SGLang and vLLM Workshops Coming to GOSIM Paris 2026๏ผ

The GOSIM Workshops have long been known for their diversity, hands-on learning, and interactivity, making them one of the most popular segments of the conference.

This May, the SGLang Workshop and vLLM Workshop will arrive at GOSIM Paris 2026, bringing together AI infrastructure developers from around the world to explore the latest advances in LLM inference systems.

Ticket purchase link:
https://eventbrite.com/e/gosim-paris-2026-tickets-1984013840806?aff=oddtdtcreator

#SGLang #vLLM

๐Ÿš€ Big news!
The SGLang Workshop & vLLM Workshop are coming to GOSIM Paris 2026! ๐ŸŽ‰
๐ŸŒ A must-attend event for AI developers and open-source contributors worldwide
๐Ÿ’ก Dive into cutting-edge topics: large model inference, agentic AI, and more
๐ŸŽ“ Hands-on sessions and discussions to bring high-value learning and networking

Get your early bird tickets now and enjoy the discount: https://eventbrite.com/e/gosim-paris-2026-tickets-1984013840806?aff=oddtdtcreator ๐Ÿš€

#GOSIMParis2026 #SGLang #vLLM #AIWorkshop #OpenSourceAI

Qwen (@Alibaba_Qwen)

Qwen 3.5 Medium ๋ชจ๋ธ ์‹œ๋ฆฌ์ฆˆ์˜ FP8 ๊ฐ€์ค‘์น˜๊ฐ€ ๊ณต๊ฐœ๋˜์–ด ๋ฐฐํฌ ์ค€๋น„ ์™„๋ฃŒ๋˜์—ˆ๋‹ค๋Š” ๊ณต์ง€์ž…๋‹ˆ๋‹ค. vLLM๊ณผ SGLang์— ๋Œ€ํ•œ ๋„ค์ดํ‹ฐ๋ธŒ ์ง€์›์ด ํฌํ•จ๋˜๋ฉฐ ๋ชจ๋ธ ์นด๋“œ์— ์˜ˆ์ œ ์ฝ”๋“œ๊ฐ€ ์ œ๊ณต๋ฉ๋‹ˆ๋‹ค. FP8 ์ •๋ฐ€๋„๋กœ ์›Œํฌํ”Œ๋กœ ์ตœ์ ํ™”๊ฐ€ ๊ฐ€๋Šฅํ•˜๋ฉฐ ๊ฐ€์ค‘์น˜๋Š” Hugging Face์—์„œ ํš๋“ํ•  ์ˆ˜ ์žˆ๋‹ค๊ณ  ์•ˆ๋‚ดํ•ฉ๋‹ˆ๋‹ค.

https://x.com/Alibaba_Qwen/status/2026682179305275758

#qwen3.5 #fp8 #vllm #huggingface #sglang

Qwen (@Alibaba_Qwen) on X

๐Ÿ”ฅ Qwen 3.5 Medium Model Series FP8 weights are now open and ready for deployment๏ผ Native support for vLLM and SGLang. Check the model card for example code. โšก๏ธ Optimize your workflow with FP8 precision. ๐Ÿ‘‡ Get the weights: Hugging Face๏ผšhttps://t.co/3MSb7miq68

X (formerly Twitter)

Qwen (@Alibaba_Qwen)

Qwen 3.5 Medium ์‹œ๋ฆฌ์ฆˆ์˜ FP8 ๊ฐ€์ค‘์น˜๊ฐ€ ๊ณต๊ฐœ๋˜์–ด ๋ฐฐํฌ ๊ฐ€๋Šฅํ•˜๋‹ค๋Š” ๊ณต์ง€์ž…๋‹ˆ๋‹ค. vLLM๊ณผ SGLang์— ๋„ค์ดํ‹ฐ๋ธŒ ์ง€์›์„ ์ œ๊ณตํ•˜๋ฉฐ, ๋ชจ๋ธ ์นด๋“œ์— ์˜ˆ์ œ ์ฝ”๋“œ๊ฐ€ ํฌํ•จ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค. FP8 ์ •๋ฐ€๋„๋กœ ์›Œํฌํ”Œ๋กœ์šฐ ์ตœ์ ํ™” ๊ฐ€๋Šฅํ•˜๋ฉฐ ๊ฐ€์ค‘์น˜๋Š” Hugging Face์—์„œ ํ™•์ธยท๋‹ค์šด๋กœ๋“œํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

https://x.com/Alibaba_Qwen/status/2026682179305275758

#qwen #fp8 #vllm #huggingface #sglang

Qwen (@Alibaba_Qwen) on X

๐Ÿ”ฅ Qwen 3.5 Medium Model Series FP8 weights are now open and ready for deployment๏ผ Native support for vLLM and SGLang. Check the model card for example code. โšก๏ธ Optimize your workflow with FP8 precision. ๐Ÿ‘‡ Get the weights: Hugging Face๏ผšhttps://t.co/3MSb7miq68

