Sunday small guessing puzzle Let's say I have 3 tokenizers:
– llama2: 32k vocab
– falcon: 65k vocab
– GPT4: 100k vocab I take ~2M random documents from the web (let’s say 10 random parquet files from RefinedWeb from https://
huggingface.co/datasets/tiiua
e/falcon-refinedweb
… roughly 1B tokens). I tokenize them
DATA
-
Tokenizer vocabulary size comparison across language models
By
–
-
Cyber Attack Prevention: Essential Backup Strategies
By
–
Plan for the Unexpected No one ever plans to be a victim of cyber attacks but it can happen to anyone so at all times you must have backup. Google users automatically have 15gb of cloud storage, use this wisely!
You can also utilize other providers like Dropbox, Onedrive, etc. -
Keep Your Apps Updated: Data-Conscious Software Maintenance
By
–
Tip 2: Keep Things Fresh. Develop a habit of updating your apps and software regularly! Don't see this as just another avenue to consume your data(this is specifically for those of us in third-world countries who have limited data)
-
Data Brokers Sell Consumer Information to Scammers
By
–
A data broker collects and sells consumer information, like phone numbers and email addresses. Scammers can buy your phone number from these companies and use it to contact you.
-
Data Breaches and Data Broking: Understanding Privacy Risks
By
–
I am sure we get one of this every day. Ever wondered how they have access to your numbers and emails knowing you don't give out your details very often? This happens due to data breaches or data reselling commonly called "data broking".
-
Why Cybersecurity Matters for Your Digital Treasure
By
–
Why Does Cybersecurity Matter? Okay, imagine your digital life is like a treasure chest. It's got everything from customer data to your personal info. As someone who uses the internet every day, protecting this treasure should be a concern.
-
Making Scientific Data Accessible and Discoverable for AI
By
–
How to make your scientific data accessible, discoverable and useful https://
go.nature.com/44WvGvx #AI #MachineLearning #DeepLearning #LLMs #DataScience -
Custom tool identifies influential designers for product feedback
By
–
@zoink built a built a custom app to find the most influential designers on Twitter and cold DM'd them to get their feedback on the early version of @Figma
: (Here's a link to the tool: https://
x.com/zoink/status/1
566566649712431105?s=20
…) -
Supply Chain Inefficiencies: $400M Hourly Costs Impact
By
–
According to WSJ, $9B of goods stuck for a day, costs $400 million per hour. Supply chains are the lifelines that connect raw materials to end consumers.
— Helen Yu (@YuHelenYu) 8 septembre 2023
Delighted to host #BizTalks on "Solving Supply Chain Inefficiencies" with @TamaraMcCleary and @bevictoryus.
Watch the… pic.twitter.com/3kS8uHFbAIAccording to WSJ, $9B of goods stuck for a day, costs $400 million per hour. Supply chains are the lifelines that connect raw materials to end consumers. Delighted to host #BizTalks on "Solving Supply Chain Inefficiencies" with @TamaraMcCleary and @bevictoryus
. Watch the -
California Delete Act Enhances Consumer Data Privacy Rights
By
–
“Rights that are too complicated to use are not really rights.” – @SenJoshBecker I spoke w/ Becker about his Delete Act bill to regulate data brokers, improve transparency, add accountability and give Californians better control of their data. #SB362 https://
digiday.com/marketing/cali
fornia-state-senator-josh-becker-on-privacy-data-brokers-and-why-hes-sponsoring-the-delete-act/
…