v1.1.2

Added support for batch embedding of multiple text strings if using the all-Mini-L6-v2 model. As it is so small it barely uses the GPU so can batch process embedding to maximize embedding performance. Runs much faster now.

v1.1.1

Added support for 2 Embedding models:
Google's EmbeddingGemma (highest quality and works with larger texts but 300Mb - 1024 token input, 768 dimensional output)
all-Mini-L6-v2 (fastest and only 30mb - 128 token input, 384 dimensional output)

v1.0.1

Fixes bug where user attempts to query vector DB with nothing stored.

v1.0.0

Initial commit and release.