Added possibility to authenticate HF Downloader for model download + …

…retrieve token from secrets in github actions
sandrohanea · Jan 12, 2025 · 3f2443f · 3f2443f
1 parent 34d8ba1
commit 3f2443f
Show file tree

Hide file tree

Showing 20 changed files with 70 additions and 42 deletions.
diff --git a/.github/workflows/build-all.yml b/.github/workflows/build-all.yml
@@ -76,6 +76,7 @@ jobs:
          - wasm
          - linux
       uses: ./.github/workflows/dotnet.yml
+      secrets: inherit
 
    dotnet-maui-build-and-test:
       needs:
@@ -85,3 +86,4 @@ jobs:
          - wasm
          - linux
       uses: ./.github/workflows/dotnet-maui.yml
+      secrets: inherit
diff --git a/.github/workflows/dotnet.yml b/.github/workflows/dotnet.yml
@@ -10,6 +10,9 @@ permissions:
   contents: read
   checks: write
 
+env:
+  HF_TOKEN: ${{ secrets.HF_TOKEN }}
+
 jobs:
   dotnet-macos:
     runs-on: macos-15

diff --git a/Whisper.net.Demo/Program.cs b/Whisper.net.Demo/Program.cs
@@ -13,7 +13,7 @@ async Task Demo(Options opt)
     if (!File.Exists(opt.ModelName))
     {
         Console.WriteLine($"Downloading Model {opt.ModelName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(opt.ModelType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(opt.ModelType);
         using var fileWriter = File.OpenWrite(opt.ModelName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/Whisper.net/Ggml/WhisperGgmlDownloader.cs b/Whisper.net/Ggml/WhisperGgmlDownloader.cs
@@ -1,12 +1,15 @@
 // Licensed under the MIT license: https://opensource.org/licenses/MIT
 
-using System.IO.Compression;
-
 namespace Whisper.net.Ggml;
 
-public static class WhisperGgmlDownloader
+public class WhisperGgmlDownloader(HttpClient httpClient)
 {
-    private static readonly Lazy<HttpClient> httpClient = new(() => new HttpClient() { Timeout = Timeout.InfiniteTimeSpan });
+    private static readonly Lazy<WhisperGgmlDownloader> defaultInstance = new
+        (
+            () => new WhisperGgmlDownloader(new() { Timeout = TimeSpan.FromHours(1) })
+        );
+
+    public static WhisperGgmlDownloader Default { get; } = defaultInstance.Value;
 
     /// <summary>
     /// Gets the download stream for the model
@@ -15,15 +18,15 @@ public static class WhisperGgmlDownloader
     /// <param name="quantization">The quantization of the model.</param>
     /// <param name="cancellationToken">A cancellation token used to cancell the request to huggingface.</param>
     /// <exception cref="ArgumentOutOfRangeException"></exception>
-    public static async Task<Stream> GetGgmlModelAsync(GgmlType type, QuantizationType quantization = QuantizationType.NoQuantization, CancellationToken cancellationToken = default)
+    public async Task<Stream> GetGgmlModelAsync(GgmlType type, QuantizationType quantization = QuantizationType.NoQuantization, CancellationToken cancellationToken = default)
     {
         var subdirectory = GetQuantizationSubdirectory(quantization);
         var modelName = GetModelName(type);
 
         var url = $"https://huggingface.co/sandrohanea/whisper.net/resolve/v3/{subdirectory}/{modelName}.bin";
 
         var request = new HttpRequestMessage(HttpMethod.Get, url);
-        var response = await httpClient.Value.SendAsync(request, HttpCompletionOption.ResponseHeadersRead, cancellationToken);
+        var response = await httpClient.SendAsync(request, HttpCompletionOption.ResponseHeadersRead, cancellationToken);
         response.EnsureSuccessStatusCode();
 
