Brainkite
@brainkite.bsky.social
Yesterday, tried to vibe code on cursor a simple benchmark to compare latency of detection modes. At one point it literally hardcoded results in the code without running the model!
Would have been faster to do it myself.
Let alone rebuilding a country-sized database 😂
Would have been faster to do it myself.
Let alone rebuilding a country-sized database 😂
March 29, 2025 at 12:54 PM
Yesterday, tried to vibe code on cursor a simple benchmark to compare latency of detection modes. At one point it literally hardcoded results in the code without running the model!
Would have been faster to do it myself.
Let alone rebuilding a country-sized database 😂
Would have been faster to do it myself.
Let alone rebuilding a country-sized database 😂
Great tool, saved my life 2-3 times
February 18, 2025 at 11:15 AM
Great tool, saved my life 2-3 times
The guy made everything pretty obvious and clear during his inauguration speech. They are acting so fast because there is some inertia for people to actually accept and realise that: it is what it is and not something else.
February 9, 2025 at 6:39 PM
The guy made everything pretty obvious and clear during his inauguration speech. They are acting so fast because there is some inertia for people to actually accept and realise that: it is what it is and not something else.
1000 languages, this is mind boggling. Very excited to see the impact it’s going to have on the open source community.
December 9, 2024 at 11:10 AM
1000 languages, this is mind boggling. Very excited to see the impact it’s going to have on the open source community.
This is great! Congrats to the FineWeb team. Dataset work is always massively under appreciated.
December 9, 2024 at 11:06 AM
This is great! Congrats to the FineWeb team. Dataset work is always massively under appreciated.