Model
LFM 1.2B
An ultra-compact language model of approximately 0.70 GB with around 1.2 billion parameters, capable of running on local devices with minimal resource consumption, suitable for edge inference scenarios.
Timeline (last 90 days)
Oct 12
The candidate lineup ranges from the 0.70GB LFM 1.2B to the 24.57GB Nemotron 3 Nano 30B-A3B.
Unverified50%
Oct 12
The smallest model, LFM 1.2B, has a median generation rate of approximately 208 tokens/s, with a current task score of only 5/16 substantive and 2/16 strict.
Unverified50%