๐ฑProduct Hunt โฒ Vercel Day โฒ ๐
Launch day - Support chaaaaa
https://t.co/BN05mTJzoE
Your daily driver ai assistant where tokens are no more limited
Ai chat, deep search and Games check it out and help me with an upvote !
Thank you for the ones who cares ๐๐ฝ
So here the 2M parameter model CLAUDE FABLE training on GPU
@TheAhmadOsman its not #BuyaGPU but #TRAINYOUROWNFABLE
Hahah so much fun the gradient loss is actually at 0.2 so its really good --- 380 epochs lefts
@TheAhmadOsman We are ร 0.0 means the model has fully understand every patterns waiting to see the generation in actiondamn i was not expecting that so early
What is the most coding efficient by tokens cost llm for the same task?
Like Minimax 3 on @kilocode is mentionned to be the less costly but itโs not the better model.
Any benchmarks on this with the last open models ?
Making my own Claude Fable its over for you
No more APIs costs @OpenAI@AnthropicAI
be careful @karpathy@NoamShazeer i'm coming after you guys <3
pps: The result are so funny
@robj3d3@AnthropicAI Yo every post is about fable, i did not use it back in the days. Its a drug or something ? ๐๐๐ผ You scare me man, stay healthy