sacred apply

Dockerfile

@@ -5,6 +5,7 @@ FROM ubuntu:latest
 RUN export PATH="$PATH:/root/.local/bin"
 RUN apt update
 RUN apt-get update
+RUN apt install -y git 
 RUN apt install -y figlet
 RUN export PATH=”$PATH:/usr/local/bin/python”
 RUN apt install python3-pip -y
@@ -16,6 +17,8 @@ RUN pip3 install pillow
 RUN pip3 install scikit-learn
 RUN pip3 install matplotlib
 RUN pip3 install torchvision
+RUN pip3 install sacred
+RUN pip3 install pymongo
 
 # Args
 ARG KAGGLE_USERNAME
@@ -29,4 +32,4 @@ WORKDIR /app
 COPY . .
 
 # Create kaggle catalog for authenticate
-RUN mkdir /.kaggle/ && chmod o+w /.kaggle
+RUN mkdir /.kaggle/ && chmod o+w /.kaggle

Jenkinsfile-training

@@ -32,12 +32,13 @@ pipeline {
         stage('Train model') {
             steps {
                 sh "chmod u+x ./neutral_network.py"
-                sh "python3 neutral_network.py -e ${params.EPOCHS} -b ${params.BATCHSIZE}"
+                sh "python3 neutral_network.py with 'epochs=${params.EPOCHS}' 'batch_size=${params.BATCHSIZE}'"
             }
         }
         stage('Archive model') {
             steps {
-                archiveArtifacts artifacts: "model.zip", onlyIfSuccessful: true
+                archiveArtifacts artifacts: 'model.zip', onlyIfSuccessful: true
+                archiveArtifacts artifacts: 'sacred_runs/**', onlyIfSuccessful: true
             }
         }
     }

neutral_network.py

@@ -5,13 +5,14 @@ import torch
 import argparse
 from torch import nn
 from torch.utils.data import DataLoader, Dataset
+from sacred import Experiment
+from sacred.observers import FileStorageObserver, MongoObserver
 
 default_batch_size = 64
 default_epochs = 4
 
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
-
 class AtpDataset(Dataset):
     def __init__(self, file_name):
         df = pd.read_csv(file_name, usecols=["AvgL", "AvgW"])
@@ -83,35 +84,51 @@ def setup_args():
     args_parser.add_argument("-e", "--epochs", type=int, default=default_epochs)
     return args_parser.parse_args()
 
+def main(batch_size, epochs):
+    print(f"Using {device} device")
 
-print(f"Using {device} device")
+    plant_test = AtpDataset("atp_test.csv")
+    plant_train = AtpDataset("atp_train.csv")
 
-args = setup_args()
-batch_size = args.batchSize
+    train_dataloader = DataLoader(plant_train, batch_size=batch_size)
+    test_dataloader = DataLoader(plant_test, batch_size=batch_size)
 
-plant_test = AtpDataset("atp_test.csv")
-plant_train = AtpDataset("atp_train.csv")
+    for i, (data, labels) in enumerate(train_dataloader):
+        print(data.shape, labels.shape)
+        print(data, labels)
+        break
 
-train_dataloader = DataLoader(plant_train, batch_size=batch_size)
-test_dataloader = DataLoader(plant_test, batch_size=batch_size)
+    model = MLP()
+    print(model)
 
-for i, (data, labels) in enumerate(train_dataloader):
-    print(data.shape, labels.shape)
-    print(data, labels)
-    break
+    loss_fn = nn.MSELoss()
+    optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)
 
-model = MLP()
-print(model)
+    for t in range(epochs):
+        print(f"Epoch {t + 1}\n-------------------------------")
+        train(train_dataloader, model, loss_fn, optimizer)
+        test(test_dataloader, model, loss_fn)
+    print("Finish!")
 
-loss_fn = nn.MSELoss()
-optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)
-epochs = args.epochs
+    torch.save(model.state_dict(), "./model.zip")
+    print("Model saved in ./model.zip file.")
 
