Hugging Face's logo

sparse-encoder
/
example-splade-distilbert-base-uncased-nq

Feature Extraction
sentence-transformers
Safetensors
English
distilbert
sparse-encoder
sparse
splade
Generated from Trainer
dataset_size:99000
loss:SpladeLoss
loss:SparseMultipleNegativesRankingLoss
loss:FlopsLoss
Eval Results
text-embeddings-inference
Model card Files Community
arthurbresnu HF Staff commited on
Commit
240a804
·
verified ·
1 Parent(s): b124ad7

Add new SparseEncoder model

Browse files
Files changed (11) hide show
  1. 1_SpladePooling/config.json +5 -0
  2. README.md +1762 -0
  3. config.json +23 -0
  4. config_sentence_transformers.json +11 -0
  5. model.safetensors +3 -0
  6. modules.json +14 -0
  7. sentence_bert_config.json +4 -0
  8. special_tokens_map.json +7 -0
  9. tokenizer.json +0 -0
  10. tokenizer_config.json +56 -0
  11. vocab.txt +0 -0
1_SpladePooling/config.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "pooling_strategy": "max",
3
+ "activation_function": "relu",
4
+ "word_embedding_dimension": 30522
5
+ }
README.md ADDED
@@ -0,0 +1,1762 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ license: apache-2.0
5
+ tags:
6
+ - sentence-transformers
7
+ - sparse-encoder
8
+ - sparse
9
+ - splade
10
+ - generated_from_trainer
11
+ - dataset_size:99000
12
+ - loss:SpladeLoss
13
+ - loss:SparseMultipleNegativesRankingLoss
14
+ - loss:FlopsLoss
15
+ base_model: distilbert/distilbert-base-uncased
16
+ widget:
17
+ - text: Rollin' (Limp Bizkit song) The music video was filmed atop the South Tower
18
+ of the former World Trade Center in New York City. The introduction features Ben
19
+ Stiller and Stephen Dorff mistaking Fred Durst for the valet and giving him the
20
+ keys to their Bentley Azure. Also making a cameo is break dancer Mr. Wiggles.
21
+ The rest of the video has several cuts to Durst and his bandmates hanging out
22
+ of the Bentley as they drive about Manhattan. The song Ben Stiller is playing
23
+ at the beginning is "My Generation" from the same album. The video also features
24
+ scenes of Fred Durst with five girls dancing in a room. The video was filmed around
25
+ the same time as the film Zoolander, which explains Stiller and Dorff's appearance.
26
+ Fred Durst has a small cameo in that film.
27
+ - text: 'Maze Runner: The Death Cure On April 22, 2017, the studio delayed the release
28
+ date once again, to February 9, 2018, in order to allow more time for post-production;
29
+ months later, on August 25, the studio moved the release forward two weeks.[17]
30
+ The film will premiere on January 26, 2018 in 3D, IMAX and IMAX 3D.[18][19]'
31
+ - text: who played the dj in the movie the warriors
32
+ - text: Lionel Messi Born and raised in central Argentina, Messi was diagnosed with
33
+ a growth hormone deficiency as a child. At age 13, he relocated to Spain to join
34
+ Barcelona, who agreed to pay for his medical treatment. After a fast progression
35
+ through Barcelona's youth academy, Messi made his competitive debut aged 17 in
36
+ October 2004. Despite being injury-prone during his early career, he established
37
+ himself as an integral player for the club within the next three years, finishing
38
+ 2007 as a finalist for both the Ballon d'Or and FIFA World Player of the Year
39
+ award, a feat he repeated the following year. His first uninterrupted campaign
40
+ came in the 2008–09 season, during which he helped Barcelona achieve the first
41
+ treble in Spanish football. At 22 years old, Messi won the Ballon d'Or and FIFA
42
+ World Player of the Year award by record voting margins.
