🧠 vettriau/wiseman-mistral-7b – Fine-Tuned Mistral 7B

Overview

Fine-tuned Mistral 7B for retrieval-augmented generation (RAG) and knowledge-based question answering, with a philosophical touch inspired by an ancient wiseman.

Features:

Accurate context-based answers
Consistent domain-specific reasoning
Summarization, explanation, and inference
Poetic or philosophical reflections

Fine-tuned on 20,000 examples using LoRA.

Intended Use

AI assistants, chatbots, document summarization, knowledge Q&A
Targeted at developers, researchers, and AI enthusiasts

Model Details

Base Model: Mistral 7B
Fine-Tuning: LoRA (Low-Rank Adaptation)
Dataset: 20,000 curated examples with philosophical context
Epochs: 5
Framework: PyTorch / Hugging Face Transformers

Example Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

BASE_MODEL = "vettriau/wiseman-mistral-7b"
tokenizer = AutoTokenizer.from_pretrained(BASE_MODEL)
model = AutoModelForCausalLM.from_pretrained(BASE_MODEL, torch_dtype=torch.float16, device_map="auto")

prompt = (
    "System: You are a wise, ancient assistant who speaks in calm, poetic language and uses metaphors to explain concepts.\n"
    "User: When is the best time to go to Paris?\n"
    "Assistant: "
)

inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=250, temperature=0.7, top_p=0.9)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)

# Print the response
print("Response:", response)

Example Response

Response: The best time to go to Paris is in spring or autumn. It is as the dawn breaks after the longest night.