Search results

Recipe request: cyankiwi/Qwen3.5-4B-AWQ-INT8-INT4

Status: Open.

#560 In vllm-project/recipes;

· Poyzi06 opened on Jun 17, 2026

RFC: Add Ascend NPU hardware support

Status: Open.

#498 In vllm-project/recipes;

· robellliu-dev opened on Jun 1, 2026

[Bug] DeepSeek-V4-Pro multi_node_tep renders invalid TP=16 command for FP8 weights

Status: Open.

#497 In vllm-project/recipes;

· observerw opened on May 31, 2026

Folder names don't match provider names

Status: Open.

#491 In vllm-project/recipes;

· Karim-siala opened on May 28, 2026

Gemma4: max_num_batched_tokens error for 26B MoE on 1× A100/H100 (BF16) command

Status: Open.

#441 In vllm-project/recipes;

· suleymanzahi opened on May 5, 2026

DEEPSEEK-V4-FLASH ON H20-141G*8 SINGLE NODE ERROR

Status: Open.

#429 In vllm-project/recipes;

· cxljkb1994 opened on Apr 30, 2026

mimo-v2.5 request must chat_template_kwargs bug

Status: Open.

#425 In vllm-project/recipes;

· yiminghub2024 opened on Apr 29, 2026

vllm kvcache print error

Status: Open.

#421 In vllm-project/recipes;

· yiminghub2024 opened on Apr 29, 2026

performance-mode vllm console print is error

Status: Open.

#404 In vllm-project/recipes;

· yiminghub2024 opened on Apr 27, 2026

why deepseek-v4-pro can deploy with dp=8 on H20?

Status: Open.

#402 In vllm-project/recipes;

· yiminghub2024 opened on Apr 27, 2026

DeepSeek V4 pro cannot support TP16

Status: Open.

#400 In vllm-project/recipes;

· mpjlu opened on Apr 27, 2026

DEEPSEEK-V4-PRO ON H20-141G*8 SINGLE NODE ERROR

Status: Open.

#390 In vllm-project/recipes;

· yiminghub2024 opened on Apr 25, 2026