Tag
#fine-tuning
8 posts
Agent Core: Database-Native Tool Calling Baked Into Weights
An open-source LoRA that teaches small language models to think in SQL before grep. 9 tools. 3 databases. 61.3% tool selection accuracy. One consumer GPU.
Agent Core v5.4: When Data Augmentation Backfires
We 6.5x'd our training data and made the model safer but less capable. Here's the root cause, the numbers, and the v5.5 fix plan.
Agent Core v5.3: What We Learned Training a Tool-Calling LoRA on Real Data
Concrete training results comparing v5.2 and v5.3 of Agent Core — a universal tool-calling LoRA for 8B models. What improved, what regressed, and why SFT has a ceiling.
Agent Core: 9 Tools Is All You Need
How we trained a universal LoRA standard — 9 tools, 3 databases, a vault boundary — into a Qwen3-8B model that beats 23-tool fine-tuning with 87% fewer prompt tokens.
Why Character Belongs in Weights, Not Prompts
What Anthropic's leaked soul document teaches us about training AI personality — and why prompt-only approaches hit a ceiling.
Measuring Personality Depth: A Consensus Eval for AI Agents
A reproducible method for testing whether your AI agent's personality is truly internalized or just a fragile prompt — adapted from Anthropic's soul document extraction.
Self-Distillation: Mining Your AI Conversations for LoRA Data
Your real coding agent interactions are the highest-quality training data you'll ever have. Here's the 4-stage pipeline that converts conversation histories into structured LoRA training data — zero cost until Stage 3.
Schema on the Inside: Training an 8B Model to Recall Tool Schemas From Memory
How we trained Qwen3-8B to call 23 tools without any schemas in the prompt — and beat BFCL benchmarks in the process.