Atlassian & AI Tutorials with Code Examples

How to fine-tune Qwen3.8-27B with LoRA on a Mac: MLX fine-tuning from scratch

Qwen3.8-27B and Qwen3.5-9B were taught Forge, Jira and Confluence on a single Mac Studio with 96 GB. This is the whole procedure with the seven scripts that do it, printed in full: a pinned MLX environment with two patches, an 8-bit base with its speculative-decoding head kept as a sidecar, a chat-JSONL dataset built from a folder of Markdown, a segmented trainer with full-state checkpoints, an adapter expander, a per-module merge into the 8-bit shards. The kit was run end to end for this page on the 9B, in 21 minutes, and its real output follows each step, including the one where the validation loss turns the wrong way.

Gabriela PerdumSep 15, 202638 min read

AI & Atlassian Tutorials

Tutorials 59

JSM Portal Request Create Property Panel Submit: Types Won't Catch a Bad Payload

More Tutorials

Convert a Fine-Tuned MLX Model to GGUF: Proving Quantization Parity with KL Divergence

CSV Import Summary Field: Why It's Required and How It Breaks Your Data

Confluence nested bodied macros before a Cloud migration: find them, triage them, fix them

Confluence heap size for a Cloud migration: Data Center memory settings, Synchrony, CCMA

Jira Data Center heap size and JVM tuning for a Cloud migration: setenv.sh and JCMA

Jira 32,767 characters limit: find the description and comment data a JCMA migration lost

How to fine-tune Qwen3.8-27B with LoRA on a Mac: MLX fine-tuning from scratch

Attach an API spec to a Jira workflow validator: what the CogniRunner AI rule never reads

Lock attachments in Confluence Cloud with Sentinel Vault, step by step

Jira API rate limit: read the headers, then see what CogniRunner spends each hour

Forge App REST API with OAuth 2.0 (3LO): Scripting LeanZero Management Plans with curl

CogniRunner Coder: a Jira coding agent that commits to GitHub from the issue

CogniRunner: Jira OpenAI, Claude or Forge LLMs behind one workflow validator