Skip to content

Commit 06ebcd1

Browse files
NirantKNirantK
committed
Remove unused models from supported_onnx_models
1 parent 558a837 commit 06ebcd1

File tree

4 files changed

+52
-154
lines changed

4 files changed

+52
-154
lines changed

‎docs/examples/Supported_Models.ipynb

+50-96
Original file line numberDiff line numberDiff line change
@@ -1,15 +1,5 @@
11
{
22
"cells": [
3-
{
4-
"cell_type": "code",
5-
"execution_count": 1,
6-
"metadata": {},
7-
"outputs": [],
8-
"source": [
9-
"%load_ext autoreload\n",
10-
"%autoreload 2"
11-
]
12-
},
133
{
144
"cell_type": "code",
155
"execution_count": 4,
@@ -46,94 +36,70 @@
4636
" <tbody>\n",
4737
" <tr>\n",
4838
" <th>0</th>\n",
49-
" <td>BAAI/bge-base-en</td>\n",
50-
" <td>768</td>\n",
51-
" <td>Base English model</td>\n",
52-
" <td>0.50</td>\n",
53-
" <td>{'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en.tar.gz'}</td>\n",
54-
" </tr>\n",
55-
" <tr>\n",
56-
" <th>1</th>\n",
5739
" <td>BAAI/bge-base-en-v1.5</td>\n",
5840
" <td>768</td>\n",
5941
" <td>Base English model, v1.5</td>\n",
6042
" <td>0.44</td>\n",
6143
" <td>{'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en-v1.5.tar.gz', 'hf': 'qdrant/bge-base-en-v1.5-onnx-q'}</td>\n",
6244
" </tr>\n",
6345
" <tr>\n",
64-
" <th>2</th>\n",
65-
" <td>BAAI/bge-large-en-v1.5-quantized</td>\n",
66-
" <td>1024</td>\n",
67-
" <td>Large English model, v1.5</td>\n",
68-
" <td>1.34</td>\n",
69-
" <td>{'hf': 'qdrant/bge-large-en-v1.5-onnx-q'}</td>\n",
70-
" </tr>\n",
71-
" <tr>\n",
72-
" <th>3</th>\n",
46+
" <th>1</th>\n",
7347
" <td>BAAI/bge-large-en-v1.5</td>\n",
7448
" <td>1024</td>\n",
7549
" <td>Large English model, v1.5</td>\n",
7650
" <td>1.34</td>\n",
77-
" <td>{'hf': 'qdrant/bge-large-en-v1.5-onnx'}</td>\n",
78-
" </tr>\n",
79-
" <tr>\n",
80-
" <th>4</th>\n",
81-
" <td>BAAI/bge-small-en</td>\n",
82-
" <td>384</td>\n",
83-
" <td>Fast English model</td>\n",
84-
" <td>0.20</td>\n",
85-
" <td>{'url': 'https://storage.googleapis.com/qdrant-fastembed/BAAI-bge-small-en.tar.gz'}</td>\n",
51+
" <td>{'hf': 'qdrant/bge-large-en-v1.5-onnx-q'}</td>\n",
8652
" </tr>\n",
8753
" <tr>\n",
88-
" <th>5</th>\n",
54+
" <th>2</th>\n",
8955
" <td>BAAI/bge-small-en-v1.5</td>\n",
9056
" <td>384</td>\n",
9157
" <td>Fast and Default English model</td>\n",
9258
" <td>0.13</td>\n",
9359
" <td>{'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-small-en-v1.5.tar.gz', 'hf': 'qdrant/bge-small-en-v1.5-onnx-q'}</td>\n",
9460
" </tr>\n",
9561
" <tr>\n",
96-
" <th>6</th>\n",
62+
" <th>3</th>\n",
9763
" <td>BAAI/bge-small-zh-v1.5</td>\n",
9864
" <td>512</td>\n",
9965
" <td>Fast and recommended Chinese model</td>\n",
10066
" <td>0.10</td>\n",
10167
" <td>{'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-small-zh-v1.5.tar.gz'}</td>\n",
10268
" </tr>\n",
10369
" <tr>\n",
104-
" <th>7</th>\n",
70+
" <th>4</th>\n",
10571
" <td>sentence-transformers/all-MiniLM-L6-v2</td>\n",
10672
" <td>384</td>\n",
10773
" <td>Sentence Transformer model, MiniLM-L6-v2</td>\n",
10874
" <td>0.09</td>\n",
10975
" <td>{'url': 'https://storage.googleapis.com/qdrant-fastembed/sentence-transformers-all-MiniLM-L6-v2.tar.gz', 'hf': 'qdrant/all-MiniLM-L6-v2-onnx'}</td>\n",
11076
" </tr>\n",
11177
" <tr>\n",
112-
" <th>8</th>\n",
