Nomic Embed Text v1.5
PermissiveNomic AI · Nomic Embed · Released 2024-02
137MEmbedding
A fully open embedding model, including training data and code, with Matryoshka representation learning for variable-size, resizable embeddings.
Strengths
- +Apache 2.0 license and fully open training data/code
- +Matryoshka embeddings let you truncate dimensions to trade size for speed
- +8192-token context window
Limitations
- -Below the largest 7B-class embedders on some MTEB tasks
- -Truncated (low-dimension) Matryoshka variants lose some retrieval accuracy
License
Apache 2.0
use commercially with attribution niceties
Hardware
runs on a good consumer GPU (or Apple Silicon) with quantization
Links
Stats
nomic-ai/nomic-embed-text-v1.5— downloads·— likes
via Hugging Face