udpate form to the new datamodel

Pipfile.lock

@@ -0,0 +1,20 @@
+{
+    "_meta": {
+        "hash": {
+            "sha256": "702ad05de9bc9de99a4807c8dde1686f31e0041d7b5f6f6b74861195a52110f5"
+        },
+        "pipfile-spec": 6,
+        "requires": {
+            "python_version": "3.12"
+        },
+        "sources": [
+            {
+                "name": "pypi",
+                "url": "https://pypi.org/simple",
+                "verify_ssl": true
+            }
+        ]
+    },
+    "default": {},
+    "develop": {}
+}

assets/styles/app.css

@@ -0,0 +1,44 @@
+/* Personnalisation du thème */
+:root, :root .dark {
+  --primary-50: #fef2f2;
+  --primary-100: #fee2e2;
+  --primary-200: #fecaca;
+  --primary-300: #fca5a5;
+  --primary-400: #f87171;
+  --primary-500: #ef4444;
+  --primary-600: #dc2626;
+  --primary-700: #b91c1c;
+  --primary-800: #991b1b;
+  --primary-900: #7f1d1d;
+  --primary-950: #450a0a;
+
+  --neutral-50: #f6f2fa;
+  --neutral-100: #f2eff6;
+  --neutral-200: #e4e1e8;
+  --neutral-300: #d5d2d9;
+  --neutral-400: #bfbcc2;
+  --neutral-500: #78777a;
+  --neutral-600: #5a595c;
+  --neutral-700: #444345;
+  --neutral-800: #282829;
+  --neutral-900: #1c1b1c;
+  --neutral-950: #121112;
+}
+
+/* Changer la couleur de fond grise par défaut */
+div {
+  background : white;
+}
+
+#mandatory_part-button{
+  color: var(--primary-600);
+  font-weight: bold;
+}
+.mandatory_field label > span{
+  color: var(--primary-600);
+  font-weight: bold !important;
+}
+.mandatory_field div > span{
+  color: var(--primary-600);
+  font-weight: bold !important;
+}

{utils → assets/utils}/validation.py

@@ -1,4 +1,4 @@
-from config import OBLIGATORY_FIELDS
+from src.services.util import OBLIGATORY_FIELDS
 
 
 def validate_obligatory_fields(data):
@@ -20,6 +20,7 @@ def validate_obligatory_fields(data):
         return None
 
     missing_fields = []
+
     for field in OBLIGATORY_FIELDS:
         # if the field is mandatory, check if it is inside a mandatory section
 

main.py

@@ -0,0 +1,4 @@
+from src.app import app
+
+print("Launching BoAmps form")
+app.launch()

