metadata

tags:
  - t5
  - fine-tuned
  - information-extraction

My Fine-Tuned T5 Model

This is a fine-tuned T5 model for extracting structured information from automobile part descriptions.

Usage

from transformers import T5Tokenizer, T5ForConditionalGeneration

model_name = "your-username/your-repo-name"
tokenizer = T5Tokenizer.from_pretrained(model_name)
model = T5ForConditionalGeneration.from_pretrained(model_name)

input_text = "Brand: Ford, Model: Mustang, Year: 2018, Color: Red"
inputs = tokenizer(input_text, return_tensors="pt", max_length=128, truncation=True)
outputs = model.generate(**inputs)
decoded_output = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(decoded_output)

---
license: apache-2.0
---