README.md

metadata

title: Prometh-MOEM-V.01 Model Showcase
emoji: 👁
colorFrom: red
colorTo: pink
sdk: gradio
pinned: false
license: apache-2.0
language:
  - en

Prometh-MOEM-V.01 Model Card 👁

Prometh-MOEM-V.01 is a pioneering Mixture of Experts (MoE) model, blending the capabilities of multiple foundational models to enhance performance across a variety of tasks. This model leverages the collective strengths of its components, achieving unparalleled accuracy, speed, and versatility.

🚀 Model Sources and Components

This MoE model amalgamates specialized models including:

• Wtzwho/Prometh-merge-test2
• mistralai/Mistral-7B-Instruct-v0.2
• Wtzwho/Prometh-merge-test3
• meta-math/MetaMath-Mistral-7B

🌟 Key Features

• Enhanced Performance: Tailored for peak accuracy and efficiency.
• Versatility: Exceptionally adaptable across a wide range of NLP tasks.
• State-of-the-Art Integration: Incorporates the latest in AI research for effective model integration.

📈 Application Areas

Prometh-MOEM-V.01 excels in:

• Text generation
• Sentiment analysis
• Language translation
• Question answering

💻 Usage Instructions

To utilize Prometh-MOEM-V.01 in your projects:

pip install -qU transformers bitsandbytes accelerate

from transformers import AutoTokenizer, pipeline
import torch

model = "Wtzwho/Prometh-MOEM-V.01"
tokenizer = AutoTokenizer.from_pretrained(model)

# Setup pipeline
pipeline = pipeline(
    "text-generation",
    model=model,
    model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
)

# Example query
messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])