-for t in range(epochs):
-    print(f"Epoch {t + 1}\n-------------------------------")
-    train(train_dataloader, model, loss_fn, optimizer)
-    test(test_dataloader, model, loss_fn)
-print("Finish!")
+def setup_experiment():
+    ex = Experiment('Simple Experiment')
+    ex.observers.append(FileStorageObserver('sacred_runs'))
+    # ex.observers.append(MongoObserver(url='mongodb://mongo_user:mongo_password_IUM_2022@localhost:27017',
+    #                                   db_name='sacred'))
+    return ex
 
-torch.save(model.state_dict(), "./model.zip")
-print("Model saved in ./model.zip file.")
+
+ex = setup_experiment()
+
+@ex.config
+def experiment_config():
+    batch_size = 64
+    epochs = 5
+
+@ex.automain
+def run(batch_size, epochs):
+    main(batch_size, epochs)
+    ex.add_artifact('model.zip')

sacred_runs/1/config.json

@@ -0,0 +1,5 @@
+{
+  "batch_size": 64,
+  "epochs": 5,
+  "seed": 171403633
+}

sacred_runs/1/metrics.json

@@ -0,0 +1 @@
+{}

sacred_runs/1/run.json

@@ -0,0 +1,74 @@
+{
+  "artifacts": [
+    "model.zip"
+  ],
+  "command": "run",
+  "experiment": {
+    "base_dir": "/home/wirus/dev/ium_444498",
+    "dependencies": [
+      "numpy==1.22.3",
+      "pandas==1.4.1",
+      "sacred==0.8.2",
+      "torch==1.11.0"
+    ],
+    "mainfile": "neutral_network.py",
+    "name": "Simple Experiment",
+    "repositories": [
+      {
+        "commit": "b0deed97f63ff54e9f4df87346315b37506fcf3b",
+        "dirty": true,
+        "url": "https://git.wmi.amu.edu.pl/s444498/ium_444498.git"
+      }
+    ],
+    "sources": [
+      [
+        "neutral_network.py",
+        "_sources/neutral_network_e5d851342f3b663b0ccbb114284aae02.py"
+      ]
+    ]
+  },
+  "heartbeat": "2022-05-09T16:01:55.707975",
+  "host": {
+    "ENV": {},
+    "cpu": "Intel(R) Core(TM) i3-2310M CPU @ 2.10GHz",
+    "hostname": "wirusowylapek",
+    "os": [
+      "Linux",
+      "Linux-5.10.105-1-MANJARO-x86_64-with-glibc2.35"
+    ],
+    "python_version": "3.10.2"
+  },
+  "meta": {
+    "command": "run",
+    "options": {
+      "--beat-interval": null,
+      "--capture": null,
+      "--comment": null,
+      "--debug": false,
+      "--enforce_clean": false,
+      "--file_storage": null,
+      "--force": false,
+      "--help": false,
+      "--loglevel": null,
+      "--mongo_db": null,
+      "--name": null,
+      "--pdb": false,
+      "--print-config": false,
+      "--priority": null,
+      "--queue": false,
+      "--s3": null,
+      "--sql": null,
+      "--tiny_db": null,
+      "--unobserved": false,
+      "COMMAND": null,
+      "UPDATE": [],
+      "help": false,
+      "with": false
+    }
+  },
+  "resources": [],
+  "result": null,
+  "start_time": "2022-05-09T16:01:42.424966",
+  "status": "COMPLETED",
+  "stop_time": "2022-05-09T16:01:55.705969"
+}

