# BBuf/tvm_mlir_learn

compiler learning resources collect.

Repository: https://github.com/BBuf/tvm_mlir_learn
Canonical: https://ross.abutalabs.com/products/tvm_mlir_learn
Language: Python
License Family: other
Last push: 2026-05-20T08:30:37+00:00

## Health v2 (maintenance only)
Score: 70/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 83, release rhythm 35, longevity 100
- inputs: {"age_days": 1996, "days_push": 105, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2768, forks 370 (observed 2026-08-28T04:07:18.832967+00:00)

## What it is
A curated collection of learning notes, examples, and experiments for deep learning compilers, covering TVM, MLIR, LLVM, Relay, TorchScript, and compiler-guided kernel optimization. It serves as a public archive of scheduler examples, code generation demos, and paper reading notes for AI compiler systems.

## Use cases
- learn how TVM scheduling and code generation work
- understand MLIR and LLVM compiler infrastructure
- study deep learning compiler papers like Ansor and PET
- optimize GEMM kernels using compiler techniques
- experiment with Relay IR and custom compiler passes
- get started with deep learning compiler internals

## When to choose
- you want hands-on examples for TVM, MLIR, or Relay
- you are studying AI compiler systems and kernel optimization
- you need paper reading notes on ML systems compilers

## When to avoid
- you need production-ready compiler tooling or maintained libraries
- you expect a structured course or up-to-date documentation
- you need a licensed, redistributable codebase

## Facets
- artifact type: learning-resource
- maturity: maintenance
- function: compiler, interpreter, developer-tools
- domain: compilers, machine-learning, tutorials, gpu-computing
- platform: python
- tags: tvm, mlir, llvm, deep-learning-compiler, code-generation, kernel-optimization, learning-notes, linux, docker

## Member repositories
- BBuf/tvm_mlir_learn (main) score 70

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:18.832967+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T08:16:41.960453+00:00, confidence not recorded.
  - readme: https://github.com/BBuf/tvm_mlir_learn (fetched 2026-08-28T04:07:18.832967+00:00, sha 609fac56e777)
- Data as of 2026-08-30T08:39:29.467469+00:00.