 #if NETSTANDARD
@@ -39,12 +42,12 @@ public static async Task<Stream> GetGgmlModelAsync(GgmlType type, QuantizationTy
     /// <param name="type">The type of the model which needs to be downloaded.</param>
     /// <param name="cancellationToken">A cancellation token used to stop the request to huggingface.</param>
     /// <returns></returns>
-    public static async Task<Stream> GetEncoderOpenVinoModelAsync(GgmlType type, CancellationToken cancellationToken = default)
+    public async Task<Stream> GetEncoderOpenVinoModelAsync(GgmlType type, CancellationToken cancellationToken = default)
     {
         var modelName = GetModelName(type);
         var url = $"https://huggingface.co/sandrohanea/whisper.net/resolve/v3/openvino/{modelName}-encoder.zip";
         var request = new HttpRequestMessage(HttpMethod.Get, url);
-        var response = await httpClient.Value.SendAsync(request, HttpCompletionOption.ResponseHeadersRead, cancellationToken);
+        var response = await httpClient.SendAsync(request, HttpCompletionOption.ResponseHeadersRead, cancellationToken);
         response.EnsureSuccessStatusCode();
 #if NETSTANDARD
         return await response.Content.ReadAsStreamAsync();
@@ -58,7 +61,7 @@ public static async Task<Stream> GetEncoderOpenVinoModelAsync(GgmlType type, Can
     /// </summary>
     /// <param name="type"> The type of the model which needs to be loaded</param>
     /// <returns></returns>
-    public static string GetOpenVinoManifestFileName(GgmlType type)
+    public string GetOpenVinoManifestFileName(GgmlType type)
     {
         var modelName = GetModelName(type);
         return $"{modelName}-encoder.xml";
@@ -73,13 +76,13 @@ public static string GetOpenVinoManifestFileName(GgmlType type)
     /// Needs to be extracted on in the same directory as the ggml model, also ggml model needs to be loaded using file path, not stream.
     /// </remarks>
     /// <exception cref="ArgumentOutOfRangeException"></exception>
-    public static async Task<Stream> GetEncoderCoreMLModelAsync(GgmlType type, CancellationToken cancellationToken = default)
+    public async Task<Stream> GetEncoderCoreMLModelAsync(GgmlType type, CancellationToken cancellationToken = default)
     {
         var modelName = GetModelName(type);
         var url = $"https://huggingface.co/sandrohanea/whisper.net/resolve/v3/coreml/{modelName}-encoder.zip";
 
         var request = new HttpRequestMessage(HttpMethod.Get, url);
-        var response = await httpClient.Value.SendAsync(request, HttpCompletionOption.ResponseHeadersRead, cancellationToken);
+        var response = await httpClient.SendAsync(request, HttpCompletionOption.ResponseHeadersRead, cancellationToken);
         response.EnsureSuccessStatusCode();
 
 #if NETSTANDARD
@@ -89,21 +92,6 @@ public static async Task<Stream> GetEncoderCoreMLModelAsync(GgmlType type, Cance
 #endif
     }
 
-    /// <summary>
-    /// Extracts the given zip stream to the given path.
-    /// </summary>
-    /// <param name="zipStream">The zip stream to be extracted.</param>
-    /// <param name="path">The path.</param>
-    /// <remarks>
-    /// In order to work, you'll need to provide the same path as the ggml model.
-    /// </remarks>
-    /// <returns></returns>
-    public static async Task ExtractToPath(this Task<Stream> zipStream, string path)
-    {
-        using var zipArchive = new ZipArchive(await zipStream, ZipArchiveMode.Read);
-        zipArchive.ExtractToDirectory(path);
-    }
-
     private static string GetModelName(GgmlType type)
     {
         return type switch

diff --git a/Whisper.net/Ggml/ZipStreamExtensions.cs b/Whisper.net/Ggml/ZipStreamExtensions.cs
@@ -0,0 +1,23 @@
+// Licensed under the MIT license: https://opensource.org/licenses/MIT
+
+using System.IO.Compression;
+
+namespace Whisper.net.Ggml;
+
+public static class ZipStreamExtensions
+{
+    /// <summary>
+    /// Extracts the given zip stream to the given path.
+    /// </summary>
+    /// <param name="zipStream">The zip stream to be extracted.</param>
+    /// <param name="path">The path.</param>
+    /// <remarks>
+    /// In order to work, you'll need to provide the same path as the ggml model.
+    /// </remarks>
+    /// <returns></returns>
+    public static async Task ExtractToPath(this Task<Stream> zipStream, string path)
+    {
+        using var zipArchive = new ZipArchive(await zipStream, ZipArchiveMode.Read);
+        zipArchive.ExtractToDirectory(path);
+    }
+}
diff --git a/examples/BlazorApp/BlazorApp.Client/Pages/WhisperComponent.razor b/examples/BlazorApp/BlazorApp.Client/Pages/WhisperComponent.razor
@@ -38,7 +38,7 @@ else
         if (whisperFactory == null)
         {
             using var memoryStream = new MemoryStream();
-            var model = await WhisperGgmlDownloader.GetGgmlModelAsync(GgmlType.Tiny);
+            var model = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(GgmlType.Tiny);
             await model.CopyToAsync(memoryStream);
             whisperFactory = WhisperFactory.FromBuffer(memoryStream.ToArray());
         }

