Açık kaynak radarı

Havuz

Takip ettiğimiz konularda yükselen açık kaynak projeler — yıldız ivmesi, gerçek geliştirme aktivitesi ve AI değerlendirmesiyle.

122

repo · sayfa 1 / 3

Sırala

Yıldız/gün (hız)Son gün (★/gün)Son hafta (★/gün)Son ay (★/gün)Toplam yıldız AI önem skoru Bana uyum Commit (28g)Yazar (28g)En yeni

Konu

Tümü AI / LLM18 Veri9 Geliştirici araçları13 Diğer3 SaaS altyapı13 Web11

Süz

Yalnız hareketli Yalnız analizli≥ 0 ≥ 100 ≥ 500 ≥ 2.000

Yalnız hareketli, son 28 günde commit almamış repoları gizler (yıldızı olan ama bırakılmış projeler). Commit sayısı henüz ölçülmemiş repolar da gizlenir.

EtiketTüm GitHub etiketleri — repo sahiplerinin koyduğu topic'lergöster

Sırala:

Etiketler yükleniyor…

Aktif vllm

1
ZhuLinsen/LiteMoE 26
A MoE inference engine you can read in an afternoon.
8,7/gün ort. 11 commit 1 yazar Python 01 Kas 2026 güncellendi dün
gpu inference llm mixture-of-experts vllm
26
2
Aevonix/FlashAlign 123
ReAlign models instantly, on demand. Per-request alignment for verified users and channels, or Always On. Use directly or with Hermes, OpenC
30,8/gün ort. +3 son gün 7 commit 1 yazar Python 31 Eki 2026 güncellendi 3 gün önce
abliteration hermes inference llama-cpp llm model-alignment openclaw sglang
123
3
Zihao-Wu06/distillops 23
Black-box LLM distillation for any verifiable task, with text or image inputs: teacher API -> rejection sampling -> SFT -> eval -> GRPO
3,3/gün ort. — commit Python 28 Eki 2026 güncellendi bugün
distillation fine-tuning grpo knowledge-distillation llm qwen reinforcement-learning rlvr
23
4
alleyxd/SecondThought 20
Think only when the tool call looks unsure: confidence-gated reasoning for small local models. Code, BFCL results and paper.
1,5/gün ort. +0 son gün — commit Python 22 Eki 2026 güncellendi 6 gün önce
bfcl function-calling llm qwen3 reasoning research tool-calling vllm
20
5
Dami-wq-k03/nano-lmcache 27
📚 An educational, step-by-step implementation of LLM KV cache reuse — learn prefix caching, chunk keys, storage & eviction, from pure Python
1,7/gün ort. +1 son gün — commit Python 19 Eki 2026 güncellendi dün
educational kv-cache llm llm-inference lmcache prefix-caching python vllm
27
6
Lowaltitudeflight/local-llm-tuning-guide 29
A decision library for running local LLMs on multi-GPU machines without NVLink: parallelism, communication, quantization, speculative decodi
1,5/gün ort. +10 son gün — commit Python 16 Eki 2026 güncellendi bugün
benchmarking decision-guide gpu-serving inference-optimization kv-cache llm llm-deployment llmops
29
7
Singularity-Compute/super-decider 19
Decisions, not essays: a choice, a yes or no, or a score with a calibrated probability. Open weights, Jev-compatible API, adaptive thinking.
0,9/gün ort. +0 son gün — commit Python 13 Eki 2026 güncellendi 9 gün önce
ai-agents calibration classification decision-model diffusion-llm guardrails jev llm-routing
19
8
T-Crypt/berth 287
Berth is a lightweight GPU resource manager written in Go. It controls which applications get access to your GPU, when they get it, and how
12/gün ort. +3 son gün 808 commit 2 yazar Go 11 Eki 2026 güncellendi dün
comfyui golang gpu inference llama-cpp llama-swap llm local-ai
287
9
DominikBucko/qwen38-flash-next-3090 13
Run Qwen3.8-Flash-Next on ONE RTX 3090 (24 GB) + 64 GB RAM: 128K context, up to 2,100 tok/s prefill, 43–51 tok/s decode. vLLM runtime with h