43
+ - text: 'Send In the Clowns "Send In the Clowns" is a song written by Stephen Sondheim
44
+ for the 1973 musical A Little Night Music, an adaptation of Ingmar Bergman''s
45
+ film Smiles of a Summer Night. It is a ballad from Act Two, in which the character
46
+ Desirée reflects on the ironies and disappointments of her life. Among other things,
47
+ she looks back on an affair years earlier with the lawyer Fredrik, who was deeply
48
+ in love with her but whose marriage proposals she had rejected. Meeting him after
49
+ so long, she realizes she is in love with him and finally ready to marry him,
50
+ but now it is he who rejects her: he is in an unconsummated marriage with a much
51
+ younger woman. Desirée proposes marriage to rescue him from this situation, but
52
+ he declines, citing his dedication to his bride. Reacting to his rejection, Desirée
53
+ sings this song. The song is later reprised as a coda after Fredrik''s young wife
54
+ runs away with his son, and Fredrik is finally free to accept Desirée''s offer.[1]'
55
+ datasets:
56
+ - sentence-transformers/natural-questions
57
+ pipeline_tag: feature-extraction
58
+ library_name: sentence-transformers
59
+ metrics:
60
+ - dot_accuracy@1
61
+ - dot_accuracy@3
62
+ - dot_accuracy@5
63
+ - dot_accuracy@10
64
+ - dot_precision@1
65
+ - dot_precision@3
66
+ - dot_precision@5
67
+ - dot_precision@10
68
+ - dot_recall@1
69
+ - dot_recall@3
70
+ - dot_recall@5
71
+ - dot_recall@10
72
+ - dot_ndcg@10
73
+ - dot_mrr@10
74
+ - dot_map@100
75
+ - query_active_dims
76
+ - query_sparsity_ratio
77
+ - corpus_active_dims
78
+ - corpus_sparsity_ratio
79
+ co2_eq_emissions:
80
+ emissions: 2.229070129979357
81
+ energy_consumed: 0.0397771218255029
82
+ source: codecarbon
83
+ training_type: fine-tuning
84
+ on_cloud: false
85
+ cpu_model: AMD Ryzen 9 6900HX with Radeon Graphics
86
+ ram_total_size: 30.6114501953125
87
+ hours_used: 0.322
88
+ hardware_used: 1 x NVIDIA GeForce RTX 3070 Ti Laptop GPU
89
+ model-index:
90
+ - name: splade-distilbert-base-uncased trained on Natural Questions
91
+ results:
92
+ - task:
93
+ type: sparse-information-retrieval
94
+ name: Sparse Information Retrieval
95
+ dataset:
96
+ name: NanoMSMARCO
97
+ type: NanoMSMARCO
98
+ metrics:
99
+ - type: dot_accuracy@1
100
+ value: 0.28
101
+ name: Dot Accuracy@1
102
+ - type: dot_accuracy@3
103
+ value: 0.5
104
+ name: Dot Accuracy@3
105
+ - type: dot_accuracy@5
106
+ value: 0.66
107
+ name: Dot Accuracy@5
108
+ - type: dot_accuracy@10
109
+ value: 0.74
110
+ name: Dot Accuracy@10
111
+ - type: dot_precision@1
112
+ value: 0.28
113
+ name: Dot Precision@1
114
+ - type: dot_precision@3
115
+ value: 0.16666666666666663
116
+ name: Dot Precision@3
117
+ - type: dot_precision@5
118
+ value: 0.132
119
+ name: Dot Precision@5
120
+ - type: dot_precision@10
121
+ value: 0.07400000000000001
122
+ name: Dot Precision@10
123
+ - type: dot_recall@1
124
+ value: 0.28
125
+ name: Dot Recall@1
126
+ - type: dot_recall@3
127
+ value: 0.5
128
+ name: Dot Recall@3
129
+ - type: dot_recall@5
130
+ value: 0.66
131
+ name: Dot Recall@5
132
+ - type: dot_recall@10
133
+ value: 0.74
134
+ name: Dot Recall@10
135
+ - type: dot_ndcg@10
136
+ value: 0.49577037509991184
137
+ name: Dot Ndcg@10
138
+ - type: dot_mrr@10
139
+ value: 0.4185238095238094
140
+ name: Dot Mrr@10
141
+ - type: dot_map@100
142
+ value: 0.4294303031277172
143
+ name: Dot Map@100
144
+ - type: query_active_dims
145
+ value: 52.560001373291016
146
+ name: Query Active Dims
147
+ - type: query_sparsity_ratio
148
+ value: 0.9982779633912164
149
+ name: Query Sparsity Ratio
150
+ - type: corpus_active_dims
151
+ value: 106.13404846191406
152
+ name: Corpus Active Dims
153
