🐼

senior dapp/blockchain engr
cats 🐱 | @vielski_ ❤️

[Deno @deno_land]

Programming should be simple

[🐼 @bmaaaatttt]

🤞 🤞

[AK @_akhaliq]

JPMorgan announces DocLLM

A layout-aware generative language model for multimodal document understanding

paper page: https://t.co/azTKT5jZjH

Enterprise documents such as forms, invoices, receipts, reports, contracts, and other similar records, often carry rich semantics at the intersection of textual and spatial modalities. The visual cues offered by their complex layouts play a crucial role in comprehending these documents effectively. In this paper, we present DocLLM, a lightweight extension to traditional large language models (LLMs) for reasoning over visual documents, taking into account both textual semantics and spatial layout. Our model differs from existing multimodal LLMs by avoiding expensive image encoders and focuses exclusively on bounding box information to incorporate the spatial layout structure. Specifically, the cross-alignment between text and spatial modalities is captured by decomposing the attention mechanism in classical transformers to a set of disentangled matrices. Furthermore, we devise a pre-training objective that learns to infill text segments. This approach allows us to address irregular layouts and heterogeneous content frequently encountered in visual documents. The pre-trained model is fine-tuned using a large-scale instruction dataset, covering four core document intelligence tasks. We demonstrate that our solution outperforms SotA LLMs on 14 out of 16 datasets across all tasks, and generalizes well to 4 out of 5 previously unseen datasets.

[Watcher.Guru @WatcherGuru]

JUST IN: 🇵🇭 Philippines to block access to #Binance in the country.

[Elon Musk @elonmusk]

This letter about OpenAI was just sent to me.

These seem like concerns worth investigating.

https://t.co/hnAepyfLE6

[Jim Fan @DrJimFan]

OAI’s economy of scale is a killer advantage. Some back of envelope calculations:

It only costs $15 to read or $45 to write the ENTIRE Harry Potter series, all 7 books, with GPT-4-turbo.

It costs $180 for GPT-4-V to watch ALL 8 HP movies, 1 frame/s, at 360p resolution.

[Rowan Cheung @rowancheung]

🚨 BREAKING: OpenAI just released 128k context GPT-4 Turbo at 1/3rd the price.

This means GPT-4 is cheaper, faster, and can now read a prompt as long as an entire book.

The announcement comes with additional huge news for developers:

- New Assistants API for Code Interpreter and Retrieva
- GPT-4 Turbo updated knowledge cutoff to April 2023
- New GPT-4 Turbo Vision API
- New Dall-E 3 API
- New text-to-speech API (6 preset voices to choose from & insanely realistic voice)
- A new Whisper V3 model coming soon
- Slashed rate limits (cheaper development costs)

Prepare for a whole new wave of AI tools.

[🐼 @bmaaaatttt]

@cindeeeeeh @angelo_agliam hahahaha set na @angelo_agliam

[🐼 @bmaaaatttt]

@cindeeeeeh @angelo_agliam not sure pa pero pplano din SG

[🐼 @bmaaaatttt]

@cindeeeeeh @angelo_agliam san kayo manonood?

[🐼 @bmaaaatttt]

dfq haha 🤡🤡

[Elon Musk @elonmusk]

As promised

[🐼 @bmaaaatttt]

vibing to quinn xcii just feels right...

[🐼 @bmaaaatttt]

@cindeeeeeh same experience haha parang ayaw ka bigyan ng trabaho

[🐼 @bmaaaatttt]

@cindeeeeeh

[🐼 @bmaaaatttt]

@angelo_agliam ok po

[🐼 @bmaaaatttt]

@vielski_ iloveyouuu 🥰 tulala man baga ako dyan haha

[🐼 @bmaaaatttt]

@vielski_ miss na miss na din kita 🥺🥺

[🐼 @bmaaaatttt]

@vielski_

[trash @trashh_dev]

no loyalty.

[trash @trashh_dev]

startups go brrr