Updated portal

README.md

@@ -9,4 +9,64 @@ app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Random Baseline Model for Climate Disinformation Classification
+
+## Model Description
+
+This is a random baseline model for the Frugal AI Challenge 2024, specifically for the text classification task of identifying climate disinformation. The model serves as a performance floor, randomly assigning labels to text inputs without any learning.
+
+### Intended Use
+
+- **Primary intended uses**: Baseline comparison for climate disinformation classification models
+- **Primary intended users**: Researchers and developers participating in the Frugal AI Challenge
+- **Out-of-scope use cases**: Not intended for production use or real-world classification tasks
+
+## Training Data
+
+The model uses the QuotaClimat/frugalaichallenge-text-train dataset:
+- Size: ~6000 examples
+- Split: 80% train, 20% test
+- 8 categories of climate disinformation claims
+
+### Labels
+0. No relevant claim detected
+1. Global warming is not happening
+2. Not caused by humans
+3. Not bad or beneficial
+4. Solutions harmful/unnecessary
+5. Science is unreliable
+6. Proponents are biased
+7. Fossil fuels are needed
+
+## Performance
+
+### Metrics
+- **Accuracy**: ~12.5% (random chance with 8 classes)
+- **Environmental Impact**:
+  - Emissions tracked in gCO2eq
+  - Energy consumption tracked in Wh
+
+### Model Architecture
+The model implements a random choice between the 8 possible labels, serving as the simplest possible baseline.
+
+## Environmental Impact
+
+Environmental impact is tracked using CodeCarbon, measuring:
+- Carbon emissions during inference
+- Energy consumption during inference
+
+This tracking helps establish a baseline for the environmental impact of model deployment and inference.
+
+## Limitations
+- Makes completely random predictions
+- No learning or pattern recognition
+- No consideration of input text
+- Serves only as a baseline reference
+- Not suitable for any real-world applications
+
+## Ethical Considerations
+
+- Dataset contains sensitive topics related to climate disinformation
+- Model makes random predictions and should not be used for actual classification
+- Environmental impact is tracked to promote awareness of AI's carbon footprint
+```

app.py

@@ -22,50 +22,47 @@ print(HF_TOKEN)
 if not HF_TOKEN:
     print("Warning: HF_TOKEN not found in environment variables. Submissions will not work.")
 
+# Initialize carbon emissions tracker with CodeCarbon
 tracker = EmissionsTracker(allow_multiple_runs=True)
 
-# Function to get space username and URL
-def get_space_info():
-    space_name = os.getenv("SPACE_ID", "")
-    if space_name:
-        try:
-            username = space_name.split("/")[0]
-            space_url = f"https://huggingface.co/spaces/{space_name}"
-            return username, space_url
-        except Exception as e:
-            print(f"Error getting space info: {e}")
-    return "local-user", "local-development"
-
-def clean_emissions_data(emissions_data):
-    """Remove unwanted fields from emissions data"""
-    data_dict = emissions_data.__dict__
-    fields_to_remove = ['timestamp', 'project_name', 'experiment_id', 'latitude', 'longitude']
-    return {k: v for k, v in data_dict.items() if k not in fields_to_remove}
+#--------------------------------------------------------------------------------------------
+# FUNCTION TO UPDATE WITH YOUR MODEL SUBMISSION
+#--------------------------------------------------------------------------------------------
 
 @spaces.GPU
-def evaluate():
+def evaluate(model_description):
     # Get space info
     username, space_url = get_space_info()
     
     # Initialize tracker
     tracker.start()
     tracker.start_task("inference")
+
+    #--------------------------------------------------------------------------------------------
+    # YOUR MODEL INFERENCE CODE HERE
+    # Update the code below to replace the random baseline by your model inference within the inference pass where the energy consumption and emissions are tracked.
+    #--------------------------------------------------------------------------------------------   
     
