diff --git a/.gitattributes b/.gitattributes
index 1f65aa8130ce5180ad8b39b896ef29309512178c..90c1938abbe8be5ee71fd6d9e64097cf2dd06e6a 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -1,18 +1,3 @@
-embeddings filter=lfs diff=lfs merge=lfs -text
-model/tokenizer.json filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.128.fp32.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.512.fp16.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/tokenizer.json.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.1024.fp16.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.256.fp16.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.384.fp16.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.512.fp32.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.1024.fp32.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.128.fp16.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.128.int8.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.384.fp32.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.1024.int8.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.256.fp32.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.256.int8.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.384.int8.npy.zst filter=lfs diff=lfs merge=lfs -text
-model/static-embeddings.512.int8.npy.zst filter=lfs diff=lfs merge=lfs -text
+models/**/*.npy filter=lfs diff=lfs merge=lfs -text
+models/**/*.zst filter=lfs diff=lfs merge=lfs -text
+models/**/*.json filter=lfs diff=lfs merge=lfs -text
diff --git a/README.md b/README.md
index 351046e1683a919631f36a606d6e69905f6047ed..e8178e5f02df660feda151af058abf76edd50693 100644
--- a/README.md
+++ b/README.md
@@ -1,66 +1,22 @@
----
-language:
-- en
-- ar
-- bg
-- ca
-- cs
-- da
-- de
-- el
-- es
-- et
-- fa
-- fi
-- fr
-- gl
-- gu
-- he
-- hi
-- hu
-- hy
-- id
-- it
-- ja
-- ka
-- ko
-- ku
-- lt
-- lv
-- mk
-- mn
-- mr
-- ms
-- my
-- nb
-- nl
-- pl
-- pt
-- ro
-- ru
-- sk
-- sl
-- sq
-- sr
-- sv
-- th
-- tr
-- uk
-- ur
-- vi
-- zh
-- hr
-license: apache-2.0
----
# Static Embeddings
This project contains multilingual static embeddings that are appropriate for generating
quick embeddings in edge devices. They are re-packaged from other projects in production
ready assets.
-## Current model
+## Models
-The current model is using [sentence-transformers/static-similarity-mrl-multilingual-v1](https://huggingface.co/sentence-transformers/static-similarity-mrl-multilingual-v1)
-as it is multi-lingual and uses Matryoshka Loss, which allows for the arbitrary truncation
-of the embedding vectors. The length of the vector can be tuned for its use, and still
-retain the semantic meaning.
+* [minishlab/potion-retrieval-32M/](models/minishlab/potion-retrieval-32M/README.md)
+* [minishlab/potion-multilingual-128M/](models/minishlab/potion-multilingual-128M/README.md)
+* [sentence-transformers/static-retrieval-mrl-en-v1/](models/sentence-transformers/static-retrieval-mrl-en-v1/README.md)
+* [sentence-transformers/static-similarity-mrl-multilingual-v1](models/sentence-transformers/static-similarity-mrl-multilingual-v1/README.md)
+
+## Updating
+
+Add models to `scripts/build_models.py`.
+
+```sh
+pipx install huggingface_hub
+huggingface-cli login
+uv run scripts/build_models.py
+```
diff --git a/models/minishlab/potion-multilingual-128M/README.md b/models/minishlab/potion-multilingual-128M/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..4b4c468b957e08fd592e218bc6bc407f36097b2c
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/README.md
@@ -0,0 +1,128 @@
+# [minishlab/potion-multilingual-128M](https://huggingface.co/minishlab/potion-multilingual-128M)
+
+License: [mit](https://choosealicense.com/licenses/mit/)
+
+A multilingual embedder. The details are a bit scant on how it's trained as
+there is no source code for it. However, it's likely a close architecture
+to the potion-retrieval-32M model, but trained on Common Crawl data.
+
+The 128M references the number of parameters in the embeddings:
+
+256 dimensions * 500,353 vocab.
+
+## Model Stats
+
+Stats that describe the embeddings tensor shapes and value distribution.
+
+| item | metric | value |
+| --------------| ----------------------- | ----- |
+| vocab | size | 500,353 |
+| embedding | dimensions | 256 |
+| vector length | mean | 12.73 |
+| vector length | median | 11.94 |
+| vector length | stddev | 5.12 |
+| values | mean | -0.00 |
+| values | median | -0.00 |
+| values | stddev | 0.86 |
+
+## Mean Pooled Quantization Loss
+
+This test roundtrips the vectors through quantization, but performs the
+mean pooling arithmetic in float32 space. The quantized and unquantized
+mean pooled vectors are compared to each other to determine their cosine
+similarity, to show how much the meaning of the vector has changed due
+to quantization.
+
+| Precision | Cosine Similarity |
+| ------------- | ----------------- |
+| fp16 | 1.00000 |
+| fp8 e4m3 | 0.99993 |
+| fp8 e5m2 | 0.99973 |
+
+## Quantization Loss Per Vector
+
+While ultimately the embedding vectors will be mean pooled together, it's
+still useful to look at the loss per-vector in the embedding table to see
+which quantization strategies retain the most vector meaning.
+
+- **Cosine Similarity** — measures how well the *direction* of embedding vectors
+is preserved after quantization, independent of scale. This is especially
+relevant when embeddings are used for similarity search or retrieval.
+- **MSE (Mean Squared Error)** — emphasizes large errors by squaring the
+differences. Useful for detecting whether any values are badly distorted.
+- **MAE (Mean Absolute Error)** — the average absolute difference between
+original and quantized values. Easier to interpret, less sensitive to outliers.
+
+| Precision | Metric | Value |
+| ------------- | ------ | ----- |
+| fp16 | cosine similarity | 1.00000 |
+| fp8 e4m3 | cosine similarity | 0.99965 |
+| fp8 e5m2 | cosine similarity | 0.99863 |
+| fp16 | MSE | 0.00000 |
+| fp8 e4m3 | MSE | 0.00052 |
+| fp8 e5m2 | MSE | 0.00205 |
+| fp16 | MAE | 0.00011 |
+| fp8 e4m3 | MAE | 0.01364 |
+| fp8 e5m2 | MAE | 0.02717 |
+
+## Tokenizer Examples
+
+**Input:** This is an example of encoding
+**Tokens**: `▁This` `▁is` `▁an` `▁example` `▁of` `▁encoding`
+
+**Input:** The quick brown fox jumps over the lazy dog.
+**Tokens**: `▁The` `▁quick` `▁brown` `▁fox` `▁jumps` `▁over` `▁the` `▁lazy` `▁dog` `▁` `.`
+
+**Input:** Curaçao, naïve fiancé, jalapeño, déjà vu.
+**Tokens**: `▁Cura` `ça` `o` `▁` `,` `▁na` `ï` `ve` `▁fiancé` `▁` `,` `▁ja` `lap` `eño` `▁` `,` `▁déjà` `▁vu` `▁` `.`
+
+**Input:** Привет, как дела?
+**Tokens**: `▁При` `вет` `▁` `,` `▁как` `▁дела` `▁?`
+
+**Input:** Бързата кафява лисица прескача мързеливото куче.
+**Tokens**: `▁Бър` `за` `та` `▁кафяв` `а` `▁лис` `ица` `▁пре` `ска` `ча` `▁` `мър` `зе` `ливо` `то` `▁куче` `▁` `.`
+
+**Input:** Γρήγορη καφέ αλεπού πηδάει πάνω από τον τεμπέλη σκύλο.
+**Tokens**: `▁Γ` `ρή` `γο` `ρη` `▁καφέ` `▁α` `λε` `πού` `▁` `πη` `δά` `ει` `▁πάνω` `▁από` `▁τον` `▁τε` `μπ` `έλη` `▁σκύλο` `▁` `.`
+
+**Input:** اللغة العربية جميلة وغنية بالتاريخ.
+**Tokens**: `▁اللغة` `▁العربية` `▁جميلة` `▁وغ` `نية` `▁بال` `تاريخ` `▁` `.`
+
+**Input:** مرحبا بالعالم!
+**Tokens**: `▁مرحبا` `▁بالعالم` `▁!`
+
+**Input:** Simplified: 快速的棕色狐狸跳过懒狗。
+**Tokens**: `▁Simp` `l` `ified` `▁:` `▁` `快速` `的` `棕` `色` `狐` `狸` `跳` `过` `懒` `狗` `。`
+
+**Input:** Traditional: 快速的棕色狐狸跳過懶狗。
+**Tokens**: `▁Tradition` `al` `▁:` `▁` `快速` `的` `棕` `色` `狐` `狸` `跳` `過` `懶` `狗` `。`
+
+**Input:** 素早い茶色の狐が怠け者の犬を飛び越える。
+**Tokens**: `▁素` `早い` `茶` `色` `の` `狐` `が` `怠` `け` `者の` `犬` `を` `飛び` `越` `える` `。`
+
+**Input:** コンピュータープログラミング
+**Tokens**: `▁` `コンピュータ` `ー` `プロ` `グラ` `ミ` `ング`
+
+**Input:** 빠른 갈색 여우가 게으른 개를 뛰어넘습니다.
+**Tokens**: `▁빠른` `▁갈` `색` `▁여` `우` `가` `▁게` `으` `른` `▁` `개를` `▁뛰어` `넘` `습니다` `▁` `.`
+
+**Input:** तेज़ भूरी लोमड़ी आलसी कुत्ते के ऊपर कूदती है।
+**Tokens**: `▁तेज़` `▁भू` `री` `▁लो` `म` `ड़ी` `▁आ` `ल` `सी` `▁कुत्ते` `▁के` `▁ऊपर` `▁` `कू` `द` `ती` `▁है` `।`
+
+**Input:** দ্রুত বাদামী শিয়াল অলস কুকুরের উপর দিয়ে লাফ দেয়।
+**Tokens**: `▁দ্রুত` `▁বাদাম` `ী` `▁শি` `য়াল` `▁অ` `ল` `স` `▁কু` `কুর` `ের` `▁উপর` `▁দিয়ে` `▁লা` `ফ` `▁দেয়` `।`
+
+**Input:** வேகமான பழுப்பு நரி சோம்பேறி நாயின் மேல் குதிக்கிறது.
+**Tokens**: `▁வேக` `மான` `▁பழ` `ு` `ப்பு` `▁ந` `ரி` `▁சோ` `ம்` `பே` `றி` `▁நா` `யின்` `▁மேல்` `▁கு` `தி` `க்கிறது` `▁` `.`
+
+**Input:** สุนัขจิ้งจอกสีน้ำตาลกระโดดข้ามสุนัขขี้เกียจ.
+**Tokens**: `▁` `สุนัข` `จิ` `้ง` `จอ` `ก` `สีน้ําตาล` `กระโดด` `ข้าม` `สุนัข` `ขี้` `เกีย` `จ` `▁` `.`
+
+**Input:** ብሩክ ቡናማ ቀበሮ ሰነፍ ውሻን ተዘልሏል።
+**Tokens**: `▁` `ብሩ` `ክ` `▁ቡና` `ማ` `▁` `ቀበ` `ሮ` `▁ሰ` `ነፍ` `▁` `ው` `ሻ` `ን` `▁ተ` `ዘ` `ል` `ሏል` `።`
+
+**Input:** Hello 世界 مرحبا 🌍
+**Tokens**: `▁Hello` `▁世界` `▁مرحبا` `▁🌍`
+
+**Input:** 123, αβγ, абв, العربية, 中文, हिन्दी.
