讲任何设备组合到一起，建设成Ai集群
Python · active 2025-05-05 → 2026-08-07 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 460 days from 2025-05-05 to 2026-08-07. Pushes: 2,632 total, peak 84 in a day. Pull requests: 335 total, peak 12 in a day. Issues: 393 total, peak 136 in a day. Comments: 1,042 total, peak 109 in a day. Stars: 7,204 total, peak 593 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
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| Evanev7 | 1.3K | 792 | 48 | 258 |
| rltakashige | 1.1K | 885 | 48 | 94 |
| AlexCheema | 687 | 362 | 80 | 164 |
| JakeHillion | 257 | 148 | 39 | 34 |
| ciaranbor | 211 | 179 | 22 | 4 |
| AndreiCravtov | 170 | 166 | 0 | 3 |
| davidmcc73 | 80 | 80 | 0 | 0 |
| samiamjidkhan | 31 | 23 | 4 | 3 |
| Drifter4242 | 30 | 0 | 4 | 13 |
| DevKiDCosmo | 20 | 0 | 0 | 13 |
| diffray-bot | 17 | 0 | 0 | 17 |
| nightguarder | 16 | 0 | 1 | 11 |
| rafipatel | 13 | 0 | 1 | 6 |
| southkorea2013 | 9 | 0 | 0 | 1 |
| tinybluedev | 9 | 0 | 1 | 5 |
| frenchtoasters | 9 | 0 | 0 | 5 |
| lucasajackson | 9 | 0 | 0 | 4 |
| mustafalpyilmaz | 8 | 0 | 7 | 0 |
| Wysie | 8 | 0 | 0 | 5 |
| joanwhitegt-art | 7 | 0 | 0 | 7 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#957shoemoney2026-07-29 21:39[MEDIUM] Better placement algorithm for pipeline parallelism using memory bandwidth + latency
- Pull request#2144Evanev72026-06-03 15:33
- Issue comment#19795F0jd2vLq54RerYW2026-06-01 00:44feat(api): /events accepts since and limit query params with X-EXO-Last-Idx header
- Issue comment#21345F0jd2vLq54RerYW2026-06-01 00:29fix(security): block SSRF in fetch_image_url via scheme allowlist and private IP rejection
- Issue comment#2076Evanev72026-05-28 22:45libp2p -> zenoh pure
- Issue comment#2107raajkumars2026-05-23 13:29fix(runner_supervisor): wrap blocking pipe/join ops in to_thread to prevent event loop stall
- Pull request#2107raajkumars2026-05-22 22:07
- Issue comment#2103Winston-95272026-05-22 05:17feat: Add Native CUDA Support for Linux GPU Inference
- Pull request#2103Winston-95272026-05-22 05:03
- Issue#2086Forfar-20202026-05-16 07:21How to config and load the DeepSeek v3.2 model
- Issue#2088PixelMyoos2026-05-12 22:12macmon causes MLX Metal GPU crash during inference — recommend mactop instead
- Issue#2085atlasshrugged10002026-05-11 21:45[BUG] Kimi-K2.6 fails to load after upstream HuggingFace repo removed tokenization_kimi.py
- Issue#2082AlexCheema2026-05-10 20:30[BUG] Streaming chat completions emit null-valued delta fields, breaking some OpenAI clients
- Issue comment#2077Evanev72026-05-10 17:39[BUG] 2-node Mac cluster: asymmetric topology blocks 2-node placements
- Issue comment#2061Evanev72026-05-10 16:57Map GLM 4.7 stop tokens to GLM 4 IDs
- Issue comment#2059team-wcv2026-05-10 09:14Stabilize master election and event replay
- Pull request#2079team-wcv2026-05-10 09:13
- Issue#1850jagmarques2026-05-08 21:14KV cache compression for distributed inference memory savings
- Issue comment#1850jagmarques2026-05-08 21:14KV cache compression for distributed inference memory savings
- Issue#2069130513325982026-05-08 10:54[Performance] Significant latency when calling local Exo models via OpenClaw compared to WebUI
- Pull request#2064team-wcv2026-05-07 00:11
- Pull request#2063team-wcv2026-05-07 00:06
- Issue comment#2052orbisai0security2026-05-06 01:55fix: the rust networking layer uses mdns for distrib... in discovery.rs
- Issue comment#1102Evanev72026-05-04 22:27Mac Studio + DGX Spark
- Issue comment#2007bobheldt2026-05-04 22:19[BUG] Unstable node count and DHCP issues on Mac Studio M3 Ultra cluster via Thunderbolt
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 7,204 stars here means stars gained during the window, not the repo's star count.