this post was submitted on 21 Nov 2023
1 points (100.0% liked)

LocalLLaMA

1 readers
1 users here now

Community to discuss about Llama, the family of large language models created by Meta AI.

founded 10 months ago
MODERATORS
communick@poweruser.forum
1
Discrepancy between TheBloke_Orca-2-13B-GPTQ and the original one with the tested logic question (alien.top)
submitted 10 months ago by Longjumping-Bake-557@alien.top to c/localllama@poweruser.forum
2 comments fedilink hide all child comments
 
top 2 comments
sorted by: hot top controversial new old
[โ€“] panchovix@alien.top 1 points 10 months ago

Try at FP16 or 8bit at most, probably a 13B models suffers too much at 4bits.

[โ€“] dbinokc@alien.top 1 points 10 months ago

I was not impressed with Orca2 13B GPTQ for java and javascript coding type questions. It almost seemed reluctant to give answers. I had to explicitly prompt to print out the code and even then it did not do a very good job.