Recent Posts
VjJ/WN
No.475170
Abey lindu suar, 105b parameters 600b ko hara diya? Pindu haramkhor
>“At 105 billion parameters, on most benchmarks this model beats DeepSeek R1 released a year ago, which was a 600-billion-parameter model."
>"It is cheaper than something like a Gemini Flash, but outperforms it in many benchmarks”
L7x0XI
No.475177
>>475170(OP)
Yaar bhosadpillet
JKz0Pc
No.475179
>>475170(OP)
It's good but the reason for doing it and the way its being done is wrong.
India shouldn't let the driving force be wanting to impress others.
It should be well thought out and fulfill a ujecase for Indians in the first place and also keep in mind the resources available to the country and indians.
Sharing Sarvam for free with the world is retarded.
VjJ/WN
No.475183
>>475177
Look at these shitty comparisons. Talk me Deepseek bhaukega lekin benchmark sab Gotgolpa tier model se karega.
+tiufK
No.475186
+tiufK
No.475187
+tiufK
No.475188
CtNKuF
No.475189
>>475179
yaar haggu mulle
VjJ/WN
No.475195
>>475179
Voice models are not that important idk why they are focusing on Voice model and keep on saying "indian languages"
JKz0Pc
No.475197
>>475195
they most likely used this as a ujecase to sell it to the older people in GoI
old people think languages are very difficult and indian languages even more so so telling them we can do indian languages better than phoren probably sold it to GoI
even if its true its still stupid to leave it open and let non Indians use Indian resources
if Sarvam was a private company using its own funds that would be a different story tho
z9HfSE
No.475199
>>475197
Why is this mulla saying all thism
PbC4Vn
No.475205
>>475202
Picrel yourself?
