metadata
license: apache-2.0
embeddings
embeddings is a repo for quantized bert models in ggml format.
usage
You can utilize bert.cpp as usual or use our new api to quickly test it out.
api sample output
{
"time": {
"timestamp": 1708287985.356056,
"prediction_time": 0.07016229629516602
},
"texts": {
"txt1": "A person ",
"txt2": "A cat"
},
"similarity": {
"semantic": 0.035270897102758283,
"literal": 0.22219999999999998
}
}