve Model Selector responsiveness

* chore: update webui build output

* refactor: Cleanup

* refactor: Cleanup

* refactor: Formatters

* chore: update webui build output

* refactor: Copy To Clipboard Icon component

* chore: update webui build output

* refactor: Cleanup

* chore: update webui build output

* refactor: UI badges

* chore: update webui build output

* refactor: Cleanup

* refactor: Cleanup

* chore: update webui build output

* add --models-allow-extra-args for security

* nits

* add stdin_file

* fix merge

* fix: Retrieve lost setting after resolving merge conflict

* refactor: DatabaseStore -> DatabaseService

* refactor: Database, Conversations & Chat services + stores architecture improvements (WIP)

* refactor: Remove redundant settings

* refactor: Multi-model business logic WIP

* chore: update webui build output

* feat: Switching models logic for ChatForm or when regenerating messges + modality detection logic

* chore: update webui build output

* fix: Add `untrack` inside chat processing info data logic to prevent infinite effect

* fix: Regenerate

* feat: Remove redundant settigns + rearrange

* fix: Audio attachments

* refactor: Icons

* chore: update webui build output

* feat: Model management and selection features WIP

* chore: update webui build output

* refactor: Improve server properties management

* refactor: Icons

* chore: update webui build output

* feat: Improve model loading/unloading status updates

* chore: update webui build output

* refactor: Improve API header management via utility functions

* remove support for extra args

* set hf_repo/docker_repo as model alias when posible

* refactor: Remove ConversationsService

* refactor: Chat requests abort handling

* refactor: Server store

* tmp webui build

* refactor: Model modality handling

* chore: update webui build output

* refactor: Processing state reactivity

* fix: UI

* refactor: Services/Stores syntax + logic improvements

Refactors components to access stores directly instead of using exported getter functions.

This change centralizes store access and logic, simplifying component code and improving maintainability by reducing the number of exported functions and promoting direct store interaction.

Removes exported getter functions from `chat.svelte.ts`, `conversations.svelte.ts`, `models.svelte.ts` and `settings.svelte.ts`.

* refactor: Architecture cleanup

* feat: Improve statistic badges

* feat: Condition available models based on modality + better model loading strategy & UX

* docs: Architecture documentation

* feat: Update logic for PDF as Image

* add TODO for http client

* refactor: Enhance model info and attachment handling

* chore: update webui build output

* refactor: Components naming

* chore: update webui build output

* refactor: Cleanup

* refactor: DRY `getAttachmentDisplayItems` function + fix UI

* chore: update webui build output

* fix: Modality detection improvement for text-based PDF attachments

* refactor: Cleanup

* docs: Add info comment

* refactor: Cleanup

* re

* refactor: Cleanup

* refactor: Cleanup

* feat: Attachment logic & UI improvements

* refactor: Constants

* feat: Improve UI sidebar background color

* chore: update webui build output

* refactor: Utils imports + move types to `app.d.ts`

* test: Fix Storybook mocks

* chore: update webui build output

* test: Update Chat Form UI tests

* refactor: Tooltip Provider from core layout

* refactor: Tests to separate location

* decouple server_models from server_routes

* test: Move demo test  to tests/server

* refactor: Remove redundant method

* chore: update webui build output

* also route anthropic endpoints

* fix duplicated arg

* fix invalid ptr to shutdown_handler

* server : minor

* rm unused fn

* add ?autoload=true|false query param

* refactor: Remove redundant code

* docs: Update README documentations + architecture & data flow diagrams

* fix: Disable autoload on calling server props for the model

* chore: update webui build output

* fix ubuntu build

* fix: Model status reactivity

* fix: Modality detection for MODEL mode

* chore: update webui build output

---------

Co-authored-by: Aleksander Grygier <aleksander.grygier@gmail.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2025-12-01 19:41:04 +01:00

