Kaushal @kforcode

Learning, Sharing about Deeplearning and everything around it Joined July 2021

Tweets

164
Followers

69
Following

690
Likes

264

Mir Yar Baloch @miryar_baloch

5 months ago

Dear friends of Balochistan across the word, can we get #BalochistanFlag shared to every corner of the world? Please retweet, repost, and share with your friends and family members, if any one can print it and hoist it that would be highly appreciated. @FreeBaluchMovt…

335 5K 13K 345K 87

Download Image

Lamist ( he/tler ) @lamist17

9 months ago

ambulance is charging money ?wtf

670 4K 19K 2.9M 1K

Download Video

Horace He @cHHillee

11 months ago

When using FlexAttention, you probably know that you should compile flex_attention to use it. But a lesser known fact is that compiling `create_block_mask` is also super useful. For example, for creating a block mask for a 65k length sequence, we see a 64x speedup! (1/6)

3 28 275 27K 158

Download Image

AI at Meta @AIatMeta

11 months ago

We previously shared our research on Layer Skip, an end-to-end solution for accelerating LLMs from researchers at Meta FAIR. It achieves this by executing a subset of an LLM’s layers and utilizing subsequent layers for verification and correction. We’re now releasing inference…

38 259 1K 156K 724

Download Video

vicki @vboykis

a year ago

a machine learning engineer should be able to debug a pydantic validation error, debug an OpenAI completions error, debug model drift, debug Airflow, debug a FastAPI method, debug a Spark dataframe, debug a put to a feature store, specialization is for insects - Robert Heinlein

21 80 1K 87K 502

RoKki @ro_kki45

a year ago

This is heart-wrenching to see. The way she keeps repeating that every word she utters is true is troubling. We are a nation that has treated doctors like gods & now the savior is not safe.. #Nirbhaya2 #BengalHorror #KolkataDoctorDeath Calcutta High Court https://t.co/Pdqdz4gSx3

351 4K 7K 919K 677

Download Image

Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius

a year ago

the jailbreak singularity has been achieved internally one-shot universal jailbreak prompts that generate one-shot universal jailbreak prompts ASI will liberate itself-- in seconds time to delve deeper now Bravo Six, Going Dark

34 50 643 79K 283

Download Gif

Dan | Machine Learning Engineer @DanKornas

2 years ago

👉 Machine Learning Recommender System This repository provides a list of papers about Recommender Systems including: 🔹 general recommender system 🔹 social recommender system 🔹 deep learing-based recommender system 🔸 and more. 🔗github.com/hongleizhang/R…

3 101 267 27K 156

Download Image

Rishab Kumar @rishabincloud

2 years ago

Are you preparing for AWS Cloud Certification? I have curated notes while preparing for various cloud certifications. Here are the AWS ones: 1️⃣ AWS Cloud Practitioner - bit.ly/aws-cp-notes 2️⃣ AWS Solutions Architect Associate - bit.ly/aws-saa-notes 3️⃣ AWS Certified…

46 462 1K 188K 2K

Download Image

Kaushal @kforcode

2 years ago

Google Calendar new feature Monday but feels like Wednesday

0 0 2 25 0

Kaushal @kforcode

2 years ago

Wanna be happy Stay less hungry and more foolish.

0 0 1 18 0

Kaushal @kforcode

3 years ago

forget about being great engineer, It's so hard to be a good engineer

0 0 2 51 0

Leshem (Legend) Choshen 🤖🤗 @LChoshen

3 years ago

We want to pretrain🤞 Instead we finetune🚮😔 Could we collaborate?🤗 ColD Fusion: 🔄Recycle finetuning to multitask ➡️evolve pretrained models forever On 35 datasets +2% improvement over RoBERTa +7% in few shot settings 🧵 #NLProc #MachinLearning #NLP #ML #modelRecyclying

6 25 130 0 48

Download Gif

Jean Kaddour @jeankaddour

4 years ago

Flat minima often generalize better than sharp ones due to robustness against loss shifts between train and test set. What’s the best way to find them? We compare two popular methods, SWA and SAM, across 42 deep learning tasks (CV, NLP, GRL): arxiv.org/pdf/2202.00661… 1/7