Archives
Trending
Support
Login
clear text
XML
Django
JavaScript
MATLAB
C
C++
C#
Python
SQL
Shell
Bash
Markdown
YAML
JSON
HTML
CSS
PHP
Java
Ruby
Go
Rust
Swift
Kotlin
Arduino
TypeScript
Perl
Autohotkey
Lua
SQF
R
Scala
Haskell
Groovy
Dart
Clojure
VB.NET
Objective-C
PowerShell
Bash
CoffeeScript
Verilog
#!/bin/bash python -m vllm.entrypoints.openai.api_server \ --model qwen/Qwen3-32B-Instruct \ --max-model-len 65536 \ --kv-cache-dtype fp8 \ --rope-scaling '{"rope_type":"yarn","factor":2.0,"original_max_position_embeddings":32768}' \ --max-num-seqs 8 \ --max-num-batched-tokens 131072 \ --gpu-memory-utilization 0.90 \ --speculative-config '{"method":"ngram","num_speculative_tokens":5}' \ --host 0.0.0.0 \ --port 8000
Mark as private (unlisted)
for 30 minutes
for 6 hours
for 1 day
for 1 week
for 1 month
for 1 year
everlasting (like CCCP)
