I will finetune a custom llm for your use case

I will fine-tune a custom LLM or VLM for your use case

Looking for a production-ready language model tailored to your data, language, or domain? You're in the right place.

I'm a Machine Learning Engineer specializing in LLMs and VLMs, with proven experience building SOTA models including a fine-tuned 24B Russian-language model and a 4B agentic model for tool use and OpenClaw-compatible frameworks.

What I can do for you:

• Fine-tune models from 1B to 70B (Mistral, LLaMA, Qwen, etc.)
• Synthetic dataset creation and cleaning
• SMPO / DPO alignment for better instruction-following
• Agentic tool-use fine-tuning
• Benchmarking and evaluation reports
• Deployment-ready weights + inference setup

Why work with me:

• Real SOTA results, not just experiments
• PyTorch + Transformers native
• Clear communication in English, Russian, or German
• End-to-end ownership from dataset to deployment

Not sure which package fits? Message me I'm happy to scope your project before you order.