78+
" <th>5</th>\n",
11379
" <td>intfloat/multilingual-e5-large</td>\n",
11480
" <td>1024</td>\n",
11581
" <td>Multilingual model, e5-large. Recommend using this model for non-English languages</td>\n",
11682
" <td>2.24</td>\n",
11783
" <td>{'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-multilingual-e5-large.tar.gz', 'hf': 'qdrant/multilingual-e5-large-onnx'}</td>\n",
11884
" </tr>\n",
11985
" <tr>\n",
120-
" <th>9</th>\n",
86+
" <th>6</th>\n",
12187
" <td>sentence-transformers/paraphrase-multilingual-mpnet-base-v2</td>\n",
12288
" <td>768</td>\n",
12389
" <td>Sentence-transformers model for tasks like clustering or semantic search</td>\n",
12490
" <td>1.11</td>\n",
12591
" <td>{'hf': 'xenova/paraphrase-multilingual-mpnet-base-v2'}</td>\n",
12692
" </tr>\n",
12793
" <tr>\n",
128-
" <th>10</th>\n",
94+
" <th>7</th>\n",
12995
" <td>jinaai/jina-embeddings-v2-base-en</td>\n",
13096
" <td>768</td>\n",
13197
" <td>English embedding model supporting 8192 sequence length</td>\n",
13298
" <td>0.55</td>\n",
13399
" <td>{'hf': 'xenova/jina-embeddings-v2-base-en'}</td>\n",
134100
" </tr>\n",
135101
" <tr>\n",
136-
" <th>11</th>\n",
102+
" <th>8</th>\n",
137103
" <td>jinaai/jina-embeddings-v2-small-en</td>\n",
138104
" <td>512</td>\n",
139105
" <td>English embedding model supporting 8192 sequence length</td>\n",
@@ -145,61 +111,49 @@
145111
"</div>"
146112
],
147113
"text/plain": [
148-
" model dim \\\n",
149-
"0 BAAI/bge-base-en 768 \n",
150-
"1 BAAI/bge-base-en-v1.5 768 \n",
151-
"2 BAAI/bge-large-en-v1.5-quantized 1024 \n",
152-
"3 BAAI/bge-large-en-v1.5 1024 \n",
153-
"4 BAAI/bge-small-en 384 \n",
154-
"5 BAAI/bge-small-en-v1.5 384 \n",
155-
"6 BAAI/bge-small-zh-v1.5 512 \n",
156-
"7 sentence-transformers/all-MiniLM-L6-v2 384 \n",
157-
"8 intfloat/multilingual-e5-large 1024 \n",
158-
"9 sentence-transformers/paraphrase-multilingual-mpnet-base-v2 768 \n",
159-
"10 jinaai/jina-embeddings-v2-base-en 768 \n",
160-
"11 jinaai/jina-embeddings-v2-small-en 512 \n",
114+
" model dim \\\n",
115+
"0 BAAI/bge-base-en-v1.5 768 \n",
116+
"1 BAAI/bge-large-en-v1.5 1024 \n",
117+
"2 BAAI/bge-small-en-v1.5 384 \n",
118+
"3 BAAI/bge-small-zh-v1.5 512 \n",
119+
"4 sentence-transformers/all-MiniLM-L6-v2 384 \n",
120+
"5 intfloat/multilingual-e5-large 1024 \n",
121+
"6 sentence-transformers/paraphrase-multilingual-mpnet-base-v2 768 \n",
122+
"7 jinaai/jina-embeddings-v2-base-en 768 \n",
123+
"8 jinaai/jina-embeddings-v2-small-en 512 \n",
161124
"\n",
162-
" description \\\n",
163-
"0 Base English model \n",
164-
"1 Base English model, v1.5 \n",
165-
"2 Large English model, v1.5 \n",
166-
"3 Large English model, v1.5 \n",
167-
"4 Fast English model \n",
168-
"5 Fast and Default English model \n",
169-
"6 Fast and recommended Chinese model \n",
170-
"7 Sentence Transformer model, MiniLM-L6-v2 \n",
171-
"8 Multilingual model, e5-large. Recommend using this model for non-English languages \n",
172-
"9 Sentence-transformers model for tasks like clustering or semantic search \n",
173-
"10 English embedding model supporting 8192 sequence length \n",
174-
"11 English embedding model supporting 8192 sequence length \n",
125+
" description \\\n",
126+
"0 Base English model, v1.5 \n",
127+
"1 Large English model, v1.5 \n",
