fixes for modelsmith on local mode

utils/setup_environment.sh → backend/setup_environment.sh

@@ -1,3 +1,5 @@
+#!/bin/bash
+
 #    Copyright 2024 Cisco Systems, Inc. and its affiliates
 
 #   Licensed under the Apache License, Version 2.0 (the "License");
@@ -14,8 +16,6 @@
 
 #   SPDX-License-Identifier: Apache-2.0
 
-#!/bin/bash
-
 cd "$(dirname "$0")/.."
 PROJECT_ROOT="$(pwd)"
 backend_path="$PROJECT_ROOT/backend"
@@ -60,19 +60,24 @@ create_update_env_file() {
     mkdir -p "$backend_path"
 
     local machine_learning_core_path="../machine_learning_core"
-    echo -e "${YELLOW}Enter CONDA_SH_PATH:${NC}"
-    read -p "(e.g., ~/miniconda3/etc/profile.d/conda.sh): " conda_sh_path
+    echo -e "${YELLOW}Enter CONDA_SH_PATH (press Enter for default):${NC}"
+    read -p "(default: ~/miniconda3/etc/profile.d/conda.sh): " conda_sh_path
+    conda_sh_path=${conda_sh_path:-~/miniconda3/etc/profile.d/conda.sh}
 
     if [[ $env_choice == "2" ]]; then
         echo -e "${YELLOW}Enter MACHINE_LEARNING_CORE_PATH for VM:${NC}"
         read -p "(e.g., /path/to/machine_learning_core on VM): " machine_learning_core_vm_path
         machine_learning_core_path="$machine_learning_core_vm_path"
     fi
 
+    echo -e "${YELLOW}Enter HUGGING_FACE_ACCESS_TOKEN:${NC}"
+    read -p "(e.g., your_hugging_face_access_token): " hugging_face_access_token
+
     {
         echo "PORT=3000"
         echo "MACHINE_LEARNING_CORE_PATH=$machine_learning_core_path"
         echo "CONDA_SH_PATH=$conda_sh_path"
+        echo "HUGGING_FACE_ACCESS_TOKEN=$hugging_face_access_token"
         echo ""
     } > "$env_file"
 

backend/src/facades/algorithmExecutionFacade.js

@@ -32,6 +32,7 @@ class ScriptExecutor {
 		if (this.type === ALGORITHM_TYPES.MULTIFLOW) {
 			return `bash -c "source ${process.env.CONDA_SH_PATH} && cd ${process.env.MACHINE_LEARNING_CORE_PATH}/multiflow && conda activate modelsmith && python3 ${algorithm} ${args}"`;
 		}
+
 		return `bash -c "source ${process.env.CONDA_SH_PATH} && conda activate modelsmith && python3 ${scriptPath}${algorithm} ${args}"`;
 	}
 

backend/src/router/modelsRouter.js

@@ -128,7 +128,14 @@ router.get('/models-list/:type', checkSshConnection, (req, res) => {
 
 			res.status(OK).send(modelsWithTrainingInfo);
 		},
-		() => {},
+		() => {
+			const modelsAreNotTrained = models.map((modelName) => ({
+				name: modelName,
+				isTrained: false
+			}));
+
+			res.status(OK).send(modelsAreNotTrained);
+		},
 		(error) => {
 			if (error.includes('No such file or directory')) {
 				const modelsWithTrainingInfo = models.map((modelName) => ({

docs/configure-modelsmith-on-vm.md

@@ -100,7 +100,15 @@ conda --version
 
 ## Phase 2: Configure the Environment
 
-On your local machine, run the `setup_environment.sh` script located in the `utils` directory of the repository. This script configures the `.env` file for the backend server. When prompted for the environment for model training, select "2. VM" and provide the necessary SSH details for the primary and, optionally, backup VMs.
+After configuring the Python environment, the next step involves setting up the project environment. This is done by executing the `setup_environment.sh` script located in the `backend` directory. The script performs several tasks, including setting up the `.env` file for the backend server.
+
+Execute the script by running:
+
+```bash
+bash backend/setup_environment.sh
+```
+
+When asked about "Select the environment for model training:" -> choose 2. VM
 
 ### Example .env Configuration for VM Usage
 

docs/configure-pruning-quant-ml.md

@@ -1,5 +1,7 @@
 # Configure ModelSmith with Pruning, Quantization, and Machine Unlearning
 
+**Note**: The training of models can also be performed from the frontend page after starting the app - [Model Training](http://localhost:4200/model-training).
+
 This guide provides detailed steps to configure and train models using ModelSmith for pruning, quantization, and machine unlearning.
 