services/json_generator.py

@@ -1,398 +0,0 @@
-import json
-import tempfile
-from datetime import datetime
-from utils.validation import validate_obligatory_fields
-
-
-def generate_json(
-    # Header
-    licensing, formatVersion, formatVersionSpecificationUri, reportId, reportDatetime, reportStatus,
-    publisher_name, publisher_division, publisher_projectName, publisher_confidentialityLevel, publisher_publicKey,
-    # Task
-    taskType, taskFamily, taskStage, algorithmName, framework, frameworkVersion, classPath, tuning_method,
-    hyperparameter_names, hyperparameter_values, quantization, dataType, fileType, volume, volumeUnit, items,
-    shape_item, nbRequest, nbTokensInput, nbWordsInput, nbTokensOutput, nbWordsOutput, contextWindowSize, cache,
-    source, sourceUri, owner, measuredAccuracy, estimatedAccuracy,
-    # Measures
-    measurementMethod, manufacturer, version, cpuTrackingMode, gpuTrackingMode, averageUtilizationCpu,
-    averageUtilizationGpu, serverSideInference, unit, powerCalibrationMeasurement, durationCalibrationMeasurement,
-    powerConsumption, measurementDuration, measurementDateTime,
-    # System
-    os, distribution, distributionVersion,
-    # Software
-    language, version_software,
-    # Infrastructure
-    infraType, cloudProvider, cloudInstance, componentName, componentType, nbComponent, memorySize, manufacturer_infra, family, series, share,
-    # Environment
-    country, latitude, longitude, location, powerSupplierType, powerSource, powerSourceCarbonIntensity,
-    # Quality
-    quality,
-    # Hash
-    hashAlgorithm, cryptographicAlgorithm, value_hash
-):
-    """Generate JSON data from form inputs."""
-    # TO CHANGE
-    # Process hyperparameters
-    hyperparameters = []
-    max_length = max(len(hyperparameter_names), len(hyperparameter_values))
-    for i in range(max_length):
-        hyperparameters.append({
-            "name": hyperparameter_names[i] if i < len(hyperparameter_names) and hyperparameter_names[i] else "",
-            "value": hyperparameter_values[i] if i < len(hyperparameter_values) and hyperparameter_values[i] else ""
-        })
-
-    # Process inference properties
-    inference_props_list = []
-    max_length = max(len(nbRequest), len(nbTokensInput), len(nbWordsInput), len(
-        nbTokensOutput), len(nbWordsOutput), len(contextWindowSize), len(cache))
-    for i in range(max_length):
-        inference_props = {}
-        if i < len(nbRequest) and nbRequest[i]:
-            inference_props["nbRequest"] = nbRequest[i]
-        if i < len(nbTokensInput) and nbTokensInput[i]:
-            inference_props["nbTokensInput"] = nbTokensInput[i]
-        if i < len(nbWordsInput) and nbWordsInput[i]:
-            inference_props["nbWordsInput"] = nbWordsInput[i]
-        if i < len(nbTokensOutput) and nbTokensOutput[i]:
-            inference_props["nbTokensOutput"] = nbTokensOutput[i]
-        if i < len(nbWordsOutput) and nbWordsOutput[i]:
-            inference_props["nbWordsOutput"] = nbWordsOutput[i]
-        if i < len(contextWindowSize) and contextWindowSize[i]:
-            inference_props["contextWindowSize"] = contextWindowSize[i]
-        if i < len(cache) and cache[i]:
-            inference_props["cache"] = cache[i]
-        inference_props_list.append(inference_props)
-
-    # Process components
-    components_list = []
-    max_length = max(len(componentName), len(componentType), len(nbComponent), len(memorySize), len(
-        manufacturer_infra), len(family), len(series), len(share))
-    for i in range(max_length):
-        component = {}
-        if i < len(componentName) and componentName[i]:
-            component["componentName"] = componentName[i]
-        if i < len(componentType) and componentType[i]:
-            component["componentType"] = componentType[i]
-        if i < len(nbComponent) and nbComponent[i]:
-            component["nbComponent"] = nbComponent[i]
-        if i < len(memorySize) and memorySize[i]:
-            component["memorySize"] = memorySize[i]
-        if i < len(manufacturer_infra) and manufacturer_infra[i]:
-            component["manufacturer"] = manufacturer_infra[i]
-        if i < len(family) and family[i]:
-            component["family"] = family[i]
-        if i < len(series) and series[i]:
-            component["series"] = series[i]
-        if i < len(share) and share[i]:
-            component["share"] = share[i]
-        components_list.append(component)
-
-    # process report
-    report = {}
-
-    # Process header
-    header = {}
-    if licensing:
-        header["licensing"] = licensing
-    if formatVersion:
-        header["formatVersion"] = formatVersion
-    if formatVersionSpecificationUri:
-        header["formatVersionSpecificationUri"] = formatVersionSpecificationUri
-    if reportId:
-        header["reportId"] = reportId
-    if reportDatetime:
-        header["reportDatetime"] = reportDatetime or datetime.now().isoformat()
-    if reportStatus:
-        header["reportStatus"] = reportStatus
-
-    publisher = {}
-    if publisher_name:
-        publisher["name"] = publisher_name
-    if publisher_division:
-        publisher["division"] = publisher_division
-    if publisher_projectName:
-        publisher["projectName"] = publisher_projectName
-    if publisher_confidentialityLevel:
-        publisher["confidentialityLevel"] = publisher_confidentialityLevel
-    if publisher_publicKey:
-        publisher["publicKey"] = publisher_publicKey
-
-    if publisher:
-        header["publisher"] = publisher
-
-    if header:
-        report["header"] = header
-
-    # proceed task
-
-    # proceed algorithm
-    algorithm = {}
-    if algorithmName:
-        algorithm["algorithmName"] = algorithmName
-    if framework:
-        algorithm["framework"] = framework
-    if frameworkVersion:
-        algorithm["frameworkVersion"] = frameworkVersion
-    if classPath:
-        algorithm["classPath"] = classPath
-    if hyperparameters:
-        algorithm["hyperparameters"] = hyperparameters
-    if quantization:
-        algorithm["quantization"] = quantization
-
-    # proceed dataset
-    dataset = {}
-    if dataType:
-        dataset["dataType"] = dataType
-    if fileType:
-        dataset["fileType"] = fileType
-    if volume:
-        dataset["volume"] = volume
-    if volumeUnit:
-        dataset["volumeUnit"] = volumeUnit
-    if items:
-        dataset["items"] = items
-    if shape_item:
-        dataset["shape"] = [{"item": shape_item}]
-    if inference_props_list:
-        dataset["inferenceProperties"] = inference_props_list
-    if source:
-        dataset["source"] = source
-    if sourceUri:
-        dataset["sourceUri"] = sourceUri
-    if owner:
-        dataset["owner"] = owner
-
-    # proceed all task
-    task = {}
-    if taskType:
-        task["taskType"] = taskType
-    if taskFamily:
-        task["taskFamily"] = taskFamily
-    if taskStage:
-        task["taskStage"] = taskStage
-    if algorithm:
-        task["algorithms"] = [algorithm]
-    if dataset:
-        task["dataset"] = [dataset]
-    if measuredAccuracy:
-        task["measuredAccuracy"] = measuredAccuracy
-    if estimatedAccuracy:
-        task["estimatedAccuracy"] = estimatedAccuracy
-    report["task"] = task
-
-    # proceed measures
-    measures = {}
-    if measurementMethod:
-        measures["measurementMethod"] = measurementMethod
-    if manufacturer:
-        measures["manufacturer"] = manufacturer
-    if version:
-        measures["version"] = version
-    if cpuTrackingMode:
-        measures["cpuTrackingMode"] = cpuTrackingMode
-    if gpuTrackingMode:
-        measures["gpuTrackingMode"] = gpuTrackingMode
-    if averageUtilizationCpu:
-        measures["averageUtilizationCpu"] = averageUtilizationCpu
-    if averageUtilizationGpu:
-        measures["averageUtilizationGpu"] = averageUtilizationGpu
-    if serverSideInference:
-        measures["serverSideInference"] = serverSideInference
-    if unit:
-        measures["unit"] = unit
-    if powerCalibrationMeasurement:
-        measures["powerCalibrationMeasurement"] = powerCalibrationMeasurement
-    if durationCalibrationMeasurement:
-        measures["durationCalibrationMeasurement"] = durationCalibrationMeasurement
-    if powerConsumption:
-        measures["powerConsumption"] = powerConsumption
-    if measurementDuration:
-        measures["measurementDuration"] = measurementDuration
-    if measurementDateTime:
-        measures["measurementDateTime"] = measurementDateTime
-    report["measures"] = [measures]
-
-    # proceed system
-    system = {}
-    if os:
-        system["os"] = os
-    if distribution:
-        system["distribution"] = distribution
-    if distributionVersion:
-        system["distributionVersion"] = distributionVersion
-    if system:
-        report["system"] = system
-
-    # proceed software
-    software = {}
-    if language:
-        software["language"] = language
-    if version_software:
-        software["version"] = version_software
-    if software:
-        report["software"] = software
-
-    # proceed infrastructure
-    infrastructure = {}
-    if infraType:
-        infrastructure["infraType"] = infraType
-    if cloudProvider:
-        infrastructure["cloudProvider"] = cloudProvider
-    if cloudInstance:
-        infrastructure["cloudInstance"] = cloudInstance
-    if components_list:
-        infrastructure["components"] = components_list
-    report["infrastructure"] = infrastructure
-
-    # proceed environment
-    environment = {}
-    if country:
-        environment["country"] = country
-    if latitude:
-        environment["latitude"] = latitude
-    if longitude:
-        environment["longitude"] = longitude
-    if location:
-        environment["location"] = location
-    if powerSupplierType:
-        environment["powerSupplierType"] = powerSupplierType
-    if powerSource:
-        environment["powerSource"] = powerSource
-    if powerSourceCarbonIntensity:
-        environment["powerSourceCarbonIntensity"] = powerSourceCarbonIntensity
-    report["environment"] = environment
-
-    # proceed quality
-    if quality:
-        report["quality"] = quality
-
-    # proceed hash
-    hash = {}
-    if hashAlgorithm:
-        hash["hashAlgorithm"] = hashAlgorithm
-    if cryptographicAlgorithm:
-        hash["cryptographicAlgorithm"] = cryptographicAlgorithm
-    if value_hash:
-        hash["value"] = value_hash
-    if hash:
-        report["$hash"] = hash
-
-    """
-    data = {
-        "header": {
-            "licensing": licensing,
-            "formatVersion": formatVersion,
-            "formatVersionSpecificationUri": formatVersionSpecificationUri,
-            "reportId": reportId,
-            "reportDatetime": reportDatetime or datetime.now().isoformat(),
-            "reportStatus": reportStatus,
-            "publisher": {
-                "name": publisher_name,
-                "division": publisher_division,
-                "projectName": publisher_projectName,
-                "confidentialityLevel": publisher_confidentialityLevel,
-                "publicKey": publisher_publicKey
-            }
-        },
-        "task": {
-            "taskType": taskType,
-            "taskFamily": taskFamily,
-            "taskStage": taskStage,
-            "algorithms": [
-                {
-                    "algorithmName": algorithmName,
-                    "framework": framework,
-                    "frameworkVersion": frameworkVersion,
-                    "classPath": classPath,
-                    "hyperparameters": {
-                        "tuning_method": tuning_method,
-                        "values": hyperparameters,
-                    },
-                    "quantization": quantization
-                }
-            ],
-            "dataset": [
-                {
-                    "dataType": dataType,
-                    "fileType": fileType,
-                    "volume": volume,
-                    "volumeUnit": volumeUnit,
-                    "items": items,
-                    "shape": [
-                        {
-                            "item": shape_item
-                        }
-                    ],
-                    "inferenceProperties": inference_props_list,
-                    "source": source,
-                    "sourceUri": sourceUri,
-                    "owner": owner
-                }
-            ],
-            "measuredAccuracy": measuredAccuracy,
-            "estimatedAccuracy": estimatedAccuracy
-        },
-        "measures": [
-            {
-                "measurementMethod": measurementMethod,
-                "manufacturer": manufacturer,
-                "version": version,
-                "cpuTrackingMode": cpuTrackingMode,
-                "gpuTrackingMode": gpuTrackingMode,
-                "averageUtilizationCpu": averageUtilizationCpu,
-                "averageUtilizationGpu": averageUtilizationGpu,
-                "serverSideInference": serverSideInference,
-                "unit": unit,
-                "powerCalibrationMeasurement": powerCalibrationMeasurement,
-                "durationCalibrationMeasurement": durationCalibrationMeasurement,
-                "powerConsumption": powerConsumption,
-                "measurementDuration": measurementDuration,
-                "measurementDateTime": measurementDateTime
-            }
-        ],
-        "system": {
-            "os": os,
-            "distribution": distribution,
-            "distributionVersion": distributionVersion
-        },
-        "software": {
-            "language": language,
-            "version": version_software
-        },
-        "infrastructure": {
-            "infraType": infraType,
-            "cloudProvider": cloudProvider,
-            "cloudInstance": cloudInstance,
-            "components": components_list
-        },
-        "environment": {
-            "country": country,
-            "latitude": latitude,
-            "longitude": longitude,
-            "location": location,
-            "powerSupplierType": powerSupplierType,
-            "powerSource": powerSource,
-            "powerSourceCarbonIntensity": powerSourceCarbonIntensity
-        },
-        "quality": quality,
-        "$hash": {
-            "hashAlgorithm": hashAlgorithm,
-            "cryptographicAlgorithm": cryptographicAlgorithm,
-            "ecryptedValue": value_hash
-        }
-    }
-    """
-
-    # Validate obligatory fields
-    is_valid, message = validate_obligatory_fields(report)
-    if not is_valid:
-        return message, None, ""
-
-    # Create the JSON string
-    json_str = json.dumps(report)
-    print(json_str)
-    # Create and save the JSON file
-    with tempfile.NamedTemporaryFile(mode='w', prefix="report", delete=False, suffix='.json') as file:
-        json.dump(report, file, indent=4)
-        return message, file.name, json_str