0,4/gün ort. +0 son gün — commit Python 30 Eyl 2026 güncellendi 30 Eyl 2026
128k-context 24gb-vram consumer-gpu cpu-offloading int4 kv-cache-quantization llm llm-inference
13
10
StartLuxLabs/StartLux-Decision 1.644
Open decision models from 0.8B to 27B: a calibrated probability for every option in one pass, and a short thought only when unsure. Engine f
45,7/gün ort. +0 son gün 22 commit 2 yazar Python 29 Eyl 2026 güncellendi 12 gün önce
agents calibration classification decision-making gguf inference llama-cpp llm
1.644
11
SerendipityOneInc/instinct 17
Instinct: single-forward-pass decision models (choice / true-false / score) returning a probability per candidate
0,5/gün ort. +0 son gün — commit Python 29 Eyl 2026 güncellendi 30 Eyl 2026
calibration classification decision-model llm qwen transformers vllm
17
12
KickItLikeShika/composimplex 16
Composable LLM decoding on the probability simplex: library and paper code.
0,4/gün ort. +0 son gün — commit Python 28 Eyl 2026 güncellendi 01 Eki 2026
decoding llm mirror-descent optimization pytorch sampling text-generation translation
16
13
modelsphere/modelsphere 172
ModelSphere is an open-source LLM inference infrastructure designed to make production-grade model serving simple, efficient, and continuous
4,5/gün ort. +0 son gün 36 commit 4 yazar Shell 27 Eyl 2026 güncellendi bugün
ansible ascend gpu helmfile inference kubernetes llm modelsphere
172
14
manjunathshiva/opendecider 38
Open, calibrated System 1 decision models: typed choice / score / yes-no answers with a probability for every option. 400M to 80B; CPU, NVID
1/gün ort. +1 son gün — commit Python 27 Eyl 2026 güncellendi dün
apple-silicon calibration classification decision-model huggingface llm pytorch system-one
38
15
caiovicentino/eikos 43
Open, calibrated, single-pass typed-decision models (4B & 27B) for finance and trading
1/gün ort. +0 son gün — commit Python 23 Eyl 2026 güncellendi 02 Eki 2026
calibration decision-making finance llm mlx trading vllm
43
16
nokia-applied-research/AnyJev önem 55gpt6-luna uyum 45gpt6-luna 1.581
AnyJev, open LLM’lerden typed kararlar ve eşiklenebilir olasılıklar üretmek için calibration ve karar katmanları sunan bir Python kütüphanesidir.
Neden yükseliyor · Yedi günde 802 yıldız almış; dikkat çekmesinde etiket gerektirmeyen ilk seviye, az etiketle kalibrasyon iddiası ve modeli fine-tune etmeden karar endpoint’i sunma yaklaşımı etkili görünüyor. Bu ilgi benimseme kanıtı değil; verilen sonuçlar README’deki sınırlı benchmark’a dayanıyor.
35,9/gün ort. +13 son gün 9 commit 3 yazar Python 21 Eyl 2026 güncellendi 14 gün önce
calibration decision-model jev jev-model llm system-one transformers vllm
1.581
17
IST-DASLab/disaggregated-quantization 32
Disaggregated quantization official codebase
0,7/gün ort. +0 son gün — commit Jupyter Notebook 21 Eyl 2026 güncellendi 14 gün önce
llm quantization quantization-aware-training vllm
32
18
lukeckprobierts/Hemmingway-1 46
Hemmingway-1: a 27B model that writes the way a person writes. Open weights, free for non-commercial use (CC BY-NC 4.0); commercial use by a
1/gün ort. +0 son gün — commit 20 Eyl 2026 güncellendi 02 Eki 2026
creative-writing language-model llm open-weights qwen ai-writing gguf humanlike-ai
46
19
ikermoel/open-alternative-jev 76