9.3 KiB

sequenceDiagram
    participant UI as 🧩 ModelsSelector
    participant Hooks as 🪝 useModelChangeValidation
    participant modelsStore as 🗄️ modelsStore
    participant serverStore as 🗄️ serverStore
    participant convStore as 🗄️ conversationsStore
    participant ModelsSvc as ⚙️ ModelsService
    participant PropsSvc as ⚙️ PropsService
    participant API as 🌐 llama-server

    Note over modelsStore: State:<br/>models: ModelOption[]<br/>routerModels: ApiModelDataEntry[]<br/>selectedModelId, selectedModelName<br/>loading, updating, error<br/>modelLoadingStates (Map)<br/>modelPropsCache (Map)<br/>propsCacheVersion

    %% ═══════════════════════════════════════════════════════════════════════════
    Note over UI,API: 🚀 INITIALIZATION (MODEL mode)
    %% ═══════════════════════════════════════════════════════════════════════════

    UI->>modelsStore: fetch()
    activate modelsStore
    modelsStore->>modelsStore: loading = true

    alt serverStore.props not loaded
        modelsStore->>serverStore: fetch()
        Note over serverStore: → see server-flow.mmd
    end

    modelsStore->>ModelsSvc: list()
    ModelsSvc->>API: GET /v1/models
    API-->>ModelsSvc: ApiModelListResponse {data: [model]}

    modelsStore->>modelsStore: models = $state(mapped)
    Note right of modelsStore: Map to ModelOption[]:<br/>{id, name, model, description, capabilities}

    Note over modelsStore: MODEL mode: Get modalities from serverStore.props
    modelsStore->>modelsStore: modelPropsCache.set(model.id, serverStore.props)
    modelsStore->>modelsStore: models[0].modalities = props.modalities

    modelsStore->>modelsStore: Auto-select single model
    Note right of modelsStore: selectedModelId = models[0].id
    modelsStore->>modelsStore: loading = false
    deactivate modelsStore

    %% ═══════════════════════════════════════════════════════════════════════════
    Note over UI,API: 🚀 INITIALIZATION (ROUTER mode)
    %% ═══════════════════════════════════════════════════════════════════════════

    UI->>modelsStore: fetch()
    activate modelsStore
    modelsStore->>ModelsSvc: list()
    ModelsSvc->>API: GET /v1/models
    API-->>ModelsSvc: ApiModelListResponse
    modelsStore->>modelsStore: models = $state(mapped)
    deactivate modelsStore

    Note over UI: After models loaded, layout triggers:
    UI->>modelsStore: fetchRouterModels()
    activate modelsStore
    modelsStore->>ModelsSvc: listRouter()
    ModelsSvc->>API: GET /models
    API-->>ModelsSvc: ApiRouterModelsListResponse
    Note right of API: {data: [{id, status, path, in_cache}]}
    modelsStore->>modelsStore: routerModels = $state(data)

    modelsStore->>modelsStore: fetchModalitiesForLoadedModels()
    loop each model where status === "loaded"
        modelsStore->>PropsSvc: fetchForModel(modelId)
        PropsSvc->>API: GET /props?model={modelId}
        API-->>PropsSvc: ApiLlamaCppServerProps
        modelsStore->>modelsStore: modelPropsCache.set(modelId, props)
    end
    modelsStore->>modelsStore: propsCacheVersion++
    deactivate modelsStore

    %% ═══════════════════════════════════════════════════════════════════════════
    Note over UI,API: 🔄 MODEL SELECTION (ROUTER mode)
    %% ═══════════════════════════════════════════════════════════════════════════

    UI->>Hooks: useModelChangeValidation({getRequiredModalities, onSuccess?, onValidationFailure?})
    Note over Hooks: Hook configured per-component:<br/>ChatForm: getRequiredModalities = usedModalities<br/>ChatMessage: getRequiredModalities = getModalitiesUpToMessage(msgId)

    UI->>Hooks: handleModelChange(modelId, modelName)
    activate Hooks
    Hooks->>Hooks: previousSelectedModelId = modelsStore.selectedModelId
    Hooks->>modelsStore: isModelLoaded(modelName)?

    alt model NOT loaded
        Hooks->>modelsStore: loadModel(modelName)
        Note over modelsStore: → see LOAD MODEL section below
    end