app.py → src/app.py

@@ -1,7 +1,8 @@
 import gradio as gr
-from services.huggingface import init_huggingface, update_dataset
-from services.json_generator import generate_json
-from ui.form_components import (
+from os import path
+from src.services.huggingface import init_huggingface, update_dataset
+from src.services.json_generator import generate_json
+from src.ui.form_components import (
     create_header_tab,
     create_task_tab,
     create_measures_tab,
@@ -12,6 +13,7 @@ from ui.form_components import (
     create_quality_tab,
     create_hash_tab
 )
+css_path = path.join(path.dirname(__file__), "../assets/styles/app.css")
 
 # Initialize Hugging Face
 init_huggingface()
@@ -31,7 +33,7 @@ def handle_submit(*inputs):
 
 
 # Create Gradio interface
-with gr.Blocks() as demo:
+with gr.Blocks(css_paths=css_path) as app:
     gr.Markdown("## Data Collection Form")
     gr.Markdown("Welcome to this Huggingface space, where you can create a report on the energy consumption of an AI task in BoAmps format, by filling in a form.")
 
@@ -69,6 +71,3 @@ with gr.Blocks() as demo:
         ],
         outputs=[output, file_output, json_output]
     )
-
-if __name__ == "__main__":
-    demo.launch()

{services → src/services}/huggingface.py

@@ -1,7 +1,6 @@
-from huggingface_hub import login
-from datasets import load_dataset, Dataset, concatenate_datasets
 import json
-from config import HF_TOKEN, DATASET_NAME
+from huggingface_hub import login
+from src.services.util import HF_TOKEN
 
 
 def init_huggingface():
@@ -44,22 +43,6 @@ def update_dataset(json_data):
 
 def create_flattened_data(data):
     """Create a flattened data structure for the dataset."""
-    # Handle hyperparameters
-    hyperparameters = data.get("task", {}).get("algorithms", [{}])[
-        0].get("hyperparameters", {}).get("values", [])
-
-    # Process hyperparameters
-    hyperparameter_names = []
-    hyperparameter_values = []
-    for hp in hyperparameters:
-        if "name" in hp and "value" in hp:  # Match the keys used in JSON
-            hyperparameter_names.append(hp["name"])
-            hyperparameter_values.append(str(hp["value"]))
-
-    hyperparameter_name_str = ", ".join(
-        hyperparameter_names) if hyperparameter_names else None
-    hyperparameter_value_str = ", ".join(
-        hyperparameter_values) if hyperparameter_values else None
 
     # Handle inference properties
     inference_props = data.get("task", {}).get(
@@ -139,16 +122,12 @@ def create_flattened_data(data):
         "publisher_publicKey": [data["header"]["publisher"]["publicKey"]],
 
         # Task
-        "taskType": [data["task"]["taskType"]],
         "taskFamily": [data["task"]["taskFamily"]],
         "taskStage": [data["task"]["taskStage"]],
         "algorithmName": [data["task"]["algorithms"][0]["algorithmName"]],
         "framework": [data["task"]["algorithms"][0]["framework"]],
         "frameworkVersion": [data["task"]["algorithms"][0]["frameworkVersion"]],
         "classPath": [data["task"]["algorithms"][0]["classPath"]],
-        "tuning_method": [data["task"]["algorithms"][0]["hyperparameters"]["tuning_method"]],
-        "hyperparameterName": [hyperparameter_name_str],
-        "hyperparameterValue": [hyperparameter_value_str],
         "quantization": [data["task"]["algorithms"][0]["quantization"]],
         "dataType": [data["task"]["dataset"][0]["dataType"]],
         "fileType": [data["task"]["dataset"][0]["fileType"]],

