Anul 3 Semestrul 1

2025-02-06 20:33:26 +02:00
parent 0b130ee18c
commit 184f3bd92e
313 changed files with 348499 additions and 0 deletions
@@ -0,0 +1,135 @@
+
+using BERTTokenizers;
+using Microsoft.EntityFrameworkCore;
+using Microsoft.EntityFrameworkCore.Metadata.Internal;
+using Microsoft.ML;
+using Microsoft.ML.Data;
+using Microsoft.ML.Transforms.Onnx;
+using System.Collections.Generic;
+using Tokenizers.DotNet;
+using WebApi.Models;
+
+namespace WebApi.Services
+{
+    public class ModelInput
+    {
+        [VectorType(1, 128)]
+        [ColumnName("input_ids")]
+        public long[] InputIds { get; set; }
+
+        [VectorType(1, 128)]
+        [ColumnName("attention_mask")]
+        public long[] AttentionMask { get; set; }
+
+        [VectorType(1, 128)]
+        [ColumnName("token_type_ids")]
+        public long[] TokenTypeIds { get; set; }
+    }
+
+    public class ModelOutput
+    {
+        [VectorType(1, 6)]
+        [ColumnName("output")]
+        public float[] Output { get; set; }
+    }
+
+    public class StoryPointEstimator
+    {
+        private readonly MLContext _mlContext;
+        private readonly OnnxScoringEstimator estimator;
+        private readonly String ModelPath = "./AIModels/storypoint_estimator.onnx";
+        private readonly String GithubRepo = "Not-Atlassian/StoryPointEstimator";
+        private readonly String ModelFilename = "storypoint_estimator.onnx";
+
+        public StoryPointEstimator()
+        {
+            _mlContext = new MLContext();
+            EnsureModelExists().Wait();
+            estimator = _mlContext.Transforms.ApplyOnnxModel("./AIModels/storypoint_estimator.onnx");
+        }
+
+
+        private async Task EnsureModelExists()
+        {
+            if (!File.Exists(ModelPath))
+            {
+                Console.WriteLine("Model file not found locally. Downloading...");
+                await DownloadModelFromGithub();
+            }
+
+        }
+
+        private async Task DownloadModelFromGithub()
+        {
+            using var httpClient = new HttpClient();
+            var apiUrl = $"https://api.github.com/repos/{GithubRepo}/releases/latest";
+
+            httpClient.DefaultRequestHeaders.UserAgent.ParseAdd("Mozilla/5.0 (compatible; StoryPointEstimator)");
+
+            var response = await httpClient.GetAsync(apiUrl);
+            response.EnsureSuccessStatusCode();
+
+            var releaseData = await response.Content.ReadAsStringAsync();
+
+            var releaseJson = System.Text.Json.JsonDocument.Parse(releaseData);
+            var assets = releaseJson.RootElement.GetProperty("assets");
+
+            string? downloadUrl = null;
+            foreach (var asset in assets.EnumerateArray())
+            {
+                if (asset.GetProperty("name").GetString() == ModelFilename)
+                {
+                    downloadUrl = asset.GetProperty("browser_download_url").GetString();
+                    break;
+                }
+            }
+
+            if (downloadUrl == null)
+            {
+                throw new FileNotFoundException("Model file not found in the latest GitHub release.");
+            }
+
+            Console.WriteLine($"Downloading model from {downloadUrl}...");
+
+            var modelData = await httpClient.GetByteArrayAsync(downloadUrl);
+
+            Directory.CreateDirectory(Path.GetDirectoryName(ModelPath)!);
+
+            await File.WriteAllBytesAsync(ModelPath, modelData);
+
+            Console.WriteLine($"Model downloaded and saved to {ModelPath}");
+        }
+
+
+        public async Task<int> EstimateStoryPoints(string title, string description)
+        {
+            var tokenizer = new BertUncasedBaseTokenizer();
+
+            var encoded = tokenizer.Encode(128, title + " " + description);
+
+            var bertInput = new ModelInput()
+            {
+                InputIds = encoded.Select(t => t.InputIds).ToArray(),
+                AttentionMask = encoded.Select(t => t.AttentionMask).ToArray(),
+                TokenTypeIds = encoded.Select(t => t.TokenTypeIds).ToArray()
+            };
+            var data = _mlContext.Data.LoadFromEnumerable(new List<ModelInput>());
+            var transformedData = estimator.Fit(data);
+            var predictionEngine = _mlContext.Model.CreatePredictionEngine<ModelInput, ModelOutput>(transformedData);
+            var prediction = predictionEngine.Predict(bertInput);
+            var index = prediction.Output.ToList().IndexOf(prediction.Output.Max());
+            return index switch
+            {
+                0 => 1,
+                1 => 2,
+                2 => 3,
+                3 => 5,
+                4 => 8,
+                5 => 13,
+                6 => int.MaxValue,
+                _ => 1
+            };
+        }
+    }
+
+}