GRGracana·Aug 3, 2026
Linking here since the pruning method you're talking about wasn't the top search result when I looked: https://github.com/CerebrasResearch/reap
In Qwen3.8-Max: A New Bar for Coding and CoworkCerebrasResearch
reap
Python2026–2026
CerebrasResearch / reap
reap
REAP: Router-weighted Expert Activation Pruning for SMoE compression
2Distinct people
2Mentions
0Explicit recommenders
541GitHub stars
What to know before exploring
What it is
REAP: Router-weighted Expert Activation Pruning for SMoE compression
GitHub
Where it is used
No documented product use has been verified yet.
What people are saying
No recommendation phrase was found in the indexed comments.
No classified warning is attached to this page. Read the discussions before treating silence as approval.
Also in these discussions
No other repository is repeatedly linked in the same discussions.
This week on Hacker News
Nov 7–Nov 14 UTC · Direct repository links and separately labeled related project stories. Indexed through 11/14/2026, 5:04:57 PM UTC.
No direct repository links were found in the indexed HN discussions over the past seven days. Older discussions are available below.
Mentions over time
HN mentions by week2 dated mentions
Observed mentions: 2026-06-22 20:03:54 UTC to 2026-08-03 12:21:59 UTC. Weeks start Monday, UTC.
Counts come from the repository’s dated source evidence. Index cutoff: 2026-11-02 08:16:59 UTC.
View mention counts and date ranges
| Period | Start (UTC) | End (exclusive, UTC) | Mentions | Submissions | Comments |
|---|---|---|---|---|---|
| Jun 22 | 2026-06-22 00:00:00 UTC | 2026-06-29 00:00:00 UTC | 1 | 0 | 1 |
| Jun 29 | 2026-06-29 00:00:00 UTC | 2026-07-06 00:00:00 UTC | 0 | 0 | 0 |
| Jul 6 | 2026-07-06 00:00:00 UTC | 2026-07-13 00:00:00 UTC | 0 | 0 | 0 |
| Jul 13 | 2026-07-13 00:00:00 UTC | 2026-07-20 00:00:00 UTC | 0 | 0 | 0 |
| Jul 20 | 2026-07-20 00:00:00 UTC | 2026-07-27 00:00:00 UTC | 0 | 0 | 0 |
| Jul 27 | 2026-07-27 00:00:00 UTC | 2026-08-03 00:00:00 UTC | 0 | 0 | 0 |
| Aug 3 | 2026-08-03 00:00:00 UTC | 2026-08-10 00:00:00 UTC | 1 | 0 | 1 |
What Hacker News says
2 sourcesInspired by [https://github.com/antirez/ds4](ds4), [https://github.com/CerebrasResearch/reap](REAP) and [https://huggingface.co/0xSero/DeepSeek-V4-Flash-162B](OxSero's Deepseek v4 reap) I wanted to push and see how much performance can be extracted from a single DGX Spark. It should also work day one (hopefully) on the upcoming Spark devices.
I made three versions, one with 128 experts kept, one with 150 and the biggest (borderline fitting one) with 180 experts out of 256. Experts kept are based around coding / agentic / research workloads.
Goal is to have a higher-precision (NVFP4) option to run the model, the original full ds4 already runs the IQ2XXS version. Custom CUDA kernels are written to try and best align the NVFP4 models to the Spark.
The K180 runs at around 119/122GB ram usage at the full 1M context, tested up to 32k prefill an
Read the rest
d was stable. For best memory efficiency, you might need DS4_CUDA_MANAGED_MODEL=1 DS4_KV_TURBO=1. More memory/bandwidth optimizations are coming, after that I plan on tackling re-adjusting the MTP heads (which would require re-training them on the new architectures).
Benchmarking hasn't been done yet, as I have mostly been busy with the CUDA. Treat as experimental.
Model links:
https://huggingface.co/sleepyeldrazi/DeepSeek-v4-Flash-REAP-...
https://huggingface.co/sleepyeldrazi/DeepSeek-v4-Flash-REAP-...
https://huggingface.co/sleepyeldrazi/DeepSeek-v4-Flash-REAP-...