resolve comments

🎨 Auto-generated directory tree for repository in Architecture.md

.github/Architecture.md

@@ -6,6 +6,11 @@
 📦 training
 |  |- 📂 training:
 |  |  |- 📂 routes:
+|  |  |  |- 📂 training:
+|  |  |  |  |- 📂 results:
+|  |  |  |  |  |- 📜 results.py
+|  |  |  |  |  |- 📜 __init__.py
+|  |  |  |  |  |- 📜 schemas.py
 |  |  |  |- 📂 tabular:
 |  |  |  |  |- 📜 tabular.py
 |  |  |  |  |- 📜 __init__.py
@@ -30,18 +35,24 @@
 |  |  |  |- 📜 __init__.py
 |  |  |  |- 📜 health_check_middleware.py
 |  |  |- 📂 core:
-|  |  |  |- 📜 trainer.py
-|  |  |  |- 📜 criterion.py
-|  |  |  |- 📜 dl_model.py : torch model based on user specifications from drag and drop
-|  |  |  |- 📜 dataset.py : read in the dataset through URL or file upload
+|  |  |  |- 📂 celery:
+|  |  |  |  |- 📜 trainer.py
+|  |  |  |  |- 📜 criterion.py
+|  |  |  |  |- 📜 dl_model.py : torch model based on user specifications from drag and drop
+|  |  |  |  |- 📜 dataset.py : read in the dataset through URL or file upload
+|  |  |  |  |- 📜 __init__.py
+|  |  |  |  |- 📜 worker.py
+|  |  |  |  |- 📜 optimizer.py : what optimizer to use (ie: SGD or Adam for now)
 |  |  |  |- 📜 __init__.py
 |  |  |  |- 📜 authenticator.py
-|  |  |  |- 📜 optimizer.py : what optimizer to use (ie: SGD or Adam for now)
 |  |  |- 📜 asgi.py
+|  |  |- 📜 constants.py : list of helpful constants
+|  |  |- 📜 celery_app.py
 |  |  |- 📜 settings.py
 |  |  |- 📜 __init__.py
 |  |  |- 📜 wsgi.py
 |  |  |- 📜 urls.py
+|  |  |- 📜 celeryconfig.py
 |  |- 📜 README.md
 |  |- 📜 docker-compose.yml
 |  |- 📜 cli.py
@@ -59,15 +70,11 @@
 
 ```
 📦 frontend
-|  |- 📂 layer_docs:
-|  |  |- 📜 Linear.md : Doc for Linear layer
-|  |  |- 📜 Softmax.md : Doc for Softmax layer
-|  |  |- 📜 softmax_equation.png : PNG file of Softmax equation
-|  |  |- 📜 ReLU.md : Doc for ReLU later
 |  |- 📂 public:
 |  |  |- 📂 images:
 |  |  |  |- 📂 logos:
 |  |  |  |  |- 📂 dlp_branding:
+|  |  |  |  |  |- 📜 dlp-logo.png : DLP Logo, duplicate of files in public, but essential as the frontend can't read public
 |  |  |  |  |  |- 📜 dlp-logo.svg : DLP Logo, duplicate of files in public, but essential as the frontend can't read public
 |  |  |  |  |- 📜 dsgt-logo-white-back.png
 |  |  |  |  |- 📜 python-logo.png
@@ -214,6 +221,7 @@
 |  |  |- 📂 pages:
 |  |  |  |- 📂 train:
 |  |  |  |  |- 📜 [train_space_id].tsx
+|  |  |  |  |- 📜 metrics_to_charts.tsx
 |  |  |  |  |- 📜 index.tsx
 |  |  |  |- 📜 _app.tsx
 |  |  |  |- 📜 forgot.tsx

dlp-terraform/ecs/sqs.tf

@@ -2,6 +2,24 @@ resource "aws_sqs_queue" "training_queue" {
   name       = "training-queue.fifo"
   fifo_queue = true
   message_retention_seconds = 60*24
+
+  redrive_policy = jsonencode({
+    deadLetterTargetArn = aws_sqs_queue.training_queue_deadletter.arn
+    maxReceiveCount     = 4
+  })
+}
+
+resource "aws_sqs_queue" "training_queue_deadletter" {
+  name = "training-deadletter-queue"
+}
+
+resource "aws_sqs_queue_redrive_allow_policy" "training_queue_redrive_allow_policy" {
+  queue_url = aws_sqs_queue.training_queue_deadletter.id
+
+  redrive_allow_policy = jsonencode({
+    redrivePermission = "byQueue",
+    sourceQueueArns   = [aws_sqs_queue.training_queue.arn]
+  })
 }
 
 output "sqs_queue_url" {

frontend/src/pages/train/[train_space_id].tsx

@@ -7,13 +7,13 @@ import { DetailedTrainResultsData } from "@/features/Train/types/trainTypes";
 import Container from "@mui/material/Container";
 import Grid from "@mui/material/Grid";
 import Paper from "@mui/material/Paper";
-import dynamic from "next/dynamic";
 import { useRouter } from "next/router";
-import { Data, XAxisName, YAxisName } from "plotly.js";
 import React, { useEffect } from "react";
-const Plot = dynamic(() => import("react-plotly.js"), { ssr: false });
-
-const LINE_CHART_COLORS = ["red", "blue", "green"];
+import {
+  mapMetricToLinePlot,
+  mapMetricToAucRocPlot,
+  mapMetricToConfusionMatrixPlot,
+} from "./metrics_to_charts";
 
 const mapTrainResultsDataToCharts = (
   detailedTrainResultsData: DetailedTrainResultsData
@@ -27,120 +27,11 @@ const mapTrainResultsDataToCharts = (
   while (i < sortedData.length) {
     const metric = sortedData[i];
     if (metric.chart_type === "LINE") {
-      const data = [];