src/services/json_generator.py

@@ -0,0 +1,235 @@
+import json
+import tempfile
+from datetime import datetime
+from assets.utils.validation import validate_obligatory_fields
+
+
+def generate_json(
+    # Header
+    licensing, formatVersion, formatVersionSpecificationUri, reportId,
+    reportDatetime, reportStatus, publisher_name, publisher_division,
+    publisher_projectName, publisher_confidentialityLevel, publisher_publicKey,
+    # Task
+    taskFamily, taskStage, nbRequest,
+    trainingType, algorithmType, algorithmName, algorithmUri, foundationModelName, foundationModelUri, parametersNumber, framework,  frameworkVersion, classPath, layersNumber, epochsNumber, optimizer, quantization,
+    dataUsage, dataType, dataFormat, dataSize, dataQuantity, shape, source, sourceUri,  owner,
+    measuredAccuracy, estimatedAccuracy,
+    # Measures
+    measurementMethod, manufacturer, version, cpuTrackingMode, gpuTrackingMode,
+    averageUtilizationCpu, averageUtilizationGpu, powerCalibrationMeasurement,
+    durationCalibrationMeasurement, powerConsumption,
+    measurementDuration, measurementDateTime,
+    # System
+    os, distribution, distributionVersion,
+    # Software
+    language, version_software,
+    # Infrastructure
+    infraType, cloudProvider, cloudInstance, cloudService, componentName, componentType,
+    nbComponent, memorySize, manufacturer_infra, family,
+    series, share,
+    # Environment
+    country, latitude, longitude, location,
+    powerSupplierType, powerSource, powerSourceCarbonIntensity,
+    # Quality
+    quality,
+    # Hash
+    hashAlgorithm, cryptographicAlgorithm, value_hash
+):
+    """Generate JSON data from form inputs."""
+    # Process algorithms
+    algorithms_list = []
+    algorithm_fields = {"trainingType": trainingType, "algorithmType": algorithmType, "algorithmName": algorithmName, "algorithmUri": algorithmUri, "foundationModelName": foundationModelName, "foundationModelUri": foundationModelUri,
+                        "parametersNumber": parametersNumber, "framework": framework,  "frameworkVersion": frameworkVersion, "classPath": classPath, "layersNumber": layersNumber, "epochsNumber": epochsNumber, "optimizer": optimizer, "quantization": quantization}
+    nb_algo = 0
+    # ça ça marche pas
+    for f in algorithm_fields:
+        nb_algo = max(nb_algo, len(algorithm_fields[f]))
+    for i in range(nb_algo):
+        algortithm = {}
+        for f in algorithm_fields:
+            if i < len(algorithm_fields[f]) and algorithm_fields[f][i]:
+                algortithm[f] = algorithm_fields[f][i]
+        algorithms_list.append(algortithm)
+
+    # Process dataset
+    dataset_list = []
+    dataset_fields = {"dataUsage": dataUsage, "dataType": dataType, "dataFormat": dataFormat, "dataSize": dataSize,
+                      "dataQuantity": dataQuantity, "shape": shape, "source": source, "sourceUri": sourceUri, "owner": owner}
+    nb_data = 0
+    for f in dataset_fields:
+        nb_data = max(nb_data, len(dataset_fields[f]))
+    for i in range(nb_data):
+        data = {}
+        for f in dataset_fields:
+            if i < len(dataset_fields[f]) and dataset_fields[f][i]:
+                data[f] = dataset_fields[f][i]
+        dataset_list.append(data)
+
+    # Process measures
+    measures_list = []
+    measure_fields = {"measurementMethod": measurementMethod, "manufacturer": manufacturer, "version": version, "cpuTrackingMode": cpuTrackingMode,
+                      "gpuTrackingMode": gpuTrackingMode, "averageUtilizationCpu": averageUtilizationCpu, "averageUtilizationGpu": averageUtilizationGpu,
+                      "powerCalibrationMeasurement": powerCalibrationMeasurement,  "durationCalibrationMeasurement": durationCalibrationMeasurement,
+                      "powerConsumption": powerConsumption, "measurementDuration": measurementDuration, "measurementDateTime": measurementDateTime}
+    nb_measures = 0
+    for f in measure_fields:
+        nb_measures = max(nb_measures, len(measure_fields[f]))
+    for i in range(nb_measures):
+        measure = {}
+        for f in measure_fields:
+            if i < len(measure_fields[f]) and measure_fields[f][i]:
+                measure[f] = measure_fields[f][i]
+        measures_list.append(measure)
+
+    # Process components
+    components_list = []
+    component_fields = {"componentName": componentName, "componentType": componentType, "nbComponent": nbComponent,
+                        "memorySize": memorySize, "manufacturer_infra": manufacturer_infra, "family": family,
+                        "series": series, "share": share}
+    nb_components = 0
+    for f in component_fields:
+        nb_components = max(nb_components, len(component_fields[f]))
+    for i in range(nb_components):
+        component = {}
+        for f in component_fields:
+            if i < len(component_fields[f]) and component_fields[f][i]:
+                component[f] = component_fields[f][i]
+        components_list.append(component)
+
+    # process report
+    report = {}
+
+    # Process header
+    header = {}
+    if licensing:
+        header["licensing"] = licensing
+    if formatVersion:
+        header["formatVersion"] = formatVersion
+    if formatVersionSpecificationUri:
+        header["formatVersionSpecificationUri"] = formatVersionSpecificationUri
+    if reportId:
+        header["reportId"] = reportId
+    if reportDatetime:
+        header["reportDatetime"] = reportDatetime or datetime.now().isoformat()
+    if reportStatus:
+        header["reportStatus"] = reportStatus
+
+    publisher = {}
+    if publisher_name:
+        publisher["name"] = publisher_name
+    if publisher_division:
+        publisher["division"] = publisher_division
+    if publisher_projectName:
+        publisher["projectName"] = publisher_projectName
+    if publisher_confidentialityLevel:
+        publisher["confidentialityLevel"] = publisher_confidentialityLevel
+    if publisher_publicKey:
+        publisher["publicKey"] = publisher_publicKey
+
+    if publisher:
+        header["publisher"] = publisher
+
+    if header:
+        report["header"] = header
+
+    # proceed task
+    task = {}
+    if taskStage:
+        task["taskStage"] = taskStage
+    if taskFamily:
+        task["taskFamily"] = taskFamily
+    if nbRequest:
+        task["nbRequest"] = nbRequest
+    if algorithms_list:
+        task["algorithms"] = algorithms_list
+    if dataset_list:
+        task["dataset"] = dataset_list
+    if measuredAccuracy:
+        task["measuredAccuracy"] = measuredAccuracy
+    if estimatedAccuracy:
+        task["estimatedAccuracy"] = estimatedAccuracy
+    report["task"] = task
+
+    # proceed measures
+    if measures_list:
+        report["measures"] = measures_list
+
+    # proceed system
+    system = {}
+    if os:
+        system["os"] = os
+    if distribution:
+        system["distribution"] = distribution
+    if distributionVersion:
+        system["distributionVersion"] = distributionVersion
+    if system:
+        report["system"] = system
+
+    # proceed software
+    software = {}
+    if language:
+        software["language"] = language
+    if version_software:
+        software["version"] = version_software
+    if software:
+        report["software"] = software
+
+    # proceed infrastructure
+    infrastructure = {}
+    if infraType:
+        infrastructure["infraType"] = infraType
+    if cloudProvider:
+        infrastructure["cloudProvider"] = cloudProvider
+    if cloudInstance:
+        infrastructure["cloudInstance"] = cloudInstance
+    if cloudService:
+        infrastructure["cloudService"] = cloudService
+    if components_list:
+        infrastructure["components"] = components_list
+    report["infrastructure"] = infrastructure
+
+    # proceed environment
+    environment = {}
+    if country:
+        environment["country"] = country
+    if latitude:
+        environment["latitude"] = latitude
+    if longitude:
+        environment["longitude"] = longitude
+    if location:
+        environment["location"] = location
+    if powerSupplierType:
+        environment["powerSupplierType"] = powerSupplierType
+    if powerSource:
+        environment["powerSource"] = powerSource
+    if powerSourceCarbonIntensity:
+        environment["powerSourceCarbonIntensity"] = powerSourceCarbonIntensity
+    if environment:
+        report["environment"] = environment
+
+    # proceed quality
+    if quality:
+        report["quality"] = quality
+
+    # proceed hash
+    hash = {}
+    if hashAlgorithm:
+        hash["hashAlgorithm"] = hashAlgorithm
+    if cryptographicAlgorithm:
+        hash["cryptographicAlgorithm"] = cryptographicAlgorithm
+    if value_hash:
+        hash["value_hash"] = value_hash
+    if hash:
+        report["hash"] = hash
+
+    # Validate obligatory fields
+    is_valid, message = validate_obligatory_fields(report)
+    if not is_valid:
+        return message, None, ""
+        # Create the JSON string
+    json_str = json.dumps(report)
+    print(json_str)
+    # Create and save the JSON file
+    with tempfile.NamedTemporaryFile(mode='w', prefix="report", delete=False, suffix='.json') as file:
+        json.dump(report, file, indent=4)
+        return message, file.name, json_str

config.py → src/services/util.py

@@ -5,30 +5,31 @@ HF_TOKEN = os.environ.get("HF_TOKEN")
 DATASET_NAME = "soprasteria/BoAmps_leaderboard"
 
 # Form Field Configurations
-MANDATORY_SECTIONS = ["task", "measures", "infrastructure", "environment"]
+# not used and verified for now
+MANDATORY_SECTIONS = ["task", "measures", "infrastructure"]
 OBLIGATORY_FIELDS = [
-    "formatVersion", "reportId", "reportStatus", "confidentialityLevel",
-    "taskType", "taskFamily", "taskStage", "algorithmName", "dataType",
-    "volume", "volumeUnit", "nbRequest", "measurementMethod", "unit",
-    "powerConsumption", "os", "language", "infraType", "componentType",
-    "nbComponent", "country", "hashAlgorithm", "cryptographicAlgorithm", "value"
+    "taskStage", "taskFamily", "dataUsage", "dataType",
+    "measurementMethod", "powerConsumption", "infraType", "componentType",
+    "nbComponent"
 ]
 
