We just opened sourced SQLCoder-70B! It outperforms all publicly accessible LLMs for Postgres text-to-SQL generation by a very wide margin.

SQLCoder is finetuned on @AIatMeta's CodeLlama-70B model that was released yesterday on less than 20,000 hand-curated prompt completion pairs.

You can find it on @huggingface at . This follows our 15B, 7B, and 34B models – and is the most capable of them all.

The model has a cc-by-sa-4 license, which means that you are free to use it as is for any use (including commercial) as long as you also open-source any changes to you make to it (i.e., if you fine-tune it further).huggingface.co/defog/sqlcoder… Some nerd stuff! 🤓

- If you want a very simple way to play with it, check out our Github repo:

- You'll need a TON of VRAM to this fast. We've found doing AWQ quantizations a really good to keep accuracy high while keeping latency and VRAM low. Would recommend!

- Larger models have so much more "intelligence" than smaller ones. This had a very similar dataset as our 34B model, but performs significantly better. It does have some know flaws and weirdly trips up on certain words (like "domain", strangely) – but we'll fix that soongithub.com/defog-ai/sqlco…

Sigh. Delhi's air is going to crap again. Many reasons for this, but stubble burning is a massive contributor, and it is mostly concentrated in Punjab

Penalizing farmers is not the solution to this – this is actually a legislative and *groundwater* issue (more below) Stubble burning has gone up a lot in the last 30 days, as it does around this time ever year, because farmers have to clear their fields and burning stubble is the most efficient way to do this.

@ShekharGupta did a great video about this 3 years ago:

Finished a 14 day experiment with a continuous glucose monitor in my arm. Learnt a lot about how my body reacts to food!

🧵w graphs below. May be useful for folks with high blood sugar

All food in this thread was home cooked. I wrote some python to extract & clean the data A surprising learning for me – carb heavy meals after not eating for a while cause a huge sugar spike!

Rajma + a whole meal wrap after 22 hours of fasting led to this. When intermittent fasting, will avoid a carb-heavy lunch from now [2/]

After doing more research on this (caveat – I'm not a doctor):

1. The AIIMS director seems to have acknowledged the presence of a mutant variant in India – but didn't say if this was a new variant or a previously known variant

timesofindia.indiatimes.com/city/delhi/mut…

https://twitter.com/rishdotblog/status/1380039664266924033

2. Some variants can reinfect those who were previously infected. The South Africa variant escaped antibodies in 48% of those who had previously recovered from Covid (caveat: small sample size of 44)

nejm.org/doi/full/10.10…

Reflecting on our decision to shut down Popper.ai last year. The product consistently served 100M+ unique monthly IPs, and produced content that was always on the first page of Google

But it failed as a business, largely because of my management failures. A long🧵 Popper.ai started when @pysamarth and I realised that many of the articles published in the news industry were repetitive, and could be automated

We started experimenting with automated stories about the economy, and eventually generated 100k stories/month [1/]

My dad in India was recently in the ICU for 8 days, with a severe case of Covid+Dengue. He is 57 years old and also has diabetes. He's now recovering and is out of ICU, but is still hospitalized

Sharing what I learnt through the process in case it's useful for someone else [1/n] He had extremely high fever (104+°F) when he first started presenting symptoms on Nov 9, and his platelets were falling rapidly. My mom got him tested for both Dengue and Covid, where he tested negative for Dengue and positive for Covid [2/n]