Open-source alternative to TypeSafe's Jev: a System One style model layer that gives typed, calibrated decisions from any open-weights LLM i
1,6/gün ort. +0 son gün 0 commit 0 yazar Python 18 Eyl 2026 güncellendi 25 Eyl 2026
calibration classification llm transformers vllm jev logprobs open-weights
76
20
kikoncuo/jevfire 76
JEV-inspired parallel decisions for CUDA LLMs. One context, many decisions. vLLM API, game-agent examples, and reproducible benchmarks.
1,6/gün ort. +0 son gün — commit JavaScript 16 Eyl 2026 güncellendi 18 Eyl 2026
cuda game-ai inference jev llm parallel-decoding structured-output vllm
76
21
maximpri/mlxtop 144
A top for local LLMs: memory, paging, GPU and tokens/s for oMLX, Ollama, llama.cpp, vLLM and more
2,7/gün ort. +1 son gün — commit Rust 11 Eyl 2026 güncellendi dün
apple-silicon llama-cpp llm local-llm mlx monitoring ollama rust
144
22
Multicorewareinc/M-Stack 22
Declarative Python SDK for building AI inference platforms on Kubernetes — bare machines to a served model, composed rather than hand-assemb
0,4/gün ort. +0 son gün — commit Python 11 Eyl 2026 güncellendi 22 Eyl 2026
ai-infrastructure gateway-api gpu inference kserve kubernetes llm llmops
22
23
plotarmordev/thegrill 14
Did your change make your LLM server faster? Measure it: faster, slower, or too close to call. Linux and Apple Silicon.
0,3/gün ort. +1 son gün — commit Rust 09 Eyl 2026 güncellendi dün
apple-silicon benchmark cli dgx-spark inference llm llm-inference rust
14
24
Datura-ai/lium-localmaxxing 36
Reproduce Lium's LocalMaxxing leaderboard runs (7 boards, 20 Verified) on a rented Lium GPU in one script — vLLM/llama.cpp recipes, measured
0,6/gün ort. +0 son gün — commit Shell 06 Eyl 2026 güncellendi 26 Eyl 2026
gpu lium reproducible benchmark llama-cpp llm-inference vllm
36
25
alexellis/rigmark 26
RigMark benchmarks local AI the way coding agents actually use it.
0,4/gün ort. +0 son gün — commit Python 05 Eyl 2026 güncellendi 10 gün önce
benchmark dgx-spark inference llm vllm
26
26
HarvardMadSys/chutes_workload 40
one-year LLM serving trace from Chutes
0,7/gün ort. +1 son gün — commit Python 04 Eyl 2026 güncellendi 18 gün önce
inference llm sglang traces vllm
40
27
vcruz305/GLM-5.3-Flash-EXL3-K2-DGX-Spark-recipe 26
vLLM recipe: GLM-5.3-Flash EXL3 K2 on one DGX Spark GB10. Native MTP k=2. Measured tok/s.
0,4/gün ort. +0 son gün — commit Python 29 Ağu 2026 güncellendi 10 Eyl 2026
dgx-spark exl3 glm llm quantization vllm
26
28
Reederey87/glm53-flash-exl3-2x-dgx-spark 64
GLM-5.3-Flash EXL3 (320B MoE) on 2x NVIDIA DGX Spark — production serving kit, 1M context, 97%+ multi-session prefix caching, DFlash2 spec d
0,9/gün ort. +1 son gün — commit Python 27 Ağu 2026 güncellendi bugün
dgx-spark flashinfer gb10 glm llm-inference self-hosted sparse-attention tensor-parallel
64
29
peonist-ai/crucible 13
MoE expert compression toolkit — prune, merge, and forge specialist models for local inference.
0,2/gün ort. +0 son gün — commit Python 27 Ağu 2026 güncellendi dün
expert-pruning gguf llama-cpp llm llm-inference local-llm machine-learning mixture-of-experts
13
30
whitecircle/halo 841
Halo is an open-source framework built by White Circle for training large language and multimodal models
12/gün ort. +1 son gün 5 commit 2 yazar Python 26 Ağu 2026 güncellendi dün
agent async-rl distributed-training fine-tuning grpo large-language-models llm lora
