vibe MLing - e2e finetunes on arbitrary domains with claude code via carefully synthesized data | Cologne .

Members-Only

Recent Talks & Demos are for members only

Exclusive feed

You must be an AI Tinkerers active member to view these talks and demos.

January 21, 2026 · Cologne

Finetuning with Claude Synthetic Data

This talk details three new SKILLs for end-to-end model finetuning on arbitrary domains using Claude-generated data, covering extraction, synthesis, and evaluation.

Overview
Links
Tech stack
  • Claude
    Claude is Anthropic's flagship family of large language models (LLMs): a high-performance, Constitutional AI system built for safety, complex reasoning, and expert-level collaboration.
    Claude is a next-generation AI assistant developed by Anthropic, a research firm prioritizing AI safety. The models (including Opus, Sonnet, and Haiku) leverage Constitutional AI to ensure helpful, honest, and harmless outputs, a key differentiator from competitors. Claude excels at complex enterprise tasks: processing massive context windows for in-depth data analysis, generating and reviewing code, and providing expert-level summarization for documents up to 200,000 tokens. It is deployed as a conversational chatbot and via API, offering scalable AI solutions for developers and businesses.
  • GPT-5
    OpenAI's GPT-5: The unified, multimodal foundation model delivering PhD-level reasoning and state-of-the-art coding performance.
    Copy that: GPT-5 is OpenAI's flagship multimodal model, launched August 7, 2025, as the successor to GPT-4. This is a major architectural shift: it unifies advanced reasoning capabilities (like the 'o-series' models) and rapid response times into a single system, eliminating the need for manual model switching (Source: OpenAI, August 2025). The model demonstrates state-of-the-art performance across technical benchmarks (math, programming, finance) and features a massive 272,000-token context window (Source: Jagran Josh, Voiceflow). Developers access it via the API in variants—including `gpt-5-mini` and `gpt-5-nano`—optimized for latency and cost trade-offs, making this frontier intelligence accessible across all ChatGPT tiers (Source: Botpress, Jagran Josh, Voiceflow).
  • Gemini 3
    Gemini 3 Pro: Google's most intelligent, state-of-the-art multimodal AI model, engineered for advanced reasoning and complex agentic coding workflows.
    This is the third-generation model, the new intelligence foundation for Google's ecosystem. Gemini 3 Pro delivers state-of-the-art reasoning and world-leading multimodal understanding: it seamlessly processes text, images, video, audio, and code. It is our most powerful agentic and vibe-coding model yet, designed to handle long-horizon, multi-step tasks; it features an industry-leading 1 million token context window. The model's performance is validated by its breakthrough score of 1501 Elo on the LMArena Leaderboard, confirming its superior capability for developers building on platforms like Vertex AI and the new Google Antigravity.
  • Qwen-3 14B
    Qwen3-14B is a 14.8B parameter causal language model (CLM) featuring dynamic 'thinking' and 'non-thinking' modes for adaptive performance across complex reasoning and general dialogue.
    This is Qwen3-14B: a dense 14.8B parameter CLM from Alibaba Cloud, engineered for dual-mode operation. It seamlessly shifts to a 'thinking' mode for high-stakes analytical tasks (math, code generation, logical inference) and defaults to an efficient 'non-thinking' mode for general dialogue. The model supports a native 32,768-token context window, expandable to 131,072 tokens via YaRN (Yet another RoPE N). We've confirmed its enhanced reasoning capabilities and superior human alignment, plus robust support for over 100 languages: this model is a versatile, high-performance asset.
  • Hugging Face
    Hugging Face is the central, open-source platform and community for building AI applications, hosting over 300,000 models and datasets via the popular Transformers library.
    Hugging Face functions as the 'GitHub for machine learning,' providing a massive, collaborative Hub for AI assets (models, datasets, and demos). Its core technology is the open-source **Transformers** Python library, which simplifies the use of state-of-the-art models (e.g., BERT, GPT) for various tasks: natural language processing, computer vision, and audio. The platform hosts over 300,000 models and thousands of datasets, streamlining the entire ML workflow from research to deployment via **Spaces** (interactive demos). This ecosystem makes advanced AI accessible, efficient, and reproducible for developers and enterprises globally.

Related projects