Compare all-mpnet-base-v2 With BGE and CLIP Embeddings
all-mpnet-base-v2 is the dependable default in our text embedding API comparison, versus BGE-Large's retrieval quality and BGE-M3's multilingual modes.
replicate/all-mpnet-base-v2 📝 → ❓
Performance
1.1sTypical run time
~29sCold start (first call)
2.4MTotal runs
About
This is a language model that can be used to obtain document embeddings suitable for downstream tasks like semantic search and clustering.
Example Output
Output
[object Object]
Performance Metrics
1.08s
Prediction Time
28.56s
Total Time
Input Parameters
- text
- A single string to encode.
- text_batch
- A JSON-formatted list of strings to encode.
Output Schema
Output
Version Details
- Version ID
b6b7585c9640cd7a9572c6e129c9549d79c9c31f0d3fdce7baac7c67ca38f305- Version Created
- May 30, 2023