Reworked the Hugging Face provider

app/MindWork AI Studio/Assistants/I18N/allTexts.lua

@@ -5944,6 +5944,9 @@ UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGMETHODDIALOG::T662524223"] = "A lin
 -- Cancel
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGMETHODDIALOG::T900713019"] = "Cancel"
 
+-- Hugging Face Inference Provider
+UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGPROVIDERDIALOG::T1085481431"] = "Hugging Face Inference Provider"
+
 -- Failed to store the API key in the operating system. The message was: {0}. Please try again.
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGPROVIDERDIALOG::T1122745046"] = "Failed to store the API key in the operating system. The message was: {0}. Please try again."
 
@@ -5968,6 +5971,9 @@ UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGPROVIDERDIALOG::T1847791252"] = "Up
 -- Failed to load the API key from the operating system. The message was: {0}. You might ignore this message and provide the API key again.
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGPROVIDERDIALOG::T1870831108"] = "Failed to load the API key from the operating system. The message was: {0}. You might ignore this message and provide the API key again."
 
+-- Hugging Face offers embeddings through a few of its inference providers only, which is why this list is shorter than the one for chatting.
+UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGPROVIDERDIALOG::T194295715"] = "Hugging Face offers embeddings through a few of its inference providers only, which is why this list is shorter than the one for chatting."
+
 -- Model
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::EMBEDDINGPROVIDERDIALOG::T2189814010"] = "Model"
 
@@ -6361,6 +6367,9 @@ UI_TEXT_CONTENT["AISTUDIO::DIALOGS::PROVIDERDIALOG::T1870831108"] = "Failed to l
 -- Speech input
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::PROVIDERDIALOG::T1874348907"] = "Speech input"
 
+-- Choose which inference provider should answer your requests. When you pick one of the automatic options instead, Hugging Face selects a provider for you and switches to another one when your choice is unavailable.
+UI_TEXT_CONTENT["AISTUDIO::DIALOGS::PROVIDERDIALOG::T1889879830"] = "Choose which inference provider should answer your requests. When you pick one of the automatic options instead, Hugging Face selects a provider for you and switches to another one when your choice is unavailable."
+
 -- Please enter a model name.
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::PROVIDERDIALOG::T1936099896"] = "Please enter a model name."
 
@@ -7987,6 +7996,9 @@ UI_TEXT_CONTENT["AISTUDIO::DIALOGS::SINGLEINPUTDIALOG::T4030229154"] = "Your Inp
 -- Cancel
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::SINGLEINPUTDIALOG::T900713019"] = "Cancel"
 
+-- Hugging Face Inference Provider
+UI_TEXT_CONTENT["AISTUDIO::DIALOGS::TRANSCRIPTIONPROVIDERDIALOG::T1085481431"] = "Hugging Face Inference Provider"
+
 -- Failed to store the API key in the operating system. The message was: {0}. Please try again.
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::TRANSCRIPTIONPROVIDERDIALOG::T1122745046"] = "Failed to store the API key in the operating system. The message was: {0}. Please try again."
 
@@ -8029,6 +8041,9 @@ UI_TEXT_CONTENT["AISTUDIO::DIALOGS::TRANSCRIPTIONPROVIDERDIALOG::T2810182573"] =
 -- Instance Name
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::TRANSCRIPTIONPROVIDERDIALOG::T2842060373"] = "Instance Name"
 
+-- Hugging Face transcribes audio through a few of its inference providers only, which is why this list is shorter than the one for chatting.
+UI_TEXT_CONTENT["AISTUDIO::DIALOGS::TRANSCRIPTIONPROVIDERDIALOG::T3397943774"] = "Hugging Face transcribes audio through a few of its inference providers only, which is why this list is shorter than the one for chatting."
+
 -- Please enter a transcription model name.
 UI_TEXT_CONTENT["AISTUDIO::DIALOGS::TRANSCRIPTIONPROVIDERDIALOG::T3703662664"] = "Please enter a transcription model name."
 