 # Dropdown Options
-REPORT_STATUS_OPTIONS = ["draft", "final", "corrective", "$other"]
+REPORT_STATUS_OPTIONS = ["draft", "final", "corrective", "other"]
 CONFIDENTIALITY_LEVELS = ["public", "internal", "confidential", "secret"]
-DATA_TYPES = ["tabular", "audio", "boolean",
-              "image", "video", "object", "text", "$other"]
-DATA_UNITS = ['kilobyte', 'megabyte', 'gigabyte', 'terabyte',
-              'petabyte', 'exabyte', 'zettabyte', 'yottabyte']
+DATA_USAGE_OPTIONS = ["input", "output"]
+DATA_FORMAT = ["3gp", "3gpp", "3gpp2", "8svx", "aa", "aac", "aax", "act", "afdesign", "afphoto", "ai", "aiff", "alac", "amr", "amv", "ape", "arrow", "asf", "au", "avi", "avif", "awb", "bmp", "bpg", "cd5", "cda", "cdr", "cgm", "clip", "cpt", "csv", "deep", "dirac", "divx", "drawingml", "drw", "dss", "dvf", "ecw", "eps", "fits", "flac", "flif", "flv", "flvf4v", "gem", "gerber", "gif", "gle", "gsm", "heif", "hp-gl", "html", "hvif", "ico", "iklax", "ilbm", "img", "ivs", "jpeg", "json", "kra", "lottie", "m4a", "m4b", "m4p", "m4v", "mathml", "matroska", "mdp", "mmf", "movpkg", "mp3", "mpc", "mpeg1",
+               "mpeg2", "mpeg4", "msv", "mxf", "naplps", "netpbm", "nmf", "nrrd", "nsv", "odg", "ods", "ogg", "opus", "pam", "parquet", "pbm", "pcx", "pdf", "pdn", "pgf", "pgm", "pgml", "pict", "plbm", "png", "pnm", "postscript", "ppm", "psd", "psp", "pstricks", "qcc", "quicktime", "ra", "raw", "realmedia", "regis", "rf64", "roq", "sai", "sgi", "sid", "sql", "sln", "svg", "svi", "swf", "text", "tga", "tiff", "tinyvg", "tta", "vicar", "vivoactive", "vml", "vob", "voc", "vox", "wav", "webm", "webp", "wma", "wmf", "wmv", "wv", "xaml", "xar", "xcf", "xisf", "xls", "xlsx", "xml", "xps", "yaml", "other"]
+DATA_TYPES = ["tabular", "audio", "boolean", "image",
+              "video", "object", "text", "token", "word", "other"]
+DATA_SOURCE = ["public", "private", "other"]
 ACCURACY_LEVELS = ["veryPoor", "poor", "average", "good", "veryGood"]
 MEASUREMENT_UNITS = ["Wh", "kWh", "MWh", "GWh", "kJoule", "MJoule", "GJoule", "TJoule", "PJoule",
                      "BTU", "kiloFLOPS", "megaFLOPS", "gigaFLOPS", "teraFLOPS", "petaFLOPS",
                      "exaFLOPS", "zettaFLOPS", "yottaFLOPS"]
-INFRA_TYPES = ["publicCloud", "privateCloud", "onPremise", "$other"]
-POWER_SUPPLIER_TYPES = ["public", "private", "internal", "$other"]
+INFRA_TYPES = ["publicCloud", "privateCloud", "onPremise", "other"]
+POWER_SUPPLIER_TYPES = ["public", "private", "internal", "other"]
 POWER_SOURCES = ["solar", "wind", "nuclear",
-                 "hydroelectric", "gas", "coal", "$other"]
+                 "hydroelectric", "gas", "coal", "other"]
 QUALITY_LEVELS = ["high", "medium", "low"]
 HASH_ALGORITHMS = ["MD5", "RIPEMD-128", "RIPEMD-160", "RIPEMD-256", "RIPEMD-320",
                    "SHA-1", "SHA-224", "SHA256", "SHA-384", "SHA-512"]

{ui → src/ui}/form_components.py

@@ -1,9 +1,10 @@
 import gradio as gr
-from config import (
-    REPORT_STATUS_OPTIONS, CONFIDENTIALITY_LEVELS, DATA_TYPES,
-    DATA_UNITS, ACCURACY_LEVELS, MEASUREMENT_UNITS, INFRA_TYPES,
+from src.services.util import (
+    REPORT_STATUS_OPTIONS, CONFIDENTIALITY_LEVELS, DATA_USAGE_OPTIONS, DATA_FORMAT,
+    DATA_TYPES, DATA_SOURCE,
+    ACCURACY_LEVELS, INFRA_TYPES,
     POWER_SUPPLIER_TYPES, POWER_SOURCES, QUALITY_LEVELS,
-    HASH_ALGORITHMS, CRYPTO_ALGORITHMS, CACHE_OPTIONS
+    HASH_ALGORITHMS, CRYPTO_ALGORITHMS
 )
 
 
@@ -42,9 +43,11 @@ def create_dynamic_section(section_name, fields_config, initial_count=1, layout=
 
                 for field_idx, config in enumerate(fields_config):
                     component = config["type"](
-                        label=f"{config['label']} {i + 1}",
+                        label=f"{config['label']} ({section_name}{i + 1})",
                         info=config.get("info", ""),
-                        **config.get("kwargs", {})
+                        value=config.get("value", ""),
+                        **config.get("kwargs", {}),
+                        elem_classes=config.get("elem_classes", "")
                     )
                     row_components.append(component)
                     field_refs.append(component)
@@ -86,17 +89,17 @@ def create_header_tab():
         licensing = gr.Textbox(
             label="Licensing", info="(the type of licensing applicable for the sharing of the report)")
         formatVersion = gr.Textbox(
-            label="Format Version", info="Required field<br>(the version of the specification of this set of schemas defining the report's fields)")
+            label="Format Version", info="(the version of the specification of this set of schemas defining the report's fields)")
         formatVersionSpecificationUri = gr.Textbox(
             label="Format Version Specification URI", info="(the URI of the present specification of this set of schemas)")
         reportId = gr.Textbox(
-            label="Report ID", info="Required field<br>(the unique identifier of this report, preferably as a uuid4 string)")
+            label="Report ID", info="(the unique identifier of this report, preferably as a uuid4 string)")
         reportDatetime = gr.Textbox(
-            label="Report Datetime", info="(the publishing date of this report in format YYYY-MM-DD HH:MM:SS)")
+            label="Report Datetime", info="(Required field<br>the publishing date of this report in format YYYY-MM-DD HH:MM:SS)", elem_classes="mandatory_field")
         reportStatus = gr.Dropdown(value=None,
                                    label="Report Status",
                                    choices=REPORT_STATUS_OPTIONS,
-                                   info="Required field<br>(the status of this report)"
+                                   info="(the status of this report)"
                                    )
 
