Daemontatox/PathFinderAi3.0

PathFinderAI 3.0

PathFinderAI 3.0 is a high-performance language model designed for advanced reasoning, real-time text analysis, and decision support. Fine-tuned for diverse applications, it builds upon the capabilities of Qwen2, optimized with cutting-edge tools for efficiency and performance.

Features

• Advanced Reasoning: Fine-tuned for real-time problem-solving and logic-driven tasks.
• Enhanced Performance: Trained 2x faster with Unsloth and the Hugging Face TRL library.
• Multi-domain Capability: Excels in education, research, business, legal, and healthcare applications.
• Optimized Architecture: Leverages Qwen2 for robust language understanding and generation.

Training Details

• Base Model: Daemontatox/PathFinderAI3.0
• Training Frameworks: Unsloth and Hugging Face’s TRL library.
• Optimization: Quantization-aware training for faster inference and deployment on resource-constrained environments.

Deployment

This model is ideal for deployment on both cloud platforms and edge devices, including Raspberry Pi, utilizing efficient quantization techniques.

License

The model is open-sourced under the Apache 2.0 license.

Usage

To load the model:

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "Daemontatox/PathFinderAI3.0"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Example usage
input_text = "What is the capital of France?"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.decode(outputs[0]))

Model Applications PathFinderAI 3.0 is designed for:

Real-time reasoning and problem-solving Text generation and comprehension Legal and policy analysis Educational tutoring Healthcare decision support

Open LLM Leaderboard Evaluation Results

Detailed results can be found here! Summarized results can be found here!

Metric	Value (%)
Average	40.11
IFEval (0-Shot)	42.71
BBH (3-Shot)	55.54
MATH Lvl 5 (4-Shot)	48.34
GPQA (0-shot)	21.14
MuSR (0-shot)	20.05
MMLU-PRO (5-shot)	52.86