Hugging Face's logo

Datasets:
canada-quant
/
dsv4-flash-w4a16-spark-image

Tags:
deepseek-v4
vllm
dgx-spark
blackwell
sm120
sm121
License:
Dataset card Files
xet
 Community
dsv4-flash-w4a16-spark-image / README.md
pastapaul's picture
pastapaul
Initial README
70d7ec5 verified
preview code
|
raw
history blame contribute delete
2.94 kB
metadata 
license: apache-2.0
tags:
  - deepseek-v4
  - vllm
  - dgx-spark
  - blackwell
  - sm120
  - sm121
  - docker-image
  - gb10

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) — SM12x DSV4 support + experimental superset (split-KV decode, GB10 fused-MoE config aliases, tuned MLA graph defaults)
  • Cherry-pick f910a73a93 from neuralmagic/vllm@kylesayrs/deepseek-ct (vLLM PR #41276 — DSV4 compressed-tensors attention)
  • Local patch patch_v4_packed_mapping.py — adds packed_modules_mapping to DeepseekV4ForCausalLM
  • 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:exp

Once 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-build

The 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