GLM-5.2 QuantTrio TP=4+DCP2 on 4x NVIDIA DGX Spark (GB10) — tuned and measured under a real agent workload. Adaptive MTP, the indexer law, and the negative results.
How fast joesinvestments/GLM-5.2-QuantTrio-TP4-DCP2-4x-DGX-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
Nov 4 – Nov 20, 2026
Stars +1in 17 days
01
1
Forks +0in 17 days
01
Merged PRs 0in 17 days
01
Trendshift Trendingwhere the repository ranked each day it made the top 25
Not in the Trendshift top 25 in this period
Nov 4Nov 7Nov 10Nov 13Nov 16Nov 19
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
Day
Stars gained
Forks gained
Merged PRs
Wed, Nov 4, 2026
+1
+0
0
Thu, Nov 5, 2026
+0
+0
0
Fri, Nov 6, 2026
+0
+0
0
Sat, Nov 7, 2026
+0
+0
0
Sun, Nov 8, 2026
+0
+0
0
Mon, Nov 9, 2026
+0
+0
0
Tue, Nov 10, 2026
+0
+0
0
Wed, Nov 11, 2026
+0
+0
0
Thu, Nov 12, 2026
+0
+0
0
Fri, Nov 13, 2026
+0
+0
0
Sat, Nov 14, 2026
+0
+0
0
Sun, Nov 15, 2026
+0
+0
0
Mon, Nov 16, 2026
+0
+0
0
Tue, Nov 17, 2026
+0
+0
0
Wed, Nov 18, 2026
+0
+0
0
Thu, Nov 19, 2026
+0
+0
0
Fri, Nov 20, 2026
+0
+0
0
Total
+1
+0
0
1#1 on the list
7Rank that day
Not trending
Social mentions
Recent discussions about this repository across the web
2 posts
Adopted your decode-aware prefill mod - the one your Dockerfile doesn’t build. Per-request decode at concurrency 12 went 0.9 -> 6.3 tok/s for us. ~6.8×. Biggest win in the repo for anyone running…
GLM-5.3-Flash (320B MoE) at TP4 across four DGX Sparks, same day as the model drop: 36 tok/s, 1.26M-token FP8 KV pool, 262K context, MTP spec decode. First TP4 glm5_next outside B200 hardware. Full patched-image recipe + the GB10 memory study.