abhinavkulkarni

  Location: India (H-1B)
  Remote: Yes
  Willing to relocate: Yes
  Resume: https://abhinavkulkarni.github.io/assets/Resume.pdf
  Technologies: Machine Learning, NLP, LLM, Generative AI
  Email: abhinavkulkarni[at]gmail[dot]com

Machine Learning engineer with expertise in Deep Learning focussed on NLP and LLMs. I have experience dealing with web-scale data. I recently helped integrate AWQ 4-bit quantization to HuggingFace's TGI: https://bit.ly/3QuZmKk

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

SEEKING WORK | REMOTE | WORLDWIDE | FLEXIBLE SCHEDULE

Location: India (H1-B approved)

Remote: Yes

Willing to relocate: Yes

Machine Learning engineer with expertise in Deep Learning focussed on NLP and LLMs. I have experience dealing with web-scale data.

I recently helped integrate AWQ 4-bit quantization to HuggingFace's TGI: https://bit.ly/3QuZmKk

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

  Location: India (H-1B)
  Remote: Yes
  Willing to relocate: Yes
  Resume: https://abhinavkulkarni.github.io/assets/Resume.pdf
  Technologies: Machine Learning, NLP, LLM, Generative AI
  Email: abhinavkulkarni[at]gmail[dot]com

Machine Learning engineer with expertise in Deep Learning focussed on NLP and LLMs. I have experience dealing with web-scale data.

I recently helped integrate AWQ 4-bit quantization to HuggingFace's TGI: https://bit.ly/3QuZmKk

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

SEEKING WORK | REMOTE | WORLDWIDE | FLEXIBLE SCHEDULE

Machine Learning engineer with expertise in Deep Learning focussed on NLP and LLMs. I have experience dealing with web-scale data.

I recently helped integrate AWQ 4-bit quantization to HuggingFace's TGI: https://bit.ly/3QuZmKk

I have a keen eye for product development. 12+ years of experience working for SV mid/large-sized companies.

While my focus is Machine Learning, I'm a swiss army knife in overall backend development, System Design, and dealing with scale.

My LinkedIn is the bio.

There already are techniques to guade LLMs into producing output that adhere to a schema. For e.g. forcing LLMs to stick to a Context-Free Grammar: https://matt-rickard.com/context-free-grammar-parsing-with-l...

Just like many similar methods, this is based on logit biasing, so it may have an impact on quality.

Here's one: https://huggingface.co/spaces/mike-ravkine/can-ai-code-resul...

Hey @minxomat,

HyDE sounds like an interesting approach. All dense retrieval approaches suffer from the problem you outlined in the blogpost. Have you looked at keyword-based or late-interaction models for retrieval such as ColBERTv2[1]? I find that late-interaction methods seem to offer best trade-off between semantic intelligence (precision) and retriving relevant documents (recall).

[1] https://github.com/stanford-futuredata/ColBERT

S&P’s projection hinges on the continuation of India’s trade and financial liberalization, labor market reform, as well as investment in India’s infrastructure and human capital.

I am surprised nobody has mentioned Flutter yet. It obviates the need to learn HTML and CSS (for the most part). Yes, Flutter is not very web-friendly (SEO is a pain, etc.), but I have found it to be the best way to prototype an idea and put it out there.

Do give it a try. There are plenty of YouTube tutorials out there.

Yeah, that was great. Here are John's other appearances on other podcasts: https://jkstream.com/#/personality-john-carmack-Q92605

No, speaker diarization is not part of Whisper. There are open source projects - such as Kaldi [1], but it's hard to get them running if you are not an area expert.

[1] https://kaldi-asr.org/

We are trying to remove some of that noise by building a better podcast discovery and search product at JKStream (https://jkstream.com). For e.g., here's a cleaner feed for Malcolm Gladwell: https://jkstream.com/#/personality-malcolm-gladwell-Q318429. You, as a user choose to subscribe to your favorite thought leaders and personalities and you get a much cleaner, personalized podcast feed.