 ## Prerequisites

docs/configure-the-project-locally.md

@@ -26,41 +26,75 @@ node -v
 
 To get started, you need to set up the Python environment for ModelSmith. This involves creating a new Conda environment and installing the necessary packages. Follow these steps:
 
-1. **Create and Activate a New Conda Environment**
+1. **Install Miniconda**: Miniconda is a minimal installer for Conda, which is an open-source package management system and environment management system. Installing Miniconda on the VM will help manage the Python environment and dependencies required for the ModelSmith project. Here's how to download and install Miniconda:
+
+   1.1. Download the Miniconda installer for Linux:
+
+   ```bash
+   wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
+   ```
+
+   1.2. Make the installer script executable:
+
+   ```bash
+   chmod +x Miniconda3-latest-Linux-x86_64.sh
+   ```
+
+   1.3. Run the installer script:
+
+   ```bash
+   ./Miniconda3-latest-Linux-x86_64.sh
+   ```
+
+   Follow the on-screen prompts to complete the installation. It is generally safe to accept the default installation settings. Make sure to agree to initialize Miniconda by running `conda init`.
+
+   1.4. After installation, restart your terminal or source your `.bashrc` file to make the `conda` command available:
+
+   ```bash
+   source ~/.bashrc
+   ```
+
+   1.5. Verify the installation by checking the Conda version:
+
+   ```bash
+   conda --version
+   ```
+
+2. **Create and Activate a New Conda Environment**
    Open a terminal and run the following commands:
 
    ```bash
    conda create -n modelsmith python=3.9
    conda activate modelsmith
    ```
 
-2. **Install PyTorch and Related Libraries**
+3. **Install PyTorch and Related Libraries**
    With the environment activated, install PyTorch, torchvision, and torchaudio for CUDA 11.8 by running:
 
    ```bash
    pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
    ```
 
-3. **Configure Pruning, Quantization, Machine Unlearning Python Projects**:
+4. **Configure Pruning, Quantization, Machine Unlearning Python Projects**:
 
    Please refer to the [Pruning, Quantization, Machine Unlearning Configuration Guide](configure-pruning-quant-ml.md) for detailed instructions.
 
-4. **Configure AutoAWQ for Model Quantization**:
+5. **Configure AutoAWQ for Model Quantization**:
 
    Please refer to the [AutoAWQ Configuration Guide](configure-autoawq.md) for detailed instructions.
 
-5. **Configure Multiflow Python Project**:
+6. **Configure Multiflow Python Project**:
 
    Please refer to the [Multiflow Configuration Guide](configure-multiflow.md) for detailed instructions.
 
 ## Phase 2: Run the Environment Setup Script
 
-After configuring the Python environment, the next step involves setting up the project environment. This is done by executing the `setup_environment.sh` script located in the `utils` directory. The script performs several tasks, including setting up the `.env` file for the backend server.
+After configuring the Python environment, the next step involves setting up the project environment. This is done by executing the `setup_environment.sh` script located in the `backend` directory. The script performs several tasks, including setting up the `.env` file for the backend server.
 
 Execute the script by running:
 
 ```bash
-bash utils/setup_environment.sh
+bash backend/setup_environment.sh
 ```
 
 When asked about "Select the environment for model training:" -> choose 1. Local

frontend/src/app/modules/algorithm-comparison/components/run-drawer-actions/run-drawer-actions.component.html

@@ -51,7 +51,7 @@
 				<mat-expansion-panel-header>
 					<mat-panel-title class="paragraph-bold-p3-small-bold"> Statistics </mat-panel-title>
 				</mat-expansion-panel-header>
-				@if (summarizedRecord?.statistics | keyvalue; as statisticsItems) { @for (item of statisticsItems; track item;
+				@if (summarizedRecord.statistics | keyvalue; as statisticsItems) { @for (item of statisticsItems; track item;
 				let last = $last) {
 				<div class="key-value-container single">
 					<div class="key-value-pair space-between">
@@ -78,7 +78,7 @@
 				<mat-expansion-panel-header>
 					<mat-panel-title class="paragraph-bold-p3-small-bold"> Model training details </mat-panel-title>
 				</mat-expansion-panel-header>
-				@if (summarizedRecord?.modelTrainingDetails | hasKeys) { @for (item of summarizedRecord?.modelTrainingDetails |
+				@if (summarizedRecord.modelTrainingDetails | hasKeys) { @for (item of summarizedRecord.modelTrainingDetails |
 				keyvalue; track item; let last = $last) {
 				<div class="key-value-container single">
 					<div class="key-value-pair space-between">

machine_learning_core/bash/kill_script.sh

@@ -26,7 +26,7 @@ fi
 PROCESS_NAME="$1"
 
 # Fetch all processes that have the name $PROCESS_NAME
-PIDS=$(ps aux | grep $PROCESS_NAME | grep -v grep | awk '{print $2}')
+PIDS=$(ps -ef | grep "$PROCESS_NAME" | grep -v grep | awk '{print $2}')
 
 if [ -z "$PIDS" ]; then
     echo "No process found for $PROCESS_NAME"

start_modelsmith.sh

@@ -96,7 +96,7 @@ env_file="$PROJECT_ROOT/backend/.env"
 # Check if .env file exists and call setup_environment.sh if needed
 if [ ! -f "$env_file" ]; then
     # Call the setup script from the utils directory
-    bash "$PROJECT_ROOT/utils/setup_environment.sh"
+    bash "$PROJECT_ROOT/backend/setup_environment.sh"
 fi
 
 # Start backend and frontend servers

utils/deploy_vm_modelsmith_options.sh

@@ -1,3 +1,5 @@
+#!/bin/bash
+
 #    Copyright 2024 Cisco Systems, Inc. and its affiliates
 
 #   Licensed under the Apache License, Version 2.0 (the "License");
@@ -14,7 +16,6 @@
 
 #   SPDX-License-Identifier: Apache-2.0
 
-#!/bin/bash
 
 # ANSI color codes
 GREEN="\033[0;32m"