@@ -9274,6 +9289,21 @@ UI_TEXT_CONTENT["AISTUDIO::PROVIDER::CONFIDENCELEVELEXTENSIONS::T3188327965"] =
 -- Very Low
 UI_TEXT_CONTENT["AISTUDIO::PROVIDER::CONFIDENCELEVELEXTENSIONS::T786675843"] = "Very Low"
 
+-- Automatic: the cheapest provider
+UI_TEXT_CONTENT["AISTUDIO::PROVIDER::HUGGINGFACE::HFINFERENCEPROVIDEREXTENSIONS::T1680748563"] = "Automatic: the cheapest provider"
+
+-- Automatic: your preferred order
+UI_TEXT_CONTENT["AISTUDIO::PROVIDER::HUGGINGFACE::HFINFERENCEPROVIDEREXTENSIONS::T2027398472"] = "Automatic: your preferred order"
+
+-- Automatic: the fastest provider
+UI_TEXT_CONTENT["AISTUDIO::PROVIDER::HUGGINGFACE::HFINFERENCEPROVIDEREXTENSIONS::T997045984"] = "Automatic: the fastest provider"
+
+-- No Hugging Face inference provider offers the selected model. Please check the model name and whether it is still available on Hugging Face.
+UI_TEXT_CONTENT["AISTUDIO::PROVIDER::HUGGINGFACE::PROVIDERHUGGINGFACE::T1055093108"] = "No Hugging Face inference provider offers the selected model. Please check the model name and whether it is still available on Hugging Face."
+
+-- The Hugging Face inference provider '{0}' does not offer the selected model. Please select another inference provider, or let Hugging Face choose one for you.
+UI_TEXT_CONTENT["AISTUDIO::PROVIDER::HUGGINGFACE::PROVIDERHUGGINGFACE::T3314840969"] = "The Hugging Face inference provider '{0}' does not offer the selected model. Please select another inference provider, or let Hugging Face choose one for you."
+
 -- Self-hosted
 UI_TEXT_CONTENT["AISTUDIO::PROVIDER::LLMPROVIDERSEXTENSIONS::T146444217"] = "Self-hosted"
 
@@ -11197,9 +11227,15 @@ UI_TEXT_CONTENT["AISTUDIO::TOOLS::VALIDATION::PROVIDERVALIDATION::T3550629491"]
 -- Please enter an instance name.
 UI_TEXT_CONTENT["AISTUDIO::TOOLS::VALIDATION::PROVIDERVALIDATION::T3999823516"] = "Please enter an instance name."
 
+-- This Hugging Face inference provider does not transcribe audio. Please select another one.
+UI_TEXT_CONTENT["AISTUDIO::TOOLS::VALIDATION::PROVIDERVALIDATION::T4142849031"] = "This Hugging Face inference provider does not transcribe audio. Please select another one."
+
 -- Please select an Hugging Face inference provider.
 UI_TEXT_CONTENT["AISTUDIO::TOOLS::VALIDATION::PROVIDERVALIDATION::T497939286"] = "Please select an Hugging Face inference provider."
 
+-- This Hugging Face inference provider does not create embeddings. Please select another one.
+UI_TEXT_CONTENT["AISTUDIO::TOOLS::VALIDATION::PROVIDERVALIDATION::T649507886"] = "This Hugging Face inference provider does not create embeddings. Please select another one."
+
 -- Please select a model.
 UI_TEXT_CONTENT["AISTUDIO::TOOLS::VALIDATION::PROVIDERVALIDATION::T818893091"] = "Please select a model."
 

app/MindWork AI Studio/Components/Settings/SettingsPanelEmbeddings.razor.cs

@@ -88,6 +88,7 @@ private async Task EditEmbeddingProvider(EmbeddingProvider embeddingProvider)
             { x => x.IsSelfHosted, embeddingProvider.IsSelfHosted },
             { x => x.IsEditing, true },
             { x => x.DataHost, embeddingProvider.Host },
+            { x => x.HFInferenceProviderId, embeddingProvider.HFInferenceProvider },
             { x => x.IsEnterpriseConfiguration, embeddingProvider.IsEnterpriseConfiguration },
         };
 

