Mesurer l'empreinte carbone des modèles d'IA avec la méthode Ecologits.ai
Ceci est unAI Summarization, Multimodal AIworkflow d'automatisation du domainecontenant 7 nœuds.Utilise principalement des nœuds comme Set, ManualTrigger, ChainLlm, LmChatOpenAi. Mesurer l'empreinte carbone des modèles d'IA avec la méthode Ecologits.ai
- •Clé API OpenAI
Nœuds utilisés (7)
Catégorie
{
"nodes": [
{
"id": "e374f8b8-ff4a-4b98-af50-d609338ec38f",
"name": "Lors du clic sur 'Exécuter le workflow'",
"type": "n8n-nodes-base.manualTrigger",
"position": [
0,
-160
],
"parameters": {},
"typeVersion": 1
},
{
"id": "6cce6b66-bd1a-419b-86c1-b76aa257e96c",
"name": "Chaîne LLM de base",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
608,
-160
],
"parameters": {
"text": "Enter here your user prompt",
"batching": {},
"messages": {
"messageValues": [
{
"message": "Enter here the system prompt"
}
]
},
"promptType": "define"
},
"typeVersion": 1.7
},
{
"id": "d0ea8139-307d-4de6-9f29-11216958f362",
"name": "OpenAI Chat Model",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
672,
64
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4o",
"cachedResultName": "gpt-4o"
},
"options": {}
},
"credentials": {
"openAiApi": {
"id": "dMiSy27YCK6c6rra",
"name": "Duv's OpenAI"
}
},
"typeVersion": 1.2
},
{
"id": "0fca2f27-8a0b-46d0-9dfc-27967afe2ae5",
"name": "Calculer gCO₂e",
"type": "n8n-nodes-base.set",
"position": [
960,
-160
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "cc17f2be-ce12-488f-89c7-de200b4c4869",
"name": "AI output",
"type": "string",
"value": "={{ $json.text }}"
},
{
"id": "c396e3b8-f07f-4153-9892-1b499a724dbc",
"name": "AI output gCO₂e",
"type": "number",
"value": "={{ Math.ceil($json.text.length / 4) * $('Conversion factor').item.json['Conversion factor (in gCO₂e/token)'] }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "5c25ded0-c24d-455b-82fb-d54d267ca591",
"name": "Note adhésive",
"type": "n8n-nodes-base.stickyNote",
"position": [
-624,
-384
],
"parameters": {
"width": 560,
"height": 672,
"content": "# Measure Your AI's Carbon Footprint\n\nThis workflow demonstrates a technique to calculate the gCO₂e (grams of CO₂ equivalent) of an AI model's output, based on the methodology from **Ecologits.ai**.\n\n## How it works\n\nA dedicated **Conversion factor** node makes it easy to set your parameters. The **Calculate gCO₂e** node then uses this factor and the AI's text output to estimate the carbon footprint.\n\n## How to use this snippet\n\n1. **Set your conversion factor (Important!):** The default factor is for **GPT-4o in the US**. You **must** visit **ecologits.ai/latest** to find the correct factor for *your model and server region* and update the value in the **\"Conversion factor\"** node.\n2. **Connect the snippet:** Place the **\"Conversion factor\"** node before your AI node and the **\"Calculate gCO₂e\"** node after it.\n3. **Update the calculation:** Modify the **\"Calculate gCO₂e\"** node to use the output text from *your* AI node.\n\n**Pro-Tip:** For higher accuracy, use the direct `output_tokens` value from your AI node's data if it's available."
},
"typeVersion": 1
},
{
"id": "941043b0-01ee-4553-87ec-1246a4cb2f2b",
"name": "Facteur de conversion",
"type": "n8n-nodes-base.set",
"position": [
304,
-160
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "a2c5484b-173e-4647-8dc1-23c32a899f75",
"name": "Conversion factor (in gCO₂e/token)",
"type": "number",
"value": 0.0612
}
]
}
},
"typeVersion": 3.4
},
{
"id": "430fc390-50b7-4feb-8c8f-be196a342d60",
"name": "Note adhésive1",
"type": "n8n-nodes-base.stickyNote",
"position": [
224,
-240
],
"parameters": {
"color": 5,
"width": 272,
"height": 336,
"content": "### Adapt this value to your model & settings\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nUse the expert mode here to find the factor that fits best:\nhttps://huggingface.co/spaces/genai-impact/ecologits-calculator"
},
"typeVersion": 1
}
],
"connections": {
"6cce6b66-bd1a-419b-86c1-b76aa257e96c": {
"main": [
[
{
"node": "0fca2f27-8a0b-46d0-9dfc-27967afe2ae5",
"type": "main",
"index": 0
}
]
]
},
"941043b0-01ee-4553-87ec-1246a4cb2f2b": {
"main": [
[
{
"node": "6cce6b66-bd1a-419b-86c1-b76aa257e96c",
"type": "main",
"index": 0
}
]
]
},
"d0ea8139-307d-4de6-9f29-11216958f362": {
"ai_languageModel": [
[
{
"node": "6cce6b66-bd1a-419b-86c1-b76aa257e96c",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"e374f8b8-ff4a-4b98-af50-d609338ec38f": {
"main": [
[
{
"node": "941043b0-01ee-4553-87ec-1246a4cb2f2b",
"type": "main",
"index": 0
}
]
]
}
}
}Comment utiliser ce workflow ?
Copiez le code de configuration JSON ci-dessus, créez un nouveau workflow dans votre instance n8n et sélectionnez "Importer depuis le JSON", collez la configuration et modifiez les paramètres d'authentification selon vos besoins.
Dans quelles scénarios ce workflow est-il adapté ?
Intermédiaire - Résumé IA, IA Multimodale
Est-ce payant ?
Ce workflow est entièrement gratuit et peut être utilisé directement. Veuillez noter que les services tiers utilisés dans le workflow (comme l'API OpenAI) peuvent nécessiter un paiement de votre part.
Workflows recommandés
Partager ce workflow