Ternlight – 7 MB embedding model that runs in browser (WASM)

soycaporal1 pts1 comments

ternlight · semantic search · React docs

Loading the engine…

0 / 0

CHUNKS

— emb/sec

ON YOUR CPU

0.0 s

ELAPSED

ASK<br>share state across components<br>skip a re-render<br>run code after render<br>avoid prop drilling<br>form input value

EMBED LATENCY

—ms

single embed() call (@ternlight/mini)

Copy query embedding

MODEL

Embedding —<br>Linear weights ternary (BitLinear)<br>Transformer layers—<br>Attention heads —<br>d_model —<br>Output —<br>Max input —

PACKAGE

Runs on @ternlight/mini ↗<br>Bundle —<br>Engine init —<br>Throughput —<br>License MIT<br>After first loadcached (0 network)

ternlight embedding model runs engine render

Related Articles