+ - type: corpus_sparsity_ratio
154
+ value: 0.9965227033463759
155
+ name: Corpus Sparsity Ratio
156
+ - type: dot_accuracy@1
157
+ value: 0.28
158
+ name: Dot Accuracy@1
159
+ - type: dot_accuracy@3
160
+ value: 0.5
161
+ name: Dot Accuracy@3
162
+ - type: dot_accuracy@5
163
+ value: 0.66
164
+ name: Dot Accuracy@5
165
+ - type: dot_accuracy@10
166
+ value: 0.74
167
+ name: Dot Accuracy@10
168
+ - type: dot_precision@1
169
+ value: 0.28
170
+ name: Dot Precision@1
171
+ - type: dot_precision@3
172
+ value: 0.16666666666666663
173
+ name: Dot Precision@3
174
+ - type: dot_precision@5
175
+ value: 0.132
176
+ name: Dot Precision@5
177
+ - type: dot_precision@10
178
+ value: 0.07400000000000001
179
+ name: Dot Precision@10
180
+ - type: dot_recall@1
181
+ value: 0.28
182
+ name: Dot Recall@1
183
+ - type: dot_recall@3
184
+ value: 0.5
185
+ name: Dot Recall@3
186
+ - type: dot_recall@5
187
+ value: 0.66
188
+ name: Dot Recall@5
189
+ - type: dot_recall@10
190
+ value: 0.74
191
+ name: Dot Recall@10
192
+ - type: dot_ndcg@10
193
+ value: 0.49577037509991184
194
+ name: Dot Ndcg@10
195
+ - type: dot_mrr@10
196
+ value: 0.4185238095238094
197
+ name: Dot Mrr@10
198
+ - type: dot_map@100
199
+ value: 0.4294303031277172
200
+ name: Dot Map@100
201
+ - type: query_active_dims
202
+ value: 52.560001373291016
203
+ name: Query Active Dims
204
+ - type: query_sparsity_ratio
205
+ value: 0.9982779633912164
206
+ name: Query Sparsity Ratio
207
+ - type: corpus_active_dims
208
+ value: 106.13404846191406
209
+ name: Corpus Active Dims
210
+ - type: corpus_sparsity_ratio
211
+ value: 0.9965227033463759
212
+ name: Corpus Sparsity Ratio
213
+ - task:
214
+ type: sparse-information-retrieval
215
+ name: Sparse Information Retrieval
216
+ dataset:
217
+ name: NanoNFCorpus
218
+ type: NanoNFCorpus
219
+ metrics:
220
+ - type: dot_accuracy@1
221
+ value: 0.38
222
+ name: Dot Accuracy@1
223
+ - type: dot_accuracy@3
224
+ value: 0.46
225
+ name: Dot Accuracy@3
226
+ - type: dot_accuracy@5
227
+ value: 0.52
228
+ name: Dot Accuracy@5
229
+ - type: dot_accuracy@10
230
+ value: 0.62
231
+ name: Dot Accuracy@10
232
+ - type: dot_precision@1
233
+ value: 0.38
234
+ name: Dot Precision@1
235
+ - type: dot_precision@3
236
+ value: 0.31999999999999995
237
+ name: Dot Precision@3
238
+ - type: dot_precision@5
239
+ value: 0.292
240
+ name: Dot Precision@5
241
+ - type: dot_precision@10
242
+ value: 0.236
243
+ name: Dot Precision@10
244
+ - type: dot_recall@1
245
+ value: 0.0242331024704017
246
+ name: Dot Recall@1
247
+ - type: dot_recall@3
248
+ value: 0.053060546044216006
249
+ name: Dot Recall@3
250
+ - type: dot_recall@5
251
+ value: 0.07273890139350063
252
+ name: Dot Recall@5
253
+ - type: dot_recall@10
254
+ value: 0.09593681264940912
255
+ name: Dot Recall@10
256
+ - type: dot_ndcg@10
257
+ value: 0.2784960942139155
258
+ name: Dot Ndcg@10
259
+ - type: dot_mrr@10
260
+ value: 0.4393888888888889
261
+ name: Dot Mrr@10
262
+ - type: dot_map@100
263
+ value: 0.11744499575471842
264
+ name: Dot Map@100
265
+ - type: query_active_dims
266
+ value: 62.5
267
+ name: Query Active Dims
268
+ - type: query_sparsity_ratio
269
+ value: 0.9979522967040168
270
+ name: Query Sparsity Ratio
271
+ - type: corpus_active_dims
272
+ value: 126.24652862548828
273
+ name: Corpus Active Dims
274
+ - type: corpus_sparsity_ratio
275
+ value: 0.9958637530756345
276
+ name: Corpus Sparsity Ratio
277
+ - type: dot_accuracy@1
278
+ value: 0.38
279
+ name: Dot Accuracy@1
280
+ - type: dot_accuracy@3
281
+ value: 0.46
282
+ name: Dot Accuracy@3
283
+ - type: dot_accuracy@5
284
+ value: 0.52
285
+ name: Dot Accuracy@5
286
+ - type: dot_accuracy@10
287
+ value: 0.62
288
+ name: Dot Accuracy@10
289
+ - type: dot_precision@1
