This model is a fine-tuned version of distilbert-base-uncased on the "King-8/help-request-messages" dataset. It achieves the following results on the evaluation set:

Loss: 1.3083

🤖 CIC Help Classifier Model

Overview

This model is a fine-tuned text classification model designed to identify the type of help a user needs within the Coding in Color (CIC) ecosystem.

It enables AI systems to understand user challenges and provide structured support.

🧠 Model Details

Base model: distilbert-base-uncased
Task: Text classification
Training data: CIC Help Classification Dataset
Framework: Hugging Face Transformers

📊 Labels

learning_help
project_help
attendance_issue
technical_issue
general_guidance

⚙️ Training

Epochs: 3
Dataset size: 100 samples
Train/Validation/Test split used

📈 Performance Notes

Training and validation loss decreased across epochs
Model performs well on common help scenarios
Accuracy is limited due to small dataset size

🧪 Example Usage

predict("I'm stuck on my project and don't know what to do")

Output:

json

{
  "type": "project_help",
  "confidence": 0.82
}

🔗 Use Case

This model is designed to be integrated into:

MCP server tools
Slack-based support systems
AI assistants for CIC students

🚀 Future Improvements

Fine-tune on larger CIC dataset
Add real-time feedback learning
Integrate with response generation models
Improve classification accuracy with more edge cases

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
1.3887	1.0	9	1.4495
1.2613	2.0	18	1.3350
1.1704	3.0	27	1.3083

Framework versions

Transformers 5.0.0
Pytorch 2.10.0+cpu
Datasets 4.0.0
Tokenizers 0.22.2

King-8/help-classifier

help_classifier