Loading video...

Video Failed to Load

Go Home

Advice for AI engineers 💡 A small Visual Language Model fine-tuned on your custom dataset is as accurate as GPT-5... ... and costs 50 times less. For example, LFM2-VL-3B by ↓

paulabartabajo_'s profile picture

Pau Labarta Bajo

70,233 subscribers

112,571 views • 8 months ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Advice for AI engineers 💡 A small Audio Language Model transcribes audio in real-time on your device, without internet connection or cloud costs. For example, LFM2-Audio-1.5B by ↓
0:14

Advice for AI engineers 💡 A small Audio Language Model transcribes audio in real-time on your device, without internet connection or cloud costs. For example, LFM2-Audio-1.5B by ↓

Pau Labarta Bajo

31,993 views • 8 months ago

Today Meta released "Code Llama", a large language model fine-tuned for coding tasks. It's publicly available and can be used for commercial use! It outperforms GPT 3.5 and you can even run it locally on your Macbook using Ollama.
0:11

Today Meta released "Code Llama", a large language model fine-tuned for coding tasks. It's publicly available and can be used for commercial use! It outperforms GPT 3.5 and you can even run it locally on your Macbook using Ollama.

Marcel Pociot 🧪

50,094 views • 2 years ago

Next week we're open sourcing Agent Builder - a visual n8n-style workflow builder for AI agents. Build AI agent workflows with a visual canvas by connecting , LLMs, logic nodes, and MCPs, then deploy as an API. Stay tuned for this one 👀
0:16

Next week we're open sourcing Agent Builder - a visual n8n-style workflow builder for AI agents. Build AI agent workflows with a visual canvas by connecting , LLMs, logic nodes, and MCPs, then deploy as an API. Stay tuned for this one 👀

Caleb Peffer (Hiring!)

251,825 views • 9 months ago

Fine-tune DeepSeek-OCR on your own language! (100% local) DeepSeek-OCR is a 3B-parameter vision model that achieves 97% precision while using 10× fewer vision tokens than text-based LLMs. It handles tables, papers, and handwriting without killing your GPU or budget. Why it matters: Most vision models treat documents as massive sequences of tokens, making long-context processing expensive and slow. DeepSeek-OCR uses context optical compression to convert 2D layouts into vision tokens, enabling efficient processing of complex documents. The best part? You can easily fine-tune it for your specific use case on a single GPU. I used Unsloth to run this experiment on Persian text and saw an 88.26% improvement in character error rate. ↳ Base model: 149% character error rate (CER) ↳ Fine-tuned model: 60% CER (57% more accurate) ↳ Training time: 60 steps on a single GPU Persian was just the test case. You can swap in your own dataset for any language, document type, or specific domain you're working with. I've shared the complete guide in the next tweet - all the code, notebooks, and environment setup ready to run with a single click. Everything is 100% open-source!
0:14

Fine-tune DeepSeek-OCR on your own language! (100% local) DeepSeek-OCR is a 3B-parameter vision model that achieves 97% precision while using 10× fewer vision tokens than text-based LLMs. It handles tables, papers, and handwriting without killing your GPU or budget. Why it matters: Most vision models treat documents as massive sequences of tokens, making long-context processing expensive and slow. DeepSeek-OCR uses context optical compression to convert 2D layouts into vision tokens, enabling efficient processing of complex documents. The best part? You can easily fine-tune it for your specific use case on a single GPU. I used Unsloth to run this experiment on Persian text and saw an 88.26% improvement in character error rate. ↳ Base model: 149% character error rate (CER) ↳ Fine-tuned model: 60% CER (57% more accurate) ↳ Training time: 60 steps on a single GPU Persian was just the test case. You can swap in your own dataset for any language, document type, or specific domain you're working with. I've shared the complete guide in the next tweet - all the code, notebooks, and environment setup ready to run with a single click. Everything is 100% open-source!