         with gr.Accordion("Publisher"):
@@ -109,7 +112,8 @@ def create_header_tab():
             publisher_confidentialityLevel = gr.Dropdown(value=None,
                                                          label="Confidentiality Level",
                                                          choices=CONFIDENTIALITY_LEVELS,
-                                                         info="Required field<br>(the confidentiality of the report)"
+                                                         info="Required field<br>(the confidentiality of the report)",
+                                                         elem_classes="mandatory_field"
                                                          )
             publisher_publicKey = gr.Textbox(
                 label="Public Key", info="(the cryptographic public key to check the identity of the publishing organization)")
@@ -123,125 +127,157 @@ def create_header_tab():
 
 def create_task_tab():
     """Create the task tab components."""
-    with gr.Tab("Task"):
-        taskType = gr.Textbox(
-            label="Task Type", info="Required field<br>(type of the computing task of machine learning, example : datacreation, preprocessing, supervisedLearning, unsupervisedLearning, semiSupervisedLearning ...)")
-        taskFamily = gr.Textbox(
-            label="Task Family", info="Required field<br>(the family of task performed, example : classification, regression, chatbot, summarization, keyword extraction, image recognition...)")
+    with gr.Tab("Task", elem_id="mandatory_part"):
         taskStage = gr.Textbox(
-            label="Task Stage", info="Required field<br>(stage of the task, example: training, finetuning, reinforcement, inference, rag...)")
+            label="Task Stage", info="Required field<br>(stage of the task, example: datacreation, preprocessing, training, finetuning, inference, retraining..., add a + between stages if several but we do recommand to measure each step independantly)", elem_classes="mandatory_field")
+        taskFamily = gr.Textbox(
+            label="Task Family", info="Required field<br>(the family of task you are running, e.g. text classification, image generation, speech recognition, robotics navigation...)", elem_classes="mandatory_field")
+        nbRequest = gr.Number(
+            label="Number of Requests", info="(if inference stage, the number of requests the measure corresponds to, 0 or empty if you're not measuring the inference stage)",
+            value=lambda: None, minimum=0)
 
         with gr.Accordion("Algorithms"):
-            algorithmName = gr.Textbox(
-                label="Algorithm Name", info="Required field<br>(the case-sensitive common name of the algorithm, example: randomForest, svm, xgboost...)")
-            framework = gr.Textbox(
-                label="Framework", info="(the common name of the software framework implementing the algorithm)")
-            frameworkVersion = gr.Textbox(
-                label="Framework Version", info="(the version of the software framework)")
-            classPath = gr.Textbox(
-                label="Class Path", info="(the full class path of the algorithm within the framework)")
-            tuning_method = gr.Textbox(
-                label="Tuning Method", info="(the method of hyperparameters tuning used (if any), example: gridSearch, randomizedSearch...)")
-
-            with gr.Accordion("Hyperparameters"):
-                _, hyperparameter_names, hyperparameter_values, add_btn = create_dynamic_section(
-                    section_name="Hyperparameter",
-                    fields_config=[
-                        {
-                            "type": gr.Textbox,
-                            "label": "Hyperparameter Name",
-                            "info": "(name of the hyperparameter)",
-                            "kwargs": {"interactive": True}
-                        },
-                        {
-                            "type": gr.Textbox,
-                            "label": "Hyperparameter Value",
-                            "info": "(value of the hyperparameter)",
-                            "kwargs": {"placeholder": "Enter value..."}
-                        }
-                    ],
-                    initial_count=0,
-                )
-
-            quantization = gr.Textbox(
-                label="Quantization", info="(the data weights (in bits) obtained thanks to the quantization, example: 2, 8, 16...)")
+            _, trainingType, algorithmType, algorithmName, algorithmUri, foundationModelName, foundationModelUri, parametersNumber, framework,  frameworkVersion, classPath, layersNumber, epochsNumber, optimizer, quantization, add_algorithm_btn = create_dynamic_section(
+                section_name="Algorithms",
+                fields_config=[
+                    {
+                        "type": gr.Textbox,
+                        "label": "Type of training",
+                        "info": "(if applicable, type of training (if the stage corresponds to a training) : supervisedLearning, unsupervisedLearning, semiSupervisedLearning, reinforcementLearning, transferLearning ...)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Type of algorithm",
+                        "info": "(the type of algorithm used, example : embeddings creation, rag, nlp, neural network, llm...)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Algorithm Name",
+                        "info": "(the case-sensitive common name of the algorithm, example: randomForest, svm, xgboost...)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Algorithm Uri",
+                        "info": "(the URI of the model, if publicly available)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Foundation Model Name",
+                        "info": "(if a foundation model is used, its case-sensitive common name, example: llama3.1-8b, gpt4-o...)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Foundation Model Uri",
+                        "info": "(the URI of the foundation model, if publicly available)",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Number of parameters",
+                        "info": "(if applicable, number of billions of total parameters of your model, e.g. 8 for llama3.1-8b)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Framework",
+                        "info": "(the common name of the software framework implementing the algorithm, if any)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "frameworkVersion",
+                        "info": "(the version of the software framework implementing the algorithm, if any)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "classPath",
+                        "info": "(the full class path of the algorithm within the framework, with elements separated by dots)",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Number of layers in the network",
+                        "info": "(if deep learning, precise the number of layers in your network)",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Number of epochs",
+                        "info": "(if training, the number of complete passes through the training dataset)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "optimizer",
+                        "info": "(the algorithm used to optimize the models weights, e.g. gridSearch, lora, adam)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "quantization",
+                        "info": "(the type of quantization used : fp32, fp16, b16, int8 ...)",
+                    }
+                ],
+                initial_count=0,
+                layout="column"
+            )
 
