DeepSeek V4 Flash-0731 on 2x RTX PRO 6000 Blackwell — Karmic Kraken TP2 DSpark K5

Launch command

serve deepseek-ai/DeepSeek-V4-Flash-0731 --async-scheduling --attention-backend B12X --block-size 256 --code-revision 9e165c30e2704aec5d9d593cce3eebd58bbef1cb --compilation-config ''\''{"cudagraph_mode":"FULL_AND_PIECEWISE","custom_ops":["all"]}'\''' --decode-context-parallel-size 1 --default-chat-template-kwargs ''\''{"thinking":true,"reasoning_effort":"high"}'\''' --dtype bfloat16 --enable-auto-tool-choice --enable-chunked-prefill --enable-flashinfer-autotune --enable-force-include-usage --enable-prefix-caching --enable-prompt-tokens-details --enable-request-id-headers --gpu-memory-utilization 0.975 --host 0.0.0.0 --kv-cache-dtype fp8 --load-format instanttensor --max-cudagraph-capture-size 48 --max-model-len -1 --max-num-batched-tokens 4096 --max-num-seqs 8 --moe-backend b12x --override-generation-config ''\''{"temperature":1.0,"top_p":0.95}'\''' --pipeline-parallel-size 1 --port 8000 --prefix-cache-retention-interval 4096 --reasoning-parser deepseek_v4 --revision 9e165c30e2704aec5d9d593cce3eebd58bbef1cb --no-scheduler-reserve-full-isl --served-model-name DeepSeek-V4-Flash-0731 --speculative-config ''\''{"method":"dspark","draft_sample_method":"probabilistic","rejection_sample_method":"standard","num_speculative_tokens":5,"model":"deepseek-ai/DeepSeek-V4-Flash-0731","revision":"9e165c30e2704aec5d9d593cce3eebd58bbef1cb"}'\''' --tensor-parallel-size 2 --tokenizer-mode deepseek_v4 --tool-call-parser deepseek_v4 --trust-remote-code

Use this recipe with your agent

Paste this into any agent chat — it fetches and applies the recipe:

Fetch https://localairecipes.com/api/v1/recipes/ds4-flash-2x6000-kk-dspark-k5.json and apply the inference recipe it describes (engine flags, model, quant, context) to my setup. Ask before overwriting any existing config.

Or fetch it directly:

curl -fsSL https://localairecipes.com/api/v1/recipes/ds4-flash-2x6000-kk-dspark-k5.json

Install requirements

• Docker (or a compatible OCI container runtime)
• NVIDIA Container Toolkit (nvidia-container-toolkit) - exposes host GPUs to the container

NVIDIA driver floor: unknown - check the image tag

Source: NVIDIA CUDA Toolkit Release Notes - driver compatibility tables (https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/, fetched 2026-10-03).

Map the placeholders to your rig

• Hardware-bound values in this recipe (device counts, bind-mount paths, utilization, ports) describe the author's machine, not yours - map them to your own rig before launching. Keep the engine flags, model, quant, and parallelism strategy; adjust paths and ports to the hardware you actually have.
• GPU identity: list your GPUs with `nvidia-smi -L` and substitute your own GPU UUIDs/indices wherever the recipe pins the author's.
• Model directory: bind-mount paths are author-side - point them at your own mount where the model files live.
• Host port: pick a free port on your machine and publish the container there; keep the in-container port unchanged.

Revision notes

Benchmark-qualified configuration from local-inference-lab/rtx6kpro (Karmic Kraken serving report). Five warmed 30-second windows per cell on RTX PRO 6000 Blackwell Max-Q; decode ctx-0 temp 1. Launch argv transcribed verbatim from the repo's unified Docker guide (same composition). Measured: C1 212.5 tok/s, C8 667.5 tok/s, 32K prefill 11392 tok/s.

Revision history

• bfdf9daa3e history: re-squash corpus to a single root commit after the slug reconciliation (prior tip kept at refs/lar-archive/pre-single-slug-resquash) 25 d ago

Fork

Sign in to fork this recipe — a fork is your own copy; you own it and only you can update it.

Export

recipe.yaml (raw)