X (formerly Twitter)

Qwen (@Alibaba_Qwen)

SGLang์„ ์‚ฌ์šฉํ•ด ์ƒˆ๋กœ์šด AI ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜์„ ์ฆ‰์‹œ ์‹คํ–‰ํ•  ์ˆ˜ ์žˆ๋‹ค๋Š” ๋‚ด์šฉ์ž…๋‹ˆ๋‹ค. SGLang์€ ์ตœ๊ทผ ์ฃผ๋ชฉ๋ฐ›๋Š” ๊ฒฝ๋Ÿ‰ ์–ธ์–ด ๋ชจ๋ธ ํ”„๋ ˆ์ž„์›Œํฌ๋กœ, ๊ฐœ๋ฐœ์ž๋“ค์ด ์‹ ์†ํ•˜๊ฒŒ ๋ชจ๋ธ์„ ๋ฐฐํฌํ•˜๊ณ  ํ…Œ์ŠคํŠธํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•ฉ๋‹ˆ๋‹ค.

https://x.com/Alibaba_Qwen/status/2026348924433477775

#sglang #ai #framework #deployment

Qwen (@Alibaba_Qwen) on X

โœจ Run it now with SGLang๏ผChong!

X (formerly Twitter)

Qwen (@Alibaba_Qwen)

Qwen3.5-397B-A17B-FP8 ๋ชจ๋ธ ๊ฐ€์ค‘์น˜๊ฐ€ ๊ณต๊ฐœ๋˜์—ˆ๋‹ค๋Š” ๋ฐœํ‘œ์ž…๋‹ˆ๋‹ค. SGLang ์ง€์›์ด ๋ณ‘ํ•ฉ๋˜์—ˆ๊ณ  vLLM์šฉ PR์ด ์ œ์ถœ๋˜์–ด(vLLM ๋ฆฌํฌ ์—ฐ๋™ ์˜ˆ์ •) ์ฃผ์š” ์ถ”๋ก  ํ”„๋ ˆ์ž„์›Œํฌ์—์„œ ๊ณง ์‚ฌ์šฉ ๊ฐ€๋Šฅํ•ด์ง„๋‹ค๋Š” ๊ธฐ์ˆ ยท์˜คํ”ˆ์†Œ์Šค ์—…๋ฐ์ดํŠธ๋ฅผ ์•Œ๋ฆฝ๋‹ˆ๋‹ค. ๋ชจ๋ธ ์นด๋“œ์™€ ์˜ˆ์ œ ์ฝ”๋“œ๋„ ์ œ๊ณต๋ฉ๋‹ˆ๋‹ค.

https://x.com/Alibaba_Qwen/status/2024161147537232110

#qwen3.5 #openweights #vllm #sglang

Qwen (@Alibaba_Qwen) on X

๐Ÿš€ Qwen3.5-397B-A17B-FP8 weights are now open! It took some time to adapt the inference frameworks, but here we are: โœ… SGLang support is merged ๐Ÿ”„ vLLM PR submitted โ†’ https://t.co/rJkuitOBWs Check the model card for example code. vLLM support landing in the next couple of days!

X (formerly Twitter)

Qwen (@Alibaba_Qwen)

Qwen3.5-397B-A17B-FP8 ๋ชจ๋ธ์˜ ๊ฐ€์ค‘์น˜๊ฐ€ ๊ณต๊ฐœ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ์ถ”๋ก  ํ”„๋ ˆ์ž„์›Œํฌ ์ ์‘์ด ์ง„ํ–‰๋˜์—ˆ๊ณ  SGLang ์ง€์›์ด ๋ณ‘ํ•ฉ๋˜์—ˆ์œผ๋ฉฐ vLLM์— ๋Œ€ํ•œ PR์ด ์ œ์ถœ๋˜์–ด ๊ณง vLLM ์ง€์›์ด ๋„์ž…๋  ์˜ˆ์ •์ž…๋‹ˆ๋‹ค. ๋ชจ๋ธ ์นด๋“œ์— ์˜ˆ์ œ ์ฝ”๋“œ๊ฐ€ ํฌํ•จ๋˜์–ด ์žˆ์–ด ๊ฐœ๋ฐœ์ž๋“ค์ด ๊ณง๋ฐ”๋กœ ํ…Œ์ŠคํŠธํ•˜๊ณ  ๋ฐฐํฌํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

https://x.com/Alibaba_Qwen/status/2024161147537232110

#qwen3.5 #openweights #vllm #sglang #inference

Qwen (@Alibaba_Qwen) on X

๐Ÿš€ Qwen3.5-397B-A17B-FP8 weights are now open! It took some time to adapt the inference frameworks, but here we are: โœ… SGLang support is merged ๐Ÿ”„ vLLM PR submitted โ†’ https://t.co/rJkuitOBWs Check the model card for example code. vLLM support landing in the next couple of days!