     # Make random predictions
     true_labels = test_dataset["label"]
     predictions = [random.randint(0, 7) for _ in range(len(true_labels))]
-    
+
+    #--------------------------------------------------------------------------------------------
+    # YOUR MODEL INFERENCE STOPS HERE
+    #--------------------------------------------------------------------------------------------   
+
+    # Stop tracking emissions
+    emissions_data = tracker.stop_task()    
+
     # Calculate accuracy
     accuracy = accuracy_score(true_labels, predictions)
     
-    # Stop tracking emissions
-    emissions_data = tracker.stop_task()
-    
     # Prepare complete results
     results = {
         "username": username,
         "space_url": space_url,
         "submission_timestamp": datetime.now().isoformat(),
+        "model_description": model_description if model_description else "No description provided",
         "accuracy": float(accuracy),
         "energy_consumed_wh": emissions_data.energy_consumed * 1000,
         "emissions_gco2eq": emissions_data.emissions * 1000,
@@ -80,40 +77,76 @@ def evaluate():
         json.dumps(results, indent=2)
     ]
 
+
+#--------------------------------------------------------------------------------------------
+# HELPER FUNCTIONS
+#--------------------------------------------------------------------------------------------
+
+# Function to get space username and URL
+def get_space_info():
+    space_name = os.getenv("SPACE_ID", "")
+    if space_name:
+        try:
+            username = space_name.split("/")[0]
+            space_url = f"https://huggingface.co/spaces/{space_name}"
+            return username, space_url
+        except Exception as e:
+            print(f"Error getting space info: {e}")
+    return "local-user", "local-development"
+
+def clean_emissions_data(emissions_data):
+    """Remove unwanted fields from emissions data"""
+    data_dict = emissions_data.__dict__
+    fields_to_remove = ['timestamp', 'project_name', 'experiment_id', 'latitude', 'longitude']
+    return {k: v for k, v in data_dict.items() if k not in fields_to_remove}
+
+
 def submit_results(results_json):
     if not results_json:
         return gr.Warning("No results to submit")
     
-    if not HF_TOKEN:
-        return gr.Warning("HF_TOKEN not found. Please set up your Hugging Face token.")
+    # Check if we're in a Space or have admin dev rights
+    space_name = os.getenv("SPACE_ID")
+    is_admin_dev = os.getenv("ADMIN_DEV") == "true"
     
-    # try:
-    # results_json is already a string, no need to load it
-    results_str = json.dumps(results_json)  # Parse the JSON string to get the data
+    if not space_name and not is_admin_dev:
+        message = "You cannot submit your model locally, you need to deploy it as a Hugging Face Space first, and then submit it."
+        return gr.Warning(message)
     
-    # Create a temporary file with the results
-    with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.json') as f:
-        # Write the original JSON string to file
-        f.write(results_str)
-        temp_path = f.name
-    
-    # Upload to the dataset
-    api = HfApi(token=HF_TOKEN)
-    path_in_repo = f"submissions/{results_json['username']}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
-    api.upload_file(
-        path_or_fileobj=temp_path,
-        path_in_repo=path_in_repo,
-        repo_id="frugal-ai-challenge/public-leaderboard-text",
-        repo_type="dataset",
-        token=HF_TOKEN
-    )
+    if not HF_TOKEN:
+        return gr.Warning("HF_TOKEN not found. Please set up your Hugging Face token.")
     
-    # Clean up
-    os.unlink(temp_path)
+    try:
+        # results_json is already a dict from gr.JSON
+        results_str = json.dumps(results_json)
+        
+        # Create a temporary file with the results
+        with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.json') as f:
+            f.write(results_str)
+            temp_path = f.name
+        
+        # Upload to the dataset
+        api = HfApi(token=HF_TOKEN)
+        path_in_repo = f"submissions/{results_json['username']}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
+        api.upload_file(
+            path_or_fileobj=temp_path,
+            path_in_repo=path_in_repo,
+            repo_id="frugal-ai-challenge/public-leaderboard-text",
+            repo_type="dataset",
+            token=HF_TOKEN
+        )
+        
+        # Clean up
+        os.unlink(temp_path)
+        
+        return gr.Info("Results submitted successfully to the leaderboard! 🎉")
+    except Exception as e:
+        return gr.Warning(f"Error submitting results: {str(e)}")
     
