Phi-4 = gpt-4o-mini
#4
by
maxbn
- opened
Thanks for the model!
I tested my set of questions for the test and Phi-4 showed results similar to GPT-4o-mini.
I ran this model on my PC:
- Ryzen 5-3600
- RTX3050 (8Gb)
- 16Gb RAM
I'm looking forward to the official release of the model to see the results of professional tests!
How many token speed you get?
How many token speed you get?
I use rtx4090, 5.5bpw is less than 60tokens/s at the fastest.
26.7 tokens/s with RTX 3060 gaming OC 12GBvram
Ryzen 5600x
32GB ram
Wow for 3060 is giving too good performance 👍