All issues

Search results

DeepSeek V4 Flash 0731 on vLLM-MoET: Single-GPU Validation Report

Status: Open.

#22 In kacper-daftcode/vLLM-Moet;

· jpezzulli opened on Jul 31, 2026

Proposed fix for #20: route DS4 MTP drafter experts to Mxfp4MoEMethod; guard against uninitialized drafter scales

Status: Open.

#21 In kacper-daftcode/vLLM-Moet;

· macdad222 opened on Jul 28, 2026

DS4 MTP drafter loaded with wrong quant method (ignored_layers 'mtp.*' never matches) + uninitialized per-tensor scales: 0% MTP acceptance under W2, allocation-dependent degradation on stock FP4

Status: Open.

#20 In kacper-daftcode/vLLM-Moet;

· macdad222 opened on Jul 28, 2026

Will this work with the DSPARK version?

Status: Open.

#19 In kacper-daftcode/vLLM-Moet;

· dicksondickson opened on Jul 24, 2026

mtp guided expert weight prefetch

Status: Open.

#18 In kacper-daftcode/vLLM-Moet;

· ycui7 opened on Jul 24, 2026

Request: Support for nvidia/MiniMax-M3-NVFP4

Status: Open.

#17 In kacper-daftcode/vLLM-Moet;

· DikiyLifter opened on Jul 21, 2026

Validation report: Qwen3.5-122B-A10B-NVFP4 on 1× RTX PRO 6000

Status: Open.

#15 In kacper-daftcode/vLLM-Moet;

· jpezzulli opened on Jul 13, 2026

check_patch_files.py fails when a release tag is also resolved as a path

Status: Open.

#14 In kacper-daftcode/vLLM-Moet;

· jpezzulli opened on Jul 13, 2026

Question / Feature Request: Support for 8-GPU configurations (8tp)

Status: Open.

#13 In kacper-daftcode/vLLM-Moet;

· yxopen opened on Jul 13, 2026