@bxa.hm1: #xhbuon

bxa hơm

Open In TikTok:

Region: VN

Tuesday 25 August 2026 21:02:10 GMT

39494

4338

21

270

Music

Download

No Watermark .mp4 (3.82MB) No Watermark(HD) .mp4 (5.07MB) Watermark .mp4 (4.19MB) Music .mp3

Comments

Hoàng Đại (Bo nhỏ) :

gáng mà đợi lỡ có kì tích roi sao tus oi

2026-08-26 14:00:09

5

✡️minh trọng🈯 :

này giống Khánh an vậy

2026-08-26 14:05:21

0

Cả Đời Không Hạnh Phúc ✈️ :

sao mà h họ cx muon bỏ mih roi mấy bạn ạ

2026-08-28 12:46:01

1

Chí Nguyên :

giống chợ ở Tân Uyên dị

2026-08-26 23:35:47

0

kimthu :

họ không về thật ☺️

2026-08-26 16:23:58

0

* Lộc Cá Koi * 🙋6⁴ :

đợi đại đại đi mà 😁

2026-08-26 17:56:46

0

Minh Khang 💛 :

2026-08-27 05:46:14

0

🔥 :

2026-08-26 10:45:42

0

tuấn anh :

tim hộ với

2026-08-26 03:37:28

2

Phu Pham6606 :

2026-08-26 05:57:09

0

THỪA KỸ NĂNG :

😁😁

2026-08-26 05:33:53

0

Nhi le :

hộ mik 4vd vs ạ

2026-08-26 05:51:09

1

Ah ba :

🥰🥰🥰

2026-08-26 02:57:19

1

Trường ( Chuối ) 🍌. :

❤️❤️❤️❤️

2026-08-26 03:17:32

0

To see more videos from user @bxa.hm1, please go to the Tikwm homepage.

Other Videos

LangSmith in 5 Minutes (Clearly Explained) In this video, Cloud X Berry explains LangSmith in a simple and beginner-friendly way. You ask an AI assistant where your order is, and it tells you. Behind that one sentence the model was called, a tool was chosen, an order-tracking service was queried, and the model was called a second time to write the reply you actually read. You saw a question and an answer. Everything that decided that answer is invisible and that is fine right up until the day it answers wrong. We start with the version where you have no visibility at all. The answer comes back wrong and you have exactly two things to look at: the question that went in and the response that came out. Did the model misunderstand? Did your RAG system retrieve the wrong documents? Did the agent call the wrong tool? You cannot say, because the middle was never recorded. And a wrong answer is not the only failure, sometimes the answer is perfectly correct and the application suddenly took 15 seconds, and you still cannot point at the step that spent them. From there we build the idea up piece by piece: a worked example of a single request as it moves through a model, a tool and back again, why one question to a user is four steps internally, and what developers already do about this in traditional software with logs and monitoring. Then what AI adds to that list — the prompt, the model, the tokens, the latency, the cost and the decisions an agent made. That is observability, and tracing is how LangSmith delivers it: an operation is a run, the runs together are a trace, and a trace is what lets you see where the time went and where things broke. Then the half most explanations skip. Understanding what happened is only one part of building a good AI application, the other is knowing whether your changes helped. You swap a prompt, switch a model, change how documents are retrieved, and the app still works. But is it better? We go through evaluation: building a set of test cases, running the same tests again after a change, and comparing results, because a new prompt might give better answers while a different model is faster but less accurate, and a new retrieval strategy might improve answers and increase your cost. Measuring instead of hoping. We finish by placing LangSmith next to LangChain and LangGraph — which one builds the application, which one builds the workflow, which one tells you what happened when it ran and an honest note on when you do not need any of this at all. CHAPTERS 0:00 One question, one answer 0:09 Then your application starts growing 0:25 It gives the wrong answer — now what? 0:31 You see both ends, never the middle 0:43 Or it was correct, and took 15 seconds 0:55 Where LangSmith comes in 0:58 A worked example:

LangSmith in 5 Minutes (Clearly Explained) In this video, Cloud X Berry explains LangSmith in a simple and beginner-friendly way. You ask an AI assistant where your order is, and it tells you. Behind that one sentence the model was called, a tool was chosen, an order-tracking service was queried, and the model was called a second time to write the reply you actually read. You saw a question and an answer. Everything that decided that answer is invisible and that is fine right up until the day it answers wrong. We start with the version where you have no visibility at all. The answer comes back wrong and you have exactly two things to look at: the question that went in and the response that came out. Did the model misunderstand? Did your RAG system retrieve the wrong documents? Did the agent call the wrong tool? You cannot say, because the middle was never recorded. And a wrong answer is not the only failure, sometimes the answer is perfectly correct and the application suddenly took 15 seconds, and you still cannot point at the step that spent them. From there we build the idea up piece by piece: a worked example of a single request as it moves through a model, a tool and back again, why one question to a user is four steps internally, and what developers already do about this in traditional software with logs and monitoring. Then what AI adds to that list — the prompt, the model, the tokens, the latency, the cost and the decisions an agent made. That is observability, and tracing is how LangSmith delivers it: an operation is a run, the runs together are a trace, and a trace is what lets you see where the time went and where things broke. Then the half most explanations skip. Understanding what happened is only one part of building a good AI application, the other is knowing whether your changes helped. You swap a prompt, switch a model, change how documents are retrieved, and the app still works. But is it better? We go through evaluation: building a set of test cases, running the same tests again after a change, and comparing results, because a new prompt might give better answers while a different model is faster but less accurate, and a new retrieval strategy might improve answers and increase your cost. Measuring instead of hoping. We finish by placing LangSmith next to LangChain and LangGraph — which one builds the application, which one builds the workflow, which one tells you what happened when it ran and an honest note on when you do not need any of this at all. CHAPTERS 0:00 One question, one answer 0:09 Then your application starts growing 0:25 It gives the wrong answer — now what? 0:31 You see both ends, never the middle 0:43 Or it was correct, and took 15 seconds 0:55 Where LangSmith comes in 0:58 A worked example: "Where is my order?" 1:23 One question outside, four steps inside 1:43 What happens when something goes wrong 1:56 Logs and monitoring, and what AI adds to them 2:18 Observability, and tracing as the way in 2:43 Runs, traces, and what a trace lets you inspect 3:17 Understanding is only half the job 3:22 You change the prompt, the model, the retrieval 3:32 But did it actually get better? 3:37 Evaluation: the same tests, run twice 3:55 Measure instead of hoping 4:19 Where it fits alongside LangChain and LangGraph 4:55 Build, build, then see inside 5:03 When you do not need any of this 5:23 Recap: see it, understand it, measure it WHO THIS IS FOR Developers who keep seeing LangSmith mentioned alongside LangChain and LangGraph and want to know what it actually does, anyone whose AI application has grown past a single LLM call into RAG, tools and agents, and engineers who are changing prompts and models on instinct and would rather measure whether those changes helped. #LangSmith #LangChain #LangGraph

BONGKAR RAHASIA DUKUN SANTET DAN ISINYAA#tiktoktainment #masukberanda #fakesituation #dikiriminpaketsantet #filmpaketsantet #paketsantet #bts #dukun #dukunsantet

BONGKAR RAHASIA DUKUN SANTET DAN ISINYAA#tiktoktainment #masukberanda #fakesituation #dikiriminpaketsantet #filmpaketsantet #paketsantet #bts #dukun #dukunsantet

Bringing comedy that we can laugh together Back one stand up at a time let's go fam BigTex is back #fyp #foryou #mrbigtex23 #BigTexComedy #bigtexhellbus

Bringing comedy that we can laugh together Back one stand up at a time let's go fam BigTex is back #fyp #foryou #mrbigtex23 #BigTexComedy #bigtexhellbus

Part 7/7 ~ Your Life as Every Rank in Software engineering From writing your first lines of code as a Junior Developer to becoming a Senior Engineer, Tech Lead, Staff Engineer, and beyond, this video breaks down the different levels of a software engineering career — including the skills, responsibilities, projects, salary progression, and challenges that come with each rank. We’ll go through 10 levels, showing how your work changes as you gain experience, take on bigger responsibilities, and move from simply writing code to designing systems and leading engineering teams. Whether you're learning to code, studying computer science, or already working as a developer, this video gives you a realistic picture of what the journey can look like. 💻 Which level are you currently at? If you enjoyed the video, subscribe for more

Part 7/7 ~ Your Life as Every Rank in Software engineering From writing your first lines of code as a Junior Developer to becoming a Senior Engineer, Tech Lead, Staff Engineer, and beyond, this video breaks down the different levels of a software engineering career — including the skills, responsibilities, projects, salary progression, and challenges that come with each rank. We’ll go through 10 levels, showing how your work changes as you gain experience, take on bigger responsibilities, and move from simply writing code to designing systems and leading engineering teams. Whether you're learning to code, studying computer science, or already working as a developer, this video gives you a realistic picture of what the journey can look like. 💻 Which level are you currently at? If you enjoyed the video, subscribe for more "Your Life at Every Level" videos! 00:52 junior developer 02:29 software engineer (mid-level) 04:05 senior software engineer 05:03 staff engineer 06:13 engineering manager 07:12 engineering consultant 08:15 CTO #SoftwareEngineering #Programming #Coding #ComputerScience #SoftwareDeveloper

this could be us🤷🏼♀️

this could be us🤷🏼♀️

About

Robot
API

Legal

Privacy Policy