Frank 9971bdef0f feat: initial vLLM Qwen3.6-35B-A3B-FP8 serve script with DFlash drafting on dankserver ...

- Podman container with CUDA GPU access
- DFlash speculative decoding via --spec-method dflash
- FP8 KV cache, 4 parallel streams, 262K context
- CPU expert offloading via --cpu-offload-gb
- systemd service for persistent runtime

2026-08-07 18:28:38 -07:00

..

Dankclaw System Specs.md

docs: fix dankclaw specs (30GiB RAM, 96GiB GPU-visible, ROCm 7.2.4) + add validate.sh

2026-08-06 22:40:44 -07:00

Dankserver System Specs.md

feat: initialize vLLM scripts knowledge base

2026-08-05 23:49:02 -07:00

Hybrid CPU GPU Processing Dankserver.md

feat: initialize vLLM scripts knowledge base

2026-08-05 23:49:02 -07:00

index.md

chore: fix typo in dankclaw specs, format index table

2026-08-06 22:29:23 -07:00

RAM CPU Offloading in VLLM.md

feat: initialize vLLM scripts knowledge base

2026-08-05 23:49:02 -07:00

Running Qwen3.6 27B FP8 on Dankclaw.md

docs: fix dankclaw specs (30GiB RAM, 96GiB GPU-visible, ROCm 7.2.4) + add validate.sh

2026-08-06 22:40:44 -07:00

Running Qwen3.6 35B A3B FP8 on Dankserver.md

feat: initialize vLLM scripts knowledge base

2026-08-05 23:49:02 -07:00

Shared Memory Architecture Dankclaw.md

docs: fix dankclaw specs (30GiB RAM, 96GiB GPU-visible, ROCm 7.2.4) + add validate.sh

2026-08-06 22:40:44 -07:00

VLLM Index.md

feat: initialize vLLM scripts knowledge base

2026-08-05 23:49:02 -07:00

VLLM Overview.md

feat: initialize vLLM scripts knowledge base

2026-08-05 23:49:02 -07:00

vllm_research_findings.md

feat: initial vLLM Qwen3.6-35B-A3B-FP8 serve script with DFlash drafting on dankserver

2026-08-07 18:28:38 -07:00