Pulse
2026-09-18-v3.0
Dataset
2026-10-27-catalog-ai-v2
Ruleset
b2e8889c66f7
Scanned
2026-10-27
Window
February–September 15, 2026
Catalog tracking observation. This repository is measured with the same fixed window and ruleset, but it is not part of the published benchmark or its aggregate claims.

🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support

01Result

huggingface/accelerate ↗

huggingface/accelerate AI coding statistics: 3.3%

of 120 commits in February–September 15, 2026 were committed by an AI agent — a floor, not an estimate
3.3% agent · top contributor 75%2 contributors
individual

One person is driving this. The headline number reflects a single contributor's workflow, not a project-wide shift.

4
agent commits
116
other considered commits
2
agent contributors
75%
largest contributor

02Drivers

What shapes this result

Mode, timing, and detected signatures show whether the result reflects broad adoption or one concentrated workflow.

Assisted or autonomous

4assisted
0autonomous

Assisted — a developer’s identity authored the commit and the agent added itself as co-author. Autonomous — a bot identity authored the commit without a human identity attached.

Agent share by month

Agent-attributed commit share by month
MonthAgent shareConsidered commits
2026-020%12
2026-030%21
2026-047.1%14
2026-050%6
2026-0613%23
2026-070%9
2026-080%6
2026-090%29

Peak month: 13%. Bars are scaled to that peak, not to 100%.

Which agents

AgentVendorCommitsShare
Claude CodeAnthropic43.3%

03Evidence

Inspect the matches

Public commit samples let you inspect the signatures behind the aggregate result.

Matched commit samples

  • 61e64209 on GitHubFix to-fsdp2: actually drop REMOVED / NOT_YET_IMPLEMENTED FSDP1 keys (#406Claude Code
  • bd208073 on GitHubRaise ValueError (not a bare string) in ParallelismConfig.get_device_mesh Claude Code
  • 0dca40ab on GitHubFix dtype_byte_size for FP8 fnuz / e8m0fnu dtypes (#4063)Claude Code

04Method

What entered the denominator

The result begins with commit history, removes merges and routine automation, and reports literal agent signatures as a lower bound.

Exclusion ledger

Commits in window127
Merge commits removed−0
CI / dependency bots removed−7
Considered120

Bots are excluded from the denominator entirely. Agent attribution requires a literal supported signature; no code-style inference is used.

Continue exploring

Claude Code commitsBrowse repositories with detected Claude Code signatures →Python AI coding statisticsCompare evidence across Python repositories →GitHub AI commit analyzerInspect another repository with the same evidence model →
Reproduce this observation

Run the same detector locally.

Deterministic signature matching; no code-style inference. Squash merges can remove trailers, so this result remains a measured floor.

npx qmmit-cli huggingface/accelerate

Scan another repository

Try