Resources

Not longer compatible with the newest llama.cpp

👍 5

#22 opened almost 2 years ago by

bero1985

Update README.md

#21 opened over 2 years ago by

SalmanFaroz

Update README to fix install of huggingfacecli command

#20 opened over 2 years ago by

JohanDL

Hardware Requirements for Q4_K_M

#19 opened over 2 years ago by

ShivanshMathur007

function calling

#18 opened over 2 years ago by

ybsid

Can not load by ctransformers

#17 opened over 2 years ago by

hattran

I love the mixtral-8x7b-instruct-v0.1.Q5_K_M.gguf

#16 opened over 2 years ago by

johanteekens

My cpu is only using 50% of its cores.

#15 opened almost 3 years ago by

jeffwadsworth

Would this run on a 32GM RAM & 8GB VRAM?

#14 opened almost 3 years ago by

Troyanovsky

Weird. Ooga is still not loading after a fresh pull from release.

#13 opened almost 3 years ago by

moona99

Anyone else seeing similar behavior? I especially like the start "Death, ..." plus some gobblygook.

#12 opened almost 3 years ago by

BigDeeper

Could this model deploy with fastchat? :)

#10 opened almost 3 years ago by

FlameHunter

How many tokens per second?

#9 opened almost 3 years ago by

Hoioi

KCPP frankenstein experimental release for Mixtral

❤️ 1

#8 opened almost 3 years ago by

Nexesenex

Not finding blk.0.ffn_gate.weight. I checked sha256sum, matches the Q6_K version. Any thoughts on how to fix this?

#6 opened almost 3 years ago by

BigDeeper

For the time being that mode with unofficial llamacpp works terrible - bad bad in answering - Instruct version is the best all of llm ever so far.

#5 opened almost 3 years ago by

mirek190

create_tensor: tensor 'blk.0.ffn_gate.weight' not found

#4 opened almost 3 years ago by

Althenwolf

It works.

👍🤗 2

#3 opened almost 3 years ago by

Yuuru

mixtral instruct too?

#2 opened almost 3 years ago by

nbilla

Other quant types.

👍 2

#1 opened almost 3 years ago by

dog3-l0ver