app/MindWork AI Studio/Components/Settings/SettingsPanelTranscription.razor.cs

@@ -86,6 +86,7 @@ private async Task EditTranscriptionProvider(TranscriptionProvider transcription
             { x => x.IsSelfHosted, transcriptionProvider.IsSelfHosted },
             { x => x.IsEditing, true },
             { x => x.DataHost, transcriptionProvider.Host },
+            { x => x.HFInferenceProviderId, transcriptionProvider.HFInferenceProvider },
             { x => x.IsEnterpriseConfiguration, transcriptionProvider.IsEnterpriseConfiguration },
         };
 

app/MindWork AI Studio/Dialogs/EmbeddingProviderDialog.razor

@@ -1,4 +1,5 @@
 @using AIStudio.Provider
+@using AIStudio.Provider.HuggingFace
 @using AIStudio.Provider.SelfHosted
 @inherits MSGComponentBase
 
@@ -64,6 +65,24 @@
                 </MudSelect>
             }
 
+            @if (this.DataLLMProvider.IsHFInstanceProviderNeeded())
+            {
+                <MudSelect T="HFInferenceProvider" Value="@this.HFInferenceProviderId" ValueChanged="@this.OnHFInferenceProviderChanged" Label="@T("Hugging Face Inference Provider")" Class="mb-3" OpenIcon="@Icons.Material.Filled.Dns" AdornmentColor="Color.Info" Adornment="Adornment.Start" Disabled="@this.IsEnterpriseConfiguration" Validation="@this.providerValidation.ValidatingHFInstanceProviderForEmbeddings">
+                    @foreach (HFInferenceProvider inferenceProvider in Enum.GetValues(typeof(HFInferenceProvider)))
+                    {
+                        @if (inferenceProvider.SupportsEmbeddings())
+                        {
+                            <MudSelectItem Value="@inferenceProvider">
+                                @inferenceProvider.ToName()
+                            </MudSelectItem>
+                        }
+                    }
+                </MudSelect>
+                <MudJustifiedText Class="mb-3">
+                    @T("Hugging Face offers embeddings through a few of its inference providers only, which is why this list is shorter than the one for chatting.")
+                </MudJustifiedText>
+            }
+
             <MudField FullWidth="true" Label="@T("Model selection")" Variant="Variant.Outlined" Class="mb-3">
                 <MudStack Row="@true" AlignItems="AlignItems.Center" StretchItems="StretchItems.End">
                     @if (this.DataLLMProvider.IsEmbeddingModelProvidedManually(this.DataHost))

app/MindWork AI Studio/Dialogs/EmbeddingProviderDialog.razor.cs

@@ -1,5 +1,6 @@
 using AIStudio.Components;
 using AIStudio.Provider;
+using AIStudio.Provider.HuggingFace;
 using AIStudio.Settings;
 using AIStudio.Tools.Services;
 using AIStudio.Tools.Validation;
@@ -68,6 +69,12 @@ public partial class EmbeddingProviderDialog : MSGComponentBase, ISecretId
     /// </summary>
     [Parameter]
     public Model DataModel { get; set; }
+
+    /// <summary>
+    /// The Hugging Face inference provider to use.
+    /// </summary>
+    [Parameter]
+    public HFInferenceProvider HFInferenceProviderId { get; set; } = HFInferenceProvider.NONE;
 
     /// <summary>
     /// Should the dialog be in editing mode?
@@ -151,6 +158,7 @@ private EmbeddingProvider CreateEmbeddingProviderSettings()
             IsEnterpriseConfiguration = this.IsEnterpriseConfiguration,
             EnterpriseConfigurationPluginId = Guid.Empty,
             CustomIconDataUrl = this.DataCustomIconDataUrl,
+            HFInferenceProvider = this.HFInferenceProviderId,
         };
     }
 
@@ -316,6 +324,22 @@ private void OnHostChanged(Host selectedHost)
         this.dataLoadingModelsIssue = string.Empty;
     }
 
