Virtuoso Medium V2

🚀 Virtuoso-Medium-v2 (32B)

Virtuoso-Medium-v2 (32B) is our next - generation, 32 - billion - parameter language model. It builds upon the original Virtuoso - Medium architecture, distilled from Deepseek - v3 with an expanded dataset of 5B+ tokens worth of logits. It outperforms our previous release in benchmark scores and surpasses Arcee - Nova 2024 in certain tasks.

🚀 Quick Start

Below is a sample code snippet using transformers:

from transformers import AutoTokenizer, AutoModelForCausalLM

model_name = "arcee-ai/Virtuoso-Medium-v2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

prompt = "Provide a concise summary of quantum entanglement."
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=150)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

✨ Features

• High - Performance Benchmark Scores: Achieves higher scores than previous releases and surpasses Arcee - Nova 2024 in certain tasks.
• Advanced Distillation: Distilled from Deepseek - v3 with a large dataset of 5B+ tokens worth of logits.
• Cross - Architecture Compatibility: Uses specialized “tokenizer surgery” for cross - architecture compatibility.

📦 Installation

The library used for this model is transformers. You can install it using pip:

pip install transformers

📚 Documentation

GGUF

Quantizations are available here.

Model Details

Background on Deepseek Distillation

Deepseek - v3 serves as the teacher model. Instead of standard supervised fine - tuning, we apply full logit - level replication, which ensures more precise knowledge transference, including advanced reasoning in technical and scientific queries, complex code generation, and mathematical problem - solving.

Intended Use Cases

• Advanced Chatbots & Virtual Assistants
• Enterprise Data Analysis & Workflow Automation
• Research Simulations & Natural Language Understanding
• Educational Tools for STEM Fields

Evaluations

Training & Fine - Tuning

• Initial Training: Started with Qwen - 32B, calibrated for large - scale text ingestion.
• Distillation & Merging: Trained on ~1.1B tokens worth of Deepseek - v3 logits. Employed “fusion merging” to retain teacher expertise. Final step included DPO to improve alignment and reduce model hallucinations.
• Continuous Development: Additional R1 distillations are in progress to enhance performance and specialization.

Performance

Thanks to a larger parameter count and a richer training corpus, Virtuoso - Medium - v2 delivers high scores across multiple benchmarks (BBH, MMLU - PRO, MATH, etc.). It often outperforms other 30B+ models and some 70B+ architectures in specific tasks.

Limitations

• Context Length: 128k Tokens
• Knowledge Cut - off: Training data may not reflect the latest events or developments after June 2024.

Ethical Considerations

⚠️ Important Note

Like any language model, Virtuoso - Medium - v2 can potentially generate harmful or biased content if prompted in certain ways.

📄 License

Virtuoso - Medium - v2 (32B) is released under the [Apache - 2.0 License](https://www.apache.org/licenses/LICENSE - 2.0). You are free to use, modify, and distribute this model in both commercial and non - commercial applications, subject to the terms and conditions of the license.

If you have questions or would like to share your experiences using these models, please connect with us on social media. We’re excited to see what you build—and how these models help you innovate!