         with gr.Accordion("Dataset"):
-            dataType = gr.Dropdown(value=None,
-                                   label="Data Type",
-                                   choices=DATA_TYPES,
-                                   info="Required field<br>(the nature of the data)"
-                                   )
-            fileType = gr.Textbox(
-                label="File Type", info="(the file type of the dataset)")
-            volume = gr.Number(value=lambda: None,
-                               label="Volume", info="Required field<br>(the size of the dataset)")
-            volumeUnit = gr.Dropdown(value=None,
-                                     label="Volume Unit",
-                                     choices=DATA_UNITS,
-                                     info="Required field<br>(the unit of the size)")
-            items = gr.Number(value=lambda: None,
-                              label="Items", info="(the number of items in the dataset)")
-            shape_item = gr.Number(value=lambda: None,
-                                   label="Shape Item", info="(the shape of each dataset item)")
-
-            with gr.Accordion("Inference Properties"):
-                _, nbRequest, nbTokensInput, nbWordsInput,  nbTokensOutput, nbWordsOutput, contextWindowSize, cache, add_inference_btn = create_dynamic_section(
-                    section_name="Inference Property",
-                    fields_config=[
-                        {
-                            "type": gr.Number,
-                            "value": lambda: None,
-                            "label": "Number of Requests",
-                            "info": "Required field<br>(the number of requests the measure corresponds to)",
-                        },
-                        {
-                            "type": gr.Number,
-                            "value": lambda: None,
-                            "label": "Number of Tokens Input",
-                            "info": "(the number of tokens in the input)",
-                        },
-                        {
-                            "type": gr.Number,
-                            "value": lambda: None,
-                            "label": "Number of Words Input",
-                            "info": "(the number of words in the input)",
-                        },
-                        {
-                            "type": gr.Number,
-                            "value": lambda: None,
-                            "label": "Number of Tokens Output",
-                            "info": "(the number of tokens in the output)",
-                        },
-                        {
-                            "type": gr.Number,
-                            "value": lambda: None,
-                            "label": "Number of Words Output",
-                            "info": "(the number of words in the output)",
-                        },
-                        {
-                            "type": gr.Number,
-                            "value": lambda: None,
-                            "label": "Context Window Size",
-                            "info": "(the number of tokens kept in memory)",
-                        },
-                        {
-                            "type": gr.Dropdown,
-                            "label": "Cache",
-                            "info": "(the presence of a cache function)",
-                            "kwargs": {"choices": CACHE_OPTIONS, "value": None}
-                        }
-                    ],
-                    initial_count=0,
-                    layout="column"
-                )
-
-            source = gr.Textbox(
-                label="Source", info="(the kind of source of the dataset)")
-            sourceUri = gr.Textbox(
-                label="Source URI", info="(the URI of the dataset)")
-            owner = gr.Textbox(
-                label="Owner", info="(the owner of the dataset)")
+            _, dataUsage, dataType, dataFormat, dataSize, dataQuantity, shape, source, sourceUri,  owner, add_dataset_btn = create_dynamic_section(
+                section_name="Dataset",
+                fields_config=[
+                    {
+                        "type": gr.Dropdown,
+                        "label": "Data Usage",
+                        "info": "Required field<br>(the use of the dataset: is it used as model input or output ?)",
+                        "value": None,
+                        "kwargs": {"choices": DATA_USAGE_OPTIONS},
+                        "elem_classes": "mandatory_field",
+                    },
+                    {
+                        "type": gr.Dropdown,
+                        "label": "Data Type",
+                        "info": "Required field<br>(the nature of the data used)",
+                        "value": None,
+                        "kwargs": {"choices": DATA_TYPES},
+                        "elem_classes": "mandatory_field",
+                    },
+                    {
+                        "type": gr.Dropdown,
+                        "label": "Data Format",
+                        "info": "(if the data is passed in the form of a file, what format is the data in?)",
+                        "value": None,
+                        "kwargs": {"choices": DATA_FORMAT}
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Data Size",
+                        "info": "(the size of the dataset (in Go), if small quantity just fill the field quantity)",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Data Quantity",
+                        "info": "(the number of data in the dataset, e.g. 3 (images, audio or tokens))",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Data shape",
+                        "info": "(the shape of your dataset, can be found with X.shape with dataframes, e.g. (12, 1000) for a 2D table with 12 columns and 1000 rows)",
+                    },
+                    {
+                        "type": gr.Dropdown,
+                        "label": "Data source",
+                        "info": "(the kind of source of the dataset)",
+                        "value": None,
+                        "kwargs": {"choices": DATA_SOURCE}
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Source Uri",
+                        "info": "(the URI of the dataset if available)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Owner",
+                        "info": "(the owner of the dataset if available)",
+                    }
+                ],
+                initial_count=0,
+                layout="column"
+            )
 
         with gr.Row():
             measuredAccuracy = gr.Number(value=lambda: None,
@@ -252,65 +288,105 @@ def create_task_tab():
                                             info="(estimated accuracy assessment)"
                                             )
 
-        return [
-            taskType, taskFamily, taskStage, algorithmName, framework,
-            frameworkVersion, classPath, tuning_method, hyperparameter_names, hyperparameter_values,
-            quantization, dataType, fileType, volume,
-            volumeUnit, items, shape_item, nbRequest, nbTokensInput,
-            nbWordsInput, nbTokensOutput, nbWordsOutput, contextWindowSize,
-            cache, source, sourceUri, owner, measuredAccuracy, estimatedAccuracy
-        ]
+    return [
+        taskFamily, taskStage, nbRequest,
+        trainingType, algorithmType, algorithmName, algorithmUri, foundationModelName, foundationModelUri, parametersNumber, framework,  frameworkVersion, classPath, layersNumber, epochsNumber, optimizer, quantization,
+        dataUsage, dataType, dataFormat, dataSize, dataQuantity, shape, source, sourceUri,  owner,
+        measuredAccuracy, estimatedAccuracy
+    ]
 
 
 def create_measures_tab():
     """Create the measures tab components."""
-    with gr.Tab("Measures"):
-        measurementMethod = gr.Textbox(
-            label="Measurement Method", info="Required field<br>(the method used to perform the energy or FLOPS measure)")
-        manufacturer = gr.Textbox(
-            label="Manufacturer", info="(the builder of the measuring tool)")
-        version = gr.Textbox(
-            label="Version", info="(the version of the measuring tool)")
-        cpuTrackingMode = gr.Textbox(
-            label="CPU Tracking Mode", info="(the method used to track CPU consumption)")
-        gpuTrackingMode = gr.Textbox(
-            label="GPU Tracking Mode", info="(the method used to track GPU consumption)")
-        averageUtilizationCpu = gr.Number(value=lambda: None,
-                                          label="Average Utilization CPU", info="(the average percentage of CPU use)")
-        averageUtilizationGpu = gr.Number(value=lambda: None,
-                                          label="Average Utilization GPU", info="(the average percentage of GPU use)")
-        serverSideInference = gr.Textbox(
-            label="Server Side Inference", info="(inference server consumption estimation)")
-        unit = gr.Dropdown(value=None,
-                           label="Unit",
-                           choices=MEASUREMENT_UNITS,
-                           info="Required field<br>(the unit of power consumption measure)"
-                           )
-        powerCalibrationMeasurement = gr.Number(value=lambda: None,
-                                                label="Power Calibration Measurement", info="(power consumed during calibration)")
-        durationCalibrationMeasurement = gr.Number(value=lambda: None,
-                                                   label="Duration Calibration Measurement", info="(duration of calibration in seconds)")
-        powerConsumption = gr.Number(value=lambda: None,
-                                     label="Power Consumption", info="Required field<br>(the power consumption measure)")
-        measurementDuration = gr.Number(value=lambda: None,
-                                        label="Measurement Duration", info="(the duration of measurement in seconds)")
-        measurementDateTime = gr.Textbox(
-            label="Measurement DateTime", info="(when measurement began)")
+    with gr.Tab("Measures", elem_id="mandatory_part"):
+        with gr.Accordion("Measures"):
+            _, measurementMethod, manufacturer, version, cpuTrackingMode, gpuTrackingMode, averageUtilizationCpu, averageUtilizationGpu, powerCalibrationMeasurement,  durationCalibrationMeasurement, powerConsumption, measurementDuration, measurementDateTime, add_measurement_btn = create_dynamic_section(
+                section_name="Measures",
+                fields_config=[
+                    {
+                        "type": gr.Textbox,
+                        "label": "Method of measurement",
+                        "info": "Required field<br>(the energy measure obtained from software and/or hardware tools, for a computing task)",
+                        "elem_classes": "mandatory_field",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Manufacturer",
+                        "info": "(the builder of the measuring tool, if the measurement method is wattmeter)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Version of the measurement tool",
+                        "info": "(the version of the measuring tool, if any)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "CPU tracking mode",
+                        "info": "(the method used to track the consumption of the CPU, example: constant, rapl...)",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "GPU tracking mode",
+                        "info": "(the method used to track the consumption of the GPU, example: constant, nvml...)",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Average CPU Utilization",
+                        "info": "(the average percentage of use of the CPU during the task, for example: 0.5 if your CPU load was 50% on average)",
+                        "minimum": 0,
+                        "maximum": 1
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Average GPU Utilization",
+                        "info": "(the average percentage of use of the GPU during the task, for example: 0.8 if your GPU load was 80% on average)",
+                        "minimum": 0,
+                        "maximum": 1
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Power calibration measurement",
+                        "info": "(the power consumed (in kWh) during the calibration measure if any (to isolate the initial consumption of the hardware))",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Duration calibration measurement",
+                        "info": "(the duration of the calibration if any (in seconds))",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Power consumption",
+                        "info": "Required field<br>(the power consumption measure of the computing task (in kWh))",
+                        "elem_classes": "mandatory_field",
+                    },
+                    {
+                        "type": gr.Number,
+                        "label": "Measurement Duration",
+                        "info": "(the duration of the measurement (in seconds))",
+                    },
+                    {
+                        "type": gr.Textbox,
+                        "label": "Measurement date time",
+                        "info": "(the date when the measurement began, in format YYYY-MM-DD HH:MM:SS)",
+                    }
+                ],
+                initial_count=0,
+                layout="column"
+            )
 