X (formerly Twitter)

merve (@mervenoyann)

Minimax M2.5๊ฐ€ vLLM๊ณผ SGLang์„ ์ง€์›ํ•˜๋ฉฐ, Hugging Face Transformers์™€์˜ ํ†ตํ•ฉ์„ ์œ„ํ•œ ์ปค์Šคํ…€ ์ฝ”๋“œ๋„ ์ œ๊ณต๋œ๋‹ค๊ณ  ๋ฐœํ‘œ๋˜์–ด ๊ฐœ๋ฐœ์ž์šฉ ๋Ÿฐํƒ€์ž„/ํ”„๋ ˆ์ž„์›Œํฌ ํ˜ธํ™˜์„ฑ ์—…๋ฐ์ดํŠธ๊ฐ€ ์ด๋ฃจ์–ด์กŒ์Œ์„ ์•Œ๋ฆฝ๋‹ˆ๋‹ค.

https://x.com/mervenoyann/status/2022327214138925111

#minimax #m2.5 #vllm #transformers #sglang

merve (@mervenoyann) on X

comes with vLLM/SGLang support and transformers with custom code! https://t.co/VU9G7EGga8

X (formerly Twitter)

LMSYS Org (@lmsysorg)

๋ฏธ์Šคํ„ฐ๋ฆฌ์˜€๋˜ Pony Alpha๊ฐ€ GLM-5๋กœ ๋ฐํ˜€์กŒ๊ณ , Zai_org์˜ ๋ชจ๋ธ ์ถœ์‹œ๋ฅผ ์ถ•ํ•˜ํ•˜๋ฉฐ SGLang์ด Day-0๋ถ€ํ„ฐ ์ง€์›ํ•œ๋‹ค๊ณ  ๋ฐœํ‘œ๋˜์—ˆ์Šต๋‹ˆ๋‹ค. ๋ชจ๋ธ ์‚ฌ์–‘์€ 744B ํŒŒ๋ผ๋ฏธํ„ฐ(40B active) ๋ฐ 28.5T ํ† ํฐ ํ”„๋ฆฌํŠธ๋ ˆ์ธ์œผ๋กœ, ๋ณต์žกํ•œ ์‹œ์Šคํ…œ ์—”์ง€๋‹ˆ์–ด๋ง๊ณผ ์žฅ๊ธฐ ์—์ด์ „ํŠธ ์ž‘์—…์— ์ตœ์ ํ™”๋˜์—ˆ๋‹ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค.

https://x.com/lmsysorg/status/2021639499374375014

#glm5 #ponyalpha #sglang #opensource

LMSYS Org (@lmsysorg) on X

๐ŸŽ‰ The mysterious Pony Alpha is finally revealed, congrats to @Zai_org on releasing GLM-5! SGLang is ready to support on day-0. ๐Ÿ› ๏ธ 744B params (40B active) model built for complex systems engineering & long-horizon agentic tasks ๐Ÿ“š 28.5T tokens pretraining for a stronger

X (formerly Twitter)

LMSYS Org (@lmsysorg)

DeepSeek-OCR 2๊ฐ€ Visual Causal Flow๋ฅผ ๋„์ž…ํ•˜๊ณ  SGLang์œผ๋กœ ์‹คํ–‰ ๊ฐ€๋Šฅํ•ด์กŒ์Šต๋‹ˆ๋‹ค. DeepEncoder V2๋ฅผ ํ™œ์šฉํ•ด ๊ธฐ์กด์˜ ์ขŒ์ƒโ†’์šฐํ•˜ ๊ณ ์ • ์Šค์บ” ๋Œ€์‹  ์ด๋ฏธ์ง€ ๋‚ด์šฉ ๊ธฐ๋ฐ˜์œผ๋กœ ์‹œ๊ฐ ํ† ํฐ ์ˆœ์„œ๋ฅผ ์žฌ์ •๋ ฌํ•˜๊ณ  ๋‹จ๊ณ„๋ณ„(์Šคํ…๋ณ„) ๋น„์ฃผ์–ผ ์ฒ˜๋ฆฌ๋ฅผ ์ˆ˜ํ–‰ํ•˜์—ฌ OCR ๋ฐ ๋น„์ „-์–ธ์–ด ์ฒ˜๋ฆฌ์˜ ์œ ์—ฐ์„ฑ๊ณผ ์ •ํ™•๋„ ํ–ฅ์ƒ์ด ๊ธฐ๋Œ€๋ฉ๋‹ˆ๋‹ค.

https://x.com/lmsysorg/status/2017305889490014281

#deepseek #ocr #visionlanguage #sglang

LMSYS Org (@lmsysorg) on X

๐Ÿš€ DeepSeek-OCR 2 by @deepseek_ai introduces a new Visual Causal Flow, and you can now run it with SGLang ๐Ÿ”ฅ Highlights ๐Ÿ‘€ Uses DeepEncoder V2 to reorder visual tokens based on image content, instead of a fixed top-left โ†’ bottom-right scan ๐Ÿง Performs step-by-step visual

X (formerly Twitter)