sacred_runs/_sources/neutral_network_e5d851342f3b663b0ccbb114284aae02.py

@@ -0,0 +1,135 @@
+from ast import arg
+import numpy as np
+import pandas as pd
+import torch
+import argparse
+from torch import nn
+from torch.utils.data import DataLoader, Dataset
+from sacred import Experiment
+from sacred.observers import FileStorageObserver, MongoObserver
+
+default_batch_size = 64
+default_epochs = 4
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+
+class AtpDataset(Dataset):
+    def __init__(self, file_name):
+        df = pd.read_csv(file_name, usecols=["AvgL", "AvgW"])
+        df = df.dropna()
+
+        # Loser avg and Winner avg
+        x = df.iloc[:, 1].values
+        y = df.iloc[:, 0].values
+
+        self.x_train = torch.from_numpy(x)
+        self.y_train = torch.from_numpy(y)
+        self.x_train.type(torch.LongTensor)
+
+    def __len__(self):
+        return len(self.y_train)
+
+    def __getitem__(self, idx):
+        return self.x_train[idx].float(), self.y_train[idx].float()
+
+
+class MLP(nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.layers = nn.Sequential(
+            nn.Linear(1, 64),
+            nn.ReLU(),
+            nn.Linear(64, 32),
+            nn.ReLU(),
+            nn.Linear(32, 1),
+        )
+
+    def forward(self, x):
+        x = x.view(x.size(0), -1)
+        return self.layers(x)
+
+
+def train(dataloader, model, loss_fn, optimizer):
+    size = len(dataloader.dataset)
+    model.train()
+    for batch, (X, y) in enumerate(dataloader):
+        X, y = X.to(device), y.to(device)
+        pred = model(X)
+        loss = loss_fn(pred, y)
+        optimizer.zero_grad()
+        loss.backward()
+        optimizer.step()
+        if batch % 100 == 0:
+            loss, current = loss.item(), batch * len(X)
+            print(f"loss: {loss:>7f}  [{current:>5d}/{size:>5d}]")
+
+
+def test(dataloader, model, loss_fn):
+    num_batches = len(dataloader)
+    model.eval()
+    test_loss, correct = 0, 0
+    with torch.no_grad():
+        for X, y in dataloader:
+            X, y = X.to(device), y.to(device)
+            pred = model(X)
+            test_loss += loss_fn(pred, y).item()
+    test_loss /= num_batches
+    print(f"Avg loss (using {loss_fn}): {test_loss:>8f} \n")
+    return test_loss
+
+
+def setup_args():
+    args_parser = argparse.ArgumentParser(prefix_chars="-")
+    args_parser.add_argument("-b", "--batchSize", type=int, default=default_batch_size)
+    args_parser.add_argument("-e", "--epochs", type=int, default=default_epochs)
+    return args_parser.parse_args()
+
+def main(batch_size, epochs):
+    print(f"Using {device} device")
+
+    plant_test = AtpDataset("atp_test.csv")
+    plant_train = AtpDataset("atp_train.csv")
+
+    train_dataloader = DataLoader(plant_train, batch_size=batch_size)
+    test_dataloader = DataLoader(plant_test, batch_size=batch_size)
+
+    for i, (data, labels) in enumerate(train_dataloader):
+        print(data.shape, labels.shape)
+        print(data, labels)
+        break
+
+    model = MLP()
+    print(model)
+
+    loss_fn = nn.MSELoss()
+    optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)
+
+    for t in range(epochs):
+        print(f"Epoch {t + 1}\n-------------------------------")
+        train(train_dataloader, model, loss_fn, optimizer)
+        test(test_dataloader, model, loss_fn)
+    print("Finish!")
+
+    torch.save(model.state_dict(), "./model.zip")
+    print("Model saved in ./model.zip file.")
+
+def setup_experiment():
+    ex = Experiment('Simple Experiment')
+    ex.observers.append(FileStorageObserver('sacred_runs'))
+    # ex.observers.append(MongoObserver(url='mongodb://mongo_user:mongo_password_IUM_2022@localhost:27017',
+    #                                   db_name='sacred'))
+    return ex
+
+
+ex = setup_experiment()
+
+@ex.config
+def experiment_config():
+    batch_size = 64
+    epochs = 5
+
+@ex.automain
+def run(batch_size, epochs):
+    main(batch_size, epochs)
+
+ex.add_artifact('model.zip')