290
+ value: 0.38
291
+ name: Dot Precision@1
292
+ - type: dot_precision@3
293
+ value: 0.31999999999999995
294
+ name: Dot Precision@3
295
+ - type: dot_precision@5
296
+ value: 0.292
297
+ name: Dot Precision@5
298
+ - type: dot_precision@10
299
+ value: 0.236
300
+ name: Dot Precision@10
301
+ - type: dot_recall@1
302
+ value: 0.0242331024704017
303
+ name: Dot Recall@1
304
+ - type: dot_recall@3
305
+ value: 0.053060546044216006
306
+ name: Dot Recall@3
307
+ - type: dot_recall@5
308
+ value: 0.07273890139350063
309
+ name: Dot Recall@5
310
+ - type: dot_recall@10
311
+ value: 0.09593681264940912
312
+ name: Dot Recall@10
313
+ - type: dot_ndcg@10
314
+ value: 0.2784960942139155
315
+ name: Dot Ndcg@10
316
+ - type: dot_mrr@10
317
+ value: 0.4393888888888889
318
+ name: Dot Mrr@10
319
+ - type: dot_map@100
320
+ value: 0.11744499575471842
321
+ name: Dot Map@100
322
+ - type: query_active_dims
323
+ value: 62.5
324
+ name: Query Active Dims
325
+ - type: query_sparsity_ratio
326
+ value: 0.9979522967040168
327
+ name: Query Sparsity Ratio
328
+ - type: corpus_active_dims
329
+ value: 126.24652862548828
330
+ name: Corpus Active Dims
331
+ - type: corpus_sparsity_ratio
332
+ value: 0.9958637530756345
333
+ name: Corpus Sparsity Ratio
334
+ - task:
335
+ type: sparse-information-retrieval
336
+ name: Sparse Information Retrieval
337
+ dataset:
338
+ name: NanoNQ
339
+ type: NanoNQ
340
+ metrics:
341
+ - type: dot_accuracy@1
342
+ value: 0.36
343
+ name: Dot Accuracy@1
344
+ - type: dot_accuracy@3
345
+ value: 0.58
346
+ name: Dot Accuracy@3
347
+ - type: dot_accuracy@5
348
+ value: 0.66
349
+ name: Dot Accuracy@5
350
+ - type: dot_accuracy@10
351
+ value: 0.76
352
+ name: Dot Accuracy@10
353
+ - type: dot_precision@1
354
+ value: 0.36
355
+ name: Dot Precision@1
356
+ - type: dot_precision@3
357
+ value: 0.2
358
+ name: Dot Precision@3
359
+ - type: dot_precision@5
360
+ value: 0.14
361
+ name: Dot Precision@5
362
+ - type: dot_precision@10
363
+ value: 0.08
364
+ name: Dot Precision@10
365
+ - type: dot_recall@1
366
+ value: 0.34
367
+ name: Dot Recall@1
368
+ - type: dot_recall@3
369
+ value: 0.56
370
+ name: Dot Recall@3
371
+ - type: dot_recall@5
372
+ value: 0.64
373
+ name: Dot Recall@5
374
+ - type: dot_recall@10
375
+ value: 0.72
376
+ name: Dot Recall@10
377
+ - type: dot_ndcg@10
378
+ value: 0.5341909779287488
379
+ name: Dot Ndcg@10
380
+ - type: dot_mrr@10
381
+ value: 0.4836666666666667
382
+ name: Dot Mrr@10
383
+ - type: dot_map@100
384
+ value: 0.47986132889672267
385
+ name: Dot Map@100
386
+ - type: query_active_dims
387
+ value: 45.63999938964844
388
+ name: Query Active Dims
389
+ - type: query_sparsity_ratio
390
+ value: 0.9985046851651384
391
+ name: Query Sparsity Ratio
392
+ - type: corpus_active_dims
393
+ value: 104.37854766845703
394
+ name: Corpus Active Dims
395
+ - type: corpus_sparsity_ratio
396
+ value: 0.9965802192625498
397
+ name: Corpus Sparsity Ratio
398
+ - type: dot_accuracy@1
399
+ value: 0.36
400
+ name: Dot Accuracy@1
401
+ - type: dot_accuracy@3
402
+ value: 0.58
403
+ name: Dot Accuracy@3
404
+ - type: dot_accuracy@5
405
+ value: 0.66
406
+ name: Dot Accuracy@5
407
+ - type: dot_accuracy@10
408
+ value: 0.76
409
+ name: Dot Accuracy@10
410
+ - type: dot_precision@1
411
+ value: 0.36
412
+ name: Dot Precision@1
413
+ - type: dot_precision@3
414
+ value: 0.2
415
+ name: Dot Precision@3
416
+ - type: dot_precision@5
417
+ value: 0.14
418
+ name: Dot Precision@5
419
+ - type: dot_precision@10
420
+ value: 0.08
421
+ name: Dot Precision@10
422
+ - type: dot_recall@1
423
+ value: 0.34
424
+ name: Dot Recall@1
425
+ - type: dot_recall@3
426
+ value: 0.56
427
+ name: Dot Recall@3
428
+ - type: dot_recall@5
429
