Footer

Your go-to destination for discovering trending open-source projects and uncovering the insights that matter

Legal

  • Terms and conditions
  • Privacy policy

More

  • Pricing
  • Advertise
  • onedump
  • gosnakego
  • Contact us

Social

  • X (Twitter)
  • GitHub
Trendshift owl
What they build speaks louder than words
© 2026 Trendshift, created by Julian Li with ❤️
SignalAdvertise
Trendshift
SignalAdvertise
Sign in
Trendshift owl

Discover trends that matter

DailyWeeklyMonthlyYearly
Live mentionsTopics
RepositoriesDevelopers
InsightsStats
Busbar
Busbar
ClawBench
ClawBench
Bifrost
Bifrost

jayleaton/glm53-tensorfold-spark

GLM-5.3-Flash (abliterated EXL3) on 2x NVIDIA DGX Spark with the TensorFold engine: 1.8x faster decode than vLLM, 4x256k concurrent threads, byte-exact speculative decoding. Work in progress.

#LLM inference
Visit GitHub
Python
117
12
1 contributors
Apache License 2.0
FeaturedAdvertise to 180K+ monthly visitors
Busbar

Control what AI can do before it acts

ClawBench

a benchmark for personal agents

Bifrost

Enterprise AI Gateway Built For Scale: Route, Govern, and Secure All Your AI Traffic

Momentum

How fast jayleaton/glm53-tensorfold-spark is gaining stars, and where it lands on Trendshift and GitHub Trending

Activity and trending

Stars, forks and merged pull requests each day or month, lined up with the repository's rank on Trendshift Trending

Sep 27 – Oct 22, 2026

Stars +119 in 26 days

0204060
46

Forks +12 in 26 days

0123
3

Merged PRs 11 in 26 days

024
4

Trendshift Trendingwhere the repository ranked each day it made the top 25

Not in the Trendshift top 25 in this period

Sep 27Oct 1Oct 5Oct 9Oct 13Oct 17Oct 21
The left and right arrow keys step from day to day; Home and End go to the first and the last

Stars, forks, merged pull requests and Trendshift Trending rank, by day
DayStars gainedForks gainedMerged PRs
Sun, Sep 27, 2026+0+00
Mon, Sep 28, 2026+0+00
Tue, Sep 29, 2026+38+24
Wed, Sep 30, 2026+46+22
Thu, Oct 1, 2026+13+21
Fri, Oct 2, 2026+2+10
Sat, Oct 3, 2026+5+31
Sun, Oct 4, 2026+3+00
Mon, Oct 5, 2026+3+03
Tue, Oct 6, 2026+0+10
Wed, Oct 7, 2026+0+00
Thu, Oct 8, 2026+2+00
Fri, Oct 9, 2026+0+00
Sat, Oct 10, 2026+2+10
Sun, Oct 11, 2026+1+00
Mon, Oct 12, 2026+2+00
Tue, Oct 13, 2026+0+00
Wed, Oct 14, 2026+0+00
Thu, Oct 15, 2026+0+00
Fri, Oct 16, 2026+0+00
Sat, Oct 17, 2026+0+00
Sun, Oct 18, 2026+0+00
Mon, Oct 19, 2026+0+00
Tue, Oct 20, 2026+0+00
Wed, Oct 21, 2026+0+00
Thu, Oct 22, 2026+2+00
Total+119+1211
  • 1#1 on the list
  • 7Rank that day
  • Not trending

Social mentions

Recent discussions about this repository across the web
  1. Sep 301 post

    • Last post I showed you how to get over 100 tok/s on two DGX Sparks running Qwen 3.8 Flash. Today I confirmed the a 2x speed-up is coming for Local AI. The answer is TensorFold. I'll tell you how I…

      Read full on X, posted by @yume_arasaki

      @yume_arasaki · x.com

  2. Sep 292 posts

    • Same-weights A/B, fine print in his own README: the ~2x cells are greedy decoding; sampled T=1 is 1.2-1.5x. It's a 57-patch TensorFold fork on abliterated EXL3 weights, one pair of Sparks, three days…

      Read full on X, posted by @jimmy_longbow_

      @jimmy_longbow_ · x.com

    • GLM-5.3-Flash on 2x DGX Spark, same weights. TensorFold vs vLLM: chat 44.6 vs 22.8 tok/s code 77.6 vs 41.9 tok/s prompt processing 1,607 vs 1,448 tok/s Prompt processing was losing to vLLM yesterday…

      Read full on X, posted by @jayleaton

      @jayleaton · x.com

  3. Sep 281 post

    • TensorFold > vLLM! 2x DGX Sparks GLM 5.3 Flash EXL3 = 80% FASTER. Repo is in the comments, and I'll keep improving it. I already see another 10-15% of easy wins in PP. TLDR - Single stream 80-100%…

      Read full on X, posted by @jayleaton

      @jayleaton · x.com

You might also like

Repositories close to jayleaton/glm53-tensorfold-spark
FeaturedAdvertise to 180K+ monthly visitors
Busbar

Control what AI can do before it acts

ClawBench

a benchmark for personal agents

Bifrost

Enterprise AI Gateway Built For Scale: Route, Govern, and Secure All Your AI Traffic

MiaAI-Lab/GLM-5.3-Flash-EXL3-2x-DGX-Sparks-TensorFold

Python

GLM-5.3-Flash EXL3 on 2x DGX Spark with TensorFold

47591

MiaAI-Lab/GLM-5.3-Flash-EXL3-2x-DGX-Sparks

Python

GLM-5.3 Flash EXL3 for 2-4x DGX Sparks

737146

alexellis/glm-5.3-flash-4x-dgx-spark-switchless

Shell

GLM-5.3-Flash (NVFP4) at TP4 across 4x DGX Spark via a switchless RoCE ring + DFlash2 — reproducible recipe

1078

jayleaton/deepseek-v41-tensorfold-spark

Zig

DeepSeek-V4.1-Flash on 2x NVIDIA DGX Spark with the TensorFold engine: 1.8-1.9x code, 2.5x multi-stream, 1.7-1.95x prefill vs the vLLM kit, exact speculative decoding. Work in progress.

8119

tonyd2wild/GLM-5.3-Flash-NVFP4-DFlash2-2x-DGX-Spark

Python

GLM-5.3-Flash (NVFP4) on 2x NVIDIA DGX Spark - vLLM TP2, 262K context, MTP. World-first deploy recipe: 7 day-0 bugs found and fixed, patched sm121 image, probes and full report.

23727

kindlingai/glm-5.3-flash-gx10

Python

GLM-5.3-Flash (NVFP4) at TP=4 across 4x ASUS GX10 / GB10 - vLLM + mentat instead of Ray, 524k context, DFlash2. Minimal working recipe.

9015

MiaAI-Lab/DeepSeek-v4.1-Flash-1017-2x-DGX-Sparks

Shell

DeepSeek-V4.1-Flash-1017 EXL3 on 2x DGX Sparks with TensorFold

23236

MiaAI-Lab/GLM-5.3-Flash-NVFP4-Dual-DGX-Spark

Shell

2x DGX Spark recipe for LibertAIDAI/GLM-5.3-Flash-NVFP4 (NVFP4, Ray TP=2, multimodal)

9011

MiaAI-Lab/Qwen3.8-Flash-Next-Single-DGX-Spark-TensorFold

Shell

Qwen3.8 Flash Next on one DGX Spark (TensorFold)

29150