fog_glutamine Profile Banner
FogDong Profile
FogDong

@fog_glutamine

Followers
1K
Following
271
Media
32
Statuses
159

CNCF Ambassador & KubeVela Maintainer She / Her

Santa Clara
Joined May 2011
Don't wanna be here? Send us removal request.
@fog_glutamine
FogDong
1 month
Super impressed by this detailed summary of my KubeCon talk on cold starts and model serving — thank you @AntoineGrondin! 🙌.
@AntoineGrondin
Antoine
1 month
Zero-Extraction Cold Starts: How FUSE-Streaming Slashed ComfyUI Cold Starts by 10x, by @fog_glutamine (@bentomlai). whats the problem with model serving?. tradional approach:.- high cost of constant overprovisioning.- wasteful. serverless instead?.- cheaper but very high cold
Tweet media one
Tweet media two
Tweet media three
Tweet media four
1
0
4
@fog_glutamine
FogDong
1 month
Just wrapped up my talks at KubeCon + CloudNativeCon Hong Kong and KubeCon Japan!. Great to connect with the cloud native community and share some ideas. Thanks to everyone who joined! 🙌. #KubeCon #CNCF #CloudNative #Kubernetes
Tweet media one
Tweet media two
Tweet media three
0
1
17
@fog_glutamine
FogDong
4 months
太期待🌶️!.
@yetone
yetone
4 months
‌‌Almost done. SimSay!!!
0
0
3
@fog_glutamine
FogDong
4 months
Tweet media one
Tweet media two
0
1
14
@fog_glutamine
FogDong
6 months
RT @wiggitywhitney: Want to dish about #observability with @vfarcic & me?. (^▽^)/ You Choose is going live NOW! The epic recap season grand….
0
3
0
@fog_glutamine
FogDong
9 months
Neutral open source🤷‍♀️🤷‍♀️🤷‍♀️.
0
0
2
@fog_glutamine
FogDong
11 months
#kubevela kiosk on #KubeCon HongKong!
Tweet media one
3
3
69
@fog_glutamine
FogDong
1 year
Excited to share that I'll be giving two talks at #KubeCon Hong Kong! One is @bentomlai collaboration with Dragonfly: "How Fast Can Your Model Composition Run in Serverless Inference." The other is the KubeVela maintainer track. Looking forward to connecting with you there!
Tweet media one
0
3
27
@fog_glutamine
FogDong
1 year
RT @yetone: 俺们 BentoML 团队最近做的一点微小的 LLM 推理引擎 benchmarking 工作:.
Tweet card summary image
bentoml.com
Compare the Llama 3 serving performance with vLLM, LMDeploy, MLC-LLM, TensorRT-LLM, and Hugging Face TGI on BentoCloud.
0
4
0
@fog_glutamine
FogDong
1 year
RT @bentomlai: 🚀 Dive into #KubeVela (! It simplifies app deployment across clouds, powered by workflow magic. 🌟 Ku….
kubevela.io
Make shipping applications more enjoyable.
0
1
0
@fog_glutamine
FogDong
1 year
RT @kk1984_: 想安利一下自己写的加强版 `kubectl port-forward`: . krelay 主要有以下几点增强.1. 支持 UDP 协议.2. 支持真·转发给 deployment,不受 rolling….
Tweet card summary image
github.com
A better alternative to `kubectl port-forward` that can forward TCP or UDP traffic to IP/Host which is accessible inside the cluster. - knight42/krelay
0
4
0
@fog_glutamine
FogDong
1 year
RT @yetone: Thanks to my beloved team for their continuous support!.
0
2
0
@fog_glutamine
FogDong
1 year
昨天了解到一个据说是自研 50b 大模型的 AI 写小说项目,今天回来一测,emmmm……难道 fine tuning 也是一种自研?.
0
0
13
@fog_glutamine
FogDong
1 year
好高的质量🫣.
1
0
5
@fog_glutamine
FogDong
1 year
难得有 Palworld 这种破圈热点可以让各大云厂商蹭热度的,但是宣传什么自动弹性、运维成本的就大可不必了,普通游戏玩家根本看不懂也不关心这个,反而直接被劝退.
1
0
12
@fog_glutamine
FogDong
2 years
RT @_jessicasachs: @skywind3000 @trunarla If you mean no offense and have good intentions, I will explain to you why you’re being offensive….
0
47
0
@fog_glutamine
FogDong
2 years
RT @gaocegege: 朋友写的一个用 TS 写交易量化策略的开源平台,帮忙分享一下. "Excited to share an open-source platform for developing quantitati….
Tweet card summary image
github.com
Yuan - Personal Investment Operating System. Contribute to No-Trade-No-Life/Yuan development by creating an account on GitHub.
0
57
0
@fog_glutamine
FogDong
2 years
RT @bentomlai: Unlocking the secrets of how to productionize AI apps with @fog_glutamine's keynote speech at #KubeCon China this year! 🚀 Co….
0
3
0
@fog_glutamine
FogDong
2 years
Happy 1024 Day🥳!  I look forward to a future where no one is surprised by anyone's appearance in the tech industry. Being a software engineer isn't about how we look, it's all about the amazing things we create.
Tweet media one
Tweet media two
0
1
42