Deep Learning

@DeepLearning18

Deep Learning - Life School

انضم في يونيو 2019

10المنشورات 139المتابعون 13المتابَعون

قد يعجبك

@Club2030Afrique

@pbpilatre

@URPSCDNA

@MCResearchInst

@fmchevalier

@GuillaumeLabbez

@maestro2balzak

@chantal_edouard

@evinhas

@frededlens

Deep Learning أعاد

Andrew Ng

@AndrewYNg

٢٩ نوفمبر ٢٠٢٣ م

New short course on sophisticated RAG (Retrieval Augmented Generation) techniques is out! Taught by @jerryjliu0 and @datta_cs of @llama_index and @truera_ai , this teaches advanced techniques that help your LLM generate good answers. Topics include: - Sentence-window retrieval,…

Deep Learning أعاد

David Mimno

@dmimno

١٧ يونيو ٢٠٢٢ م

The Text as Data Conference (TADA) 2022 will be happening at Cornell Tech on beautiful Roosevelt Island in NYC Oct 6-7. Abstract submissions are due July 18. Please RT!

dmimno's tweet image. The Text as Data Conference (TADA) 2022 will be happening at Cornell Tech on beautiful Roosevelt Island in NYC Oct 6-7. Abstract submissions are due July 18. Please RT!

Deep Learning أعاد

Dan Jurafsky

@jurafsky

٣٠ ديسمبر ٢٠٢١ م

Here's the December 2021 release of draft chapters for Speech and Language Processing, just in time for Jim and I to wish you all a Happy New Year! Enjoy! web.stanford.edu/~jurafsky/slp3/

Deep Learning أعاد

Basil Mustafa

@_basilM

٧ يونيو ٢٠٢٢ م

Beep beep! Introducing LIMoE, the Language Image Mixture of Experts: a single model, processing both modalities for contrastive image-text modelling. Cruises straight to 84.1% 0shot ImageNet accuracy without any modality-specific architectures or pre-training. (1/10)

_basilM's tweet image. Beep beep! Introducing LIMoE, the Language Image Mixture of Experts: a single model, processing both modalities for contrastive image-text modelling. Cruises straight to 84.1% 0shot ImageNet accuracy without any modality-specific architectures or pre-training. (1/10)

Deep Learning أعاد

rohan anil

@_arohan_

١٤ سبتمبر ٢٠٢٢ م

Just remembered again how true this with learning the LR literature, there is so much compute spend in finding meta-hyper-parameters (that’s hidden away) that results look great in paper but fails on new problem.

Ben Poole

@poolio

١٣ سبتمبر ٢٠٢٢ م

"We present a hyperparameter-free method..." def beautiful_method(): return method(lr=3e-4, alpha=0.13, gamma=pi/9.1)

Deep Learning أعاد

Darek Kłeczek

@dk21

١٥ سبتمبر ٢٠٢٢ م

How train the best performing NLP model? 3 techniques to experiment with. 🧵

Deep Learning أعاد

Stanford NLP Group

@stanfordnlp

١٥ سبتمبر ٢٠٢٢ م

The SERAC model for small edits to large neural networks by @_eric_mitchell_ Charles Lin @ABosselut @chrmanning @chelseabfinn is the bottom item in @DeepLearningAI_’s The Batch this week. We’re not sure why this week—it must be a slow AI news week. 😭 deeplearning.ai/the-batch/issu…