mi300
Here are 9 public repositories matching this topic...
Agent-queryable ROCm kernel optimization knowledge base for AMD Instinct MI300/gfx942 and MI350/MI355X/gfx950, packaged for Codex CLI and Claude Code with merged-PR provenance, real-silicon validation, and a maintainer-controlled pull-request evidence pipeline.
-
Updated
Jul 27, 2026 - Python
A customized fork of Axolotl optimized for AMD GPUs running ROCm 6.2+, including support for MI250 and MI300 architectures. This version streamlines fine-tuning large language models on AMD hardware, integrating tailored dependencies and build steps for improved compatibility and performance.
-
Updated
Dec 6, 2024 - Python
A complete GEMM tuning loop on AMD MI325X β profile β hotspot ranking β demand extraction β tuning β load verification β paired A/B. +4.1~5.1% request throughput on Qwen3-8B/SGLang, plus the six checks that make the number defensible.
-
Updated
Oct 3, 2026 - Python
Add this topic to your repo
To associate your repository with the mi300 topic, visit your repo's landing page and select "manage topics."