    Note over Hooks: Always fetch props (from cache or API)
    Hooks->>modelsStore: fetchModelProps(modelName)
    modelsStore-->>Hooks: props

    Hooks->>convStore: getRequiredModalities()
    convStore-->>Hooks: {vision, audio}

    Hooks->>Hooks: Validate: model.modalities ⊇ required?

    alt validation PASSED
        Hooks->>modelsStore: selectModelById(modelId)
        Hooks-->>UI: return true
    else validation FAILED
        Hooks->>UI: toast.error("Model doesn't support required modalities")
        alt model was just loaded
            Hooks->>modelsStore: unloadModel(modelName)
        end
        alt onValidationFailure provided
            Hooks->>modelsStore: selectModelById(previousSelectedModelId)
        end
        Hooks-->>UI: return false
    end
    deactivate Hooks

    %% ═══════════════════════════════════════════════════════════════════════════
    Note over UI,API: ⬆️ LOAD MODEL (ROUTER mode)
    %% ═══════════════════════════════════════════════════════════════════════════

    modelsStore->>modelsStore: loadModel(modelId)
    activate modelsStore

    alt already loaded
        modelsStore-->>modelsStore: return (no-op)
    end

    modelsStore->>modelsStore: modelLoadingStates.set(modelId, true)
    modelsStore->>ModelsSvc: load(modelId)
    ModelsSvc->>API: POST /models/load {model: modelId}
    API-->>ModelsSvc: {status: "loading"}

    modelsStore->>modelsStore: pollForModelStatus(modelId, LOADED)
    loop poll every 500ms (max 60 attempts)
        modelsStore->>modelsStore: fetchRouterModels()
        modelsStore->>ModelsSvc: listRouter()
        ModelsSvc->>API: GET /models
        API-->>ModelsSvc: models[]
        modelsStore->>modelsStore: getModelStatus(modelId)
        alt status === LOADED
            Note right of modelsStore: break loop
        else status === LOADING
            Note right of modelsStore: wait 500ms, continue
        end
    end

    modelsStore->>modelsStore: updateModelModalities(modelId)
    modelsStore->>PropsSvc: fetchForModel(modelId)
    PropsSvc->>API: GET /props?model={modelId}
    API-->>PropsSvc: props with modalities
    modelsStore->>modelsStore: modelPropsCache.set(modelId, props)
    modelsStore->>modelsStore: propsCacheVersion++

    modelsStore->>modelsStore: modelLoadingStates.set(modelId, false)
    deactivate modelsStore

    %% ═══════════════════════════════════════════════════════════════════════════
    Note over UI,API: ⬇️ UNLOAD MODEL (ROUTER mode)
    %% ═══════════════════════════════════════════════════════════════════════════

    modelsStore->>modelsStore: unloadModel(modelId)
    activate modelsStore
    modelsStore->>modelsStore: modelLoadingStates.set(modelId, true)
    modelsStore->>ModelsSvc: unload(modelId)
    ModelsSvc->>API: POST /models/unload {model: modelId}

    modelsStore->>modelsStore: pollForModelStatus(modelId, UNLOADED)
    loop poll until unloaded
        modelsStore->>ModelsSvc: listRouter()
        ModelsSvc->>API: GET /models
    end

    modelsStore->>modelsStore: modelLoadingStates.set(modelId, false)
    deactivate modelsStore

    %% ═══════════════════════════════════════════════════════════════════════════
    Note over UI,API: 📊 COMPUTED GETTERS
    %% ═══════════════════════════════════════════════════════════════════════════

    Note over modelsStore: Getters:<br/>- selectedModel: ModelOption | null<br/>- loadedModelIds: string[] (from routerModels)<br/>- loadingModelIds: string[] (from modelLoadingStates)<br/>- singleModelName: string | null (MODEL mode only)

    Note over modelsStore: Modality helpers:<br/>- getModelModalities(modelId): {vision, audio}<br/>- modelSupportsVision(modelId): boolean<br/>- modelSupportsAudio(modelId): boolean