128+
"2 Fast and Default English model \n",
129+
"3 Fast and recommended Chinese model \n",
130+
"4 Sentence Transformer model, MiniLM-L6-v2 \n",
131+
"5 Multilingual model, e5-large. Recommend using this model for non-English languages \n",
132+
"6 Sentence-transformers model for tasks like clustering or semantic search \n",
133+
"7 English embedding model supporting 8192 sequence length \n",
134+
"8 English embedding model supporting 8192 sequence length \n",
175135
"\n",
176-
" size_in_GB \\\n",
177-
"0 0.50 \n",
178-
"1 0.44 \n",
179-
"2 1.34 \n",
180-
"3 1.34 \n",
181-
"4 0.20 \n",
182-
"5 0.13 \n",
183-
"6 0.10 \n",
184-
"7 0.09 \n",
185-
"8 2.24 \n",
186-
"9 1.11 \n",
187-
"10 0.55 \n",
188-
"11 0.13 \n",
136+
" size_in_GB \\\n",
137+
"0 0.44 \n",
138+
"1 1.34 \n",
139+
"2 0.13 \n",
140+
"3 0.10 \n",
141+
"4 0.09 \n",
142+
"5 2.24 \n",
143+
"6 1.11 \n",
144+
"7 0.55 \n",
145+
"8 0.13 \n",
189146
"\n",
190-
" sources \n",
191-
"0 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en.tar.gz'} \n",
192-
"1 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en-v1.5.tar.gz', 'hf': 'qdrant/bge-base-en-v1.5-onnx-q'} \n",
193-
"2 {'hf': 'qdrant/bge-large-en-v1.5-onnx-q'} \n",
194-
"3 {'hf': 'qdrant/bge-large-en-v1.5-onnx'} \n",
195-
"4 {'url': 'https://storage.googleapis.com/qdrant-fastembed/BAAI-bge-small-en.tar.gz'} \n",
196-
"5 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-small-en-v1.5.tar.gz', 'hf': 'qdrant/bge-small-en-v1.5-onnx-q'} \n",
197-
"6 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-small-zh-v1.5.tar.gz'} \n",
198-
"7 {'url': 'https://storage.googleapis.com/qdrant-fastembed/sentence-transformers-all-MiniLM-L6-v2.tar.gz', 'hf': 'qdrant/all-MiniLM-L6-v2-onnx'} \n",
199-
"8 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-multilingual-e5-large.tar.gz', 'hf': 'qdrant/multilingual-e5-large-onnx'} \n",
200-
"9 {'hf': 'xenova/paraphrase-multilingual-mpnet-base-v2'} \n",
201-
"10 {'hf': 'xenova/jina-embeddings-v2-base-en'} \n",
202-
"11 {'hf': 'xenova/jina-embeddings-v2-small-en'} "
147+
" sources \n",
148+
"0 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en-v1.5.tar.gz', 'hf': 'qdrant/bge-base-en-v1.5-onnx-q'} \n",
149+
"1 {'hf': 'qdrant/bge-large-en-v1.5-onnx-q'} \n",
150+
"2 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-small-en-v1.5.tar.gz', 'hf': 'qdrant/bge-small-en-v1.5-onnx-q'} \n",
151+
"3 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-bge-small-zh-v1.5.tar.gz'} \n",
152+
"4 {'url': 'https://storage.googleapis.com/qdrant-fastembed/sentence-transformers-all-MiniLM-L6-v2.tar.gz', 'hf': 'qdrant/all-MiniLM-L6-v2-onnx'} \n",
153+
"5 {'url': 'https://storage.googleapis.com/qdrant-fastembed/fast-multilingual-e5-large.tar.gz', 'hf': 'qdrant/multilingual-e5-large-onnx'} \n",
154+
"6 {'hf': 'xenova/paraphrase-multilingual-mpnet-base-v2'} \n",
155+
"7 {'hf': 'xenova/jina-embeddings-v2-base-en'} \n",
156+
"8 {'hf': 'xenova/jina-embeddings-v2-small-en'} "
203157
]
204158
},
205159
"execution_count": 4,
@@ -232,7 +186,7 @@
232186
"name": "python",
233187
"nbconvert_exporter": "python",
234188
"pygments_lexer": "ipython3",
235-
"version": "3.11.5"
189+
"version": "3.10.13"
236190
},
237191
"orig_nbformat": 4
238192
},

‎fastembed/text/onnx_embedding.py

+2-53
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
import os
22
from multiprocessing import get_all_start_methods
3-
from typing import List, Dict, Any, Tuple, Union, Iterable, Type
3+
from typing import Any, Dict, Iterable, List, Tuple, Type, Union
44

55
import numpy as np
66
import onnxruntime as ort
@@ -12,15 +12,6 @@
1212
from fastembed.text.text_embedding_base import TextEmbeddingBase
1313

1414
supported_onnx_models = [
15-
{
16-
"model": "BAAI/bge-base-en",
17-
"dim": 768,
18-
"description": "Base English model",
19-
"size_in_GB": 0.5,
20-
"sources": {
21-
"url": "https://storage.googleapis.com/qdrant-fastembed/fast-bge-base-en.tar.gz",
22-
},
23-
},
2415
{
2516
"model": "BAAI/bge-base-en-v1.5",
2617
"dim": 768,
@@ -31,44 +22,15 @@
3122
"hf": "qdrant/bge-base-en-v1.5-onnx-q",
3223
},
3324
},
34-
{
35-
"model": "BAAI/bge-large-en-v1.5-quantized",
36-
"dim": 1024,
37-
"description": "Large English model, v1.5",
38-
"size_in_GB": 1.34,
39-
"sources": {
40-
"hf": "qdrant/bge-large-en-v1.5-onnx-q",
41-
},
42-
},
4325
{
4426
"model": "BAAI/bge-large-en-v1.5",
4527
"dim": 1024,
4628
"description": "Large English model, v1.5",
4729
"size_in_GB": 1.34,
4830
"sources": {
49-
"hf": "qdrant/bge-large-en-v1.5-onnx",
50-
},
51-
},
52-
{
53-
"model": "BAAI/bge-small-en",
54-
"dim": 384,
55-
"description": "Fast English model",
56-
"size_in_GB": 0.2,
57-
"sources": {
58-
"url": "https://storage.googleapis.com/qdrant-fastembed/BAAI-bge-small-en.tar.gz",
31+
"hf": "qdrant/bge-large-en-v1.5-onnx-q",
5932
},
6033
},
61-
# {
62-
# "model": "BAAI/bge-small-en",
63-
# "dim": 384,
64-
# "description": "Fast English model",
65-
# "size_in_GB": 0.2,
66-
# "hf_sources": [],
67-
# "compressed_url_sources": [
68-
# "https://storage.googleapis.com/qdrant-fastembed/fast-bge-small-en.tar.gz",
69-
# "https://storage.googleapis.com/qdrant-fastembed/BAAI-bge-small-en.tar.gz"
70-
# ]
71-
# },
7234
{
7335
"model": "BAAI/bge-small-en-v1.5",
7436
"dim": 384,
@@ -98,19 +60,6 @@
9860
"hf": "qdrant/all-MiniLM-L6-v2-onnx",
9961
},
10062
},
101-
# {
102-
# "model": "sentence-transformers/all-MiniLM-L6-v2",
103-
# "dim": 384,
104-
# "description": "Sentence Transformer model, MiniLM-L6-v2",
105-
# "size_in_GB": 0.09,
106-
# "hf_sources": [
107-
# "qdrant/all-MiniLM-L6-v2-onnx"
108-
# ],
109-
# "compressed_url_sources": [
110-
# "https://storage.googleapis.com/qdrant-fastembed/fast-all-MiniLM-L6-v2.tar.gz",
111-
# "https://storage.googleapis.com/qdrant-fastembed/sentence-transformers-all-MiniLM-L6-v2.tar.gz"
112-
# ]
113-
# }
11463
]
11564

