simonguest/echobot-t5-small-base

EchoBot Base Model

A T5-small model fine-tuned to echo/copy its input. This model serves as a baseline for educational fine-tuning exercises.

Model Description

This model was trained to perform a simple copy task: given any input text, it outputs exactly the same text.

Intended Use: This is a base model for educational purposes, designed to be fine-tuned by students on small (~50 example) datasets to learn new behaviors like:

• Emoji responses
• Language translation
• Simple math
• Directional symbols

Training Data

Trained on ~7,000 diverse examples including:

• Single words
• Short phrases
• Simple sentences
• Numbers and math expressions
• Text with emoji
• Conversational phrases

Performance

• Exact match accuracy: 100% on validation set
• Task: Text echo/copy

Usage

from transformers import T5ForConditionalGeneration, T5Tokenizer

model = T5ForConditionalGeneration.from_pretrained("simonguest/echobot-t5-small-base")
tokenizer = T5Tokenizer.from_pretrained("simonguest/echobot-t5-small-base")

input_text = "Hello, world!"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs, max_length=128)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
# Output: "Hello, world!"

Fine-tuning for Students

Students can fine-tune this model with small datasets. Example for emoji responses:

# Create 50 training examples like:
# {"input": "Happy", "target": ":-)"}
# {"input": "Sad", "target": ":-("}
# ... then fine-tune using standard T5 fine-tuning procedures

License

Apache 2.0