|
Welcome to vLLM Forums! :wave:
|
|
1
|
2175
|
March 24, 2025
|
|
About the General category
|
|
0
|
164
|
March 17, 2025
|
|
Question about max_num_seqs
|
|
1
|
40
|
October 9, 2026
|
|
Question about --enable-chunked-prefill and --max-num-batched-tokens
|
|
2
|
40
|
October 8, 2026
|
|
Handling request timeouts and KV cache bloat during bursty async workloads in vLLM
|
|
1
|
26
|
October 7, 2026
|
|
Audio_tower module missing
|
|
1
|
33
|
October 5, 2026
|
|
Is provider KV caching sufficient for agent swarms and long run agents?
|
|
2
|
75
|
October 1, 2026
|
|
Methods for Server-Side Parameter Optimization in VLLM Performance Benchmarking
|
|
1
|
56
|
September 29, 2026
|
|
Presentation on context, expert, data and tensor parallelism in vLLM
|
|
1
|
114
|
September 22, 2026
|
|
Add LLM observability data management
|
|
2
|
72
|
September 16, 2026
|
|
Vllm推理指标如何做可视化？
|
|
2
|
326
|
September 16, 2026
|
|
[RFC & PR Stack] Application-Directed Prefix Checkpoints for Hybrid GDN/Mamba (Up to 7.6x Speedup on L40S)
|
|
6
|
121
|
September 15, 2026
|
|
vLLM-0.23.0 lora加载问题
|
|
1
|
67
|
September 15, 2026
|
|
vLLM 0.29 7900xtx or 890m 1 cpu core 100%
|
|
8
|
196
|
September 14, 2026
|
|
vLLM 0.20.1, Radeon AI 9700, 1 CPU core at 100%
|
|
6
|
688
|
September 13, 2026
|
|
Muse glimmer - how to disable thinking
|
|
1
|
173
|
September 11, 2026
|
|
Not able to run amd/Qwen3.8-27B-Quark-AWQ-INT4-W4A16 on rocm
|
|
1
|
119
|
September 9, 2026
|
|
vLLM-Ascend在银河麒麟桌面版上不开Eager就报错，开了性能暴跌，换服务器版就正常，这是系统差异导致的吗？
|
|
1
|
96
|
September 4, 2026
|
|
我用单卡跑通了deepseek-v4-flash-0731的dspark模式
|
|
1
|
246
|
August 22, 2026
|
|
Flash-0731 在 vLLM 0.27.1 + SM120 上 dspark 不可启动
|
|
1
|
81
|
August 21, 2026
|
|
Why does changing VLLM_CACHE_ROOT not reproduce cold-start latency?
|
|
1
|
107
|
August 21, 2026
|
|
Model path in compute_hash() prevents torch.compile cache reuse across identical model loads
|
|
1
|
59
|
August 21, 2026
|
|
Mooncake prefix information in storage
|
|
3
|
103
|
August 21, 2026
|
|
Vllm是否支持加载deepseek_ocr模型的lora适配器进行推理
|
|
3
|
193
|
August 21, 2026
|
|
Dlash2 無法使用 在 vllm 0.27.1
|
|
1
|
209
|
August 19, 2026
|
|
Expert Parallelism All-to-All Communication without NVLink and DeepEP
|
|
4
|
794
|
August 17, 2026
|
|
FLASHINFER is FASTER then FA2 on Ampere hardware
|
|
5
|
189
|
August 16, 2026
|
|
Support for Qwen3.6 (27B & 35B-A3B) Hybrid Gated DeltaNet Models on TPU v5e-8 with vLLM and SGLang
|
|
11
|
621
|
August 14, 2026
|
|
How fast can you run DeepSeek V4 Spark 0731 with CPU-GPU hybrid inference when VRAM is not enough?
|
|
1
|
326
|
August 13, 2026
|
|
(Possible bug) --enable-prompt-tokens-details not working?
|
|
2
|
723
|
August 13, 2026
|