Back to Tags
Efficient Inference
3 items tagged with "efficient-inference"
Filter by type:
Models3
Model
Laplace 3B
Laplace 3B is a 3-billion-parameter linear-attention model trained and served on AWS Trainium2. Its recurrent state remains fixed at 29.5 MB per sequence regardless of context length, avoiding the growing KV cache used by transformer models.
Model
LFM 2.5 2.6B
Liquid AI's LFM 2.5 2.6B is a compact long-context foundation model listed on OpenRouter with a 128k-token context window.
Model
NVIDIA Nemotron 3.5 Lightning
NVIDIA's Nemotron 3.5 Lightning expands the Nemotron 3 family as an efficient open model for long-running agentic AI workloads.