We at JKStream (https://jkstream.com) are trying to solve some of these problems by allowing users to subscribe to their favorite podcast interviewers and guests. For e.g., here's a nice, clean feed for Cal Newport: https://jkstream.com/#/personality-cal-newport-Q16733303

> His podcast is awesome too: https://www.calnewport.com/podcast/

Keep track of his activity in the podcast sphere here: https://jkstream.com/#/personality-cal-newport-Q16733303

> Randall is an interesting podcast guest

You can keep track of Randall's podcast activity here: https://jkstream.com/#/personality-randall-munroe-Q285048

You can keep track of Mike Duncan's podcast activity here: https://jkstream.com/#/personality-mike-duncan-Q361250

There are still problems with search and discovery though. A large scale transcription service needs to exist - potentially supported by client side transcription when GPUs become ubiquitous - at least for top podcasts. So many people have this gripe that they heard something in some podcast, but can't remember now.

Also, at JKStream (https://jkstream.com), we are trying to solve one aspect of search & discovery. You can subscribe to your favorite podcast interviewers and guests and get notified.

You may want to read through blog of Lyn Alden: https://www.lynalden.com. She has good explanation of why US always runs deficit.

You may be interested in Andrej Karpathy's experiment with OpenAI's latest STT Whisper model. Relevant tweet here: https://twitter.com/karpathy/status/1574474950416617472

By the way, if you are looking for a clean podcast consumption experience, do give a try to https://jkstream.com. Easy way to subscribe to your favorite podcast interviewers and guests.

A simple Google search on Redlining (https://www.google.com/search?q=new+york+times+redlining) will reveal how current housing laws and massive wealth transfer that interest mortgage deduction is from less fortunate to the wealthy. There is a near unanimity among economists that housing subsidies are a bad policy and exacerbate rent seeking behavior. If not racist, those policies are downright classist (the line can be very blurry), designed to keep a small minority of homeowners (akin to landed gentry) in power.

How is eating and drinking cheap in the Bay Area? A $4.25 latte with tips and taxes easily adds up to $6.50-7.00. That's a lot of just a cup of latte.

Add to that imminent price increases due to impending minimum wage hike in SF in July.

Well, Tesla is transitioning from luxury cars to the mainstream ones.

Secondly, how about reducing our reliance of fossil fuels and instead using electricity for driving the cars (I know, I know, electricity comes largely from fossil fuels, but net, it is beneficial to use it for driving cars than gas)? That's not going to benefit the common public?

Loosening of regulation and cheap influx of money, on the whim printing of money by the Fed. Case in point, the current situation with student loans.

It's politically beneficial to make slogans about housing for all, education for all, etc. and then institute policy to distribute cheap money. Least time we did that, you know what happened (2008).

I think we need to be careful about attributing all of the China's success and splurge of VC funding to IP theft, hostile domestic market for foreign players, etc. China is succeeding because it really wants to.

With manufacturing in US hollowed out in last couple of decades and China being the new global manufacturing powerhouse, it is not surprising that China is seeking smart (software/cloud/AI/robotics powered) solutions to industrialization, warehousing, transportation, supply chain, payments and other areas. Why hasn't US achieved the level of digital payment smoothness that China has been able to?

While the points you made about lax laws and blatant data privacy invasion are an contributing factors, I think you would have seen China making great strides in the above mentioned areas even if laws were enforced strictly. It's important to recognize China's technological prowess. They have similarly taken a lead in bioengineering.

Robotics. Not humanoids necessarily (I find them creepy to tell you the truth), but the ones that will act in the background making mundane and repetitive tasks easier - garbage collection, cleaning, watering plants, restocking grocery store aisles, further automating industrial tasks, etc.

Yep. Bioengineering and Bioinformatics, especially with recent gene editing technologies invented. We still don't know what gene patterns cause most of the diseases. Sifting through genetic data (with the help of computers) and finding the patterns responsible (so that they can edited out or modified) for these diseases is going to be a huge industry. As a result, diagnosis and medicine would become personalized to a certain extent.

> I'd like to know where you're getting those steady 10% rates of return.

Index funds?

If that's the case, then perhaps we should lay (some of the) blame on the people who would chose to 'continue' the career of the said politician, no? Lets not deny agency the people have in making informed decisions.

Non-paywalled/archive link?

Not a problem, sure California residents have this right of self determination. But then stop using my tax dollar to enable your rent seeking behavior (prop 13) and paying next to nothing towards local development (again, prop 13) while shifting that burden to the renter (higher rents being partly used to fund libraries, fire stations, etc. when really, your local taxes should have been paying for those, not my continually rising rent).