正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Add this column to the dev tools to quickly see the Vercel Cache status of your requests

Alex Sidorenko

33,808 subscribers

77,692 次观看 • 2 年前 •via X (Twitter)

科学技术教育艺术

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

A look at the in-dev V2 health bars. To use this, simply add: "citadel_unit_status_use_v2" "true" "citadel_unit_status_use_v2_for_nonplayers" "true" to your convars.

A look at the in-dev V2 health bars. To use this, simply add: "citadel_unit_status_use_v2" "true" "citadel_unit_status_use_v2_for_nonplayers" "true" to your convars.

Deadlock Intel

224,403 次观看 • 6 个月前

Dev tools coming soon to TANSTACK DB! The best client side state libs have brilliant dev tools, this is even more important for DB as you store so much state in it. The DB dev tools will give you full visibility of all collections and live queries. 🚀🎉

Dev tools coming soon to TANSTACK DB! The best client side state libs have brilliant dev tools, this is even more important for DB as you store so much state in it. The DB dev tools will give you full visibility of all collections and live queries. 🚀🎉

Sam Willis

28,100 次观看 • 1 年前

Be Sure to Try this Drill and Sandpaper Hack to Sharpen Your Garden Tools Quickly and Easily

Be Sure to Try this Drill and Sandpaper Hack to Sharpen Your Garden Tools Quickly and Easily

Dieudonne HARAGIRIMANA

65,796 次观看 • 2 个月前

A basic version of Vercel built in v0 – there's no doubt I'd start this way if I was rebuilding the dashboard from scratch shadcn has made it possible to start quickly and then make the design & identity your own

A basic version of Vercel built in v0 – there's no doubt I'd start this way if I was rebuilding the dashboard from scratch shadcn has made it possible to start quickly and then make the design & identity your own

Guillermo Rauch

156,853 次观看 • 1 年前

Introducing agent-tail: a dev tool that makes your logs easily accessible to agents. npm i agent-tail Wrap dev server commands, add the Vite/Next plugin, install the skill, and your agent has full context of your browser, server, and any other logs. Link to full docs below ↓

Introducing agent-tail: a dev tool that makes your logs easily accessible to agents. npm i agent-tail Wrap dev server commands, add the Vite/Next plugin, install the skill, and your agent has full context of your browser, server, and any other logs. Link to full docs below ↓

Kyle Gill

29,414 次观看 • 6 个月前

underused Claude Code feature: type "#" to quickly add something to the Claude MD

underused Claude Code feature: type "#" to quickly add something to the Claude MD

Thariq

43,116 次观看 • 11 个月前

We're releasing 𝚟𝚎𝚛𝚌𝚎𝚕-𝚛𝚎𝚊𝚌𝚝-𝚗𝚊𝚝𝚒𝚟𝚎-𝚜𝚔𝚒𝚕𝚕𝚜, extracted from the lessons of building v0 for iOS and the upcoming Vercel app (teaser below). ▲ ~/ npx skills add vercel-labs/agent-skills

We're releasing 𝚟𝚎𝚛𝚌𝚎𝚕-𝚛𝚎𝚊𝚌𝚝-𝚗𝚊𝚝𝚒𝚟𝚎-𝚜𝚔𝚒𝚕𝚕𝚜, extracted from the lessons of building v0 for iOS and the upcoming Vercel app (teaser below). ▲ ~/ npx skills add vercel-labs/agent-skills

Vercel Developers

489,283 次观看 • 7 个月前

v0.17.0 of the TANSTACK create-start-app brings MCP support to your application. You can add whatever tools and resources you want and expose them with an MCP StreamingHTTP endpoint. It's the perfect way to start your exploration of the Model Context Protocol!

v0.17.0 of the TANSTACK create-start-app brings MCP support to your application. You can add whatever tools and resources you want and expose them with an MCP StreamingHTTP endpoint. It's the perfect way to start your exploration of the Model Context Protocol!

Jack Herrington

23,200 次观看 • 1 年前

Day 11/90 of Inference Engineering How does vLLM work and how is it used in production? Before we discuss how vLLM works internally, it helps to understand what vLLM is. At a high level, vLLM is an inference engine that is designed to serve LLMs to thousands of concurrent users efficiently while managing scarce compute and memory. The goal for vLLM is to maximize throughput and minimize latency; optimizing for the best inference economics and experience for end users. With every request from the end user, it eventually ends up in the engine core, gets scheduled alongside other requests from other concurrent users, executes on the GPU, and updates the KV cache with the new key and value vectors, and streams the tokens back to the user. The Scheduler decides what requests should execute next while continuously batching requests together to maximize GPU utilization. Continuous batching is an inference optimization that allows new requests to join a running batch as other requests finish generating tokens. This helps with keeping the GPU utilization high instead of letting it sit idle waiting for an entire batch to complete generating. After the scheduler dispatches the selected batch to the Model Executor, the Model Executor prepares the tensors and metadata required for inference, retrieves each request’s block table from KV Cache Manager, launches the optimized transformer forward pass on the GPU, computes the logits, updates the KV cache with the new key and value vectors, and finally returns the results for sampling and streaming. The KV Cache Manager uses the PagedAttention memory layout to allocate fixed-size cache blocks on demand and maintains a Free Block Queue on the CPU that tracks which blocks in the GPU’s Paged KV Cache are currently free. When a request needs additional KV cache space, the KV Cache manager takes a free block from the queue and assigns it to that request, thus avoiding an expensive search through GPU memory for available cache blocks. All of these components form the core of vLLM’s inference engine. The Scheduler determines what requests are executed, the Model Executor determines how those requests are executed, the KV Cache Manager determines where each request’s KV cache lives using the PagedAttention Memory Layout. This architecture enables vLLM to serve thousands of concurrent requests with high throughput, low latency, and efficient GPU memory utilization. Heres a little animation that visualizes everything! - I've also completed the forward pass for my mnist.c project. I had a nice chat with shrey birmiwal, such a knowledgeable guy. Excited to learn more about vLLM and implement a tiny-vLLM one day.