+    /// <summary>
+    /// Resets the model selection when the user picks another Hugging Face inference provider.
+    /// </summary>
+    /// <remarks>
+    /// Each inference provider offers embedding models of its own, so the models loaded for the
+    /// previous one say nothing about the new one.
+    /// </remarks>
+    /// <param name="selectedInferenceProvider">The inference provider the user chose.</param>
+    private void OnHFInferenceProviderChanged(HFInferenceProvider selectedInferenceProvider)
+    {
+        this.HFInferenceProviderId = selectedInferenceProvider;
+        this.DataModel = default;
+        this.availableModels.Clear();
+        this.dataLoadingModelsIssue = string.Empty;
+    }
+
     private async Task ReloadModels()
     {
         this.dataLoadingModelsIssue = string.Empty;

app/MindWork AI Studio/Dialogs/ProviderDialog.razor

@@ -69,19 +69,20 @@
 
             @if (this.DataLLMProvider.IsHFInstanceProviderNeeded())
             {
-                <MudSelect @bind-Value="@this.HFInferenceProviderId" Label="@T("Hugging Face Inference Provider")" Class="mb-3" OpenIcon="@Icons.Material.Filled.Dns" AdornmentColor="Color.Info" Adornment="Adornment.Start" Disabled="@this.IsEnterpriseConfiguration" Validation="@this.providerValidation.ValidatingHFInstanceProvider">
+                <MudSelect T="HFInferenceProvider" Value="@this.HFInferenceProviderId" ValueChanged="@this.OnHFInferenceProviderChanged" Label="@T("Hugging Face Inference Provider")" Class="mb-3" OpenIcon="@Icons.Material.Filled.Dns" AdornmentColor="Color.Info" Adornment="Adornment.Start" Disabled="@this.IsEnterpriseConfiguration" Validation="@this.providerValidation.ValidatingHFInstanceProvider">
                     @foreach (HFInferenceProvider inferenceProvider in Enum.GetValues(typeof(HFInferenceProvider)))
                     {
-                        <MudSelectItem Value="@inferenceProvider">
-                            @inferenceProvider.ToName()
-                        </MudSelectItem>
-                    }             
+                        @if (inferenceProvider.SupportsChat())
+                        {
+                            <MudSelectItem Value="@inferenceProvider">
+                                @inferenceProvider.ToName()
+                            </MudSelectItem>
+                        }
+                    }
                 </MudSelect>
-                @* ReSharper disable Asp.Entity *@
                 <MudJustifiedText Class="mb-3">
-                    Please double-check if your model name matches the curl specifications provided by the inference provider. If it doesn't, you might get a <b>Not Found</b> error when trying to use the model. Here's a <MudLink Href="https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct?inference_api=true&inference_provider=novita&language=sh" Target="_blank">curl example</MudLink>.
+                    @T("Choose which inference provider should answer your requests. When you pick one of the automatic options instead, Hugging Face selects a provider for you and switches to another one when your choice is unavailable.")
                 </MudJustifiedText>
-                @* ReSharper restore Asp.Entity *@
             }       
 
             @if (!this.IsLLMModelSelectionHidden)
@@ -90,7 +91,7 @@
                     <MudStack Row="@true" AlignItems="AlignItems.Center" StretchItems="StretchItems.End">
                         @if (this.DataLLMProvider.IsLLMModelProvidedManually())
                         {
-                            <MudButton Variant="Variant.Filled" Size="Size.Small" StartIcon="@Icons.Material.Filled.OpenInBrowser" Href="@this.DataLLMProvider.GetModelsOverviewURL(this.HFInferenceProviderId)" Target="_blank">
+                            <MudButton Variant="Variant.Filled" Size="Size.Small" StartIcon="@Icons.Material.Filled.OpenInBrowser" Href="@this.ModelsOverviewURL" Target="_blank">
                                 @T("Show available models")
                             </MudButton>
                             <MudTextField
@@ -108,6 +109,12 @@
                         }
                         else
                         {
+                            @if (!string.IsNullOrWhiteSpace(this.ModelsOverviewURL))
+                            {
+                                <MudButton Variant="Variant.Filled" Size="Size.Small" StartIcon="@Icons.Material.Filled.OpenInBrowser" Href="@this.ModelsOverviewURL" Target="_blank">
+                                    @T("Show available models")
+                                </MudButton>
+                            }
                             <MudButton Disabled="@(this.IsEnterpriseConfiguration || !this.DataLLMProvider.CanLoadModels(this.DataHost, this.dataAPIKey))" Variant="Variant.Filled" Size="Size.Small" StartIcon="@Icons.Material.Filled.Refresh" OnClick="@this.ReloadModels">
                                 @T("Load models")
                             </MudButton>

app/MindWork AI Studio/Dialogs/ProviderDialog.razor.cs

@@ -226,7 +226,7 @@ protected override async Task OnInitializedAsync()
         {
             this.dataEditingPreviousInstanceName = this.DataInstanceName.ToLowerInvariant();
 
