HN · GitHub Signals

OPEN SOURCE, IN FOCUS

glm-5.2-quanttrio-200k-4x-dgx-spark--36tok-s

tonyd2wild/glm-5.2-quanttrio-200k-4x-dgx-spark--36tok-s

Recipe: GLM-5.2 (unpruned QuantTrio Int4-Int8Mix) at 200K ctx with MTP spec decode on a 4x NVIDIA DGX Spark (GB10) cluster

Explore on GitHub ↗

Saved projects stay in this browser. No account needed.

TAKE A CLOSER LOOK

Start with the source.

The repository is the place to check installation steps, supported platforms, the license, and current maintenance before choosing a tool.

Read the documentation on GitHub ↗

Keep exploring

Find projects in the same areas, or search the archive for a different approach.

ANOTHER GOOD FIND

Explore related projects

View the archive ↗

Connected by topics and descriptions.

tonyd2wild

deepseek-v4-flash-dgx-spark

Working recipe to serve DeepSeek-V4-Flash across two NVIDIA DGX Spark (GB10) nodes with vLLM (TP=2, FP8 KV, MTP) over a RoCE/RDMA link — Docker image, launch scripts, RDMA/NCCL setup, and the gotchas.

Take a look →
ciprianveg

gb10-vllm

vLLM inference solutions for NVIDIA GB10 (DGX Spark, sm121) — KIMI-K3 B12X_MLA + DSpark, GLM-5.2

Take a look →
raullenchai

twinspark

Production-grade recipe for DeepSeek-V4-Flash-0731 (284B MoE) on 2x NVIDIA DGX Spark: self-healing 2-node vLLM cluster, reboot-verified, tuned DSpark speculative decoding (~75 tok/s), full benchmarks, OpenAI Codex CLI integration

Take a look →