+**Tokens**: `▁123` `▁` `,` `▁α` `β` `γ` `▁` `,` `▁аб` `в` `▁` `,` `▁العربية` `▁` `,` `▁中文` `▁` `,` `▁हिन्दी` `▁` `.`
\ No newline at end of file
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d128.npy b/models/minishlab/potion-multilingual-128M/fp16.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..f46d2769823e3d1e674661e195029412546cbf34
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc887899f84cbcecb758e9a937b7af684599f3c02335c5b16571290282fff374
+size 128090496
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d128.npy.zst b/models/minishlab/potion-multilingual-128M/fp16.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..142103c52e1de4c0c0b83b0f7f645f4464a08eb2
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd2fb1e3464c24178d68712c2d4e54d60725b0fc71e3f01d247e7f5e738368b3
+size 118744364
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d256.npy b/models/minishlab/potion-multilingual-128M/fp16.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..748e47419e69db759e2cb8637852bb94d6b87183
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e48643af3158e696580c6f2f0fca2512359a70823bd3f0b1326149debb7c629
+size 256180864
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d256.npy.zst b/models/minishlab/potion-multilingual-128M/fp16.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..4be3780f8d01b21565822fa80b1e978012c76e92
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5130b79d6c3d133be5abfaa771833cce819baea7c04551a308111734e1e91e7
+size 237834592
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d32.npy b/models/minishlab/potion-multilingual-128M/fp16.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..49cd3fc496a6b6ce0f538278ec6f38499b982290
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ad56f8b6e159fe2702416bbd03a5996bcbb97c1d445090d43d511eae5b2c6ec
+size 32022720
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d32.npy.zst b/models/minishlab/potion-multilingual-128M/fp16.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..65aca300c9ef54029a435c23aeefa1a9d9cbbc00
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaa759ab9044e1ac476a9e7592fed79238fa10e6b6a143cd2ac6fc66fe8472b9
+size 29649845
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d64.npy b/models/minishlab/potion-multilingual-128M/fp16.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..e697b47d273ebeeebabee58fc01eb100b9afe671
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d30d2784a45676324a1cf641a4f57796efc735a30ab9d20539e84cbc8fb8ee80
+size 64045312
diff --git a/models/minishlab/potion-multilingual-128M/fp16.d64.npy.zst b/models/minishlab/potion-multilingual-128M/fp16.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..5bf39fad1147f941a3c40970cd4a760f191866ff
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp16.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:595743a291bb571772a6d6e37e1dda91191e5ebd24e96dfcadde3ee810d917ac
+size 59310057
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d128.npy b/models/minishlab/potion-multilingual-128M/fp32.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..d97544039527ad05d741119a6aac3f4866eb0460
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53f4486ffb084395a22b5042eb01131d30985603a123718a3337639f7b561af5
+size 256180864
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d128.npy.zst b/models/minishlab/potion-multilingual-128M/fp32.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..faabcfbf64673067fc9a7f148d81b3901d02f151
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9cf09adbfe40db684bdd54eaa5a3c0faf39d1885b60b6d6224869459e0b0e5a6
+size 238079206
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d256.npy b/models/minishlab/potion-multilingual-128M/fp32.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..c69fe53e7d565192c2acf621fb2b8717e58f9294
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6387c770f72908c601f9cb98c7dc1334c230acafb46693749fe564810b7d46e8
+size 512361600
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d256.npy.zst b/models/minishlab/potion-multilingual-128M/fp32.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..4d5a9a1cf48201a3dab686177ac6f069c8e7173d
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ba369e4e097bdb169082fd3dbfcf7f6c4b3a96ef7b398f50d917943243323db
+size 476624553
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d32.npy b/models/minishlab/potion-multilingual-128M/fp32.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..6615b9cfd7206ab2950229ff84bd04d88ccf4490
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92d9d6f564521fb0e9bdf3659720367611ee18c667d44662f1d76f5a152046f9
+size 64045312
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d32.npy.zst b/models/minishlab/potion-multilingual-128M/fp32.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..4ac23852ea7c47faa017bac0386e0fc8553645c0
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c542379a851c35da71bac51d102c74679714f0e16f1826eee841f4ade6a0c88
+size 59489696
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d64.npy b/models/minishlab/potion-multilingual-128M/fp32.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..93ae587c851b9d418355f6bb912d07c8ff148a56
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d32bc6682e92c085a05a949bb12124463f81fa8fe13070c2f02eb27cca1c6ad
+size 128090496
diff --git a/models/minishlab/potion-multilingual-128M/fp32.d64.npy.zst b/models/minishlab/potion-multilingual-128M/fp32.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..1f18a06ad7d66063643abd82b1c9a08ebca733a4
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp32.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91613a230a226d52659d4237c945b999dc4c169a4a726c2d2c24710440716caa
+size 118969845
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d128.npy b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..92325f7c0b04ca7932a8d217f63c4037acd52a36
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96f86c14faebe685d9278a0a08d2a1753331b4877f1571ec26733cd82fbc0d5c
+size 64045312
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d128.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..85e80122cd271563a60ec08df8740957ed3c60c6
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c248a4ee3b127b3e3444b3a18d4350d8d263b3b0a60adc7a1d0c6c870a6e4c51
+size 53423849
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d256.npy b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..8de45ed1c9a85fe7e99cf5e030de314804acc285
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edb6d2d54ff0bbc239efe60f94522328cb22dfc2722a6dacf367ff5148a18297
+size 128090496
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d256.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..3c4a97aa35d819e6fcffc082e9ae53f8239fd7d6
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d3a59f9eb2f29aa58975d6392c4267fd7319114840cf4ad817c0bb776672a071
+size 107325555
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d32.npy b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..8444876068bfb6cfce6e8c46b64d858a16423e1d
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:32e38e83ccf6239d1d756d4a0ad6cc5506fd3140f1ea501cc9c73f8f6853027c
+size 16011424
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d32.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..c26acd518b40f1473506cc9f49a9654b1faf31c7
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2adeecb3aa5bc959bab52e5c191f6fa0ac06abef0cc61e10f8ab228c57af6057
+size 13341806
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d64.npy b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..963a4076f8851d0743e3af5252c4018e00fad27f
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa0967f31ca4f78b0d904c734712612e68e88eeb95777ee70be0dff52680cdd2
+size 32022720
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e4m3.d64.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..1e6f6e899286f534b6d0bd1746fd01a527b106be
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e4m3.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97a1b36d41adbbb4ee5b2a5d4d5ba69547764296bba6ed2f849011de7311663c
+size 26689571
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d128.npy b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..a71b31f93172a424c0c09b5c4e2f050fd6006ec8
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5b5ec3646db046b96f8f40ae5fc9b7a4fb9734763920bbe64cddb848609bc37
+size 64045312
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d128.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..7b418a9f12e576be0c4c2e427b408accfdc0d964
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e03b8845a0e901c39d5a967b0e8dc768e89f4d10d71c044077dc0fb9cf57b1ae
+size 45893916
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d256.npy b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..4f17fac7dbba2d3c31edf6ee9a33b386fd5a94c2
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:592264948feedc3655f90340c2dfa6712286380f04c3486172c2c14eb065eacf
+size 128090496
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d256.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..aa76acbcaa2d79951f232f256e31e02a336d6ab1
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b262539251e46cce6c0b10434c22f4e8435cadf317a6db40e94c726b14bfe139
+size 92364240
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d32.npy b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..bf9722726ad3a2d5841649fb76c831c3280ee9f5
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:731d0b29eb4f9209f6a56b1a2d431d6913dc80e44db0b05448ccf79b1f3c9c69
+size 16011424
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d32.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..140c49fb75abe34f136eb1934def887fad87ee8f
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a24917c635581f9b60edf8c757d40804c61fbbdaa07d77c1083d9dacc6c65ff2
+size 11441672
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d64.npy b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..6a9eb512f26051558d583015fcaecafae1959f05
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee50d08a8ac4292e3a6e47ec64da23a053f638520df8085504f2962d57801368
+size 32022720
diff --git a/models/minishlab/potion-multilingual-128M/fp8_e5m2.d64.npy.zst b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..26f84e3ce0977809d41d6f7626d1b8c2d0b37893
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/fp8_e5m2.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2496ef9bac8c4131f43492f7eb8166b2690ac8bb57c6fd9f075c197c1e0768d
+size 22899505
diff --git a/models/minishlab/potion-multilingual-128M/tokenizer.json b/models/minishlab/potion-multilingual-128M/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5f3ddc788abc2124fc39463df937b963c94c360
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:370e450e45668b38f46c2ac780839caab94d260fe9dc448e9872cb110f9c868d
+size 34630288
diff --git a/models/minishlab/potion-multilingual-128M/tokenizer.json.zst b/models/minishlab/potion-multilingual-128M/tokenizer.json.zst
new file mode 100644
index 0000000000000000000000000000000000000000..21d6f4728417bd567ada0a216203a894336ed22a
--- /dev/null
+++ b/models/minishlab/potion-multilingual-128M/tokenizer.json.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e06d472444bdc08af554a4296cf6b29f6c4f37f29eaabbffc2df48b730442b1
+size 5382872
diff --git a/models/minishlab/potion-retrieval-32M/README.md b/models/minishlab/potion-retrieval-32M/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..b5566bd7ba03d74d3239b1258f747948626ab6d5
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/README.md
@@ -0,0 +1,128 @@
+# [minishlab/potion-retrieval-32M](https://huggingface.co/minishlab/potion-retrieval-32M)
+
+License: [mit](https://choosealicense.com/licenses/mit/)
+
+The token embeddings from a monolingual English 32M parameter model that was
+distilled from embeddings that were initialized from the the multi-domain
+[BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5)
+
+The 32M references the number of parameters in the embeddings:
+
+512 dimension * 63,091 vocab.
+
+## Model Stats
+
+Stats that describe the embeddings tensor shapes and value distribution.
+
+| item | metric | value |
+| --------------| ----------------------- | ----- |
+| vocab | size | 63,091 |
+| embedding | dimensions | 512 |
+| vector length | mean | 130.27 |
+| vector length | median | 130.39 |
+| vector length | stddev | 30.43 |
+| values | mean | 0.01 |
+| values | median | 0.01 |
+| values | stddev | 5.91 |
+
+## Mean Pooled Quantization Loss
+
+This test roundtrips the vectors through quantization, but performs the
+mean pooling arithmetic in float32 space. The quantized and unquantized
+mean pooled vectors are compared to each other to determine their cosine
+similarity, to show how much the meaning of the vector has changed due
+to quantization.
+
+| Precision | Cosine Similarity |
+| ------------- | ----------------- |
+| fp16 | 1.00000 |
+| fp8 e4m3 | 0.99970 |
+| fp8 e5m2 | 0.99887 |
+
+## Quantization Loss Per Vector
+
+While ultimately the embedding vectors will be mean pooled together, it's
+still useful to look at the loss per-vector in the embedding table to see
+which quantization strategies retain the most vector meaning.
+
+- **Cosine Similarity** — measures how well the *direction* of embedding vectors
+is preserved after quantization, independent of scale. This is especially
+relevant when embeddings are used for similarity search or retrieval.
+- **MSE (Mean Squared Error)** — emphasizes large errors by squaring the
+differences. Useful for detecting whether any values are badly distorted.
+- **MAE (Mean Absolute Error)** — the average absolute difference between
+original and quantized values. Easier to interpret, less sensitive to outliers.
+
+| Precision | Metric | Value |
+| ------------- | ------ | ----- |
+| fp16 | cosine similarity | 1.00000 |
+| fp8 e4m3 | cosine similarity | 0.99965 |
+| fp8 e5m2 | cosine similarity | 0.99862 |
+| fp16 | MSE | 0.00000 |
+| fp8 e4m3 | MSE | 0.02454 |
+| fp8 e5m2 | MSE | 0.09720 |
+| fp16 | MAE | 0.00076 |
+| fp8 e4m3 | MAE | 0.09763 |
+| fp8 e5m2 | MAE | 0.19461 |
+
+## Tokenizer Examples
+
+**Input:** This is an example of encoding
+**Tokens**: `[CLS]` `this` `is` `an` `example` `of` `encoding` `[SEP]`
+
+**Input:** The quick brown fox jumps over the lazy dog.
+**Tokens**: `[CLS]` `the` `quick` `brown` `fox` `jumps` `over` `the` `lazy` `dog` `.` `[SEP]`
+
+**Input:** Curaçao, naïve fiancé, jalapeño, déjà vu.
+**Tokens**: `[CLS]` `curacao` `,` `naive` `fiance` `,` `jalapeno` `,` `deja` `vu` `.` `[SEP]`
+
+**Input:** Привет, как дела?
+**Tokens**: `[CLS]` `п` `##р` `##и` `##в` `##е` `##т` `,` `как` `д` `##е` `##л` `##а` `?` `[SEP]`
+
+**Input:** Бързата кафява лисица прескача мързеливото куче.
+**Tokens**: `[CLS]` `б` `##ъ` `##р` `##з` `##а` `##т` `##а` `к` `##а` `##ф` `##я` `##в` `##а` `л` `##и` `##с` `##и` `##ц` `##а` `п` `##р` `##е` `##с` `##ка` `##ч` `##а` `м` `##ъ` `##р` `##з` `##е` `##л` `##и` `##в` `##о` `##т` `##о` `к` `##у` `##ч` `##е` `.` `[SEP]`
+
+**Input:** Γρήγορη καφέ αλεπού πηδάει πάνω από τον τεμπέλη σκύλο.
+**Tokens**: `[CLS]` `γ` `##ρ` `##η` `##γ` `##ο` `##ρ` `##η` `κ` `##α` `##φ` `##ε` `α` `##λ` `##ε` `##π` `##ου` `π` `##η` `##δ` `##α` `##ε` `##ι` `π` `##α` `##ν` `##ω` `α` `##π` `##ο` `τ` `##ο` `##ν` `τ` `##ε` `##μ` `##π` `##ε` `##λ` `##η` `σ` `##κ` `##υ` `##λ` `##ο` `.` `[SEP]`
+
+**Input:** اللغة العربية جميلة وغنية بالتاريخ.
