Skip to content

Add LlamaSharp embeddings generator #873

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
dluc merged 3 commits into from
Oct 30, 2024
Merged

Add LlamaSharp embeddings generator #873

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
2e3b6a9
LlamaSharp embeddings
dluc Oct 30, 2024
b352345
Merge branch 'main' into llamasharpembeddings
dluc Oct 30, 2024
147f6cd
Fix build warning
dluc Oct 30, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 16 additions & 7 deletions applications/tests/Evaluation.Tests/appsettings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -96,13 +96,22 @@
"MaxRetries": 10
},
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
}
},
"DataIngestion": {
Expand Down
23 changes: 16 additions & 7 deletions examples/001-dotnet-WebClient/file9-settings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,13 +65,22 @@
"MaxRetries": 10
},
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
},
"AzureAIDocIntel": {
// "APIKey" or "AzureIdentity".
Expand Down
23 changes: 16 additions & 7 deletions examples/002-dotnet-Serverless/appsettings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -120,13 +120,22 @@
// "Endpoint": "https://s3.amazonaws.com"
},
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
}
},
"DataIngestion": {
Expand Down
23 changes: 16 additions & 7 deletions examples/002-dotnet-Serverless/file9-settings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,13 +65,22 @@
"MaxRetries": 10
},
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
},
"AzureAIDocIntel": {
// "APIKey" or "AzureIdentity".
Expand Down
23 changes: 16 additions & 7 deletions examples/105-dotnet-serverless-llamasharp/appsettings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,13 +15,22 @@
"KernelMemory": {
"Services": {
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
},
"AzureOpenAIEmbedding": {
// "ApiKey" or "AzureIdentity"
Expand Down
23 changes: 16 additions & 7 deletions examples/210-KM-without-builder/appsettings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -351,13 +351,22 @@
"Replicas": 0
},
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
},
"MongoDbAtlas": {
"ConnectionString": "mongodb://root:root@localhost:27777/?authSource=admin",
Expand Down
23 changes: 16 additions & 7 deletions examples/401-evaluation/appsettings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -103,13 +103,22 @@
"MaxEmbeddingBatchSize": 100
},
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
}
},
"DataIngestion": {
Expand Down
47 changes: 47 additions & 0 deletions extensions/LlamaSharp/LlamaSharp.FunctionalTests/LlamaSharpTextEmbeddingGeneratorTest.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,47 @@
// Copyright (c) Microsoft. All rights reserved.

using Microsoft.KernelMemory;
using Microsoft.KernelMemory.AI.LlamaSharp;
using Microsoft.KM.TestHelpers;
using Xunit.Abstractions;

namespace Microsoft.LlamaSharp.FunctionalTests;

public sealed class LlamaSharpTextEmbeddingGeneratorTest : BaseFunctionalTestCase
{
private readonly LlamaSharpTextEmbeddingGenerator _target;

public LlamaSharpTextEmbeddingGeneratorTest(
IConfiguration cfg,
ITestOutputHelper output) : base(cfg, output)
{
this.LlamaSharpConfig.Validate();
this._target = new LlamaSharpTextEmbeddingGenerator(this.LlamaSharpConfig.EmbeddingModel, loggerFactory: null);
var modelFilename = this.LlamaSharpConfig.TextModel.ModelPath.Split('/').Last().Split('\\').Last();
Console.WriteLine($"Model in use: {modelFilename}");
}

[Fact]
[Trait("Category", "LlamaSharp")]
public async Task ItGeneratesEmbeddingVectors()
{
// Act
Embedding embedding = await this._target.GenerateEmbeddingAsync("some text");

// Assert
Console.WriteLine("Embedding size: " + embedding.Length);

// Expected result using nomic-embed-text-v1.5.Q8_0.gguf
Assert.Equal(768, embedding.Length);
}

protected override void Dispose(bool disposing)
{
if (disposing)
{
this._target.Dispose();
}

base.Dispose(disposing);
}
}
4 changes: 2 additions & 2 deletions extensions/LlamaSharp/LlamaSharp.FunctionalTests/LlamaSharpTextGeneratorTest.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,8 +21,8 @@ public LlamaSharpTextGeneratorTest(
this._timer = new Stopwatch();

this.LlamaSharpConfig.Validate();
this._target = new LlamaSharpTextGenerator(this.LlamaSharpConfig, loggerFactory: null);
var modelFilename = this.LlamaSharpConfig.ModelPath.Split('/').Last().Split('\\').Last();
this._target = new LlamaSharpTextGenerator(this.LlamaSharpConfig.TextModel, loggerFactory: null);
var modelFilename = this.LlamaSharpConfig.TextModel.ModelPath.Split('/').Last().Split('\\').Last();
Console.WriteLine($"Model in use: {modelFilename}");
}

Expand Down
8 changes: 4 additions & 4 deletions extensions/LlamaSharp/LlamaSharp.FunctionalTests/LlamaSharpTokenizerTest.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,13 +18,13 @@ public LlamaSharpTokenizerTest(
{
this.LlamaSharpConfig.Validate();

var modelFilename = this.LlamaSharpConfig.ModelPath.Split('/').Last().Split('\\').Last();
var modelFilename = this.LlamaSharpConfig.TextModel.ModelPath.Split('/').Last().Split('\\').Last();
Console.WriteLine($"Model in use: {modelFilename}");

var parameters = new ModelParams(this.LlamaSharpConfig.ModelPath)
var parameters = new ModelParams(this.LlamaSharpConfig.TextModel.ModelPath)
{
ContextSize = this.LlamaSharpConfig.MaxTokenTotal,
GpuLayerCount = this.LlamaSharpConfig.GpuLayerCount ?? 20,
ContextSize = this.LlamaSharpConfig.TextModel.MaxTokenTotal,
GpuLayerCount = this.LlamaSharpConfig.TextModel.GpuLayerCount ?? 20,
};

LLamaWeights model = LLamaWeights.LoadFromFile(parameters);
Expand Down
23 changes: 16 additions & 7 deletions extensions/LlamaSharp/LlamaSharp.FunctionalTests/appsettings.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,13 +6,22 @@
},
"Services": {
"LlamaSharp": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
// "Seed": 1337,
"TextModel": {
// path to file, e.g. "llama-2-7b-chat.Q6_K.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
},
"EmbeddingModel": {
// path to file, e.g. "nomic-embed-text-v1.5.Q8_0.gguf"
"ModelPath": "",
// Max number of tokens supported by the model
"MaxTokenTotal": 4096
// Optional parameters
// "GpuLayerCount": 32,
}
},
"SimpleVectorDb": {
// Options: "Disk" or "Volatile". Volatile data is lost after each execution.
Expand Down
Loading