-        return [
-            measurementMethod, manufacturer, version, cpuTrackingMode,
-            gpuTrackingMode, averageUtilizationCpu, averageUtilizationGpu,
-            serverSideInference, unit, powerCalibrationMeasurement,
-            durationCalibrationMeasurement, powerConsumption,
-            measurementDuration, measurementDateTime
-        ]
+    return [
+        measurementMethod, manufacturer, version, cpuTrackingMode, gpuTrackingMode,
+        averageUtilizationCpu, averageUtilizationGpu, powerCalibrationMeasurement,
+        durationCalibrationMeasurement, powerConsumption,
+        measurementDuration, measurementDateTime
+    ]
 
 
 def create_system_tab():
     """Create the system tab components."""
     with gr.Tab("System"):
         os = gr.Textbox(
-            label="OS", info="Required field<br>(name of the operating system)")
+            label="OS", info="Required field<br>(name of the operating system)", elem_classes="mandatory_field")
         distribution = gr.Textbox(
             label="Distribution", info="(distribution of the operating system)")
         distributionVersion = gr.Textbox(
@@ -323,7 +399,7 @@ def create_software_tab():
     """Create the software tab components."""
     with gr.Tab("Software"):
         language = gr.Textbox(
-            label="Language", info="Required field<br>(programming language information)")
+            label="Language", info="Required field<br>(programming language information)", elem_classes="mandatory_field")
         version_software = gr.Textbox(
             label="Version", info="(version of the programming language)")
 
@@ -332,20 +408,24 @@ def create_software_tab():
 
 def create_infrastructure_tab():
     """Create the infrastructure tab components."""
-    with gr.Tab("Infrastructure"):
+    with gr.Tab("Infrastructure", elem_id="mandatory_part"):
         infraType = gr.Dropdown(value=None,
                                 label="Infrastructure Type",
                                 choices=INFRA_TYPES,
-                                info="Required field<br>(the type of infrastructure used)"
+                                info="Required field<br>(the type of infrastructure used)",
+                                elem_classes="mandatory_field"
                                 )
         cloudProvider = gr.Textbox(
-            label="Cloud Provider", info="(name of your cloud provider)")
+            label="Cloud Provider", info="(If you are on the cloud, the name of your cloud provider, for example : aws, azure, google, ovh...)")
         cloudInstance = gr.Textbox(
-            label="Cloud Instance", info="(name of your cloud instance)")
+            label="Cloud Instance", info="(If you are on a cloud vm, the name of your cloud instance, for example : a1.large, dasv4-type2...)")
+        cloudService = gr.Textbox(
+            label="Cloud Service", info="(If you are using an AI cloud service, the name of your cloud service, for example : openAI service...)")
         with gr.Accordion("Components"):
             _, componentName, componentType, nbComponent, memorySize, manufacturer_infra, family, series, share, add_component_btn = create_dynamic_section(
                 section_name="Component",
                 fields_config=[
+
                     {
                         "type": gr.Textbox,
                         "label": "Component Name",
@@ -355,39 +435,38 @@ def create_infrastructure_tab():
                         "type": gr.Textbox,
                         "label": "Component Type",
                         "info": "Required field<br>(the type of this subsystem part of your infrastructure, example: cpu, gpu, ram, hdd, sdd...)",
+                        "elem_classes": "mandatory_field",
                     },
                     {
                         "type": gr.Number,
-                        "value": lambda: None,
                         "label": "Number of Components",
-                        "info": "Required field<br>(number of items of this component)",
+                        "info": "Required field<br>(the number of items of this component in your infrastructure, if you have 1 RAM of 32Go, fill 1 here and 32 inside memorySize)",
+                        "elem_classes": "mandatory_field",
                     },
                     {
                         "type": gr.Number,
-                        "value": lambda: None,
                         "label": "Memory Size",
-                        "info": "(size of memory in Gbytes)",
+                        "info": "(the size of the memory of the component in Gbytes, useful to detail the memory associated to ONE of your gpus for example (if we want the total memory, we will multiply the memorySize by nbComponent). If the component is CPU do not fill the RAM size here, create another component for RAM, this field is for the embeded memory of a component.)",
                     },
                     {
                         "type": gr.Textbox,
                         "label": "Manufacturer",
-                        "info": "(name of the manufacturer)",
+                        "info": "(the name of the manufacturer, example: nvidia)",
                     },
                     {
                         "type": gr.Textbox,
                         "label": "Family",
-                        "info": "(family of this component)",
+                        "info": "(the family of this component, example: geforce)",
                     },
                     {
                         "type": gr.Textbox,
                         "label": "Series",
-                        "info": "(series of this component)",
+                        "info": "(the series of this component, example: gtx1080)",
                     },
                     {
                         "type": gr.Number,
-                        "value": lambda: None,
                         "label": "Share",
-                        "info": "(percentage of equipment used)",
+                        "info": "(the percentage of the physical equipment used by the task, this sharing property should be set to 1 by default (if no share) and otherwise to the correct percentage, e.g. 0.5 if you share half-time.)",
                     }
                 ],
                 initial_count=0,
@@ -395,7 +474,7 @@ def create_infrastructure_tab():
             )
 
         return [
-            infraType, cloudProvider, cloudInstance, componentName, componentType,
+            infraType, cloudProvider, cloudInstance, cloudService, componentName, componentType,
             nbComponent, memorySize, manufacturer_infra, family,
             series, share
         ]
@@ -404,10 +483,12 @@ def create_infrastructure_tab():
 def create_environment_tab():
     """Create the environment tab components."""
     with gr.Tab("Environment"):
-        country = gr.Textbox(label="Country", info="Required field")
+        country = gr.Textbox(
+            label="Country", info="Required field", elem_classes="mandatory_field")
         latitude = gr.Number(label="Latitude", value=lambda: None)
         longitude = gr.Number(label="Longitude", value=lambda: None)
-        location = gr.Textbox(label="Location")
+        location = gr.Textbox(
+            label="Location", info="(more precise location like city, region or datacenter name)")
         powerSupplierType = gr.Dropdown(value=lambda: None,
                                         label="Power Supplier Type",
                                         choices=POWER_SUPPLIER_TYPES,
@@ -445,14 +526,16 @@ def create_hash_tab():
         hashAlgorithm = gr.Dropdown(value=None,
                                     label="Hash Algorithm",
                                     choices=HASH_ALGORITHMS,
-                                    info="Required field<br>(the hash function to apply)"
+                                    info="Required field<br>(the hash function to apply)",
+                                    elem_classes="mandatory_field"
                                     )
         cryptographicAlgorithm = gr.Dropdown(value=None,
                                              label="Cryptographic Algorithm",
                                              choices=CRYPTO_ALGORITHMS,
-                                             info="Required field<br>(the public key function to apply)"
+                                             info="Required field<br>(the public key function to apply)",
+                                             elem_classes="mandatory_field"
                                              )
         value_hash = gr.Textbox(
-            label="Value", info="Required field<br>(encrypted value of the hash)")
+            label="Value", info="Required field<br>(encrypted value of the hash)", elem_classes="mandatory_field")
 
         return [hashAlgorithm, cryptographicAlgorithm, value_hash]