+**Tokens**: `[CLS]` `ا` `##ل` `##ل` `##غ` `##ة` `ا` `##ل` `##ع` `##ر` `##ب` `##ي` `##ة` `ج` `##م` `##ي` `##ل` `##ة` `و` `##غ` `##ن` `##ي` `##ة` `با` `##ل` `##ت` `##ا` `##ر` `##ي` `##خ` `.` `[SEP]`
+
+**Input:** مرحبا بالعالم!
+**Tokens**: `[CLS]` `م` `##ر` `##ح` `##ب` `##ا` `با` `##ل` `##ع` `##ا` `##ل` `##م` `!` `[SEP]`
+
+**Input:** Simplified: 快速的棕色狐狸跳过懒狗。
+**Tokens**: `[CLS]` `simplified` `:` `[UNK]` `[UNK]` `的` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `。` `[SEP]`
+
+**Input:** Traditional: 快速的棕色狐狸跳過懶狗。
+**Tokens**: `[CLS]` `traditional` `:` `[UNK]` `[UNK]` `的` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `。` `[SEP]`
+
+**Input:** 素早い茶色の狐が怠け者の犬を飛び越える。
+**Tokens**: `[CLS]` `[UNK]` `[UNK]` `い` `[UNK]` `[UNK]` `の` `[UNK]` `か` `[UNK]` `け` `[UNK]` `の` `犬` `を` `[UNK]` `ひ` `[UNK]` `え` `##る` `。` `[SEP]`
+
+**Input:** コンピュータープログラミング
+**Tokens**: `[CLS]` `コ` `##ン` `##ヒ` `##ュ` `##ー` `##タ` `##ー` `##フ` `##ロ` `##ク` `##ラ` `##ミ` `##ン` `##ク` `[SEP]`
+
+**Input:** 빠른 갈색 여우가 게으른 개를 뛰어넘습니다.
+**Tokens**: `[CLS]` `[UNK]` `ᄀ` `##ᅡ` `##ᆯ` `##ᄉ` `##ᅢ` `##ᆨ` `ᄋ` `##ᅧ` `##ᄋ` `##ᅮ` `##ᄀ` `##ᅡ` `ᄀ` `##ᅦ` `##ᄋ` `##ᅳ` `##ᄅ` `##ᅳ` `##ᆫ` `ᄀ` `##ᅢ` `##ᄅ` `##ᅳ` `##ᆯ` `[UNK]` `.` `[SEP]`
+
+**Input:** तेज़ भूरी लोमड़ी आलसी कुत्ते के ऊपर कूदती है।
+**Tokens**: `[CLS]` `त` `##ज` `भ` `##र` `##ी` `ल` `##ो` `##म` `##ड` `##ी` `आ` `##ल` `##स` `##ी` `क` `##त` `##त` `क` `[UNK]` `क` `##द` `##त` `##ी` `ह` `।` `[SEP]`
+
+**Input:** দ্রুত বাদামী শিয়াল অলস কুকুরের উপর দিয়ে লাফ দেয়।
+**Tokens**: `[CLS]` `দ` `##র` `##ত` `ব` `##া` `##দ` `##া` `##ম` `##ী` `শ` `##ি` `##য` `##া` `##ল` `অ` `##ল` `##স` `ক` `##ক` `##র` `##ে` `##র` `উ` `##প` `##র` `দ` `##ি` `##য` `##ে` `[UNK]` `দ` `##ে` `##য` `।` `[SEP]`
+
+**Input:** வேகமான பழுப்பு நரி சோம்பேறி நாயின் மேல் குதிக்கிறது.
+**Tokens**: `[CLS]` `வ` `##ே` `##க` `##ம` `##ா` `##ன` `[UNK]` `ந` `##ர` `##ி` `[UNK]` `ந` `##ா` `##ய` `##ி` `##ன` `ம` `##ே` `##ல` `[UNK]` `.` `[SEP]`
+
+**Input:** สุนัขจิ้งจอกสีน้ำตาลกระโดดข้ามสุนัขขี้เกียจ.
+**Tokens**: `[CLS]` `[UNK]` `.` `[SEP]`
+
+**Input:** ብሩክ ቡናማ ቀበሮ ሰነፍ ውሻን ተዘልሏል።
+**Tokens**: `[CLS]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[SEP]`
+
+**Input:** Hello 世界 مرحبا 🌍
+**Tokens**: `[CLS]` `hello` `世` `[UNK]` `م` `##ر` `##ح` `##ب` `##ا` `[UNK]` `[SEP]`
+
+**Input:** 123, αβγ, абв, العربية, 中文, हिन्दी.
+**Tokens**: `[CLS]` `123` `,` `α` `##β` `##γ` `,` `а` `##б` `##в` `,` `ا` `##ل` `##ع` `##ر` `##ب` `##ي` `##ة` `,` `中` `文` `,` `ह` `##ि` `##न` `##द` `##ी` `.` `[SEP]`
\ No newline at end of file
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d128.npy b/models/minishlab/potion-retrieval-32M/fp16.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..fcda7bbf8bc7ff9a53fad6192e8d5e4461af4320
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d39bf6868e526c83f2e77e474bea989fa3cec4e8eb0bd72efa3208efdc28ee4b
+size 16151424
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d128.npy.zst b/models/minishlab/potion-retrieval-32M/fp16.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..64fc87894d7a02a5fbb8caa81e4ef4777e58ac02
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfcf9586043e994e52667dc360b2979d1400f996c8055e79be402eddf98c3766
+size 14900440
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d256.npy b/models/minishlab/potion-retrieval-32M/fp16.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..7d137d4b84e398f4754a3ec669bf586f6d04b2a0
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d599424bfb205652a334abbcb84fdcf59ce5c9c13847e617fafe4b1ec7c71fe3
+size 32302720
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d256.npy.zst b/models/minishlab/potion-retrieval-32M/fp16.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..313b96a8a8c610580c525ca134091d755e5fdc5a
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e194ae0eeaeaef1f13f913d10d8c7d04b35eca44ce2e0937b6d4cac83b27a048
+size 29836775
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d32.npy b/models/minishlab/potion-retrieval-32M/fp16.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..3864379d9e16ec94f0a71ebb2cbe126be83818b4
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c33b943f6deee8af7f5b49d96fc7dd5b31573149f6d0652b500178aef779310
+size 4037952
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d32.npy.zst b/models/minishlab/potion-retrieval-32M/fp16.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..f48c19445e2f2ec93aac03a1c482011a4615a8e7
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5318a29adc2011a4708db66e1e0e2ddf5b196b7b9b8ee3e40ff4e2cad0c79c44
+size 3724961
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d512.npy b/models/minishlab/potion-retrieval-32M/fp16.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..de02c743f40955cd2ad2b9de0371da1ea1da8094
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:307f4a3953c913da8ebee260bfc1c3eb509ebcac2e6d6ad5001e209aed0de676
+size 64605312
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d512.npy.zst b/models/minishlab/potion-retrieval-32M/fp16.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..1e2896cb9084173e4ec4b32f2ce584d19a334c07
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:660a269222ee454b3908fbf4333dcd379a3d797910d1098528e7918ad9c1f8e7
+size 59832428
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d64.npy b/models/minishlab/potion-retrieval-32M/fp16.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..86b18954e01ed2b3fb3e24056b102cd804388524
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd8d09f3d66df3b228d267eee03868179f52b0ddcb0a291781c65e6768092da1
+size 8075776
diff --git a/models/minishlab/potion-retrieval-32M/fp16.d64.npy.zst b/models/minishlab/potion-retrieval-32M/fp16.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..fe8f4be94fdbd8443b5538c112df16d4c8afc64c
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp16.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:751256692ba27fdb3a181df3af218f9bf136a97559a6e38e9fb0f9a542254ea3
+size 7449228
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d128.npy b/models/minishlab/potion-retrieval-32M/fp32.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..715d1fc7dc70b3d23ce70b309d7fe64fc4a5ca34
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8f10c71cf475e65ce14ddd2cf423f2a45804b64d71243521f0e7a5189c350e0
+size 32302720
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d128.npy.zst b/models/minishlab/potion-retrieval-32M/fp32.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..7f918c31407425fbe37cc8a56ee9108629b3a70b
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5fe0561f1e6c9fbb188a3a4cef3a2dbd754d6e0799dbe2306670d78fb32c68b
+size 29981217
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d256.npy b/models/minishlab/potion-retrieval-32M/fp32.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..0b469b266fff25d6b8a3a01320273ed7e17c3d1c
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ddbdb03591e5b5c972d8928930a32887a98a39a153ff1b324efd79303b20605
+size 64605312
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d256.npy.zst b/models/minishlab/potion-retrieval-32M/fp32.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..02323d0e65e907269101d6ce59ca1c437a10c6bb
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec2187e3591002b4415ad9905999b7a801838154b9ac68b0624b2a0dcebf23ab
+size 59971356
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d32.npy b/models/minishlab/potion-retrieval-32M/fp32.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..d3f5cb664183cdc9b3bf4986e173102c617b27a9
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:245f1c8bd91da0f3ebcd5a30d3c2ffc59910bc9007961d14fb34aa89fe8aa988
+size 8075776
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d32.npy.zst b/models/minishlab/potion-retrieval-32M/fp32.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..41bb3add29b7bf1c8f8e641cdc034c4810452a3e
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10da79de587a527f94b8cccc64b06cbf72ab6831abb704d60711a267aa212551
+size 7487459
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d512.npy b/models/minishlab/potion-retrieval-32M/fp32.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..b79e3a4dc6270494dbcec49418d93bf3e449b672
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0b6ab4be9954ca35d6fc1da5c9708923ef9623a2622f06cdf0f4ff5d83c1a71
+size 129210496
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d512.npy.zst b/models/minishlab/potion-retrieval-32M/fp32.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..518d8ab1e2b942be6dd988117e3a47a50cebaca8
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13ad7f04e52a1a63635861225c683e5829f12ad163f295dab694ecabcb03431d
+size 120018211
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d64.npy b/models/minishlab/potion-retrieval-32M/fp32.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..a88719d8a3f2d1ae8948b0c4997ee6e10b642d26
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f389e15e3d861d1bb547075cb95a996d397d195afec37f0335da4b122abe6d48
+size 16151424
diff --git a/models/minishlab/potion-retrieval-32M/fp32.d64.npy.zst b/models/minishlab/potion-retrieval-32M/fp32.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..a24c5362ef9491863479258d1a0ae03854c297b8
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp32.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9354443217c381cfa10db8d5944aec887862e3cb295ffd14c91c022202f623e
+size 14980930
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d128.npy b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..c3f3e53a850e51b5c36cb0f15c5b5cf8b58c4be3
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6221da5e940c8fa8df30e02c9a538b4480afa0ec6f20530ea1d2237be54af346
+size 8075776
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d128.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..41bc74d328b71f5013e3148173c9931c8a9ac479
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07509949b8c60aa3249e718fe9a16715583329141fb269c9c6662fbdf2c7b587
+size 6673455
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d256.npy b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..f2c239c73c01e6412f14c61d605b0fea314a1bd4
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:417f94a96c1fcb9bf5fae88de3c0801a030654ac00e9dc81ca64f445c6163bba
+size 16151424
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d256.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..dcf748a8fd1a96dbe4690fb8458eaaf6fe874dc1
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a462499c4d49ffbb7707845e08fb7d3e46e20590f6f1b45619398b0db40ee774
+size 13383541
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d32.npy b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..3085a8ae936ad99c42fb78a96f47a88c1e95dd88
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83eb4bb75abb84d65100846a53a3075f2153055a10e5ccbe0784d145b1638d2d
+size 2019040
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d32.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..5c04c343682802497a4367ee5a1e80b51fbc89ca
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d00cb04fed02e263d3565d90b6cb9fef2735bff4f2b9951cbc596374fcee830
+size 1665803
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d512.npy b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..3bd9dd40bcf5a4deb3f4b8cdbe935d5670929119
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d6df654b9e3f5729653e3d1fbd26c17ce1b3e97295edc72b4e17e79b44d64af
+size 32302720
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d512.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..9f38ea15920ca60d670ed491f1c2f26b6c03fa5d
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2001d73dbb0a8ddea9032715c5ab5be3fdbe2398241cec1a5baaa5ada8096d7
+size 26994437
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d64.npy b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..4a0490d19f7d2b8c0bedc004e94cd4bec47363e8
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7cf17584250182f72ae27205926d5a1918d53e6f372d8a2b82a9a19e2764f29
+size 4037952
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e4m3.d64.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..06f2d4aa4c9bb5f12709020f2b42db74a49a157f
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e4m3.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:867937c4f785d34bfccc1e1230092721e3b2fc63b89b105ce94f60413f6ed272
+size 3333232
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d128.npy b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..6c1897ae4ec44a0705450de22f68c4ea7d6cfdb1
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5cfe82a7d3860c0dbce6b23ab28a22e09fa34c94b9189a2a0c03942ee23f6b7
+size 8075776
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d128.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..fdd29aa6ab726908ae10d8d364cf209a9c9f2b03
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a82f374ecbaa9c2b2bc710134983bc7368e092904fafd3af16e42261015a1d27
+size 5684073
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d256.npy b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..b5d81d0b91d4b63b9532a635b9f24059706bbae2
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a61a5891b1ec43d79349011b2a7a2871d083cba17f76af8a3107f3b57cb7f210
+size 16151424
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d256.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..1c313b4270c15f9f9dfdba02564561d6f62ffe17
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7eb91969ea06eef614848716f745052563bab683b157eba21c4238848f76dcb
+size 11412744
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d32.npy b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..0b4f86a69dc4d8b4f45289e72c6abae9baf6b601
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:580867db95b3aa0b3a5b522ebd15422a2833861aa2c4bb7265d57e4b79459eaf
+size 2019040
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d32.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..b52a201b344be1cb88ec20bfcec62c4622e3d7c9
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03b8d8fee46cd3fc3da208799e61a4f370957ddd8a94818871ab5ddca0c225a9
+size 1418587
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d512.npy b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..1ef39f5a48d986e055b3c215f4d1d707da86306b
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb4164bde9d610350d1f0d015e9c9b28025174edb913070de147089f53aa2736
+size 32302720
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d512.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..b86e4f883dc7768b776ffebe192a97c10c177c33
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c2bf87917d2fb950f1e42c2747b995a72bc49bdf4b43363f5c913e5d8b0a58d6
+size 23070429
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d64.npy b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..5dd85d5de31717d2cc869f4c00d75285b6e52768
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5145523fcc9b5c977cc7feee0a3180e719f31401ca4729db4e27c82e6f58fc41
+size 4037952
diff --git a/models/minishlab/potion-retrieval-32M/fp8_e5m2.d64.npy.zst b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..1e0374c69f5de958d51ad4e1641ba601faf58241
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/fp8_e5m2.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c156355dc23fa6c4492746967c8f608f6cb463d524bd994ae8b7b1ee6efaef9
+size 2838317
diff --git a/models/minishlab/potion-retrieval-32M/tokenizer.json b/models/minishlab/potion-retrieval-32M/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d6733b9a1081ef1ac460fd0c14248d051695362
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d75cbc54318138807c401b0f0c9721117c628b39de8e8e0edb6cb17e0ee7d18
+size 1493150
diff --git a/models/minishlab/potion-retrieval-32M/tokenizer.json.zst b/models/minishlab/potion-retrieval-32M/tokenizer.json.zst
new file mode 100644
index 0000000000000000000000000000000000000000..a8af65b5b8860a1e2bb3b162345409524bcbe456
--- /dev/null
+++ b/models/minishlab/potion-retrieval-32M/tokenizer.json.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:35d5b8ffb840562c70e6ac2084b4a1dec3acc3c909340e10ccb7313ef8df39cf
+size 478156
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/README.md b/models/sentence-transformers/static-retrieval-mrl-en-v1/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..97673daebc3db3a72bed01bdc958c9435b677324
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/README.md
@@ -0,0 +1,125 @@
+# [sentence-transformers/static-retrieval-mrl-en-v1](https://huggingface.co/sentence-transformers/static-retrieval-mrl-en-v1)
+
+License: [apache-2.0](https://choosealicense.com/licenses/apache-2.0/)
+
+English-only uncased similarity embeddings that were trained with Matroyshka
+loss that allows for more effective truncation of the embedding vectors. It
+was trained on a variety of domains of monolingual datasets. I was designed
+specifically for similarity retrieval.
+
+## Model Stats
+
+Stats that describe the embeddings tensor shapes and value distribution.
+
+| item | metric | value |
+| --------------| ----------------------- | ----- |
+| vocab | size | 30,522 |
+| embedding | dimensions | 1,024 |
+| vector length | mean | 555.04 |
+| vector length | median | 573.92 |
+| vector length | stddev | 219.06 |
+| values | mean | 0.02 |
+| values | median | 0.01 |
+| values | stddev | 18.65 |
+
+## Mean Pooled Quantization Loss
+
+This test roundtrips the vectors through quantization, but performs the
+mean pooling arithmetic in float32 space. The quantized and unquantized
+mean pooled vectors are compared to each other to determine their cosine
+similarity, to show how much the meaning of the vector has changed due
+to quantization.
+
+| Precision | Cosine Similarity |
+| ------------- | ----------------- |
+| fp16 | 1.00000 |
+| fp8 e4m3 | 0.99972 |
+| fp8 e5m2 | 0.99887 |
+
+## Quantization Loss Per Vector
+
+While ultimately the embedding vectors will be mean pooled together, it's
+still useful to look at the loss per-vector in the embedding table to see
+which quantization strategies retain the most vector meaning.
+
+- **Cosine Similarity** — measures how well the *direction* of embedding vectors
+is preserved after quantization, independent of scale. This is especially
+relevant when embeddings are used for similarity search or retrieval.
+- **MSE (Mean Squared Error)** — emphasizes large errors by squaring the
+differences. Useful for detecting whether any values are badly distorted.
+- **MAE (Mean Absolute Error)** — the average absolute difference between
+original and quantized values. Easier to interpret, less sensitive to outliers.
+
+| Precision | Metric | Value |
+| ------------- | ------ | ----- |
+| fp16 | cosine similarity | 1.00000 |
+| fp8 e4m3 | cosine similarity | 0.99965 |
+| fp8 e5m2 | cosine similarity | 0.99861 |
+| fp16 | MSE | 0.00001 |
+| fp8 e4m3 | MSE | 0.24369 |
+| fp8 e5m2 | MSE | 0.96497 |
+| fp16 | MAE | 0.00244 |
+| fp8 e4m3 | MAE | 0.31206 |
+| fp8 e5m2 | MAE | 0.62205 |
+
+## Tokenizer Examples
+
+**Input:** This is an example of encoding
+**Tokens**: `[CLS]` `this` `is` `an` `example` `of` `encoding` `[SEP]`
+
+**Input:** The quick brown fox jumps over the lazy dog.
+**Tokens**: `[CLS]` `the` `quick` `brown` `fox` `jumps` `over` `the` `lazy` `dog` `.` `[SEP]`
+
+**Input:** Curaçao, naïve fiancé, jalapeño, déjà vu.
+**Tokens**: `[CLS]` `cu` `##rac` `##ao` `,` `naive` `fiance` `,` `ja` `##la` `##pen` `##o` `,` `de` `##ja` `vu` `.` `[SEP]`
+
+**Input:** Привет, как дела?
+**Tokens**: `[CLS]` `п` `##р` `##и` `##в` `##е` `##т` `,` `к` `##а` `##к` `д` `##е` `##л` `##а` `?` `[SEP]`
+
+**Input:** Бързата кафява лисица прескача мързеливото куче.
+**Tokens**: `[CLS]` `б` `##ъ` `##р` `##з` `##а` `##т` `##а` `к` `##а` `##ф` `##я` `##в` `##а` `л` `##и` `##с` `##и` `##ц` `##а` `п` `##р` `##е` `##с` `##ка` `##ч` `##а` `м` `##ъ` `##р` `##з` `##е` `##л` `##и` `##в` `##о` `##т` `##о` `к` `##у` `##ч` `##е` `.` `[SEP]`
+
+**Input:** Γρήγορη καφέ αλεπού πηδάει πάνω από τον τεμπέλη σκύλο.
+**Tokens**: `[CLS]` `γ` `##ρ` `##η` `##γ` `##ο` `##ρ` `##η` `κ` `##α` `##φ` `##ε` `α` `##λ` `##ε` `##π` `##ου` `π` `##η` `##δ` `##α` `##ε` `##ι` `π` `##α` `##ν` `##ω` `α` `##π` `##ο` `τ` `##ο` `##ν` `τ` `##ε` `##μ` `##π` `##ε` `##λ` `##η` `σ` `##κ` `##υ` `##λ` `##ο` `.` `[SEP]`
+
+**Input:** اللغة العربية جميلة وغنية بالتاريخ.
+**Tokens**: `[CLS]` `ا` `##ل` `##ل` `##غ` `##ة` `ا` `##ل` `##ع` `##ر` `##ب` `##ي` `##ة` `ج` `##م` `##ي` `##ل` `##ة` `و` `##غ` `##ن` `##ي` `##ة` `ب` `##ا` `##ل` `##ت` `##ا` `##ر` `##ي` `##خ` `.` `[SEP]`
+
+**Input:** مرحبا بالعالم!
+**Tokens**: `[CLS]` `م` `##ر` `##ح` `##ب` `##ا` `ب` `##ا` `##ل` `##ع` `##ا` `##ل` `##م` `!` `[SEP]`
+
+**Input:** Simplified: 快速的棕色狐狸跳过懒狗。
+**Tokens**: `[CLS]` `simplified` `:` `[UNK]` `[UNK]` `的` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `。` `[SEP]`
+
+**Input:** Traditional: 快速的棕色狐狸跳過懶狗。
+**Tokens**: `[CLS]` `traditional` `:` `[UNK]` `[UNK]` `的` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `。` `[SEP]`
+
+**Input:** 素早い茶色の狐が怠け者の犬を飛び越える。
+**Tokens**: `[CLS]` `[UNK]` `[UNK]` `い` `[UNK]` `[UNK]` `の` `[UNK]` `か` `[UNK]` `け` `[UNK]` `の` `犬` `を` `[UNK]` `ひ` `[UNK]` `え` `##る` `。` `[SEP]`
+
+**Input:** コンピュータープログラミング
+**Tokens**: `[CLS]` `コ` `##ン` `##ヒ` `##ュ` `##ー` `##タ` `##ー` `##フ` `##ロ` `##ク` `##ラ` `##ミ` `##ン` `##ク` `[SEP]`
+
+**Input:** 빠른 갈색 여우가 게으른 개를 뛰어넘습니다.
+**Tokens**: `[CLS]` `[UNK]` `ᄀ` `##ᅡ` `##ᆯ` `##ᄉ` `##ᅢ` `##ᆨ` `ᄋ` `##ᅧ` `##ᄋ` `##ᅮ` `##ᄀ` `##ᅡ` `ᄀ` `##ᅦ` `##ᄋ` `##ᅳ` `##ᄅ` `##ᅳ` `##ᆫ` `ᄀ` `##ᅢ` `##ᄅ` `##ᅳ` `##ᆯ` `[UNK]` `.` `[SEP]`
+
+**Input:** तेज़ भूरी लोमड़ी आलसी कुत्ते के ऊपर कूदती है।
+**Tokens**: `[CLS]` `त` `##ज` `भ` `##र` `##ी` `ल` `##ो` `##म` `##ड` `##ी` `आ` `##ल` `##स` `##ी` `क` `##त` `##त` `क` `[UNK]` `क` `##द` `##त` `##ी` `ह` `।` `[SEP]`
+
+**Input:** দ্রুত বাদামী শিয়াল অলস কুকুরের উপর দিয়ে লাফ দেয়।
+**Tokens**: `[CLS]` `দ` `##র` `##ত` `ব` `##া` `##দ` `##া` `##ম` `##ী` `শ` `##ি` `##য` `##া` `##ল` `অ` `##ল` `##স` `ক` `##ক` `##র` `##ে` `##র` `উ` `##প` `##র` `দ` `##ি` `##য` `##ে` `[UNK]` `দ` `##ে` `##য` `।` `[SEP]`
+
+**Input:** வேகமான பழுப்பு நரி சோம்பேறி நாயின் மேல் குதிக்கிறது.
+**Tokens**: `[CLS]` `வ` `##ே` `##க` `##ம` `##ா` `##ன` `[UNK]` `ந` `##ர` `##ி` `[UNK]` `ந` `##ா` `##ய` `##ி` `##ன` `ம` `##ே` `##ல` `[UNK]` `.` `[SEP]`
+
+**Input:** สุนัขจิ้งจอกสีน้ำตาลกระโดดข้ามสุนัขขี้เกียจ.
+**Tokens**: `[CLS]` `[UNK]` `.` `[SEP]`
+
+**Input:** ብሩክ ቡናማ ቀበሮ ሰነፍ ውሻን ተዘልሏል።
+**Tokens**: `[CLS]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[SEP]`
+
+**Input:** Hello 世界 مرحبا 🌍
+**Tokens**: `[CLS]` `hello` `世` `[UNK]` `م` `##ر` `##ح` `##ب` `##ا` `[UNK]` `[SEP]`
+
+**Input:** 123, αβγ, абв, العربية, 中文, हिन्दी.
+**Tokens**: `[CLS]` `123` `,` `α` `##β` `##γ` `,` `а` `##б` `##в` `,` `ا` `##ل` `##ع` `##ر` `##ب` `##ي` `##ة` `,` `中` `文` `,` `ह` `##ि` `##न` `##द` `##ी` `.` `[SEP]`
\ No newline at end of file
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d1024.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..03b30132cc4626ff15fe2e93ad746fdbfda11135
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ddd3869c835b2bd7b82cd1e4bb8746f1b78d7c8cd9db7133df283755ea54ec67
+size 62509184
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d1024.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..83852d6be142cc7baf265c87bebda36e238cc116
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3cd351889a6c6f5c59b618fd07b930a3ad56f531cbf4818cacddbfcd81649da
+size 57747862
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d128.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..f92c69862341c33ba8994acab73638b4fd2a836b
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf86cdf61efa6e14d28d04a04d4b09b7b2a01663e4d3b957388e1e90e5f144c2
+size 7813760
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d128.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..307adcfb90926dcb7b1978e44ae15c070dce331c
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9472e92a0439cfbfe28eccb00a7a0385ec61528a47d58017d1f0af95a8ef2015
+size 7232185
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d256.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..bc905b360c11076c86c7f4ed9e5c3387daa623fa
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64e61708242d167f7cad3c358b9dc37a46837d2dec5f85e70ae97b157b598e78
+size 15627392
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d256.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..7a7e74f1fb68bdd167b669c8df62829a433efe1d
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3e976051d54f0c5053d83ebb748aed8db3db3ab2068909de0f671db90eb9af3
+size 14447854
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d32.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..29afdd7709ea60ac67c5eb51e85a0198d0c5c179
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e6e01ab559b15ee2cf607ff29eb735d4a5e50a32a81687491c6a527bfd4fd16
+size 1953536
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d32.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..42995801ac3b7a2dd4d706bed5e2f3270cd4adf4
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:816afd2323e4e56778d7eb8d0e7d9af9c6afb19042859bcbaae36cca45acc2d6
+size 1816316
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d512.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..f92b2094495c78919e9b3801de060d4694e337f8
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5128869afc654037f31cc9e83adc3d658f6ec608433807351aa03a99f323953
+size 31254656
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d512.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..adb4431441a253933871522e6948705befb62dc3
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:338e2f0b2e5deacda72010ddfbd9e35e10baad5e1c2b669c1da60ca6a129c6a7
+size 28880481
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d64.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..cc4d38138dfe7153b21eab025813a65b3f3af003
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99f38812643a76cda63346ff2adbb75672dcadae5e9c13316b51cc5b290e7a4e
+size 3906944
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d64.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..25224b00c2914f9a4d73e9c12ee195ffa4de7514
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp16.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:776556fa48dbcfc064dba21362af22cb647d8b206d531b7b87fc42f65e67832e
+size 3622202
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d1024.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..aa490de484607c53356d2ee92c4774f853644c68
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d43d956308fc0df0bd6e014a180135e747ef2158728e256572ad6d7cd0c43222
+size 125018240
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d1024.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..0108bd6e7215245fa2f51ef9b7a116ba6aad6bb1
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1c4755ad98492aa91618f7e1f77968840178f3e1b9cdb79824ee246ff4058d0e
+size 115965979
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d128.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..3256a60375ce6fb283c2922f6c5064127d6cb7f5
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6129ca3dd6a63472a512aac8c569c4ad149f1eb6babbcc8d3461da0337c7968e
+size 15627392
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d128.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..1b3b31eadf7d825282e38585d952f75c742710cc
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fea30619293b15023fb66a03e0affba1f4a6ea5535250eae2e39610675e7722
+size 14514190
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d256.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..02c379cd75f7dc50424fcdccfb68f26d04676f60
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f30c8afb92ee7122e53439d4d0e12d5c70f6f2c21724a2986c582a77e58f95ed
+size 31254656
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d256.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..926aeec0b1848c68dd136010ea6b434a9a0c22a2
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbebe8f553aa5d9dd3137d35ed95c61c024d6fb94c37a6d9d6c685486801e25f
+size 29013253
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d32.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..548be34b9e28ab4cac0741ebe54f4c5424f5fd8e
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c29015f985f2b47da7df7d66a67626922b2fdafa9529e67509f7c3888597073e
+size 3906944
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d32.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..373af5bba323ad516bde5cc58958f8bc5014259e
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c70ce60302d74fc700d4736ef45792eb6939bac7df9f9e2f3dac0953a3b21015
+size 3633601
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d512.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..51d7fb2b9d861e8be6cc4271210e0e47e375c9cc
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6a3d065e897517ca4ec888c68aa4403b5fe0b11aa329cd9e3f034016fcb04a6
+size 62509184
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d512.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..56aec56e5f9aec155d510d27a8620b44d6c555ca
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f6e5bacca0788475d7585af3d403f3990ef8bdeee4a8dbf8debb546fc2d6e8a
+size 58000062
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d64.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..704bfa866cd86374c984a0fdde2b9f484acc042b
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:860ef6ee1dedfd1442f36c6ffc589d0da4e0f439e9f95d1f7cb7b2b3a9cc7afa
+size 7813760
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d64.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..2b9abc08a472cd0e8f51a0f406d0c86d299fa361
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp32.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ff3a3cfd973d7c5553f09b2d4c1208d47715a04b308762472958490ae22fbe9
+size 7262326
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d1024.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..8a2e22fbfe1ceec901fddb51d45e2e207d844e08
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5a7e6eb0dd72bc5bae24d50a0b77dbf0dfd1bbdc4a0799d345f146bff0bc2d3
+size 31254656
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d1024.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..1ba54df1178f32ae0c1b441f796d07267a0371d4
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78c8eda325bd95d8fdde2f806ffcac8a0581964ccb68d5db605af3b55ae8acd6
+size 25956724
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d128.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..1e237aed1155b5856e59d62e37da30d81195fea3
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44969be48326a0675bf3bc0e867eed6e3740f0daf968532d89d12a2da3719ca6
+size 3906944
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d128.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..be3a234b0db8814802f5ea810a54a4c3dfcf14c5
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcc596fbcf1831bc60486f93b69c9e9046b3c628ce3a8f06be630b4d01784528
+size 3266488
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d256.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..78034bb6690875bc469016fbd9ec9ab354ccb076
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b8656c609f1aae181b2a38db5e38036ef8dea6767ddafae81fa1ae86be837fd
+size 7813760
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d256.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..a0b5b7292a3e9d001d455aa09f9455ca9117dc30
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd49de2d20d0010fe4939fd89e520919de425540658d86354925d5b1e74ed5c9
+size 6503696
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d32.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..ae70463d5b77399e877d85e16b0d1c6d6d7331c0
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f2bd97aaf1de87164eddfa545a24735d094706a523b20199500b1a55863e211
+size 976832
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d32.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..70e2c76da27eafc0af29f8ebae6111395a8b2986
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcb04eac6faffbac4b371adb58a43d61a3416eefc9361444ace97529f966747b
+size 829988
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d512.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..c6bfc5714aae051eb8735ec01756b8b9fd675897
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e957dc5d1a764b04fafa61429bc34564212fd408aace5a439d976b8d9e19e966
+size 15627392
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d512.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..aefa13bcd8d29c2bba402ca88c18e432f1d66ca5
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90815ea8fcc001a1e9465ff37935a41709998f1024ab1c4b4e7ef2acc3e74af7
+size 12981125
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d64.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..92082cbdb25d30f7b1567b4e83d135f763f7f09c
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd9728e2820c4d853b761cf48f01db208906c0322d8a8f7865b4c844f263d90c
+size 1953536
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d64.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..a3f5613d526e8b954e9a3841f21142fe481b3af4
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e4m3.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4eefecb3640ed091ac2d64f622999cbb278d006bbcc2b7ec4ca99d1620ad0f72
+size 1650598
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d1024.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..4c6024fb191d28441c3341243ea69888e99f613f
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:29b102f0cb5f51c34522dbab99a4b865fa1d9f19abd1612954dd53317c0d7776
+size 31254656
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d1024.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..ac65a59149b483db67af355cc93c4846792f8b17
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:064ff941fc8e770318b5b2e54decc02359e5794b4ddec2f93a747097deed08b3
+size 22092731
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d128.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..4d1da3ae4d500b99c90221dc107f001eed4073b4
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:72ab6abd96387345914c93e0dd7fe221a217819ea29665497353f58056abdb51
+size 3906944
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d128.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..3d665690f47bb98c124c7c43a000c25c3be1ea04
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d6e9fdf7f7d9b26ddf2a72735a35d48e2a36f57761020a88747ea227e13b1b4
+size 2783769
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d256.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..6dab9ab0123558b67fdfe3f06991545dc3ca27ae
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0abaa4a0ccdc5957412e2b866d99d5a7904a0cdb08f4d86ff9bf599bdbd8dea4
+size 7813760
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d256.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..854505be1a34343c9c3b925e3bbbb61671401a43
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84ec7bc6759080154063bb3231d91653ef5c9f05c888090c0bf2d745f39190d6
+size 5539462
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d32.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..002a4b60598fd1489b56c54de4e60f739cc7305b
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f7674250ea887d3dc8e366c50fd2543b42d4925bf02ccda3772493d49d7519c
+size 976832
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d32.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..f0f31185868a3437b7aa8e7e8d72a83a11c8bff8
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3b0d5fbc317d0330bc0cc6cf1d5a953078167199e2b762bfd9879ffae26c582
+size 709321
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d512.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..59b68ad1282126ec557c46cada6163af6e94d8a5
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9950cffb04cad00e19a71bcc35b30cd95ca32bcc4bb52190cad02438c66af1f4
+size 15627392
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d512.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..c10f0af4f3031bf561d897852fe0005865302a84
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0ec0ac2eaf504910bcb45db3d7f76e18bd1a9807fb455122f56994b35727f1de
+size 11046408
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d64.npy b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..469c07cd9b164ebeef99a30e72ea87354bc8b0ad
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c0b68f168abee4bbcd3b3c595a2a6a69ecbe5c76e863f8b6090ad0ce2f62111
+size 1953536
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d64.npy.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..0c0cf043406f15590fb7724fdaf5b2a1c1d90126
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/fp8_e5m2.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0facfa00fe53eb1155d627e2dbd354ee27eb78d1de157bfb8cd2199e24022611
+size 1408950
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/tokenizer.json b/models/sentence-transformers/static-retrieval-mrl-en-v1/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..c42609a392fabf89b299525fced86e51ec1009e4
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d241a60d5e8f04cc1b2b3e9ef7a4921b27bf526d9f6050ab90f9267a1f9e5c66
+size 711396
diff --git a/models/sentence-transformers/static-retrieval-mrl-en-v1/tokenizer.json.zst b/models/sentence-transformers/static-retrieval-mrl-en-v1/tokenizer.json.zst
new file mode 100644
index 0000000000000000000000000000000000000000..8b4926c1d4c722a1eab36b36d9f6d75c2fd47c77
--- /dev/null
+++ b/models/sentence-transformers/static-retrieval-mrl-en-v1/tokenizer.json.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7046761b6603a0f60bf1593b941e9bfd493f532a10b1e848e3e41b5496e7d993
+size 219696
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/README.md b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..54cae5fcdaaa007c84c3b1100fb4ffcde2764c14
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/README.md
@@ -0,0 +1,127 @@
+# [sentence-transformers/static-similarity-mrl-multilingual-v1](https://huggingface.co/sentence-transformers/static-similarity-mrl-multilingual-v1)
+
+License: [apache-2.0](https://choosealicense.com/licenses/apache-2.0/)
+
+Multi-lingual similarity embeddings that were trained with Matroyshka loss
+that allows for more effective truncation of the embedding vectors. It
+was trained on a variety of domains of multilingual datasets.
+
+It's a general purpose model that can be used for semantic textual similarity,
+paraphrase mining, text classification, clustering, and more
+
+## Model Stats
+
+Stats that describe the embeddings tensor shapes and value distribution.
+
+| item | metric | value |
+| --------------| ----------------------- | ----- |
+| vocab | size | 105,879 |
+| embedding | dimensions | 1,024 |
+| vector length | mean | 413.61 |
+| vector length | median | 437.74 |
+| vector length | stddev | 195.51 |
+| values | mean | -0.02 |
+| values | median | -0.01 |
+| values | stddev | 14.30 |
+
+## Mean Pooled Quantization Loss
+
+This test roundtrips the vectors through quantization, but performs the
+mean pooling arithmetic in float32 space. The quantized and unquantized
+mean pooled vectors are compared to each other to determine their cosine
+similarity, to show how much the meaning of the vector has changed due
+to quantization.
+
+| Precision | Cosine Similarity |
+| ------------- | ----------------- |
+| fp16 | 1.00000 |
+| fp8 e4m3 | 0.99980 |
+| fp8 e5m2 | 0.99921 |
+
+## Quantization Loss Per Vector
+
+While ultimately the embedding vectors will be mean pooled together, it's
+still useful to look at the loss per-vector in the embedding table to see
+which quantization strategies retain the most vector meaning.
+
+- **Cosine Similarity** — measures how well the *direction* of embedding vectors
+is preserved after quantization, independent of scale. This is especially
+relevant when embeddings are used for similarity search or retrieval.
+- **MSE (Mean Squared Error)** — emphasizes large errors by squaring the
+differences. Useful for detecting whether any values are badly distorted.
+- **MAE (Mean Absolute Error)** — the average absolute difference between
+original and quantized values. Easier to interpret, less sensitive to outliers.
+
+| Precision | Metric | Value |
+| ------------- | ------ | ----- |
+| fp16 | cosine similarity | 1.00000 |
+| fp8 e4m3 | cosine similarity | 0.99965 |
+| fp8 e5m2 | cosine similarity | 0.99861 |
+| fp16 | MSE | 0.00001 |
+| fp8 e4m3 | MSE | 0.14369 |
+| fp8 e5m2 | MSE | 0.56917 |
+| fp16 | MAE | 0.00183 |
+| fp8 e4m3 | MAE | 0.23372 |
+| fp8 e5m2 | MAE | 0.46585 |
+
+## Tokenizer Examples
+
+**Input:** This is an example of encoding
+**Tokens**: `[CLS]` `this` `is` `an` `example` `of` `en` `##co` `##ding` `[SEP]`
+
+**Input:** The quick brown fox jumps over the lazy dog.
+**Tokens**: `[CLS]` `the` `quick` `brown` `fox` `jump` `##s` `over` `the` `la` `##zy` `dog` `.` `[SEP]`
+
+**Input:** Curaçao, naïve fiancé, jalapeño, déjà vu.
+**Tokens**: `[CLS]` `curacao` `,` `nai` `##ve` `fia` `##nce` `,` `ja` `##lap` `##eno` `,` `deja` `vu` `.` `[SEP]`
+
+**Input:** Привет, как дела?
+**Tokens**: `[CLS]` `при` `##вет` `,` `как` `дела` `?` `[SEP]`
+
+**Input:** Бързата кафява лисица прескача мързеливото куче.
+**Tokens**: `[CLS]` `б` `##ър` `##за` `##та` `ка` `##ф` `##ява` `ли` `##си` `##ца` `пре` `##ска` `##ча` `м` `##ър` `##зе` `##ливо` `##то` `к` `##уч` `##е` `.` `[SEP]`
+
+**Input:** Γρήγορη καφέ αλεπού πηδάει πάνω από τον τεμπέλη σκύλο.
+**Tokens**: `[CLS]` `γ` `##ρη` `##γο` `##ρη` `κ` `##α` `##φ` `##ε` `α` `##λε` `##που` `π` `##η` `##δα` `##ει` `πανω` `απο` `τον` `τ` `##ε` `##μ` `##πε` `##λη` `σ` `##κ` `##υλο` `.` `[SEP]`
+
+**Input:** اللغة العربية جميلة وغنية بالتاريخ.
+**Tokens**: `[CLS]` `اللغة` `العربية` `ج` `##ميل` `##ة` `و` `##غنية` `با` `##لت` `##اري` `##خ` `.` `[SEP]`
+
+**Input:** مرحبا بالعالم!
+**Tokens**: `[CLS]` `م` `##رح` `##با` `با` `##ل` `##عا` `##لم` `!` `[SEP]`
+
+**Input:** Simplified: 快速的棕色狐狸跳过懒狗。
+**Tokens**: `[CLS]` `simplified` `:` `快` `速` `的` `棕` `色` `狐` `狸` `跳` `过` `懒` `狗` `。` `[SEP]`
+
+**Input:** Traditional: 快速的棕色狐狸跳過懶狗。
+**Tokens**: `[CLS]` `traditional` `:` `快` `速` `的` `棕` `色` `狐` `狸` `跳` `過` `懶` `狗` `。` `[SEP]`
+
+**Input:** 素早い茶色の狐が怠け者の犬を飛び越える。
+**Tokens**: `[CLS]` `素` `早` `い` `茶` `色` `の` `狐` `か` `怠` `け` `者` `の` `犬` `を` `飛` `ひ` `越` `える` `。` `[SEP]`
+
+**Input:** コンピュータープログラミング
+**Tokens**: `[CLS]` `コ` `##ン` `##ヒ` `##ュー` `##ター` `##フロ` `##ク` `##ラ` `##ミ` `##ンク` `[SEP]`
+
+**Input:** 빠른 갈색 여우가 게으른 개를 뛰어넘습니다.
+**Tokens**: `[CLS]` `ᄈ` `##ᅡ른` `가` `##ᆯ` `##색` `ᄋ` `##ᅧ` `##우` `##가` `ᄀ` `##ᅦ` `##ᄋ` `##ᅳ` `##른` `ᄀ` `##ᅢ를` `ᄄ` `##ᅱ` `##어` `##너` `##ᆷ` `##스` `##ᆸ니다` `.` `[SEP]`
+
+**Input:** तेज़ भूरी लोमड़ी आलसी कुत्ते के ऊपर कूदती है।
+**Tokens**: `[CLS]` `त` `##ज` `भर` `##ी` `ल` `##ो` `##म` `##डी` `आल` `##सी` `क` `##तत` `क` `ऊपर` `क` `##द` `##ती` `ह` `।` `[SEP]`
+
+**Input:** দ্রুত বাদামী শিয়াল অলস কুকুরের উপর দিয়ে লাফ দেয়।
+**Tokens**: `[CLS]` `দ` `##রত` `বা` `##দা` `##মী` `শ` `##িযা` `##ল` `অ` `##ল` `##স` `ক` `##কর` `##ের` `উপর` `দিযে` `ল` `##া` `##ফ` `দেয` `।` `[SEP]`
+
+**Input:** வேகமான பழுப்பு நரி சோம்பேறி நாயின் மேல் குதிக்கிறது.
+**Tokens**: `[CLS]` `வ` `##ே` `##கம` `##ான` `ப` `##ழு` `##பபு` `நர` `##ி` `ச` `##ோ` `##ம` `##ப` `##ே` `##றி` `ந` `##ாய` `##ின` `மேல` `க` `##ு` `##தி` `##ககிறது` `.` `[SEP]`
+
+**Input:** สุนัขจิ้งจอกสีน้ำตาลกระโดดข้ามสุนัขขี้เกียจ.
+**Tokens**: `[CLS]` `[UNK]` `.` `[SEP]`
+
+**Input:** ብሩክ ቡናማ ቀበሮ ሰነፍ ውሻን ተዘልሏል።
+**Tokens**: `[CLS]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[UNK]` `[SEP]`
+
+**Input:** Hello 世界 مرحبا 🌍
+**Tokens**: `[CLS]` `hello` `世` `界` `م` `##رح` `##با` `[UNK]` `[SEP]`
+
+**Input:** 123, αβγ, абв, العربية, 中文, हिन्दी.
+**Tokens**: `[CLS]` `123` `,` `α` `##β` `##γ` `,` `аб` `##в` `,` `العربية` `,` `中` `文` `,` `हिनदी` `.` `[SEP]`
\ No newline at end of file
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d1024.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..89dff6e6b5b27fdf06014f1bdce0f57af8b27854
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:660fcf3af013c9e61d27011c244a683cf23462f52b7e591abd9beb2b8f9235bf
+size 216840320
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d1024.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..a5ed52881237d7795d8abf7f53463d8f40aa2bf1
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b67bc3307eeb0387866325ff33a41f8bdd673e309c18224e1415f65b9873eba
+size 200900906
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d128.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..9900add4d6ceb69ddf236fbc6df897477f0b9af9
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6570ef89c5dca0adf3d3de7b49f124ab152211c03fec3ca9c9305fb8b2c650de
+size 27105152
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d128.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..c9b7a4c7cce40ca8aecdcba7e8b919b1798e9e89
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f76a08d3c0862729e9d82349f06672e746dfe46fe0d7e4b673b7e9aa68f3d4b
+size 25121824
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d256.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..612e5e6c6c36c01fb4e6bf1bfa8b38f030e30a2c
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de0e08dcb3d5dfcb4df5d94acd5d08271096dfb1ee1f8c70853dae72490f8b77
+size 54210176
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d256.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..c3f65207e9dfb6fb854e8728f9147e27af18cd19
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5cf8fd3a51042d50e9b32dacade40cf845e89472bb39d73360369a5fb53951a
+size 50226236
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d32.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..7971411b10733ce6827bfee3d2bfafe301ff6645
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9d5ccc8b909517d208759685e534bd9e7905ccef81c2ba0eea7474e94d7557e5
+size 6776384
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d32.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..5ca3985c12b04d0367c92783ef78aa9b2c8a837f
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6d5813fda5d1d817351854dfa518a2029600c9f49ebb23bcfb423c1f7801e4c
+size 6287693
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d512.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..961c70c0849fb48980be5e67d387c145a723b5b9
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99d247c160114e32b46969d149909e33512afcd338d2188d61a34bf1589451b8
+size 108420224
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d512.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..2e82d74aa3e0d2bd87116292e5f03e4e8d33f458
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ac0d1fc31d750d9a229ba9ab71767f0e48f2233f7b1e0d5bd87dade726da1902
+size 100438595
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d64.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..29dca429a49ff015ab0d0311a7dbc0455df9b52f
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f74f66961f1b0fc3fe238ac3fe75f3e74d4d6b62482f1e832442557d24a52f1
+size 13552640
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d64.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..052e21b935849ecff51e2801a5d113a4e3e2ef77
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp16.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:449912dc47784d5d515f18cb8e4091d07abdcd96cccd84832daf1b00c489af36
+size 12565953
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d1024.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..6d3352ac11cd1b0fb16fc04d0c6821525c698a27
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:954fce02a7ca0565db82cfd2f3bd044b5c0407aaea54bbeb493499ea47e2b196
+size 433680512
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d1024.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..884c9faac9c92a789fef7da2d1b423c967c303fd
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef55fc97d834f27413b63d9bba3b4832b8c66b2a27bfaba7afc8d2db4c236e68
+size 402617664
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d128.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..7b1afbc3f9da8c65967d19efc92ffee6268c62f4
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d30b6cb7e1a1eff228267938f5af72d08339d08b95048ace5438f832767d7427
+size 54210176
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d128.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..d34d546deeb151f76a71008a53c2236ecaf991f7
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fd55e3cd211dcc4588ce06f36e452d52216800c7a286f568bc5d48ff2985ab2
+size 50353944
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d256.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..b685b0984f7536a9b4bbfd2e75e5da456df086fc
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03e6950f0ba6dff7615326c4acab19815cf16e20db310a2cdceaaeb00dfafe9f
+size 108420224
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d256.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..10ea7f2dbf2a23c5247435393f5656459800dd24
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f15274b16a188cdc049f704989adff94786130970fe0adbf0a83be84a699da10
+size 100677611
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d32.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..6a00e836c1f7a3749bed7db6961d018dca06f245
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:078ed8a279d8786cf626e6adb21a5c0420ac1700b4315b178e3468f15513b12c
+size 13552640
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d32.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..19f917448981beccfc9b732b0198d0956178e66f
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcecfa6d13b7bae7a7fcb822e82ecb1555294e8b644ce5c4fbd06821ab4719fd
+size 12597604
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d512.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..82c4c2b33d1491216d2cd0f7c2c08f44580ae294
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:abc5c070925225bcb0252755f65716bbac33b47240f8e67ddeebceea5825778f
+size 216840320
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d512.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..54907a019ee56485f26ac8d3342c83cc3c8b2f24
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dbb050817d681d631655a6eba991f165150c0ffb45378e966605220f7bf8fe00
+size 201316000
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d64.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..31952d589e3a3f50f3c08aa7e6711a0ff4f1cef5
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e2372abaa3781d202c93ef06c04c8f9f990c5bef9b6f332fe8d5450595e1277
+size 27105152
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d64.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..cb94686159873db67aafe1e4e264cbfeaf6ca8e8
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp32.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:50514a0bf4c6284c9cffc1d2115bf9b8ab12b5f4b4a0536bf5d162e92cde8786
+size 25185829
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d1024.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..6faf6bc37c676560186aaeb34e1cf7ceba6141af
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e18f3e5c53ca5c0b3136b36458f7fb5a7b604931942c86d12770d7e5a61b8094
+size 108420224
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d1024.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..02c2c6dde90695fa8c002bc58d9c1eaf9129c1e3
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be35fa692b60ed0a6cb259cee95d9c34080f21bf048ea211fbea606bd044a6cf
+size 90840905
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d128.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..92d02cc971c297c97c8b1356f1f2bae29d531498
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fffda612fb4339c31a59d008681a826f1620dbf009805c1e5d497877a04d4f3
+size 13552640
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d128.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..d7ff83cc810bfbd2e4c668670fe33d4975dc15c3
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bfc46db014034b9776f96adf2416a79fb54f04a5ab82715371ebae96819cd38
+size 11376038
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d256.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..0855b5f3e6a0120f050b32eb92a10af2cb1e3b76
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:872b732c7316333e2622714f863dfc68c33a192ae8a97c20cfb62ad3b09f7016
+size 27105152
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d256.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..ea9ddfc3d2e275f0685ef7a8587d6c381f585240
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45de84fa8a6f39d2363048741fe5ff82c2f5199cf75a7a640738a36cdd864e55
+size 22719354
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d32.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..b83033b1e5b7bc77c290895c345bd3197747b639
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3e4069b0a907cc69933498493ed1a36aacafeacef2afa8a00e410d0df674e88
+size 3388256
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d32.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..b27535de67b170bab4adfae62093422e77132950
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a16841d6c02b55fe0e1a3959dcb681e8b6442e36e25c17cfae5bf9f1818f287f
+size 2856148
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d512.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..e4ee1e13b953089b959bfaef4607dd8251ca37bb
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c13b7881e738ac0436057907fa19ee50b4170ed3dddab28da389bbf089e13ec8
+size 54210176
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d512.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..e162b5b4a8baa708c6dd0b78fcd45f2275537582
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a4f05e20f4884521ee62dd245291d3ba3657f01de831facaa76f8db7a404dd4
+size 45403134
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d64.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..1c3e3b613d9085920a2cfd6addc23dfa455a3914
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b50e12a72eb2c857b6feb6b0cb742a8ddb03cb47eefad4d0c85045535a02096d
+size 6776384
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d64.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..f5ed606cfbcf96cf208cf1411fae408fe8b71a69
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e4m3.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d2ae256e2b2d40ac0069c6dc6d5535a5104322c0eb93d131bfd85d42548f106
+size 5702736
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d1024.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d1024.npy
new file mode 100644
index 0000000000000000000000000000000000000000..0e223287057517b37f425dfa0d2c15df16c946bf
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d1024.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:076e26375628042330b37c267e7367248c35e2c6e696f883274b0b525ae8673b
+size 108420224
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d1024.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d1024.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..8d74c38c656a242237cf94a357891e59ca6aab0b
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d1024.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d2d4d77d94153f7e82aa10c7cc29c77e3b7717c95324a0bedb72a8f391df5fa
+size 77519454
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d128.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d128.npy
new file mode 100644
index 0000000000000000000000000000000000000000..7a6675f28158d91e1c63d29ddb02f6367b4c6d23
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d128.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5da5a021feeb65f4b4abdb016c6f40421eeeefc19c470d214615bb175ade0de
+size 13552640
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d128.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d128.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..df446a8bdf44dbcfbb732b84017010c051df5b64
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d128.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:895e1878f0e43cdbfecc28de19ea9a7d90904c2b63562c3825958566120d8157
+size 9713906
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d256.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d256.npy
new file mode 100644
index 0000000000000000000000000000000000000000..a534cdebeabc457b5cda04e17a18c541944a0cb6
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d256.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e6919585fa4882b1385d961202e6a5b2432bc055f2ad332f4192d411ed28ecb
+size 27105152
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d256.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d256.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..c8c1dcece69469c1af984e54bb29aabd9f586c9c
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d256.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c75041bb648d9e16b991da2d206bbf129fb106ab7b8b08008c572b972323a98b
+size 19390609
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d32.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d32.npy
new file mode 100644
index 0000000000000000000000000000000000000000..c03da1500b705745429211ce5f7e966eb1c98fe3
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d32.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6dca145f0b15d5adc85dc643b43f7f627bb03a8b00b3745f3af63e9798e041ce
+size 3388256
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d32.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d32.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..cbe9796eb1990d53df80a47750baabee2be8b242
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d32.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4aa8e2bd3f44b6a53e53472a27cacd03b0454f6a0293191ce8767449f7015433
+size 2438537
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d512.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d512.npy
new file mode 100644
index 0000000000000000000000000000000000000000..09c6b89f675f0fc734529d6d88878a8ac01b13f8
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d512.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5abbe6cfc7a42097d268534a785c6cc4e0c7d1536ad16b920fbe39ec0ee16de
+size 54210176
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d512.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d512.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..983df41dd7958e601f4551316b5aeaf8671ea8b9
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d512.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:800c90e7ef22a5b5102b47e28920fd1e5b253fb379ecbb0029419001661622af
+size 38743054
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d64.npy b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d64.npy
new file mode 100644
index 0000000000000000000000000000000000000000..71d88d996b293de52359f8834d1c19412fcd032f
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d64.npy
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7f83ba68bdb3a16419369377ea991119c453cca1506448bc31fa5f97a1f3ed3
+size 6776384
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d64.npy.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d64.npy.zst
new file mode 100644
index 0000000000000000000000000000000000000000..bc33765440336d247d0803db2487277b19ad238b
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/fp8_e5m2.d64.npy.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3acdfbedc086052d5eeac35c7f7b030ca5d5a495920def312022bb888e761ab1
+size 4864996
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/tokenizer.json b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..1f8c378b0d14f649b58e5284a6bee646ae8f6638
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11aaf894a4ccf3d95e8830e27c0f8152791fbbff2b988e29a265580b86edd216
+size 2563370
diff --git a/models/sentence-transformers/static-similarity-mrl-multilingual-v1/tokenizer.json.zst b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/tokenizer.json.zst
new file mode 100644
index 0000000000000000000000000000000000000000..8c9559ff02973eebfc6b8ba89f693b0c46bd2197
--- /dev/null
+++ b/models/sentence-transformers/static-similarity-mrl-multilingual-v1/tokenizer.json.zst
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba5b1f4e61ca2060fa6ec2f847a8439c17572c40f5e985039dcba0684b5f04a9
+size 830407
diff --git a/scripts/build_models.py b/scripts/build_models.py
index 426998bd81f6b4fca27ec60367f334cce386dd02..262d4d9a7e663b045126dcd97fb0269a6c4441ea 100644
--- a/scripts/build_models.py
+++ b/scripts/build_models.py
@@ -1,5 +1,6 @@
+from dataclasses import dataclass
import shutil
-from textwrap import dedent
+from textwrap import dedent, indent
from typing import Any
import numpy as np
from zstandard import ZstdCompressor
@@ -9,11 +10,30 @@ from sentence_transformers import SentenceTransformer
from torch.nn import EmbeddingBag
import torch
from model2vec import StaticModel
-from tokenizers import Tokenizer
+from tokenizers import Encoding, Tokenizer
models_path = Path("models")
+@dataclass
+class ModelCard:
+ owner: str
+ repo: str
+ # The dimensions that were applied with Matroyshka Loss.
+ matroyshka_dims: list[int]
+ description: str
+ license: str
+
+ def name(self):
+ return f"{self.owner}/{self.repo}"
+
+ def path(self):
+ return models_path / self.owner / self.repo
+
+ def get_description(self):
+ return dedent(self.description).strip()
+
+
def zst_compress_file(input: Path):
cctx = ZstdCompressor()
output = input.parent / f"{input.name}.zst"
@@ -88,40 +108,133 @@ def quantization_loss_cosine(tensor: torch.Tensor, dtype: torch.dtype):
return cos.mean().item()
-def export_embeddings(
- hf_org: str, hf_repo: str, model_path: Path, embeddings: torch.Tensor
-) -> None:
+def export_embeddings(model_card: ModelCard, embeddings: torch.Tensor) -> None:
vocab_size, dimensions = embeddings.shape
# This logic can always be adjusted for models with different shapes.
assert (
embeddings.dtype == torch.float32
), f"The embeddings {embeddings.dtype} are assumed to be float32."
- assert (
- dimensions <= 1024
- ), f"The embedding {dimensions} dimension is assumed to be at most 1024."
+ for dim in model_card.matroyshka_dims:
+ assert (
+ dim <= dimensions
+ ), f"The Matroyshka dimensions {dim} were bigger than the models dimensions of {dimensions}"
+
+ truncated = embeddings[:, :dim]
+ assert truncated.shape == torch.Size([vocab_size, dim])
+
+ save_data(model_card.path() / f"fp32.d{dim}.npy", truncated)
+ save_data(
+ model_card.path() / f"fp16.d{dim}.npy",
+ truncated.to(dtype=torch.float16),
+ )
+ save_data(
+ model_card.path() / f"fp8_e5m2.d{dim}.npy",
+ truncated.to(dtype=torch.float8_e5m2),
+ )
+ save_data(
+ model_card.path() / f"fp8_e4m3.d{dim}.npy",
+ truncated.to(dtype=torch.float8_e4m3fn),
+ )
+
+
+def normalized_mean_pooling(x: torch.Tensor) -> torch.Tensor:
+ pooled = x.mean(dim=0)
+ normalized = torch.nn.functional.normalize(pooled, dim=0)
+ return normalized
+
+
+def export_readme(
+ model_card: ModelCard,
+ embeddings: torch.Tensor,
+ tokenizer: Tokenizer,
+):
+ vocab_size, dimensions = embeddings.shape
norms = torch.norm(embeddings, dim=1) # shape: [vocab_size]
- print(f" - vocab size {vocab_size:,.0f}")
- print(f" - embedding dimension {dimensions:,.0f}")
- print(f" - vector length (mean): {norms.mean().item():.2f}")
- print(f" - vector length (median): {norms.median().item():.2f}")
- print(f" - stddev ±{norms.std().item():.2f}")
- print(f" - value (mean): {embeddings.mean().item():.2f}")
- print(f" - value (median): {embeddings.median().item():.2f}")
- print(f" - stddev ±{embeddings.std().item():.2f}")
+ phrases = [
+ "The committee approved the proposal after hours of heated discussion and several last-minute amendments."
+ "When training large neural networks, careful tuning of hyperparameters can significantly affect performance and stability."
+ "Despite the heavy rain, the concert continued as planned and the crowd stayed enthusiastic until the final encore."
+ "In ancient mythology, heroes often embarked on perilous journeys to discover hidden truths about themselves and their world."
+ "The new smartphone model features an improved camera system, faster processing, and extended battery life compared to its predecessor."
+ "He tried to explain the concept using simple analogies, but the underlying mathematics remained difficult to grasp for most listeners."
+ "After weeks of negotiations, the two countries signed a historic trade agreement aimed at reducing tariffs and boosting cooperation."
+ "She paused for a moment before answering, choosing her words carefully to avoid misunderstanding in such a delicate situation."
+ "The detective pieced together the timeline of events, realizing that the key witness had provided a contradictory statement."
+ "Remote work has changed the way teams collaborate, with online tools replacing traditional office routines and in-person meetings."
+ ]
- model_path.mkdir(exist_ok=True, parents=True)
+ cosine_similarity = {
+ torch.float16: [],
+ torch.float8_e4m3fn: [],
+ torch.float8_e5m2: [],
+ }
+
+ for phrase in phrases:
+ encoding: Encoding = tokenizer.encode(phrase)
+ embedded_phrase = embeddings[torch.tensor(encoding.ids, dtype=torch.long)]
+
+ for dtype in cosine_similarity.keys():
+ pooling_unquantized = normalized_mean_pooling(embedded_phrase)
+ pooling_roundtrip = normalized_mean_pooling(
+ embedded_phrase.to(dtype).to(torch.float32)
+ )
+ cosine = torch.dot(pooling_unquantized, pooling_roundtrip).item()
+ cosine_similarity[dtype].append(cosine)
+
+ avg_cosine_similarity = {
+ dtype: sum(values) / len(values) for dtype, values in cosine_similarity.items()
+ }
+
+ tokenizer_examples = []
+ for text in [
+ "This is an example of encoding",
+ "The quick brown fox jumps over the lazy dog.",
+ "Curaçao, naïve fiancé, jalapeño, déjà vu.",
+ "Привет, как дела?",
+ "Бързата кафява лисица прескача мързеливото куче.",
+ "Γρήγορη καφέ αλεπού πηδάει πάνω από τον τεμπέλη σκύλο.",
+ "اللغة العربية جميلة وغنية بالتاريخ.",
+ "مرحبا بالعالم!",
+ "Simplified: 快速的棕色狐狸跳过懒狗。",
+ "Traditional: 快速的棕色狐狸跳過懶狗。",
+ "素早い茶色の狐が怠け者の犬を飛び越える。",
+ "コンピュータープログラミング",
+ "빠른 갈색 여우가 게으른 개를 뛰어넘습니다.",
+ "तेज़ भूरी लोमड़ी आलसी कुत्ते के ऊपर कूदती है।",
+ "দ্রুত বাদামী শিয়াল অলস কুকুরের উপর দিয়ে লাফ দেয়।",
+ "வேகமான பழுப்பு நரி சோம்பேறி நாயின் மேல் குதிக்கிறது.",
+ "สุนัขจิ้งจอกสีน้ำตาลกระโดดข้ามสุนัขขี้เกียจ.",
+ "ብሩክ ቡናማ ቀበሮ ሰነፍ ውሻን ተዘልሏል።",
+ "Hello 世界 مرحبا 🌍",
+ "123, αβγ, абв, العربية, 中文, हिन्दी.",
+ ]:
+ encoding = tokenizer.encode(text)
+ tokens = [f"`{token}`" for token in encoding.tokens]
+
+ tokenizer_examples.append(f"**Input:** {text}
")
+ tokenizer_examples.append(f"**Tokens**: {' '.join(tokens)}")
+ tokenizer_examples.append("")
+
+ tokenizer_output = "\n".join(tokenizer_examples)
+
+ with (model_card.path() / "README.md").open("wt") as file:
+ prefix = " "
- with (model_path / "README.md").open("wt") as file:
file.write(
dedent(
f"""
- # [{hf_org}/{hf_repo}](https://huggingface.co/{hf_org}/{hf_repo})
+ # [{model_card.name()}](https://huggingface.co/{model_card.name()})
+
+ License: [{model_card.license}](https://choosealicense.com/licenses/{model_card.license}/)
- Beyond the vocab size and embedding size, these are stats for the length
- of the vectors and the distribution of the values.