-    return gr.Info("Results submitted successfully to the leaderboard! 🎉")
-    # except Exception as e:
-        # return gr.Warning(f"Error submitting results: {str(e)}")
+
+#--------------------------------------------------------------------------------------------
+# DATASET PREPARATION
+#--------------------------------------------------------------------------------------------
 
 # Define the label mapping
 LABEL_MAPPING = {
@@ -127,18 +160,6 @@ LABEL_MAPPING = {
     "7_fossil_fuels_needed": 7  # Fossil fuels are needed
 }
 
-# Reverse mapping for display purposes
-LABEL_DESCRIPTIONS = {
-    0: "No relevant claim detected",
-    1: "Global warming is not happening",
-    2: "Not caused by humans",
-    3: "Not bad or beneficial",
-    4: "Solutions harmful/unnecessary",
-    5: "Science is unreliable",
-    6: "Proponents are biased",
-    7: "Fossil fuels are needed"
-}
-
 # Load and prepare the dataset
 print("Loading dataset...")
 dataset = load_dataset("QuotaClimat/frugalaichallenge-text-train")
@@ -151,19 +172,17 @@ train_test = dataset["train"].train_test_split(test_size=0.2, seed=42)
 train_dataset = train_test["train"]
 test_dataset = train_test["test"]
 
-# Display preview
-print("\nFirst 5 rows of test set:")
-for i, example in enumerate(test_dataset.select(range(5))):
-    print(f"\nExample {i+1}:")
-    print(f"Text: {example['quote'][:100]}...")
-    print(f"Label: {example['label']} - {LABEL_DESCRIPTIONS[example['label']]}")
+#--------------------------------------------------------------------------------------------
+# GRADIO INTERFACE
+#--------------------------------------------------------------------------------------------
 
 # Create the demo interface
 with gr.Blocks() as demo:
 
+    gr.Image("./logo.png", show_label=False, container=False)
 
     gr.Markdown("""
-# Frugal AI Challenge - Text task - Submission portal
+# 📜 Frugal AI Challenge - Text task - Submission portal
 ## Climate Disinformation Classification
 """)
 
@@ -186,10 +205,26 @@ To submit your results, please follow the steps below:
 7. Step 1 - Evaluate model: Click on the button to evaluate your model. This will run you model, computes the accuracy on the test set (20% of the train set), and track the energy consumption and emissions.
 8. Step 2 - Submit to leaderboard: Click on the button to submit your results to the leaderboard. This will upload the results to the leaderboard dataset and update the leaderboard.
 9. You can see the leaderboard at https://huggingface.co/datasets/frugal-ai-challenge/public-leaderboard-text
+
+## About
+> You can find more information about the Frugal AI Challenge 2025 on the [Frugal AI Challenge website](https://frugal-ai-challenge.org/).
+> Or directly on the organization page on Hugging Face: [Frugal AI Challenge](https://huggingface.co/frugal-ai-challenge)
+                        
+This portal is a submission portal for the Frugal AI Challenge 2025. It is a simple interface to evaluate and submit your model to the leaderboard.
+The challenge is organized by Hugging Face, Data For Good, and the French Ministry of Environment. 
+
+The goal of the Frugal AI Challenge is to encourage both academic and industry actors to keep efficiency in mind when deploying AI models. By tracking both energy consumption and performance for different AI tasks, we can incentivize frugality in AI deployment while also addressing real-world challenges.
 """)
 
