epoch: 26/n, Data scientist | ML engineer.

Bengaluru, India
Joined October 2020
Using this library (from google import genai), I get info of tokens used (input/output/multimodal). Would be super useful if the response also includes the exact cost of that API request. @OfficialLoganK @googleaistudio @patloeber
1
Perfect timing to roll out the new ChatGPT pricing tailored for the Indian audience with UPI
Aditya shinde retweeted
If you’re cringing at your old work, it means you’re getting better
What's the greatest invention of mankind?
Why does millennials think genz’s are 10 years old
Can it fix SBI upi server? pj…
a nice cold shower can actually fix everything
How @Tesla’s FSD gonna work in india?
1
Aditya shinde retweeted
We've heard your feedback — Kimi K2 is SLOOOOOOOOOOOOW 😭 Especially for agentic apps, output tokens per second really matters. The main issue is the flooding traffic and huge size of the model, we are actively working on inference optimization and BUY MORE MACHINES! Speed should improve significantly in the next few days. Meanwhile, K2 is also accessible via other providers (@GroqInc @OpenRouterAI @SiliconFlowAI @togethercompute @FireworksAI_HQ) — it's fully open-source, so anyone can host it.
When will @zomato and @Swiggy launch their mcp’s (to take orders)
1
The problem with vibe coding for a non developer, thats why cursor is balanced platform to work on
1
Aditya shinde retweeted
Google DeepMind's Logan Kilpatrick says AGI will be a product experience. Not a model. His bet: whoever nails memory + context around decent model at a product level wins. Users will suddenly feel like they're talking to AGI. Not from capability breakthrough, but experience breakthrough. At some point, the AGI race is actually a UX problem.
Aditya shinde retweeted
I just got scammed on @Flipkart — not once, but TWICE.  So I ordered a ₹2.6 lakh MacBook Pro and got a used, old model inside a sealed Apple box. We filmed the entire open box delivery, you can see that in first attached video. I immediately requested a replacement, and this time I unboxed the product at the EKart office under CCTV, and again the MacBook had smudges and marks so basically not new (second unboxing video attached below). 2 times. 2 wrong deliveries. After a discussion with customer care they started a return and then after 2 days they called me and offered me a compensation of 13K which I rejected, then 18K I rejected again and finally they offered me 10% which I rejected again. And after all this negotiation they rejected the return. So I called them again and I asked for 10% compensation as it was a lot of hassle by talking to them. And moreover, I was able to clean the marks but the laptop was not new in the first place. So why am I posting about it, here’s the disturbing part: it’s not just me. This seller called "Treasure Haul Online" has been scamming people for years. Users on LinkedIn, Reddit, YouTube, and X have posted about the same seller since 2023. For more information I have made a video and link to that is given below this tweet. I genuinely have 3 questions to ask: 1. How are fake products getting packed in sealed Apple boxes? 2. How is this seller still active despite public complaints from 2023? 3. How is this “Assured” if even Flipkart can’t verify what’s inside? @flipkartsupport, this isn't a mistake, it's negligence. You are enabling fraud through your platform. This requires investigation, I am requesting action against the seller and Flipkart. @jagograhakjago @CCI_India @GoI_MeitY @consumr_rights @consumerforum_ @Apple
Yesterday i was generating a synthetic data (around 4000 requests), after sometime due to some laptop issue i lost it. Is there way to find the response traces. I was using google generative api(paid tier). @OfficialLoganK @patloeber
Cooking reasoning model in hindi. similar to how deepseek r1 or o1 thinks in english
1
3
adding to the watch list
1
Which is currently the best base model pre trained in india. Is it from sarvamai or aibharat or something else?
1