From 734e1e2965cb911295d1fafb8855de28a3a23897 Mon Sep 17 00:00:00 2001 From: fzyzcjy <5236035+fzyzcjy@users.noreply.github.com> Date: Fri, 24 Apr 2026 13:23:50 +0800 Subject: [PATCH] Further update Deepseek V4 docs (#23617) --- .../autoregressive/deepseek-v4-deployment.jsx | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/docs_new/src/snippets/autoregressive/deepseek-v4-deployment.jsx b/docs_new/src/snippets/autoregressive/deepseek-v4-deployment.jsx index 5b03e33ae..58cd02fe4 100644 --- a/docs_new/src/snippets/autoregressive/deepseek-v4-deployment.jsx +++ b/docs_new/src/snippets/autoregressive/deepseek-v4-deployment.jsx @@ -137,12 +137,11 @@ export const DeepSeekV4Deployment = () => { "b200|big": { slug: "deepseek-ai/DeepSeek-V4-Pro", tp: 8, multinode: false }, "gb300|small": { slug: "deepseek-ai/DeepSeek-V4-Flash", tp: 4, multinode: false }, "gb300|big": { slug: "deepseek-ai/DeepSeek-V4-Pro", tp: 4, multinode: false }, - // H200 needs a separate FP8-only Instruct ckpt (Flash / Pro public repos - // ship FP4-mixed weights). That ckpt is still being uploaded, so we emit a - // placeholder that fails loudly on copy-paste instead of silently pulling - // the wrong weights. Replace with the real slug once Hopper ckpts are public. - "h200|small": { slug: "", tp: 4, multinode: false }, - "h200|big": { slug: "", tp: 16, multinode: true, nnodes: 2 }, + // H200 needs an FP8-only Instruct ckpt (deepseek-ai's Flash/Pro repos ship + // FP4-mixed weights that Hopper can't run). sgl-project publishes FP8 + // repackagings; Flash is public, Pro is still being uploaded. + "h200|small": { slug: "sgl-project/DeepSeek-V4-Flash-FP8", tp: 4, multinode: false }, + "h200|big": { slug: "", tp: 16, multinode: true, nnodes: 2 }, }; // Per (hardware, modelSize) PD role TP (from allinone _PD_SPEC). const PD_TP_SPEC = {