Should we download it ?

#1
by kalota - opened

Should we download this model now, or are you going to fix 10–20 basic mistakes over the next few days and make us reinstall it every time? Or will you “magically” improve the model by increasing the quant sizes, forcing us to drop to a lower quant: with no real benchmarks, or benchmarks that are missing data points? Because I pay for a limited amount of data transfer. Thanks!

Should we download this model now, or are you going to fix 10–20 basic mistakes over the next few days and make us reinstall it every time? Or will you “magically” improve the model by increasing the quant sizes, forcing us to drop to a lower quant: with no real benchmarks, or benchmarks that are missing data points? Because I pay for a limited amount of data transfer. Thanks!

Get a ticket to Arctic, you need to chill.

It’s just a complaint about a product/service. It’s normal.
🐧

This doesn't seem to work:

image

danielhanchen changed discussion status to closed

This doesn't seem to work:

image

I get the same problem when I use --cpu-moe or --n-cpu-moe. just offloading all the layers doesn't give me that issue.

Unsloth AI org

Closed because OP has a NSFW pfp.

Unsloth AI org

Yes the GGUFs are good to go and download now.

Closed because OP has a NSFW pfp.

LMAO, cherry on top.

This comment has been hidden (marked as Graphic Content)

Sign up or log in to comment