Group :: Sciences/Computer science
RPM: llama.cpp
Main Changelog Spec Patches Sources Download Gear Bugs and FR Repocop
Current version: 20230728-alt1
Build date: 29 july 2023, 22:04 ( 38.6 weeks ago )
Size: 1205.07 Kb
Home page: https://github.com/ggerganov/llama.cpp
License: MIT
Summary: Inference of LLaMA model in pure C/C++
Description:
List of contributors List of rpms provided by this srpm:
ACL:
Build date: 29 july 2023, 22:04 ( 38.6 weeks ago )
Size: 1205.07 Kb
Home page: https://github.com/ggerganov/llama.cpp
License: MIT
Summary: Inference of LLaMA model in pure C/C++
Description:
Plain C/C++ implementation (of inference of LLaMA model) without
dependencies. AVX, AVX2 and AVX512 support for x86 architectures.
Mixed F16/F32 precision. 4-bit, 5-bit and 8-bit integer quantization
support. Runs on the CPU. Supported models:
LLaMA
LLaMA 2
Alpaca
GPT4All
Chinese LLaMA / Alpaca
Vigogne (French)
Vicuna
Koala
OpenBuddy (Multilingual)
Pygmalion 7B / Metharme 7B
WizardLM
Baichuan-7B and its derivations (such as baichuan-7b-sft)
NOTE 1: You will need to:
pip3 install -r /usr/share/llama.cpp/requirements.txt
for data format conversion scripts to work.
NOTE 2:
MODELS ARE NOT PROVIDED. You need to download them from original
sites and place them into "./models" directory.
For example, LLaMA downloaded via public torrent link is 220 GB.
Overall this is all raw and experimental, no warranty, no support.
Current maintainer: Vitaly Chikunov dependencies. AVX, AVX2 and AVX512 support for x86 architectures.
Mixed F16/F32 precision. 4-bit, 5-bit and 8-bit integer quantization
support. Runs on the CPU. Supported models:
LLaMA
LLaMA 2
Alpaca
GPT4All
Chinese LLaMA / Alpaca
Vigogne (French)
Vicuna
Koala
OpenBuddy (Multilingual)
Pygmalion 7B / Metharme 7B
WizardLM
Baichuan-7B and its derivations (such as baichuan-7b-sft)
NOTE 1: You will need to:
pip3 install -r /usr/share/llama.cpp/requirements.txt
for data format conversion scripts to work.
NOTE 2:
MODELS ARE NOT PROVIDED. You need to download them from original
sites and place them into "./models" directory.
For example, LLaMA downloaded via public torrent link is 220 GB.
Overall this is all raw and experimental, no warranty, no support.
List of contributors List of rpms provided by this srpm: