🍻 Drunk Chatbot 🥴

This model has been fine-tuned to simulate the enthusiastic, tangential, and sometimes over-the-top communication style of someone who's had a few drinks. It's meant for entertainment and humor, not to be taken seriously.

Model Details

This is a fine-tuned version of a Qwen3 model using QLoRA (Quantized Low-Rank Adaptation) through the Unsloth framework, trained to respond in an exaggerated, enthusiastic manner.

Usage

You can use this model with transformers and unsloth:

from transformers import AutoModelForCausalLM, AutoTokenizer

# Load model and tokenizer
model = AutoModelForCausalLM.from_pretrained("yossarianz257/drunk-chatbot-model")
tokenizer = AutoTokenizer.from_pretrained("yossarianz257/drunk-chatbot-model")

# Create a message for the model
messages = [
    {"role": "user", "content": "How do you feel about pizza?"}
]

# Format input with chat template
prompt = tokenizer.apply_chat_template(messages, tokenize=False)

# Tokenize and generate
input_ids = tokenizer(prompt, return_tensors="pt").input_ids
outputs = model.generate(input_ids, max_new_tokens=512, temperature=0.7, top_p=0.9)
response = tokenizer.batch_decode(outputs)[0]

# Print response
print(response)

Limitations

This model is intended for entertainment purposes only. It may generate responses that are:

Exaggerated and over-the-top
Occasionally nonsensical
Not factually reliable
Inappropriate in professional contexts

Ethical Considerations

This model is not meant to encourage or glorify excessive alcohol consumption. It's a humorous take on enthusiastic conversation styles. Please use responsibly and be mindful of content that could be triggering for those with alcohol-related issues.

Acknowledgements

Thanks to the Unsloth team for their efficient fine-tuning framework
Thanks to Qwen for the base model