11665

‎tests/test_onnx_embeddings.py

-2
Original file line numberDiff line numberDiff line change
@@ -6,10 +6,8 @@
66
from fastembed.embedding import DefaultEmbedding, JinaEmbedding
77

88
CANONICAL_VECTOR_VALUES = {
9-
"BAAI/bge-small-en": np.array([-0.0232, -0.0255, 0.0174, -0.0639, -0.0006]),
109
"BAAI/bge-small-en-v1.5": np.array([0.01522374, -0.02271799, 0.00860278, -0.07424029, 0.00386434]),
1110
"BAAI/bge-small-zh-v1.5": np.array([-0.01023294, 0.07634465, 0.0691722, -0.04458365, -0.03160762]),
12-
"BAAI/bge-base-en": np.array([0.0115, 0.0372, 0.0295, 0.0121, 0.0346]),
1311
"BAAI/bge-base-en-v1.5": np.array([0.01129394, 0.05493144, 0.02615099, 0.00328772, 0.02996045]),
1412
"BAAI/bge-large-en-v1.5": np.array([0.03434538, 0.03316108, 0.02191251, -0.03713358, -0.01577825]),
1513
"sentence-transformers/all-MiniLM-L6-v2": np.array([0.0259, 0.0058, 0.0114, 0.0380, -0.0233]),

‎tests/test_text_onnx_embeddings.py

-3
Original file line numberDiff line numberDiff line change
@@ -6,13 +6,10 @@
66
from fastembed.text.text_embedding import TextEmbedding
77

88
CANONICAL_VECTOR_VALUES = {
9-
"BAAI/bge-small-en": np.array([-0.0232, -0.0255, 0.0174, -0.0639, -0.0006]),
109
"BAAI/bge-small-en-v1.5": np.array([0.01522374, -0.02271799, 0.00860278, -0.07424029, 0.00386434]),
1110
"BAAI/bge-small-zh-v1.5": np.array([-0.01023294, 0.07634465, 0.0691722, -0.04458365, -0.03160762]),
12-
"BAAI/bge-base-en": np.array([0.0115, 0.0372, 0.0295, 0.0121, 0.0346]),
1311
"BAAI/bge-base-en-v1.5": np.array([0.01129394, 0.05493144, 0.02615099, 0.00328772, 0.02996045]),
1412
"BAAI/bge-large-en-v1.5": np.array([0.03434538, 0.03316108, 0.02191251, -0.03713358, -0.01577825]),
15-
"BAAI/bge-large-en-v1.5-quantized": np.array([0.03434538, 0.03316108, 0.02191251, -0.03713358, -0.01577825]),
1613
"sentence-transformers/all-MiniLM-L6-v2": np.array([0.0259, 0.0058, 0.0114, 0.0380, -0.0233]),
1714
"intfloat/multilingual-e5-large": np.array([0.0098, 0.0045, 0.0066, -0.0354, 0.0070]),
1815
"sentence-transformers/paraphrase-multilingual-mpnet-base-v2": np.array(

0 commit comments

Comments
 (0)