Native MLX FP16 conversion of Laya's multilingual checkpoint (mmBERT) for Apple silicon, with a 1,024-token context. The base for multilingual fine-tunes in Laya Studio.
A ModernBERT-large encoder with an option-marker head. Premise and options are packed into one sequence and each option's marker is scored in a single bidirectional pass; version 1.2 makes the scoring order-invariant.
Decides
choice, score, noul
choice, score, noul, classify
Architecture
laya
von
Fine-tuned from
convaiinnovations/laya-multilingual
answerdotai/modernbert-large
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
—
63.9%
Calibration error
—
0.045
Valid action rate
—
—
Median latency
—
18 ms
p95 latency
—
—
Evaluation suite
—
Figures are from each model’s manifest; accuracy and latency are what the publishers report, on their own suites and hardware. Add a third model.
Questions
What is the difference between laya-multilingual-mlx and von?
laya-multilingual-mlx is from aac6fef and von from wfzyx. Both have open weights you can download and run. Both answer choice, score and noul questions. Only von answers classify. von reads up to 8K tokens of state, against 1K tokens for laya-multilingual-mlx. laya-multilingual-mlx is the smaller model, at 322M parameters to 395M.
Which is more accurate, laya-multilingual-mlx or von?
Only von publishes an accuracy figure (63.9% on JevBench public standard tier); laya-multilingual-mlx does not, so there is no comparison to make without your own test.
Which is cheaper, laya-multilingual-mlx or von?
laya-multilingual-mlx: Free (open weights). von: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run laya-multilingual-mlx or von locally?
Yes, both: systemone pull aac6fef/laya-multilingual-mlx and systemone pull wfzyx/von download the weights.
