80 Downloads
Description
Hybrid architecture model intended for local use, by Liquid AI
Use cases
Minimum system memory
220MB
Tags
350M
lfm2
Last update
Updated on July 16byREADME
Liquid LFM2 350M
Liquid LFM2 models are blazingly fast models that are designed for local use. Compared to similarly sized models, this model excels at mathematics, instruction following, and multilingual understanding.
LFM2 uses a hybrid architecture that performs efficiently on both CPU and GPU.
Supports a context length of 32k.
Parameters
Custom configuration options included with this model
Min P Sampling
0.15
Repeat Penalty
1.05
Temperature
0.3
Sources
The underlying model files this model uses
Based on