Akshay 🚀

126,122 views • 8 months ago

I have had early access to OpenAI's GPT-5 for the last two weeks and it is the smartest model available as of now. Just as one example, here it created an anamorphic text illusion in minecraft that spells "SPECIAL" from one angle and "GPTFIVE" from another
0:11

I have had early access to OpenAI's GPT-5 for the last two weeks and it is the smartest model available as of now. Just as one example, here it created an anamorphic text illusion in minecraft that spells "SPECIAL" from one angle and "GPTFIVE" from another

adi

82,698 views • 11 months ago

On-chain trading of SPL tokens, such as TRUMP -> USDC, typically costs $0.1–$0.5 per transaction. XOS is set to reduce on-chain transaction fees by hundreds of times, making frequent trading much more accessible for users. Stay tuned for what’s coming next! #Solana #L2
0:15

On-chain trading of SPL tokens, such as TRUMP -> USDC, typically costs $0.1–$0.5 per transaction. XOS is set to reduce on-chain transaction fees by hundreds of times, making frequent trading much more accessible for users. Stay tuned for what’s coming next! #Solana #L2

XOS

116,821 views • 1 year ago

Went down like a sack of spuds and dropped her arse on the way down🤣🤣🤣 Setting a fine example for her two young boys as well. Real role model shit right there 🤣
0:17

Went down like a sack of spuds and dropped her arse on the way down🤣🤣🤣 Setting a fine example for her two young boys as well. Real role model shit right there 🤣

Dave 🏴󠁧󠁢󠁷󠁬󠁳󠁿

51,607 views • 3 years ago

Today we’re releasing Code Llama, a large language model built on top of Llama 2, fine-tuned for coding & state-of-the-art for publicly available coding tools. Keeping with our open approach, Code Llama is publicly-available now for both research & commercial use. More ⬇️
0:14

Today we’re releasing Code Llama, a large language model built on top of Llama 2, fine-tuned for coding & state-of-the-art for publicly available coding tools. Keeping with our open approach, Code Llama is publicly-available now for both research & commercial use. More ⬇️

AI at Meta

1,004,918 views • 2 years ago

After many days, my custom Vtuber display on Unity is finished ! → No webcam, only kbm and mic ⌨️🖱️🎙️ Fully customizable for myself but also for commissions. Looking for custom features and/or animations for your avatar, or even a full model ? Summon me by DM, sacrebleu ! 👻
0:19

After many days, my custom Vtuber display on Unity is finished ! → No webcam, only kbm and mic ⌨️🖱️🎙️ Fully customizable for myself but also for commissions. Looking for custom features and/or animations for your avatar, or even a full model ? Summon me by DM, sacrebleu ! 👻

𝑨𝒏𝒅𝒚 𝑯𝒆𝒍𝒍𝒈𝒓𝒊𝒎

213,813 views • 3 years ago

this is the BEST vision language model I have ever tried! Aria is a new model by Rhymes.AI: a 25.3B multimodal model that can take image/video inputs 🤩 They release the model with Apache-2.0 license and fine-tuning scripts as well 👏 I tested it extensively, keep reading to learn more 🧶
0:10

this is the BEST vision language model I have ever tried! Aria is a new model by Rhymes.AI: a 25.3B multimodal model that can take image/video inputs 🤩 They release the model with Apache-2.0 license and fine-tuning scripts as well 👏 I tested it extensively, keep reading to learn more 🧶

merve

176,047 views • 1 year ago

A wash and blow dry just cost me £5 in India! FIVE QUID!!!! No shade to British hairdressers as I know rents and other overhead costs for small businesses but the fact is - we are being ripped off on every level in Britain.
0:13

A wash and blow dry just cost me £5 in India! FIVE QUID!!!! No shade to British hairdressers as I know rents and other overhead costs for small businesses but the fact is - we are being ripped off on every level in Britain.

Narinder Kaur

557,683 views • 7 months ago

🚀 New Video Alert – GPT-5 in Copilot Studio! GPT-5 has officially arrived in Microsoft Copilot Studio, and I couldn’t resist putting it to the test. In this video, I explore: 🔍 GPT-5 Auto & GPT-5 Reasoning models for custom agents. 🛠 How I leveraged GPT-5 Chat inside a custom prompt for more tailored outputs. ⚖️ GPT-5 Chat Vs. GPT-4.1 to highlight improvements and trade-offs If you’ve been wondering what GPT-5 brings to Copilot Studio—this video is packed with real examples, practical tests, and insights to help you understand what these new experimental models have to offer. 🎥 Watch here: #CopilotStudio #GPT5 #AIBuilder #MicrosoftCopilot #LowCodeAI #OpenAI #AI #CopilotStudio #GPT5Auto #GPT5Reasoning #GPT5Chat
0:16

🚀 New Video Alert – GPT-5 in Copilot Studio! GPT-5 has officially arrived in Microsoft Copilot Studio, and I couldn’t resist putting it to the test. In this video, I explore: 🔍 GPT-5 Auto & GPT-5 Reasoning models for custom agents. 🛠 How I leveraged GPT-5 Chat inside a custom prompt for more tailored outputs. ⚖️ GPT-5 Chat Vs. GPT-4.1 to highlight improvements and trade-offs If you’ve been wondering what GPT-5 brings to Copilot Studio—this video is packed with real examples, practical tests, and insights to help you understand what these new experimental models have to offer. 🎥 Watch here: #CopilotStudio #GPT5 #AIBuilder #MicrosoftCopilot #LowCodeAI #OpenAI #AI #CopilotStudio #GPT5Auto #GPT5Reasoning #GPT5Chat

Daniel Christian

14,705 views • 11 months ago

How can we address the scarcity of data required for specialized AI? Learn about Simula, a framework that reframes synthetic data generation as dataset-level mechanism design. By using reasoning to architect datasets from first principles, Simula enables fine-grained control over coverage, complexity, and quality. More →
0:18

How can we address the scarcity of data required for specialized AI? Learn about Simula, a framework that reframes synthetic data generation as dataset-level mechanism design. By using reasoning to architect datasets from first principles, Simula enables fine-grained control over coverage, complexity, and quality. More →

Google Research

142,546 views • 3 months ago

Gemini 3 Flash is now available in MagicPath ⚡ From our earlier testing, it’s surprisingly strong at visual design for its size, and it’s fast! My recommendation: use it as your first-pass model to explore ideas, then switch to a bigger model for editing + refinements.
0:18

Gemini 3 Flash is now available in MagicPath ⚡ From our earlier testing, it’s surprisingly strong at visual design for its size, and it’s fast! My recommendation: use it as your first-pass model to explore ideas, then switch to a bigger model for editing + refinements.

Pietro Schirano

11,427 views • 7 months ago

GPT-5.5 just dropped. So did the Deel App for ChatGPT. Because a smarter model still needs accurate data. Powered by the world's largest proprietary compliance knowledge base, covering 150+ countries, maintained by hundreds of local legal experts, and updated as laws change. Whether you're a company hiring globally or a worker checking your pay, it's all there, right from chat. Try it 👇
0:16

GPT-5.5 just dropped. So did the Deel App for ChatGPT. Because a smarter model still needs accurate data. Powered by the world's largest proprietary compliance knowledge base, covering 150+ countries, maintained by hundreds of local legal experts, and updated as laws change. Whether you're a company hiring globally or a worker checking your pay, it's all there, right from chat. Try it 👇

Alex Bouaziz

14,144 views • 3 months ago

🤖 Get ready to automate your tasks with Taskade's AI Agents, now entering Beta, powered by GPT-4. Watch as AI Agents spring into action, autonomously researching topics, completing tasks, and orchestrating your entire workflows. Reply with 'AI Agents' for early access! ✨ 🐑
0:17

🤖 Get ready to automate your tasks with Taskade's AI Agents, now entering Beta, powered by GPT-4. Watch as AI Agents spring into action, autonomously researching topics, completing tasks, and orchestrating your entire workflows. Reply with 'AI Agents' for early access! ✨ 🐑

Taskade

16,057 views • 3 years ago

LEAK: Higgsfield has been secretly using GPT-5 for a month that viral "Earth Zoom Out" trend with 1B views? Yeah… GPT-5 made it. Higgsfield AI brought it to life and now, this insane duo is open to everyone.. for FREE step by step tutorial:
0:10

LEAK: Higgsfield has been secretly using GPT-5 for a month that viral "Earth Zoom Out" trend with 1B views? Yeah… GPT-5 made it. Higgsfield AI brought it to life and now, this insane duo is open to everyone.. for FREE step by step tutorial:

el.cine

52,431 views • 11 months ago

Seedance 2.0 Mini Is Here — 50% off for 7 days. A lightweight version that delivers excellent value. Lighter on your budget, heavy on results — as told by the content team and creators. Try it now and see for yourself.
0:10

Seedance 2.0 Mini Is Here — 50% off for 7 days. A lightweight version that delivers excellent value. Lighter on your budget, heavy on results — as told by the content team and creators. Try it now and see for yourself.

Pollo AI

196,166 views • 1 month ago