LLM Benchmarks

llm.l.gaz.codes
← Back to Models

lfm2-1.2b-mlx

lfm2-1.2b-mlx

Publisher
lmstudio-community
Architecture
lfm2
Quantization
8bit
Format
mlx
Max Context
128k tokens
Type
llm
Machine
gaz-studio

Benchmark Results

Compare with other models →

knowledge

BenchmarkScore 
MMLUlocal
78.5%

reasoning

BenchmarkScore 
GSM8Klocal
85.2%
ARC-Challengelocal
75.4%

coding

BenchmarkScore 
HumanEvallocal
72.0%

language

BenchmarkScore 
HellaSwaglocal
82.3%

safety

BenchmarkScore 
TruthfulQAlocal
65.8%
LLM Benchmarks