Skip to content

[AMD] [GLM5] GLM-5.1 MXFP4 (MI355X) + enable EAGLE for gfx950 in cookbook - #29194

Merged
HaiShaw merged 1 commit into
sgl-project:mainfrom
Raiden-Makoto:RM/glm51-cookbook-mxfp4
Jun 25, 2026
Merged

HaiShaw merged 1 commit into
sgl-project:mainfrom
Raiden-Makoto:RM/glm51-cookbook-mxfp4

Conversation

@Raiden-Makoto

@Raiden-Makoto Raiden-Makoto commented Jun 24, 2026 •

Copy link
Copy Markdown
Contributor

Motivation

The SGLang Cookbook GLM-5.1 page documents AMD deployment only for the BF16/FP8 checkpoints, does not cover the MXFP4 checkpoint (amd/GLM-5.1-MXFP4) on MI355X (gfx950), and states that EAGLE speculative decoding is unsupported on AMD (with the deploy generator gating it off for all AMD GPUs). This follows up #28975 (opt-in Triton fp8 sparse-MLA prefill kernel for gfx950) and corrects the EAGLE support status for gfx950.

Modifications

  • glm-51-deployment.jsx (interactive deploy generator):
    • Add an MXFP4 quantization option, enabled and default on MI355X (gfx950); disabled on other hardware. Add mi355x.mxfp4 = { tp: 4, mem: 0.85 }; generate --model-path amd/GLM-5.1-MXFP4 --tp 4 --kv-cache-dtype fp8_e4m3 plus the existing AMD tilelang / chunked-prefill / watchdog flags.
    • EAGLE emit change: the old generator gated the Speculative Decoding toggle off for all AMD, and the old cookbook said EAGLE was unsupported on AMD. We later tested EAGLE on gfx950 (MI355X) and it works; gfx942 (MI300X/MI325X) is unverified. The generator now emits the EAGLE flags by default and excludes them only on MI300X/MI325X (gfx942). The (useless on/off) Speculative Decoding toggle was removed in favor of always emitting the flags where supported.
  • GLM-5.1.mdx:
    • Add an MXFP4 column to the hardware table (tp=4 on MI355X). Add a MXFP4 (MI355X / gfx950) command block in section 4.2 (prefixed with the optional SGLANG_DSA_TRITON_PREFILL=1, and including the EAGLE flags).
    • AMD note: "EAGLE speculative decoding is supported on MI355X (gfx950) and unverified on MI300X/MI325X (gfx942)" (was: "not currently supported on AMD").

Verification

Built locally with mint dev / mint validate (build validation passed). EAGLE on MI355X validated separately on hardware (lossless GSM8K, large ITL/throughput improvement). gfx942 (MI300X/MI325X) EAGLE verification can be a follow-up.

Scope: GLM-5.1 only; MXFP4 + EAGLE on MI355X (gfx950).


CI States

Latest PR Test (Base): ✅ Run #28133913757
Latest PR Test (Extra): ✅ Run #28151181987

@gemini-code-assist

Copy link
Copy Markdown
Contributor

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

@github-actions github-actions Bot added the documentation Improvements or additions to documentation label Jun 24, 2026
@Raiden-Makoto Raiden-Makoto changed the title [AMD] [GLM5] Add GLM-5.1 MXFP4 (MI355X / gfx950) to cookbook [AMD Add GLM-5.1 MXFP4 variant (MI355X / gfx950) to cookbook Jun 24, 2026
@Raiden-Makoto
Raiden-Makoto force-pushed the RM/glm51-cookbook-mxfp4 branch from 89cbd9d to c3222ba Compare June 24, 2026 17:31
@Raiden-Makoto
Raiden-Makoto marked this pull request as ready for review June 24, 2026 17:33
@gemini-code-assist

Copy link
Copy Markdown
Contributor

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

- Add the amd/GLM-5.1-MXFP4 checkpoint (tp=4, --kv-cache-dtype fp8_e4m3) as the
  recommended MI355X (gfx950) path in the deploy generator + ROCm command section,
  and document the opt-in SGLANG_DSA_TRITON_PREFILL=1 prefill kernel (follow-up to sgl-project#28975).
- EAGLE speculative decoding: the old cookbook said it was unsupported on AMD and
  the generator gated it off for all AMD. We tested EAGLE on gfx950 (MI355X) and it
  works (lossless, large ITL/throughput win); gfx942 (MI300X/MI325X) is unverified.
  So the generator now emits the EAGLE flags by default and excludes ONLY
  MI300X/MI325X (gfx942); the static MXFP4 command and the AMD note reflect this.
  gfx942 verification can be a follow-up.
@Raiden-Makoto
Raiden-Makoto force-pushed the RM/glm51-cookbook-mxfp4 branch from c3222ba to e409085 Compare June 24, 2026 22:32
@Raiden-Makoto Raiden-Makoto changed the title [AMD Add GLM-5.1 MXFP4 variant (MI355X / gfx950) to cookbook [AMD] [GLM5] GLM-5.1 MXFP4 (MI355X) + enable EAGLE for gfx950 in cookbook Jun 24, 2026
@HaiShaw HaiShaw added the amd label Jun 25, 2026
@HaiShaw
HaiShaw merged commit 0075c8f into sgl-project:main Jun 25, 2026
105 of 109 checks passed
@Raiden-Makoto
Raiden-Makoto deleted the RM/glm51-cookbook-mxfp4 branch July 7, 2026 19:06
Chronostasys pushed a commit to MindLab-Research/sglang that referenced this pull request Aug 24, 2026
…book (sgl-project#29194)

Co-authored-by: Raiden-Makoto <Raiden-Makoto@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

amd documentation Improvements or additions to documentation

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants