Luke Merrick
commited on
Commit
•
a0165f3
1
Parent(s):
d3c84a2
README phrasing
Browse files
README.md
CHANGED
@@ -7610,7 +7610,7 @@ model-index:
|
|
7610 |
|
7611 |
## News
|
7612 |
|
7613 |
-
07/18/2024:
|
7614 |
|
7615 |
05/10/2024: Release of the [technical report on Arctic Embed](https://arxiv.org/abs/2405.05374)
|
7616 |
|
@@ -7619,7 +7619,7 @@ model-index:
|
|
7619 |
|
7620 |
## This Model
|
7621 |
|
7622 |
-
This model is an
|
7623 |
|
7624 |
| Model Name | MTEB Retrieval Score (NDCG @ 10) |
|
7625 |
|:------------------------------------------------------------------------------------------------|:---------------------------------|
|
|
|
7610 |
|
7611 |
## News
|
7612 |
|
7613 |
+
07/18/2024: Release of `snowflake-arctic-embed-m-v1.5`, capable of producing highly compressible embedding vectors that preserve quality even when squished as small as 128 bytes per vector.
|
7614 |
|
7615 |
05/10/2024: Release of the [technical report on Arctic Embed](https://arxiv.org/abs/2405.05374)
|
7616 |
|
|
|
7619 |
|
7620 |
## This Model
|
7621 |
|
7622 |
+
This model is an updated version of the original [snowflake-arctic-embed-m](https://huggingface.co/Snowflake/snowflake-arctic-embed-m/) designed to improve embedding vector compressibility. This model achieves a slightly higher performance overall without compression, and it is additionally capable of retaining most of its retrieval quality even down to 128 byte embedding vectors through a combination of [Matryoshka Representation Learning (MRL)](https://arxiv.org/abs/2205.13147) and uniform scalar quanitization.
|
7623 |
|
7624 |
| Model Name | MTEB Retrieval Score (NDCG @ 10) |
|
7625 |
|:------------------------------------------------------------------------------------------------|:---------------------------------|
|