+ {indent(model_card.get_description(), prefix).strip()}
+
+ ## Model Stats
+
+ Stats that describe the embeddings tensor shapes and value distribution.
| item | metric | value |
| --------------| ----------------------- | ----- |
@@ -134,127 +247,152 @@ def export_embeddings(
| values | median | {embeddings.median().item():.2f} |
| values | stddev | {embeddings.std().item():.2f} |
- ## Quantization Loss
-
- | Precision | Metric | Value |
- | ------------- | ------ | ----- |
- | fp16 | mse | {quantization_loss_mse(embeddings, torch.float16):.5f} |
- | fp8 e4m3 | mse | {quantization_loss_mse(embeddings, torch.float8_e4m3fn):.5f} |
- | fp8 e5m2 | mse | {quantization_loss_mse(embeddings, torch.float8_e5m2):.5f} |
- | fp16 | mae | {quantization_loss_mae(embeddings, torch.float16):.5f} |
- | fp8 e4m3 | mae | {quantization_loss_mae(embeddings, torch.float8_e4m3fn):.5f} |
- | fp8 e5m2 | mae | {quantization_loss_mae(embeddings, torch.float8_e5m2):.5f} |
- | fp16 | cosine | {quantization_loss_cosine(embeddings, torch.float16):.5f} |
- | fp8 e4m3 | cosine | {quantization_loss_cosine(embeddings, torch.float8_e4m3fn):.5f} |
- | fp8 e5m2 | cosine | {quantization_loss_cosine(embeddings, torch.float8_e5m2):.5f} |
-
- When embeddings are quantized to lower precision (e.g. FP8) and then dequantized
- back to `float32`, some information is inevitably lost. To measure how much the
- quantized embeddings differ from the originals, we report three complementary
- metrics:
-
+ ## Mean Pooled Quantization Loss
+
+ This test roundtrips the vectors through quantization, but performs the
+ mean pooling arithmetic in float32 space. The quantized and unquantized
+ mean pooled vectors are compared to each other to determine their cosine
+ similarity, to show how much the meaning of the vector has changed due
+ to quantization.
+
+ | Precision | Cosine Similarity |
+ | ------------- | ----------------- |
+ | fp16 | {avg_cosine_similarity[torch.float16]:.5f} |
+ | fp8 e4m3 | {avg_cosine_similarity[torch.float8_e4m3fn]:.5f} |
+ | fp8 e5m2 | {avg_cosine_similarity[torch.float8_e5m2]:.5f} |
+
+ ## Quantization Loss Per Vector
+
+ While ultimately the embedding vectors will be mean pooled together, it's
+ still useful to look at the loss per-vector in the embedding table to see
+ which quantization strategies retain the most vector meaning.
+
+ - **Cosine Similarity** — measures how well the *direction* of embedding vectors
+ is preserved after quantization, independent of scale. This is especially
+ relevant when embeddings are used for similarity search or retrieval.
- **MSE (Mean Squared Error)** — emphasizes large errors by squaring the
differences. Useful for detecting whether any values are badly distorted.
- **MAE (Mean Absolute Error)** — the average absolute difference between
original and quantized values. Easier to interpret, less sensitive to outliers.
- - **Cosine Similarity** — measures how well the *direction* of embedding vectors
- is preserved after quantization, independent of scale. This is especially
- relevant when embeddings are used for similarity search or retrieval.
- Together, these metrics provide a more complete picture of quantization quality
- than any one alone.
+ | Precision | Metric | Value |
+ | ------------- | ------ | ----- |
+ | fp16 | cosine similarity | {quantization_loss_cosine(embeddings, torch.float16):.5f} |
+ | fp8 e4m3 | cosine similarity | {quantization_loss_cosine(embeddings, torch.float8_e4m3fn):.5f} |
+ | fp8 e5m2 | cosine similarity | {quantization_loss_cosine(embeddings, torch.float8_e5m2):.5f} |
+ | fp16 | MSE | {quantization_loss_mse(embeddings, torch.float16):.5f} |
+ | fp8 e4m3 | MSE | {quantization_loss_mse(embeddings, torch.float8_e4m3fn):.5f} |
+ | fp8 e5m2 | MSE | {quantization_loss_mse(embeddings, torch.float8_e5m2):.5f} |
+ | fp16 | MAE | {quantization_loss_mae(embeddings, torch.float16):.5f} |
+ | fp8 e4m3 | MAE | {quantization_loss_mae(embeddings, torch.float8_e4m3fn):.5f} |
+ | fp8 e5m2 | MAE | {quantization_loss_mae(embeddings, torch.float8_e5m2):.5f} |
- ### Interpreting Quantization Loss
-
- - **Cosine similarity** is the most important metric for embedding use-cases
- such as similarity search, clustering, or retrieval. Values close to 1.0
- mean that embedding directions are preserved after quantization, so model
- quality is likely to hold up.
-
- - **MSE and MAE** measure raw element-wise reconstruction error. They provide
- a sense of how much the numerical values change, but these shifts often have
- limited impact on cosine similarity once embeddings are pooled and
- normalized.
-
- - **FP16** is effectively lossless and can be treated as a baseline.
-
- - **FP8 E4M3** typically offers better precision (lower MSE/MAE) when values
- stay within a moderate range, making it a strong default for static
- embeddings.
-
- - **FP8 E5M2** trades some precision for greater dynamic range. It can be
- preferable if embeddings occasionally contain very large values, but it will
- usually show higher MSE/MAE than E4M3.
-
- In practice, if cosine similarity remains very close to 1.0, quantization is
- unlikely to harm downstream tasks, even if MSE/MAE look relatively large.
+ ## Tokenizer Examples
+
+ {indent(tokenizer_output, prefix).strip()}
"""
).strip()
)
- for dim in (1024, 512, 384, 256, 128):
- if dim > dimensions:
- print(f"Skipping output of {dim} as the max dimension is {dimensions}")
- continue
-
- truncated = embeddings[:, :dim]
- assert truncated.shape == torch.Size([vocab_size, dim])
-
- save_data(model_path / f"fp32.d{dim}.npy", truncated)
- save_data(
- model_path / f"fp16.d{dim}.npy",
- truncated.to(dtype=torch.float16),
- )
- save_data(
- model_path / f"fp8_e5m2.d{dim}.npy",
- truncated.to(dtype=torch.float8_e5m2),
- )
- save_data(
- model_path / f"fp8_e4m3.d{dim}.npy",
- truncated.to(dtype=torch.float8_e4m3fn),
- )
-
-def export_tokenizer(model_path: Path, tokenizer: Tokenizer) -> None:
- tokenizer_path = model_path / "tokenizer.json"
+def export_tokenizer(model_card: ModelCard, tokenizer: Tokenizer) -> None:
+ tokenizer_path = model_card.path() / "tokenizer.json"
print(f"Exporting tokenizer: {tokenizer_path}")
tokenizer.save(str(tokenizer_path))
zst_compress_file(tokenizer_path)
-def export_sentence_transformers(hf_org: str, hf_repo: str) -> None:
+def export_sentence_transformers(model_card: ModelCard) -> None:
"""Extract the embeddings and tokenizer from SentenceTransformers"""
- model_name = f"{hf_org}/{hf_repo}"
- print("Processing", model_name)
+ print("Processing", model_card.name())
- model = SentenceTransformer(f"{hf_org}/{hf_repo}", device="cpu")
+ model = SentenceTransformer(model_card.name(), device="cpu")
embedding_bag: EmbeddingBag = model[0].embedding # type: ignore
- model_path = models_path / hf_org / hf_repo
+ model_card.path().mkdir(exist_ok=True, parents=True)
+ embeddings = torch.Tensor(embedding_bag.weight)
- export_embeddings(hf_org, hf_repo, model_path, torch.Tensor(embedding_bag.weight))
- export_tokenizer(model_path, model.tokenizer)
+ export_embeddings(model_card, embeddings)
+ export_tokenizer(model_card, model.tokenizer)
+ export_readme(model_card, embeddings, model.tokenizer)
-def export_model2vec(hf_org: str, hf_repo: str) -> None:
+def export_model2vec(model_card: ModelCard) -> None:
"""Extract the embeddings and tokenizer from model2vec"""
- model = StaticModel.from_pretrained("minishlab/potion-multilingual-128M")
- model_path = models_path / hf_org / hf_repo
- export_embeddings(hf_org, hf_repo, model_path, torch.from_numpy(model.embedding))
- export_tokenizer(model_path, model.tokenizer)
+ print("Processing", model_card.name())
+
+ model = StaticModel.from_pretrained(model_card.name())
+ model_card.path().mkdir(exist_ok=True, parents=True)
+ embeddings = torch.from_numpy(model.embedding)
+ export_embeddings(model_card, embeddings)
+ export_tokenizer(model_card, model.tokenizer)
+ export_readme(model_card, embeddings, model.tokenizer)
def main() -> None:
# Static embedders that use sentence_transformers models.
sentence_transformers_models = [
- ("sentence-transformers", "static-similarity-mrl-multilingual-v1"),
- ("sentence-transformers", "static-retrieval-mrl-en-v1"),
+ ModelCard(
+ owner="sentence-transformers",
+ repo="static-similarity-mrl-multilingual-v1",
+ description="""
+ Multi-lingual similarity embeddings that were trained with Matroyshka loss
+ that allows for more effective truncation of the embedding vectors. It
+ was trained on a variety of domains of multilingual datasets.
+
+ It's a general purpose model that can be used for semantic textual similarity,
+ paraphrase mining, text classification, clustering, and more
+ """,
+ matroyshka_dims=[32, 64, 128, 256, 512, 1024],
+ license="apache-2.0",
+ ),
+ ModelCard(
+ owner="sentence-transformers",
+ repo="static-retrieval-mrl-en-v1",
+ description="""
+ English-only uncased similarity embeddings that were trained with Matroyshka
+ loss that allows for more effective truncation of the embedding vectors. It
+ was trained on a variety of domains of monolingual datasets. I was designed
+ specifically for similarity retrieval.
+ """,
+ matroyshka_dims=[32, 64, 128, 256, 512, 1024],
+ license="apache-2.0",
+ ),
]
# Static embedders that use model2vec.
model2vec_models = [
- ("minishlab", "potion-multilingual-128M"),
- ("minishlab", "potion-retrieval-32M"),
+ ModelCard(
+ owner="minishlab",
+ repo="potion-multilingual-128M",
+ # These are assumed as their is no python reference implementation:
+ matroyshka_dims=[32, 64, 128, 256],
+ description="""
+ A multilingual embedder. The details are a bit scant on how it's trained as
+ there is no source code for it. However, it's likely a close architecture
+ to the potion-retrieval-32M model, but trained on Common Crawl data.
+
+ The 128M references the number of parameters in the embeddings:
+
+ 256 dimensions * 500,353 vocab.
+ """,
+ license="mit",
+ ),
+ ModelCard(
+ owner="minishlab",
+ repo="potion-retrieval-32M",
+ matroyshka_dims=[32, 64, 128, 256, 512],
+ description="""
+ The token embeddings from a monolingual English 32M parameter model that was
+ distilled from embeddings that were initialized from the the multi-domain
+ [BAAI/bge-base-en-v1.5](https://huggingface.co/BAAI/bge-base-en-v1.5)
+
+ The 32M references the number of parameters in the embeddings:
+
+ 512 dimension * 63,091 vocab.
+ """,
+ license="mit",
+ ),
]
if models_path.exists():
@@ -262,11 +400,11 @@ def main() -> None:
shutil.rmtree(models_path)
models_path.mkdir()
- for hf_org, hf_repo in sentence_transformers_models:
- export_sentence_transformers(hf_org, hf_repo)
+ for model_card in sentence_transformers_models:
+ export_sentence_transformers(model_card)
- for hf_org, hf_repo in model2vec_models:
- export_model2vec(hf_org, hf_repo)
+ for model_card in model2vec_models:
+ export_model2vec(model_card)
if __name__ == "__main__":