-            // When using Fireworks or Hugging Face, we must copy the model name:
+            // When using Fireworks, we must copy the model name:
             if (this.DataLLMProvider.IsLLMModelProvidedManually())
                 this.dataManuallyModel = this.DataModel.Id;
 
@@ -372,6 +372,24 @@ private void OnProviderChanged(LLMProviders selectedProvider)
         this.usesLegacySystemModelFallback = false;
     }
 
+    /// <summary>
+    /// Resets the model selection when the user picks another Hugging Face inference provider.
+    /// </summary>
+    /// <remarks>
+    /// Which models are on offer depends on the inference provider, so the models loaded for the
+    /// previous one say nothing about the new one. Keeping them would let the user pick a model
+    /// their provider does not serve, which the router answers with an error.
+    /// </remarks>
+    /// <param name="selectedInferenceProvider">The inference provider the user chose.</param>
+    private void OnHFInferenceProviderChanged(HFInferenceProvider selectedInferenceProvider)
+    {
+        this.HFInferenceProviderId = selectedInferenceProvider;
+        this.DataModel = default;
+        this.capabilityOverrides = new();
+        this.availableModels.Clear();
+        this.dataLoadingModelsIssue = string.Empty;
+    }
+
     private void OnHostChanged(Host selectedHost)
     {
         // When the host changes, reset the model selection state:
@@ -430,6 +448,11 @@ this.DataLLMProvider is LLMProviders.SELF_HOSTED &&
                                              this.DataHost is Host.LLAMA_CPP &&
                                              this.usesLegacySystemModelFallback;
 
+    /// <summary>
+    /// The catalog of the provider, where the user can read up on the models before choosing one.
+    /// </summary>
+    private string ModelsOverviewURL => this.DataLLMProvider.GetModelsOverviewURL(this.HFInferenceProviderId);
+
     private void UpdateModelSelectionAfterLoading()
     {
         if (this.DataLLMProvider is not LLMProviders.SELF_HOSTED || this.DataHost is not Host.LLAMA_CPP)

app/MindWork AI Studio/Dialogs/TranscriptionProviderDialog.razor

@@ -1,4 +1,5 @@
 @using AIStudio.Provider
+@using AIStudio.Provider.HuggingFace
 @using AIStudio.Provider.SelfHosted
 @inherits MSGComponentBase
 
@@ -64,6 +65,24 @@
                 </MudSelect>
             }
 
+            @if (this.DataLLMProvider.IsHFInstanceProviderNeeded())
+            {
+                <MudSelect T="HFInferenceProvider" Value="@this.HFInferenceProviderId" ValueChanged="@this.OnHFInferenceProviderChanged" Label="@T("Hugging Face Inference Provider")" Class="mb-3" OpenIcon="@Icons.Material.Filled.Dns" AdornmentColor="Color.Info" Adornment="Adornment.Start" Disabled="@this.IsEnterpriseConfiguration" Validation="@this.providerValidation.ValidatingHFInstanceProviderForTranscription">
+                    @foreach (HFInferenceProvider inferenceProvider in Enum.GetValues(typeof(HFInferenceProvider)))
+                    {
+                        @if (inferenceProvider.SupportsTranscription())
+                        {
+                            <MudSelectItem Value="@inferenceProvider">
+                                @inferenceProvider.ToName()
+                            </MudSelectItem>
+                        }
+                    }
+                </MudSelect>
+                <MudJustifiedText Class="mb-3">
+                    @T("Hugging Face transcribes audio through a few of its inference providers only, which is why this list is shorter than the one for chatting.")
+                </MudJustifiedText>
+            }
+
             @if (!this.DataLLMProvider.IsTranscriptionModelSelectionHidden(this.DataHost))
             {
                 <MudField FullWidth="true" Label="@T("Model selection")" Variant="Variant.Outlined" Class="mb-3">