Claude Code automatically adds itself to the commit metadata "Co-Authored-By: Claude Opus <[email protected]>".
I strongly disagree with crediting Claude as a commit co-author. It is a tool I'm paying for, not a copyright holder, not an author. We don't credit Visual Studio.
Building TamilLM from scratch has been humbling.
The hard part is not just training a model. It is everything before that: finding usable Tamil data, OCR’ing old books, separating modern prose from classical text, balancing English/Tamil translation support, avoiding copyright traps, deduping web data, and proving the training stack before spending real GPU money.
We now have enough tokens for a serious first run, but the work exposed the real challenge: Tamil is not one lane. A good Tamil model needs modern writing, spoken usage, English support, translation, literature, history, and classical grounding.
Also learning the practical side: H100 capacity is scarce, data quality matters more than raw size, and every shortcut becomes expensive later.
Still moving. TamilLM is getting closer.
The AI ponzi scheme goes like this:
Everyone is generating all these long ass docs and then passing them off for others to read
Then the person receiving is like, wtf this is way too long, and hands that into an AI to read and summarize
Then they are generating a long ass response back
and this cycle goes like that forever. and we call this work now 😅
The token lords watch this from their towers nodding and grinning.
I open my laptop.
First Warp want to be updated
Codex wants to be updated
Chrome want to Finish updates
Docker Desktop wants to be updated
IntelliJ IDEA wants to be updated
Ollama wants to be updated
After doing all these updates, I take a 15 minutes break to remember why I opened my laptop.
Why do autorickshaw bhayyas think that if the front wheel of their auto squeezes into a narrow space, their whole auto can fit too???
#Hyderabadtraffic