diff --git a/examples/CoreML/Program.cs b/examples/CoreML/Program.cs
@@ -59,7 +59,7 @@ await WhisperGgmlDownloader.GetEncoderCoreMLModelAsync(ggmlType)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/Diarization/Program.cs b/examples/Diarization/Program.cs
@@ -89,7 +89,7 @@
 static async Task DownloadModel(string fileName, GgmlType ggmlType)
 {
     Console.WriteLine($"Downloading Model {fileName}");
-    using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+    using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
     using var fileWriter = File.OpenWrite(fileName);
     await modelStream.CopyToAsync(fileWriter);
 }
diff --git a/examples/MultiRuntime/Program.cs b/examples/MultiRuntime/Program.cs
@@ -53,7 +53,7 @@ public static async Task Main(string[] args)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/NAudioMp3/Program.cs b/examples/NAudioMp3/Program.cs
@@ -54,7 +54,7 @@ public static async Task Main(string[] args)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/NAudioResampleWav/Program.cs b/examples/NAudioResampleWav/Program.cs
@@ -54,7 +54,7 @@ public static async Task Main(string[] args)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/NvidiaCuda/Program.cs b/examples/NvidiaCuda/Program.cs
@@ -45,7 +45,7 @@ public static async Task Main(string[] args)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/OpenVinoExample/Program.cs b/examples/OpenVinoExample/Program.cs
@@ -59,7 +59,7 @@ await WhisperGgmlDownloader.GetEncoderOpenVinoModelAsync(ggmlType)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/ParallelExecution/Program.cs b/examples/ParallelExecution/Program.cs
@@ -56,7 +56,7 @@ public static async Task RunInParallel(string name, string wavFileName, WhisperF
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/Simple/Program.cs b/examples/Simple/Program.cs
@@ -46,7 +46,7 @@ public static async Task Main(string[] args)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/SimpleSync/Program.cs b/examples/SimpleSync/Program.cs
@@ -48,7 +48,7 @@ public static async Task Main(string[] args)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/examples/Vulkan/Program.cs b/examples/Vulkan/Program.cs
@@ -46,7 +46,7 @@ public static async Task Main(string[] args)
     private static async Task DownloadModel(string fileName, GgmlType ggmlType)
     {
         Console.WriteLine($"Downloading Model {fileName}");
-        using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(ggmlType);
+        using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(ggmlType);
         using var fileWriter = File.OpenWrite(fileName);
         await modelStream.CopyToAsync(fileWriter);
     }

diff --git a/readme.md b/readme.md
@@ -203,7 +203,7 @@ For easier integration, Whisper.net provides a Downloader using [Hugging Face](h
 var modelName = "ggml-base.bin";
 if (!File.Exists(modelName))
 {
-    using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(GgmlType.Base);
+    using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(GgmlType.Base);
     using var fileWriter = File.OpenWrite(modelName);
     await modelStream.CopyToAsync(fileWriter);
 }

diff --git a/tests/Whisper.net.Maui.Tests/MainPage.xaml.cs b/tests/Whisper.net.Maui.Tests/MainPage.xaml.cs
@@ -16,7 +16,7 @@ private async void ContentPage_Loaded(object sender, EventArgs e)
         try
         {
             using var memoryStream = new MemoryStream();
-            using var modelStream = await WhisperGgmlDownloader.GetGgmlModelAsync(GgmlType.Tiny, QuantizationType.Q4_0);
+            using var modelStream = await WhisperGgmlDownloader.Default.GetGgmlModelAsync(GgmlType.Tiny, QuantizationType.Q4_0);
             await modelStream.CopyToAsync(memoryStream);
             using var mauiStream = await FileSystem.OpenAppPackageFileAsync("kennedy.wav");
             var audioFileStream = new MemoryStream();

diff --git a/tests/Whisper.net.Tests/ModelFixture.cs b/tests/Whisper.net.Tests/ModelFixture.cs
@@ -28,8 +28,20 @@ public Task DisposeAsync()
 
     private static async Task<string> DownloadModelAsync(GgmlType type, QuantizationType quantizationType = QuantizationType.NoQuantization)
     {
+        var huggingFaceToken = Environment.GetEnvironmentVariable("HF_TOKEN");
         var ggmlModelPath = Path.Combine(Path.GetTempPath(), $"{Guid.NewGuid()}.bin");
-        var model = await WhisperGgmlDownloader.GetGgmlModelAsync(type, quantizationType);
+        var downloader = string.IsNullOrEmpty(huggingFaceToken)
+            ? WhisperGgmlDownloader.Default
+            : new(
+                new()
+                {
+                    DefaultRequestHeaders =
+                    {
+                        { "Authorization", $"Bearer {huggingFaceToken}" }
+                    },
+                    Timeout = TimeSpan.FromHours(1)
+                });
+        var model = await downloader.GetGgmlModelAsync(type, quantizationType);
         using var fileWriter = File.OpenWrite(ggmlModelPath);
         await model.CopyToAsync(fileWriter);
         return ggmlModelPath;