841
31
ccebelenski/localagents 16
MCP server that delegates Claude Code agent work to local models (llama.cpp / vLLM) on your own hardware
0,2/gün ort. +3 son gün — commit Python 26 Ağu 2026 güncellendi 2 gün önce
agents claude-code llama-cpp local-llm mcp vllm
16
32
funkeyyou/codex-model-router 4
模型路由器：在 Codex Desktop 和 Claude Desktop 使用 Claude、第三方 API 与自定义模型。Local LLM proxy for Codex & Claude Desktop: OpenAI Responses / Chat Completi
0,1/gün ort. +0 son gün — commit JavaScript 26 Ağu 2026 güncellendi bugün
anthropic claude codex codex-desktop llm-proxy model-router openai openai-compatible
4
33
zorost/sparkduet 0
Serve DeepSeek and Qwen, run an on-demand model library, and fine-tune with Unsloth QLoRA on 2x NVIDIA DGX Spark. TP=2 vLLM lanes behind one
0/gün ort. +0 son gün — commit Python 25 Ağu 2026 güncellendi 07 Eyl 2026
coding-agents deepseek dgx-spark fine-tuning gb10 homelab llama-swap llm-inference
0
34
XHToken/Spark-X2.6 önem 52gpt6-luna uyum 35gpt6-luna 879
Spark-X2.5, yerel çalıştırma ve agent iş akışları için tasarlanmış 1.7B ve 4B boyutlarında açık dil modelleri serisidir.
Neden yükseliyor · Yeni model serisi; uzun context, küçük model boyutu ve agent/coding yeteneklerini aynı anda öne çıkarıyor. README'de 34 günlük yaş, 386 yıldız ve son 28 günde 9 commit ile 2 farklı yazar bilgisi var; bunlar erken ilgi ve güncel çalışma gösteriyor, ancak model performansını bağımsız olarak kanıtlamıyor.
12,2/gün ort. +2 son gün 3 commit 2 yazar 24 Ağu 2026 güncellendi 2 gün önce
agent huggingface llm modelscope ollama sglang spark transformer
879
35
Yzw202011/OmniSpace 530
本地优先的全栈 AI 创作工作站：AI对话 · AI漫画 · 漫剧创作 · 写作台 · 知识学习五大模块；vLLM / llama.cpp / Transformers 多引擎本地推理，显存智能调度、舱壁隔离自愈、仅供学习使用，欢迎交流指正。（最新进度（除插件系统）受到协议限制，
7,3/gün ort. -1 son gün 0 commit 0 yazar Python 23 Ağu 2026 güncellendi 20 Eyl 2026
ai-comic ai-manga fastapi knowledge-graph llama-cpp local-first local-llm novel-writing
530
36
aic0d3r/qwen38-strix-halo-harness 37
AI coding agent harness for any hardware, local or cloud (OpenAI-compatible engines). Tuned for AMD Strix Halo: Qwen3.8 125B on the iGPU, NP
0,5/gün ort. +0 son gün — commit TypeScript 23 Ağu 2026 güncellendi dün
ai-coding-assistant amd benchmark coding-agent llama-cpp llm lm-studio local-llm
37
37
anakronox/spark-dash 16
A dashboard for a home cluster of NVIDIA GB10 (DGX Spark / ASUS GX10) inference servers: GPU and system health, plus live per-model metrics
0,2/gün ort. +0 son gün — commit Python 22 Ağu 2026 güncellendi 3 gün önce
dashboard dgx-spark gb10 grafana homelab llama-cpp monitoring nvidia
16
38
amasu/dgx-spark-qwen38 13
Docker compose configs for serving Qwen3.8-27B on a DGX Spark (GB10): vLLM+MTP current, SGLang+DSPARK + FP8 rollback stacks
0,2/gün ort. +0 son gün — commit Shell 18 Ağu 2026 güncellendi 8 gün önce
blackwell dgx-spark docker docker-compose fp8 gb10 llm llm-serving
13
39
0xBakeer/Qwen3.8-27B-FP8-on-a-single-DGX-Spark 55
Serving Qwen3.8-27B-FP8 on a single DGX Spark (GB10): 7.88 to 58.5 tok/s single-stream from decode strategy alone, weights untouched. Specul
0,7/gün ort. +0 son gün — commit Python 16 Ağu 2026 güncellendi 30 Eyl 2026
benchmark dflash2 dgx-spark fp8 gb10 llama-cpp llm-inference llm-serving
55
40
0xBakeer/Qwen3.8-27B-4-bit-on-a-single-DGX-Spark 51