-      for (let i = 0; i < metric.time_series.length; i++) {
-        const time_series = metric.time_series[i];
-        data.push({
-          name: time_series.y_name,
-          x: time_series.x_values,
-          y: time_series.y_values,
-          type: "scatter",
-          mode: "markers",
-          marker: { color: LINE_CHART_COLORS[i], size: 10 },
-        });
-      }
-      charts.push(
-        <Plot
-          data={data as Data[]}
-          layout={{
-            height: 350,
-            width: 525,
-            xaxis: { title: metric.time_series[0].x_name },
-            // yaxis: { title: "Y axis" },
-            title: metric.name,
-            showlegend: true,
-            paper_bgcolor: "rgba(0,0,0,0)",
-            plot_bgcolor: "rgba(0,0,0,0)",
-          }}
-          config={{ responsive: true }}
-        />
-      );
+      charts.push(mapMetricToLinePlot(metric));
     } else if (metric.chart_type === "AUC/ROC") {
-      charts.push(
-        <Plot
-          data={[
-            {
-              name: "baseline",
-              x: [0, 1],
-              y: [0, 1],
-              type: "scatter",
-              marker: { color: "grey" },
-              line: {
-                dash: "dash",
-              },
-            },
-            ...(metric.values.map((x) => ({
-              name: `(AUC: ${x[2]})`,
-              x: x[0] as number[],
-              y: x[1] as number[],
-              type: "scatter",
-            })) as Data[]),
-          ]}
-          layout={{
-            height: 350,
-            width: 525,
-            xaxis: { title: "False Positive Rate" },
-            yaxis: { title: "True Positive Rate" },
-            title: "AUC/ROC Curves for your Deep Learning Model",
-            showlegend: true,
-            paper_bgcolor: "rgba(0,0,0,0)",
-            plot_bgcolor: "rgba(0,0,0,0)",
-          }}
-          config={{ responsive: true }}
-        />
-      );
+      charts.push(mapMetricToAucRocPlot(metric));
     } else if (metric.chart_type === "CONFUSION_MATRIX") {
-      charts.push(
-        <Plot
-          data={[
-            {
-              z: metric.values,
-              type: "heatmap",
-              colorscale: [
-                [0, "#e6f6fe"],
-                [1, "#003058"],
-              ],
-            },
-          ]}
-          layout={{
-            height: 525,
-            width: 525,
-            title: "Confusion Matrix (Last Epoch)",
-            xaxis: {
-              title: "Predicted",
-            },
-            yaxis: {
-              title: "Actual",
-              autorange: "reversed",
-            },
-            showlegend: true,
-            annotations: metric.values
-              .map((row, i) =>
-                row.map((_, j) => ({
-                  xref: "x1" as XAxisName,
-                  yref: "y1" as YAxisName,
-                  x: j,
-                  y: (i + metric.values.length - 1) % metric.values.length,
-                  text: metric.values[
-                    (i + metric.values.length - 1) % metric.values.length
-                  ][j].toString(),
-                  font: {
-                    color:
-                      metric.values[
-                        (i + metric.values.length - 1) % metric.values.length
-                      ][j] > 0
-                        ? "white"
-                        : "black",
-                  },
-                  showarrow: false,
-                }))
-              )
-              .flat(),
-            paper_bgcolor: "rgba(0,0,0,0)",
-            plot_bgcolor: "rgba(0,0,0,0)",
-          }}
-        />
-      );
+      charts.push(mapMetricToConfusionMatrixPlot(metric));
     } else {
       throw Error("Undefined chart type received");
     }
@@ -163,7 +54,7 @@ const TrainSpace = () => {
       router.replace({ pathname: "/login" });
     }
   }, [user, router.isReady]);
-  
+
   if (error) {
     setTimeout(() => refetch(), 3000);
   }

frontend/src/pages/train/metrics_to_charts.tsx

@@ -0,0 +1,135 @@
+import {
+  AucRocChart,
+  ConfusionMatrixChart,
+  TimeSeriesChart,
+} from "@/features/Train/types/trainTypes";
