GV

GeezerVison

The World Wide Web, restored to 800 pixels of glory.

Home|About|Gaming cabinet|Archives|How it works|Starter sites

On this site

  • Sign in
  • ActionsCI/CD and automation
  • CodespacesCloud development environments
  • All Features
  • Documentation
  • Training
  • Certifications
  • Build with Copilot
  • Customize Copilot
  • Integrations
  • Trending repos
  • GitHub Sponsors
  • Developer articles
  • Enterprises

This view

  • Original page
  • New address

GeezerVison is showing github.com — Issues · foundation-model-stack/fastsafetensors
Change the issues type filter

Issues

New issue
Search issues
is:issue state:open
is:issue state:open
Clear filter
Search

    Issues

    • Open
      6 (6)
    • Closed
    • Comfortable display density
    • Compact display density

    • RFC: io_uring DMA-BUF copier, NVMe straight into GPU memory

      foundation-model-stack/fastsafetensors#139 · opened on Nov 4, 2026
      5

    • [Discussion] Loader building blocks for row-gathered tables (Engram, PLE), following #127

      foundation-model-stack/fastsafetensors#132 · opened on Oct 11, 2026

    • Expert-parallel shuffle OOMs on large fused MoE tensors (full-tensor on-device clone in shuffle())

      foundation-model-stack/fastsafetensors#127 · opened on Oct 6, 2026
      5

    • feat: checkpoint writing and Spyre support

      foundation-model-stack/fastsafetensors#118 · opened on Sep 30, 2026
      1

    • SafeTensorsFileLoader retains ~2x file size in memory permanently after close(), even with nogds=True + explicit gc/empty_cache

      foundation-model-stack/fastsafetensors#94 · opened on Jul 23, 2026
      3

    • [Discussion] Sub-file batching — Decouple GPU memory usage from model file size

      foundation-model-stack/fastsafetensors#71 · opened on Apr 27, 2026
      21
    You are visitor 01402520 — GeezerVison does not store the pages you look up.
    Best viewed with Netscape Navigator 4.7 or Internet Explorer 6 at 800×600. No cookies required.