Day 11/90 of Inference Engineering How does vLLM work and how is it used in production? Before we discuss how vLLM works internally, it helps to understand what vLLM is. At a high level, vLLM is an inference engine that is designed to serve LLMs to thousands of concurrent users efficiently while managing scarce compute and memory. The goal for vLLM is to maximize throughput and minimize latency; optimizing for the best inference economics and experience for end users. With every request from the end user, it eventually ends up in the engine core, gets scheduled alongside other requests from other concurrent users, executes on the GPU, and updates the KV cache with the new key and value vectors, and streams the tokens back to the user. The Scheduler decides what requests should execute next while continuously batching requests together to maximize GPU utilization. Continuous batching is an inference optimization that allows new requests to join a running batch as other requests finish generating tokens. This helps with keeping the GPU utilization high instead of letting it sit idle waiting for an entire batch to complete generating. After the scheduler dispatches the selected batch to the Model Executor, the Model Executor prepares the tensors and metadata required for inference, retrieves each request’s block table from KV Cache Manager, launches the optimized transformer forward pass on the GPU, computes the logits, updates the KV cache with the new key and value vectors, and finally returns the results for sampling and streaming. The KV Cache Manager uses the PagedAttention memory layout to allocate fixed-size cache blocks on demand and maintains a Free Block Queue on the CPU that tracks which blocks in the GPU’s Paged KV Cache are currently free. When a request needs additional KV cache space, the KV Cache manager takes a free block from the queue and assigns it to that request, thus avoiding an expensive search through GPU memory for available cache blocks. All of these components form the core of vLLM’s inference engine. The Scheduler determines what requests are executed, the Model Executor determines how those requests are executed, the KV Cache Manager determines where each request’s KV cache lives using the PagedAttention Memory Layout. This architecture enables vLLM to serve thousands of concurrent requests with high throughput, low latency, and efficient GPU memory utilization. Heres a little animation that visualizes everything! - I've also completed the forward pass for my mnist.c project. I had a nice chat with shrey birmiwal, such a knowledgeable guy. Excited to learn more about vLLM and implement a tiny-vLLM one day.

max fu

70,543 次观看 • 1 个月前

Apex player found a dev in-game, immediately asks them to delete the nemesis and add cross-progression What would you ask a dev if they were on your team?

Apex player found a dev in-game, immediately asks them to delete the nemesis and add cross-progression What would you ask a dev if they were on your team?

AlphaIntel - GTA6 News

525,530 次观看 • 3 年前

Some of the biggest news today is Cloudflare Worker's Cache. So I built this small catalog to experience it. It allows you to purge the cache for certain pages, or the entire catalog. It also shows you the exact caching headers used. See it in action or try it yourself:

Some of the biggest news today is Cloudflare Worker's Cache. So I built this small catalog to experience it. It allows you to purge the cache for certain pages, or the entire catalog. It also shows you the exact caching headers used. See it in action or try it yourself:

Jilles Soeters

17,920 次观看 • 1 个月前

The beauty of PC modding. Being able to add the actual PPV themes to the soundtrack helps tremendously with immersion. Not to mention the beautiful arena mods. B4G.TV, heisenfick, | | PWM | | - Expand Your Game, TRIBUTE: Wrestling Game Tools yawl are absolute GOATS, man! #WWE2K25

The beauty of PC modding. Being able to add the actual PPV themes to the soundtrack helps tremendously with immersion. Not to mention the beautiful arena mods. B4G.TV, heisenfick, | | PWM | | - Expand Your Game, TRIBUTE: Wrestling Game Tools yawl are absolute GOATS, man! #WWE2K25

Wrestling Modders, Inc.™

44,039 次观看 • 9 个月前

The first long ammo, lever action rifle has entered the Hunt. Say hello to the Mako, able to deliver multiple shots quickly and over long ranges. Iron Eye and Levering to add an even sharper edge to its abilities. Add it to your arsenal during Desolation's Wake!

The first long ammo, lever action rifle has entered the Hunt. Say hello to the Mako, able to deliver multiple shots quickly and over long ranges. Iron Eye and Levering to add an even sharper edge to its abilities. Add it to your arsenal during Desolation's Wake!

Hunt: Showdown

32,783 次观看 • 2 年前

Last iteration on this one: Dev: I add a cliff to the island, because players are way too smart and will use the wolf healing ability to survive the deadly water to come to the other side

Last iteration on this one: Dev: I add a cliff to the island, because players are way too smart and will use the wolf healing ability to survive the deadly water to come to the other side

Edenfall ➡ Wishlist on Steam!

84,157 次观看 • 8 个月前