metadata

datasets:
  - bitext/Bitext-retail-banking-llm-chatbot-training-dataset
language:
  - en
base_model:
  - google-t5/t5-small
pipeline_tag: question-answering
tags:
  - fintech
  - retail-banking
  - fine-tuning
  - chatbot
  - llm
license: cdla-sharing-1.0

fintech-chatbot-t5

Model Description

This model was fine-tuned using a retail banking chatbot dataset. It is based on the T5-small architecture and is capable of answering common banking-related queries like account balances, transaction details, card activations, and more.

The model has been trained to generate responses to banking-related customer queries and is suited for use in automated customer service systems or virtual assistants.

Model Details

Model Type: T5-small
Training Dataset: retail banking chatbot dataset
Tasks: Natural Language Generation (NLG)
Languages Supported: English

Training Details

Number of Epochs: 3
Training Loss: 0.79
Evaluation Loss: 0.46
Evaluation Metric: Mean Squared Error
Batch Size: 8

How to Use the Model

You can load and use this model with the following code:

from transformers import T5Tokenizer, T5ForConditionalGeneration

tokenizer = T5Tokenizer.from_pretrained("cuneytkaya/fintech-chatbot-t5")
model = T5ForConditionalGeneration.from_pretrained("cuneytkaya/fintech-chatbot-t5")

input_text = "How can I activate my credit card?"
inputs = tokenizer.encode(input_text, return_tensors="pt")
outputs = model.generate(inputs)

print(tokenizer.decode(outputs[0]))