+import dynamic from "next/dynamic";
+import { Data, XAxisName, YAxisName } from "plotly.js";
+const Plot = dynamic(() => import("react-plotly.js"), { ssr: false });
+
+const LINE_CHART_COLORS = ["red", "blue", "green"];
+
+const mapMetricToLinePlot = (metric: TimeSeriesChart) => {
+  const data = [];
+  for (let i = 0; i < metric.time_series.length; i++) {
+    const time_series = metric.time_series[i];
+    data.push({
+      name: time_series.y_name,
+      x: time_series.x_values,
+      y: time_series.y_values,
+      type: "scatter",
+      mode: "markers",
+      marker: { color: LINE_CHART_COLORS[i], size: 10 },
+    });
+  }
+  return (
+    <Plot
+      data={data as Data[]}
+      layout={{
+        height: 350,
+        width: 525,
+        xaxis: { title: metric.time_series[0].x_name },
+        // yaxis: { title: "Y axis" },
+        title: metric.name,
+        showlegend: true,
+        paper_bgcolor: "rgba(0,0,0,0)",
+        plot_bgcolor: "rgba(0,0,0,0)",
+      }}
+      config={{ responsive: true }}
+    />
+  );
+};
+
+const mapMetricToAucRocPlot = (metric: AucRocChart) => {
+  return (
+    <Plot
+      data={[
+        {
+          name: "baseline",
+          x: [0, 1],
+          y: [0, 1],
+          type: "scatter",
+          marker: { color: "grey" },
+          line: {
+            dash: "dash",
+          },
+        },
+        ...(metric.values.map((x) => ({
+          name: `(AUC: ${x[2]})`,
+          x: x[0] as number[],
+          y: x[1] as number[],
+          type: "scatter",
+        })) as Data[]),
+      ]}
+      layout={{
+        height: 350,
+        width: 525,
+        xaxis: { title: "False Positive Rate" },
+        yaxis: { title: "True Positive Rate" },
+        title: "AUC/ROC Curves for your Deep Learning Model",
+        showlegend: true,
+        paper_bgcolor: "rgba(0,0,0,0)",
+        plot_bgcolor: "rgba(0,0,0,0)",
+      }}
+      config={{ responsive: true }}
+    />
+  );
+};
+
+const mapMetricToConfusionMatrixPlot = (metric: ConfusionMatrixChart) => {
+  <Plot
+    data={[
+      {
+        z: metric.values,
+        type: "heatmap",
+        colorscale: [
+          [0, "#e6f6fe"],
+          [1, "#003058"],
+        ],
+      },
+    ]}
+    layout={{
+      height: 525,
+      width: 525,
+      title: "Confusion Matrix (Last Epoch)",
+      xaxis: {
+        title: "Predicted",
+      },
+      yaxis: {
+        title: "Actual",
+        autorange: "reversed",
+      },
+      showlegend: true,
+      annotations: metric.values
+        .map((row, i) =>
+          row.map((_, j) => ({
+            xref: "x1" as XAxisName,
+            yref: "y1" as YAxisName,
+            x: j,
+            y: (i + metric.values.length - 1) % metric.values.length,
+            text: metric.values[
+              (i + metric.values.length - 1) % metric.values.length
+            ][j].toString(),
+            font: {
+              color:
+                metric.values[
+                  (i + metric.values.length - 1) % metric.values.length
+                ][j] > 0
+                  ? "white"
+                  : "black",
+            },
+            showarrow: false,
+          }))
+        )
+        .flat(),
+      paper_bgcolor: "rgba(0,0,0,0)",
+      plot_bgcolor: "rgba(0,0,0,0)",
+    }}
+  />;
+};
+
+export {
+  mapMetricToLinePlot,
+  mapMetricToAucRocPlot,
+  mapMetricToConfusionMatrixPlot,
+};

training/training/constants.py

@@ -0,0 +1 @@
+DLP_EXECUTIONS_BUCKET_NAME = "dlp-executions"

training/training/core/celery/worker.py

@@ -9,6 +9,7 @@
 import boto3
 
 
+from training.constants import DLP_EXECUTIONS_BUCKET_NAME
 from training.core.celery.criterion import getCriterionHandler
 from training.core.celery.dataset import SklearnDatasetCreator
 from training.core.celery.dataset import ImageDefaultDatasetCreator
@@ -34,7 +35,7 @@ def saveDetailedTrainResultsDataToS3(
 ):
     s3 = boto3.resource("s3")
     s3.Object(
-        "dlp-executions", f"{detailedTrainResultsData.basic_info.trainspaceId}.json"
+        DLP_EXECUTIONS_BUCKET_NAME, f"{detailedTrainResultsData.basic_info.trainspaceId}.json"
     ).put(Body=detailedTrainResultsData.json())