+ value: 0.64
430
+ name: Dot Recall@5
431
+ - type: dot_recall@10
432
+ value: 0.72
433
+ name: Dot Recall@10
434
+ - type: dot_ndcg@10
435
+ value: 0.5341909779287488
436
+ name: Dot Ndcg@10
437
+ - type: dot_mrr@10
438
+ value: 0.4836666666666667
439
+ name: Dot Mrr@10
440
+ - type: dot_map@100
441
+ value: 0.47986132889672267
442
+ name: Dot Map@100
443
+ - type: query_active_dims
444
+ value: 45.63999938964844
445
+ name: Query Active Dims
446
+ - type: query_sparsity_ratio
447
+ value: 0.9985046851651384
448
+ name: Query Sparsity Ratio
449
+ - type: corpus_active_dims
450
+ value: 104.37854766845703
451
+ name: Corpus Active Dims
452
+ - type: corpus_sparsity_ratio
453
+ value: 0.9965802192625498
454
+ name: Corpus Sparsity Ratio
455
+ - task:
456
+ type: sparse-nano-beir
457
+ name: Sparse Nano BEIR
458
+ dataset:
459
+ name: NanoBEIR mean
460
+ type: NanoBEIR_mean
461
+ metrics:
462
+ - type: dot_accuracy@1
463
+ value: 0.34
464
+ name: Dot Accuracy@1
465
+ - type: dot_accuracy@3
466
+ value: 0.5133333333333333
467
+ name: Dot Accuracy@3
468
+ - type: dot_accuracy@5
469
+ value: 0.6133333333333334
470
+ name: Dot Accuracy@5
471
+ - type: dot_accuracy@10
472
+ value: 0.7066666666666667
473
+ name: Dot Accuracy@10
474
+ - type: dot_precision@1
475
+ value: 0.34
476
+ name: Dot Precision@1
477
+ - type: dot_precision@3
478
+ value: 0.22888888888888884
479
+ name: Dot Precision@3
480
+ - type: dot_precision@5
481
+ value: 0.18800000000000003
482
+ name: Dot Precision@5
483
+ - type: dot_precision@10
484
+ value: 0.13
485
+ name: Dot Precision@10
486
+ - type: dot_recall@1
487
+ value: 0.21474436749013393
488
+ name: Dot Recall@1
489
+ - type: dot_recall@3
490
+ value: 0.3710201820147387
491
+ name: Dot Recall@3
492
+ - type: dot_recall@5
493
+ value: 0.45757963379783356
494
+ name: Dot Recall@5
495
+ - type: dot_recall@10
496
+ value: 0.5186456042164697
497
+ name: Dot Recall@10
498
+ - type: dot_ndcg@10
499
+ value: 0.436152482414192
500
+ name: Dot Ndcg@10
501
+ - type: dot_mrr@10
502
+ value: 0.4471931216931217
503
+ name: Dot Mrr@10
504
+ - type: dot_map@100
505
+ value: 0.3422455425930528
506
+ name: Dot Map@100
507
+ - type: query_active_dims
508
+ value: 53.56666692097982
509
+ name: Query Active Dims
510
+ - type: query_sparsity_ratio
511
+ value: 0.9982449817534571
512
+ name: Query Sparsity Ratio
513
+ - type: corpus_active_dims
514
+ value: 110.01350571216182
515
+ name: Corpus Active Dims
516
+ - type: corpus_sparsity_ratio
517
+ value: 0.9963955997080087
518
+ name: Corpus Sparsity Ratio
519
+ - type: dot_accuracy@1
520
+ value: 0.44241758241758244
521
+ name: Dot Accuracy@1
522
+ - type: dot_accuracy@3
523
+ value: 0.6319937205651491
524
+ name: Dot Accuracy@3
525
+ - type: dot_accuracy@5
526
+ value: 0.6982103610675039
527
+ name: Dot Accuracy@5
528
+ - type: dot_accuracy@10
529
+ value: 0.7922762951334378
530
+ name: Dot Accuracy@10
531
+ - type: dot_precision@1
532
+ value: 0.44241758241758244
533
+ name: Dot Precision@1
534
+ - type: dot_precision@3
535
+ value: 0.28906331763474624
536
+ name: Dot Precision@3
537
+ - type: dot_precision@5
538
+ value: 0.22349764521193094
539
+ name: Dot Precision@5
540
+ - type: dot_precision@10
541
+ value: 0.1596828885400314
542
+ name: Dot Precision@10
543
+ - type: dot_recall@1
544
+ value: 0.24762543099163964
545
+ name: Dot Recall@1
546
+ - type: dot_recall@3
547
+ value: 0.4029567497606036
548
+ name: Dot Recall@3
549
+ - type: dot_recall@5
550
+ value: 0.47356029516066417
551
+ name: Dot Recall@5
552
+ - type: dot_recall@10
553
+ value: 0.5593700517107145
554
+ name: Dot Recall@10
555
+ - type: dot_ndcg@10
556
+ value: 0.4973635297458972
557
+ name: Dot Ndcg@10
558
+ - type: dot_mrr@10