Serving 4-bit Qwen3.8-27B on a single DGX Spark (GB10): 75 tok/s single-stream, 246 tok/s aggregate at 8-way concurrency. NVFP4 vs MixedInt4
0,6/gün ort. +0 son gün — commit Python 16 Ağu 2026 güncellendi 30 Eyl 2026
autoround benchmark dgx-spark gb10 int4 llm-inference llm-serving marlin
51
41
hasso5703/alzeo 245
dgx-spark-qwen38 is an open-source installer that turns an NVIDIA DGX Spark, or any GB10 machine (ASUS, Dell, HP, Lenovo, MSI, GIGABYTE, Ace
3/gün ort. +0 son gün — commit Python 15 Ağu 2026 güncellendi dün
ai-agents asus-ascent-gx10 dgx-spark gb10 image-generation llm-inference local-ai local-llm
245
42
neilthomas89440-crypto/inferna-next 24
Self-hosted GPU cluster orchestration: deploy and serve open-source AI models on your own hardware — FastAPI control plane, worker agents, v
0,3/gün ort. +0 son gün — commit Python 15 Ağu 2026 güncellendi dün
docker fastapi gpu-cluster gpu-orchestration grpc inference-server llm-serving model-serving
24
43
axem-solutions/shaide 42
Self-hosted, distributed, multi-model LLM inference
0,5/gün ort. +0 son gün — commit Go 13 Ağu 2026 güncellendi 5 gün önce
ai-platform embeddings enterpise enterpise-ai k8s llm llm-d llms
42
44
ruashots/open-h3-ir 127
Open-source local implementation of MiniMax H3 Context-IR. Turn simple prompts into structured, validated H3 video briefs.
1,5/gün ort. +1 son gün 0 commit 0 yazar Python 12 Ağu 2026 güncellendi 23 Ağu 2026
comfyui generative-video local-llm minimax minimax-h3 prompt-compiler prompt-engineering text-to-video
127
45
KhaiTrang1995/llm-internal 19
MLOps — vLLM + LiteLLM + Portainer CE + Netdata. Case study song ngữ: kiến trúc, metrics, bài học vận hành. · Self-hosted LLM inference don
0,2/gün ort. +0 son gün — commit HTML 12 Ağu 2026 güncellendi 16 Ağu 2026
case-study case-study-project docker gpu llm netdata observability portainer
19
46
infercrane/infercrane 53
Open-source infrastructure for the full inference lifecycle: deploy, observe, scale, optimize, and safely release self-hosted models behind
0,6/gün ort. +0 son gün — commit Go 09 Ağu 2026 güncellendi dün
ai-infrastructure aws gcp golang inference kubernetes llm mlops
53
47
WestWaters/pollard-weights 33
A tiered-memory system design for workloads that don't fit in RAM: measure the working set, pin the hot tier, stream the cold tier from flas
0,4/gün ort. +0 son gün — commit Python 07 Ağu 2026 güncellendi dün
gguf llm low-resource memory memory-optimization memory-tiering mixture-of-experts quantization
33
48
azrtydxb/Fastllm-proxy 108
The lowest-overhead LLM router. Production-ready, highly available, one OpenAI-compatible endpoint in front of 80 providers and your own vLL
1,2/gün ort. +0 son gün 63 commit 2 yazar Rust 06 Ağu 2026 güncellendi 2 gün önce
ai-gateway inference litellm llm llm-gateway load-balancing observability openai-api
108
49
david-g-3654/homebench 59
Benchmark your local LLMs: speed, memory, and quality, in one command. TUI leaderboard for Ollama, LM Studio, llama.cpp, and vLLM.
0,6/gün ort. +0 son gün — commit Python 03 Ağu 2026 güncellendi 10 Ağu 2026
benchmark cli evaluation llama-cpp llm lm-studio local-ai local-llm
59
50
openfabric-systems/simllm 51
Network-faithful simulation of LLM serving and training deployments
0,5/gün ort. +0 son gün — commit Python 03 Ağu 2026 güncellendi 27 gün önce
distributed-training inference llm networking rocev2 sglang simulation vllm
51