Skip to content

Strided tensor views, kernels, permutation and einsum

active 2026-01-292026-08-24 (UTC)

Complete coverage26,566 / 26,566 hourly files (100%) · 2 absent upstream2023-08-152026-08-25 (UTC)
Events
245
Pushes
77
Pull requests
41
Issues
38
Stars
0
Forks
0

Activity over time

Daily event counts in the loaded window

Line chart, 208 days from 2026-01-29 to 2026-08-24. Pushes: 77 total, peak 10 in a day. Pull requests: 41 total, peak 8 in a day. Issues: 38 total, peak 7 in a day. Comments: 4 total, peak 2 in a day. Stars: 0 total, peak 0 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
shinaoka13263324
terasakisatoshi251390
Ryo-wtnb112100
tensor4all-ai-bot[bot]1000

Recent activity

Latest issues, pull requests and releases

  • Issue#225tensor4all-ai-bot[bot]2026-08-07 18:26
    strided-perm: copy_into/copy_into_par call copy_nonoverlapping without an aliasing/overlap check (UB from safe code)
  • Issue comment#187shinaoka2026-07-30 20:38
    Add uninitialized full-overwrite replay for reduction and indexed plans
  • Issue#139Ryo-wtnb112026-07-07 15:34
    strided-kernel: zero-alloc borrowing-view (RawStridedRef) execution + reusable prepared plans for copy/map/add/scale/reduce
  • Pull request#132shinaoka2026-06-06 22:40
  • Pull request#130shinaoka2026-06-06 14:10
  • Issue#127shinaoka2026-02-20 11:07
    Extract transpose-flip helper in bgemm_blas.rs
  • Issue#123shinaoka2026-02-20 11:06
    BufferPool: use best-fit (>=) instead of exact-size matching
  • Pull request#119shinaoka2026-02-20 03:16
  • Issue#115shinaoka2026-02-19 08:12
    einsum2 batched GEMM has 3.5x overhead vs Julia for small-m contractions
  • Pull request#113shinaoka2026-02-19 02:49
  • Pull request#112shinaoka2026-02-19 02:11
  • Pull request#112shinaoka2026-02-19 02:04
  • Pull request#111shinaoka2026-02-18 23:12
  • Issue#109shinaoka2026-02-18 23:10
    perf: always-materialize output permutation in einsum2 (Step 4)
  • Issue#103shinaoka2026-02-18 22:24
    perf: eager materialization after output permutation to avoid scattered reads in subsequent steps
  • Issue#102shinaoka2026-02-18 22:24
    perf: add rayon parallelism to copy_into_col_major
  • Issue#107shinaoka2026-02-18 22:21
    bench: permutation benchmark suite for strided-perm (Step 2)
  • Issue#97shinaoka2026-02-18 14:04
    strided-kernel: small tensor fast path bypassing build_plan_fused
  • Pull request#100shinaoka2026-02-18 14:04
  • Pull request#101shinaoka2026-02-18 14:00
  • Pull request#95terasakisatoshi2026-02-18 08:40
  • Pull request#94terasakisatoshi2026-02-18 07:34
  • Pull request#92terasakisatoshi2026-02-18 07:29
  • Pull request#91terasakisatoshi2026-02-18 07:28
  • Pull request#88terasakisatoshi2026-02-18 05:43

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 0 stars here means stars gained during the window, not the repo's star count.