559
+ value: 0.5535014203483591
560
+ name: Dot Mrr@10
561
+ - type: dot_map@100
562
+ value: 0.41852441580546024
563
+ name: Dot Map@100
564
+ - type: query_active_dims
565
+ value: 70.98613267025705
566
+ name: Query Active Dims
567
+ - type: query_sparsity_ratio
568
+ value: 0.9976742633945921
569
+ name: Query Sparsity Ratio
570
+ - type: corpus_active_dims
571
+ value: 109.8632523295962
572
+ name: Corpus Active Dims
573
+ - type: corpus_sparsity_ratio
574
+ value: 0.996400522497556
575
+ name: Corpus Sparsity Ratio
576
+ - task:
577
+ type: sparse-information-retrieval
578
+ name: Sparse Information Retrieval
579
+ dataset:
580
+ name: NanoClimateFEVER
581
+ type: NanoClimateFEVER
582
+ metrics:
583
+ - type: dot_accuracy@1
584
+ value: 0.26
585
+ name: Dot Accuracy@1
586
+ - type: dot_accuracy@3
587
+ value: 0.42
588
+ name: Dot Accuracy@3
589
+ - type: dot_accuracy@5
590
+ value: 0.48
591
+ name: Dot Accuracy@5
592
+ - type: dot_accuracy@10
593
+ value: 0.62
594
+ name: Dot Accuracy@10
595
+ - type: dot_precision@1
596
+ value: 0.26
597
+ name: Dot Precision@1
598
+ - type: dot_precision@3
599
+ value: 0.15333333333333332
600
+ name: Dot Precision@3
601
+ - type: dot_precision@5
602
+ value: 0.10800000000000001
603
+ name: Dot Precision@5
604
+ - type: dot_precision@10
605
+ value: 0.07800000000000001
606
+ name: Dot Precision@10
607
+ - type: dot_recall@1
608
+ value: 0.12499999999999999
609
+ name: Dot Recall@1
610
+ - type: dot_recall@3
611
+ value: 0.20166666666666663
612
+ name: Dot Recall@3
613
+ - type: dot_recall@5
614
+ value: 0.24
615
+ name: Dot Recall@5
616
+ - type: dot_recall@10
617
+ value: 0.32166666666666666
618
+ name: Dot Recall@10
619
+ - type: dot_ndcg@10
620
+ value: 0.26602915512735714
621
+ name: Dot Ndcg@10
622
+ - type: dot_mrr@10
623
+ value: 0.3617857142857142
624
+ name: Dot Mrr@10
625
+ - type: dot_map@100
626
+ value: 0.2080211358638896
627
+ name: Dot Map@100
628
+ - type: query_active_dims
629
+ value: 89.9000015258789
630
+ name: Query Active Dims
631
+ - type: query_sparsity_ratio
632
+ value: 0.997054583529065
633
+ name: Query Sparsity Ratio
634
+ - type: corpus_active_dims
635
+ value: 107.88761901855469
636
+ name: Corpus Active Dims
637
+ - type: corpus_sparsity_ratio
638
+ value: 0.9964652506710387
639
+ name: Corpus Sparsity Ratio
640
+ - task:
641
+ type: sparse-information-retrieval
642
+ name: Sparse Information Retrieval
643
+ dataset:
644
+ name: NanoDBPedia
645
+ type: NanoDBPedia
646
+ metrics:
647
+ - type: dot_accuracy@1
648
+ value: 0.62
649
+ name: Dot Accuracy@1
650
+ - type: dot_accuracy@3
651
+ value: 0.82
652
+ name: Dot Accuracy@3
653
+ - type: dot_accuracy@5
654
+ value: 0.86
655
+ name: Dot Accuracy@5
656
+ - type: dot_accuracy@10
657
+ value: 0.92
658
+ name: Dot Accuracy@10
659
+ - type: dot_precision@1
660
+ value: 0.62
661
+ name: Dot Precision@1
662
+ - type: dot_precision@3
663
+ value: 0.5266666666666667
664
+ name: Dot Precision@3
665
+ - type: dot_precision@5
666
+ value: 0.4640000000000001
667
+ name: Dot Precision@5
668
+ - type: dot_precision@10
669
+ value: 0.43199999999999994
670
+ name: Dot Precision@10
671
+ - type: dot_recall@1
672
+ value: 0.07037508003753436
673
+ name: Dot Recall@1
674
+ - type: dot_recall@3
675
+ value: 0.1332476350020503
676
+ name: Dot Recall@3
677
+ - type: dot_recall@5
678
+ value: 0.17834335811098734
679
+ name: Dot Recall@5
680
+ - type: dot_recall@10
681
+ value: 0.3023813591870427
682
+ name: Dot Recall@10
683
+ - type: dot_ndcg@10
684
+ value: 0.5284701506717093
685
+ name: Dot Ndcg@10
686
+ - type: dot_mrr@10
687
+ value: 0.7312222222222222
688
+ name: Dot Mrr@10
689
+ - type: dot_map@100
690
+ value: 0.39806509167251636
691
+ name: Dot Map@100
692
+ - type: query_active_dims
693
+ value: 48.779998779296875
694
+ name: Query Active Dims
695
+ - type: query_sparsity_ratio
696
+ value: 0.9984018085715453
697
+ name: Query Sparsity Ratio
698
+ - type: corpus_active_dims
699
+ value: 112.2790756225586
700
+ name: Corpus Active Dims
701
+ - type: corpus_sparsity_ratio
702
+ value: 0.9963213722684438
703
+ name: Corpus Sparsity Ratio
704
+ - task:
705
+ type: sparse-information-retrieval
706
+ name: Sparse Information Retrieval
707
+ dataset:
708
+ name: NanoFEVER
709
+ type: NanoFEVER
710
+ metrics:
711
+ - type: dot_accuracy@1
712
+ value: 0.54
713
+ name: Dot Accuracy@1
714
+ - type: dot_accuracy@3
715
+ value: 0.72
716
+ name: Dot Accuracy@3
717
+ - type: dot_accuracy@5
718
+ value: 0.82
719
+ name: Dot Accuracy@5
720
+ - type: dot_accuracy@10
721
+ value: 0.92
722
+ name: Dot Accuracy@10
723
+ - type: dot_precision@1
724
+ value: 0.54
725
+ name: Dot Precision@1
726
+ - type: dot_precision@3
727
+ value: 0.24
728
+ name: Dot Precision@3
729
+ - type: dot_precision@5
730
+ value: 0.16799999999999998
731
+ name: Dot Precision@5
732
+ - type: dot_precision@10
733
+ value: 0.09599999999999997
734
+ name: Dot Precision@10
735
+ - type: dot_recall@1
736
+ value: 0.5266666666666666
737
+ name: Dot Recall@1
738
+ - type: dot_recall@3
739
+ value: 0.6866666666666665
740
+ name: Dot Recall@3
741
+ - type: dot_recall@5
742
+ value: 0.7666666666666666
743
+ name: Dot Recall@5
744
+ - type: dot_recall@10
745
+ value: 0.8766666666666667
746
+ name: Dot Recall@10
747
+ - type: dot_ndcg@10
748
+ value: 0.696250000763901
749
+ name: Dot Ndcg@10
750
+ - type: dot_mrr@10
751
+ value: 0.6531666666666667
752
+ name: Dot Mrr@10
753
+ - type: dot_map@100
754
+ value: 0.6383785103785103
755
+ name: Dot Map@100
756
+ - type: query_active_dims
757
+ value: 82.72000122070312
758
+ name: Query Active Dims
759
+ - type: query_sparsity_ratio
760
+ value: 0.9972898236937061
761
+ name: Query Sparsity Ratio
762
+ - type: corpus_active_dims
763
+ value: 121.61109161376953
764
+ name: Corpus Active Dims
765
+ - type: corpus_sparsity_ratio
766
+ value: 0.9960156250699899
767
+ name: Corpus Sparsity Ratio
768
+ - task:
769
+ type: sparse-information-retrieval
770
+ name: Sparse Information Retrieval
771
+ dataset:
772
+ name: NanoFiQA2018
773
+ type: NanoFiQA2018
774
+ metrics:
775
+ - type: dot_accuracy@1
776
+ value: 0.22
777
+ name: Dot Accuracy@1
778
+ - type: dot_accuracy@3
779
+ value: 0.42
780
+ name: Dot Accuracy@3
781
+ - type: dot_accuracy@5
782
+ value: 0.44
783
+ name: Dot Accuracy@5
784
+ - type: dot_accuracy@10
785
+ value: 0.54
786
+ name: Dot Accuracy@10
787
+ - type: dot_precision@1
788
+ value: 0.22
789
+ name: Dot Precision@1
790
+ - type: dot_precision@3
791
+ value: 0.16666666666666663
792
+ name: Dot Precision@3
793
+ - type: dot_precision@5
794
+ value: 0.11200000000000002
795
+ name: Dot Precision@5
796
+ - type: dot_precision@10
797
+ value: 0.07200000000000001
798
+ name: Dot Precision@10
799
+ - type: dot_recall@1
800
+ value: 0.12335714285714286
801
+ name: Dot Recall@1
802
+ - type: dot_recall@3
803
+ value: 0.29043650793650794
804
+ name: Dot Recall@3
805
+ - type: dot_recall@5
806
+ value: 0.3084365079365079
807
+ name: Dot Recall@5
808
+ - type: dot_recall@10
809
+ value: 0.38043650793650796
810
+ name: Dot Recall@10
811
+ - type: dot_ndcg@10
812
+ value: 0.2842623648908474
813
+ name: Dot Ndcg@10
814
+ - type: dot_mrr@10
815
+ value: 0.3184126984126985
816
+ name: Dot Mrr@10
817
+ - type: dot_map@100
818
+ value: 0.24328509002216755
819
+ name: Dot Map@100
820
+ - type: query_active_dims
821
+ value: 52.91999816894531
822
+ name: Query Active Dims
823
+ - type: query_sparsity_ratio
824