         with gr.Tab("Model Submission"):
-            gr.Markdown("## Random Baseline Model")
+            
+            with gr.Row():
+                model_description = gr.Textbox(
+                    label="Model Description (one sentence)",
+                    placeholder="Describe your model in one sentence...",
+                    value="Random baseline",
+                    lines=2
+                )
             
             with gr.Row():
                 with gr.Column(scale=1):
@@ -199,15 +234,14 @@ To submit your results, please follow the steps below:
                     
             with gr.Row():
                 accuracy_output = gr.Number(label="Accuracy", precision=4)
-                emissions_output = gr.Number(label="Emissions (gCO2eq)", precision=12)
                 energy_output = gr.Number(label="Energy Consumed (Wh)", precision=12)
-            
+                emissions_output = gr.Number(label="Emissions (gCO2eq)", precision=12)
             with gr.Row():
                 results_json = gr.JSON(label="Detailed Results", visible=True)
         
             evaluate_btn.click(
                 evaluate,
-                inputs=None,
+                inputs=[model_description],
                 outputs=[accuracy_output, emissions_output, energy_output, results_json]
             )
             
@@ -218,9 +252,14 @@ To submit your results, please follow the steps below:
             )
         
         with gr.Tab("Model Card"):
-            with open("modelcard.md", "r") as f:
-                model_card_content = f.read()
-            gr.Markdown(model_card_content)
+            with open("README.md", "r") as f:
+                content = f.read()
+                # Remove the YAML header (content between --- markers)
+                if content.startswith("---"):
+                    second_marker = content.find("---", 3)
+                    if second_marker != -1:
+                        content = content[second_marker + 3:].strip()
+            gr.Markdown(content)
 
 if __name__ == "__main__":
     demo.launch()

modelcard.md

@@ -1,61 +0,0 @@
-# Random Baseline Model Card
-
-## Model Description
-
-**Model Type:** Random Baseline Classifier  
-**Task:** Climate Change Disinformation Classification  
-**Version:** 1.0.0  
-**Last Updated:** 2024
-
-### Overview
-This is a random baseline model for climate change disinformation classification. It randomly assigns labels to text inputs, serving as a baseline for comparing more sophisticated models.
-
-### Intended Use
-- **Primary Use:** Baseline comparison for climate disinformation classification models
-- **Intended Users:** Researchers and developers working on climate disinformation detection
-- **Out-of-Scope Uses:** Not intended for production or real-world classification tasks
-
-## Training Data
-
-**Dataset:** QuotaClimat/frugalaichallenge-text-train
-- Size: ~6000 examples
-- Split: 80% train, 20% test
-- Labels: 8 categories of climate disinformation claims
-
-### Labels
-0. No relevant claim detected
-1. Global warming is not happening
-2. Not caused by humans
-3. Not bad or beneficial
-4. Solutions harmful/unnecessary
-5. Science is unreliable
-6. Proponents are biased
-7. Fossil fuels are needed
-
-## Performance
-
-### Metrics
-- **Accuracy:** ~12.5% (random chance)
-- **Environmental Impact:**
-  - Emissions (kgCO2eq)
-  - Energy Consumed (kWh)
-
-### Limitations
-- Random predictions with no learning
-- No consideration of input text
-- Serves only as a baseline reference
-
-## Ethical Considerations
-- Model makes random predictions and should not be used for actual classification
-- Dataset contains sensitive topics related to climate disinformation
-- Environmental impact is tracked to promote awareness of AI's carbon footprint
-
-## Environmental Impact
-This model tracks its environmental impact using CodeCarbon, measuring:
-- Carbon emissions
-- Energy consumption
-
-## Caveats and Recommendations
-- Use only as a baseline comparison
-- Not suitable for production use
-- Consider environmental impact when running experiments 

sandbox.ipynb

@@ -1,37 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "6941ccb0-6ed0-45c1-9460-8f8c0bbfc288",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "from codecarbon import EmissionsTracker\n",
-    "tracker = EmissionsTracker()\n",
-    "tracker.start()"
-   ]
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.12.4"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 5
-}