jumpCastle

w zeszłym rokudiscuss

It was a good benchmark until it entered the training set.

w zeszłym rokudiscuss

But open source like aider

w zeszłym rokudiscuss

Deepseek v3 was FP8

w zeszłym rokudiscuss

2 lata temudiscuss

Stock prices cannot go up without if the planet is destroyed

2 lata temudiscuss

Attention was invented because Bengio lab had to be disciplined about a black box (google had more compute)

2 lata temudiscuss

Aren't services like runpod solve half of these concerns?

2 lata temudiscuss

Are the updates really slow?

2 lata temudiscuss

Spend more time with my side projects.

2 lata temudiscuss

2 lata temudiscuss

You can use older models with the api.

2 lata temudiscuss

Also the parameters are optimized also with loss of future tokens in the sequence.

comments