canada-quant/dsv4-flash-w4a16-spark-image
vllm-w4a16-dsv4:exp — pre-built Docker image for DGX Spark TP=2 This dataset hosts a pre-built OCI tarball of the Docker image used to serve pastapaul/DeepSeek-V4-Flash-W4A16-FP8 on dual DGX Spark GB10 (SM 12.1a) — for users who can't build the image themselves due to network constraints. The image is also fine for 2× RTX PRO 6000 Blackwell Server (SM 12.0). What's inside vLLM build pinned to: jasl/vllm@ds4-sm120-experimental (HEAD c05638d70 after cherry-pick) —… See the full description on the dataset page: https://huggingface.co/datasets/canada-quant/dsv4-flash-w4a16-spark-image.
vllm-w4a16-dsv4:exp — pre-built Docker image for DGX Spark TP=2
This dataset hosts a pre-built OCI tarball of the Docker image used to serve `pastapaul/DeepSeek-V4-Flash-W4A16-FP8` on dual DGX Spark GB10 (SM 12.1a) — for users who can't build the image themselves due to network constraints.
The image is also fine for 2× RTX PRO 6000 Blackwell Server (SM 12.0).
What's inside
vLLM build pinned to:
- `jasl/vllm@ds4-sm120-experimental` (HEAD
c05638d70after cherry-pick) — SM12x DSV4 support + experimental superset (split-KV decode, GB10 fused-MoE config aliases, tuned MLA graph defaults) - Cherry-pick
f910a73a93fromneuralmagic/vllm@kylesayrs/deepseek-ct(vLLM PR #41276 — DSV4 compressed-tensors attention) - Local patch `patch_v4_packed_mapping.py` — adds
packed_modules_mappingtoDeepseekV4ForCausalLM - Workspace pre-reservation patch is not applied — landed upstream as
jasl/vllm@1d6f5c4
Compressed tarball: ~9 GB. Uncompressed image: 20.2 GB. ARM64 only.
How to use
# Download
huggingface-cli download pastapaul/dsv4-flash-w4a16-spark-image \
vllm-w4a16-dsv4-exp.tar.gz \
--repo-type dataset \
--local-dir .
# Load on each Spark
gunzip -c vllm-w4a16-dsv4-exp.tar.gz | docker load
# Verify
docker images vllm-w4a16-dsv4:expOnce loaded on both Sparks, you can use the bootstrap script with --skip-build:
curl -fsSLO https://raw.githubusercontent.com/pasta-paul/dsv4-flash-w4a16-fp8/main/scripts/bootstrap_dsv4_spark.sh
chmod +x bootstrap_dsv4_spark.sh
./bootstrap_dsv4_spark.sh \
--head-host spark-a \
--worker-host spark-b \
--skip-buildThe script will then handle network setup, model pre-cache (no HF token needed), container launch on both nodes, and wait for /health=200.
What you get
Production canonical at 1 M-token context TP=2 graphs-ON:
- decode 12 t/s smoke / 14–15 t/s think-max sustained
- NIAH 4/4 retrieval at 200K-token haystack
- mini-suite 10/10 PASS · think-max 3/3 PASS
- GSM8K 95.00% strict / 94.92% flex (preserved from prior canonical)
- tool calling: parallel function calls with structured arguments
Reference docs
- Model: https://huggingface.co/pastapaul/DeepSeek-V4-Flash-W4A16-FP8
- Reproduction repo + raw evidence: https://github.com/pasta-paul/dsv4-flash-w4a16-fp8
- Quickstart: https://github.com/pasta-paul/dsv4-flash-w4a16-fp8/blob/main/findings/QUICKSTARTDUALSPARK.md
- Phase 4e validation report: https://github.com/pasta-paul/dsv4-flash-w4a16-fp8/blob/main/findings/sparktp2deployment.md#phase-4e--production-canonical-at-1-m-context-on-ds4-sm120-experimental-2026-05-06