+ value: 0.9982661687252163
825
+ name: Query Sparsity Ratio
826
+ - type: corpus_active_dims
827
+ value: 104.23889923095703
828
+ name: Corpus Active Dims
829
+ - type: corpus_sparsity_ratio
830
+ value: 0.9965847945996016
831
+ name: Corpus Sparsity Ratio
832
+ - task:
833
+ type: sparse-information-retrieval
834
+ name: Sparse Information Retrieval
835
+ dataset:
836
+ name: NanoHotpotQA
837
+ type: NanoHotpotQA
838
+ metrics:
839
+ - type: dot_accuracy@1
840
+ value: 0.74
841
+ name: Dot Accuracy@1
842
+ - type: dot_accuracy@3
843
+ value: 0.9
844
+ name: Dot Accuracy@3
845
+ - type: dot_accuracy@5
846
+ value: 0.9
847
+ name: Dot Accuracy@5
848
+ - type: dot_accuracy@10
849
+ value: 0.98
850
+ name: Dot Accuracy@10
851
+ - type: dot_precision@1
852
+ value: 0.74
853
+ name: Dot Precision@1
854
+ - type: dot_precision@3
855
+ value: 0.41999999999999993
856
+ name: Dot Precision@3
857
+ - type: dot_precision@5
858
+ value: 0.284
859
+ name: Dot Precision@5
860
+ - type: dot_precision@10
861
+ value: 0.16399999999999998
862
+ name: Dot Precision@10
863
+ - type: dot_recall@1
864
+ value: 0.37
865
+ name: Dot Recall@1
866
+ - type: dot_recall@3
867
+ value: 0.63
868
+ name: Dot Recall@3
869
+ - type: dot_recall@5
870
+ value: 0.71
871
+ name: Dot Recall@5
872
+ - type: dot_recall@10
873
+ value: 0.82
874
+ name: Dot Recall@10
875
+ - type: dot_ndcg@10
876
+ value: 0.7250698177423742
877
+ name: Dot Ndcg@10
878
+ - type: dot_mrr@10
879
+ value: 0.8267222222222221
880
+ name: Dot Mrr@10
881
+ - type: dot_map@100
882
+ value: 0.63152024851727
883
+ name: Dot Map@100
884
+ - type: query_active_dims
885
+ value: 69.81999969482422
886
+ name: Query Active Dims
887
+ - type: query_sparsity_ratio
888
+ value: 0.9977124697039897
889
+ name: Query Sparsity Ratio
890
+ - type: corpus_active_dims
891
+ value: 134.8498992919922
892
+ name: Corpus Active Dims
893
+ - type: corpus_sparsity_ratio
894
+ value: 0.9955818786681085
895
+ name: Corpus Sparsity Ratio
896
+ - task:
897
+ type: sparse-information-retrieval
898
+ name: Sparse Information Retrieval
899
+ dataset:
900
+ name: NanoQuoraRetrieval
901
+ type: NanoQuoraRetrieval
902
+ metrics:
903
+ - type: dot_accuracy@1
904
+ value: 0.8
905
+ name: Dot Accuracy@1
906
+ - type: dot_accuracy@3
907
+ value: 0.92
908
+ name: Dot Accuracy@3
909
+ - type: dot_accuracy@5
910
+ value: 0.94
911
+ name: Dot Accuracy@5
912
+ - type: dot_accuracy@10
913
+ value: 0.98
914
+ name: Dot Accuracy@10
915
+ - type: dot_precision@1
916
+ value: 0.8
917
+ name: Dot Precision@1
918
+ - type: dot_precision@3
919
+ value: 0.35999999999999993
920
+ name: Dot Precision@3
921
+ - type: dot_precision@5
922
+ value: 0.23599999999999993
923
+ name: Dot Precision@5
924
+ - type: dot_precision@10
925
+ value: 0.12999999999999998
926
+ name: Dot Precision@10
927
+ - type: dot_recall@1
928
+ value: 0.7240000000000001
929
+ name: Dot Recall@1
930
+ - type: dot_recall@3
931
+ value: 0.8613333333333333
932
+ name: Dot Recall@3
933
+ - type: dot_recall@5
934
+ value: 0.9059999999999999
935
+ name: Dot Recall@5
936
+ - type: dot_recall@10
937
+ value: 0.9633333333333333
938
+ name: Dot Recall@10
939
+ - type: dot_ndcg@10
940
+ value: 0.8826618022083887
941
+ name: Dot Ndcg@10
942
+ - type: dot_mrr@10
943
+ value: 0.8673809523809524
944
+ name: Dot Mrr@10
945
+ - type: dot_map@100
946
+ value: 0.8507124011593389
947
+ name: Dot Map@100
948
+ - type: query_active_dims
949
+ value: 49.619998931884766
950
+ name: Query Active Dims
951
+ - type: query_sparsity_ratio
952
+ value: 0.9983742874342479
953
+ name: Query Sparsity Ratio
954
+ - type: corpus_active_dims
955
+ value: 54.11692428588867
956
+ name: Corpus Active Dims
957
+ - type: corpus_sparsity_ratio
958
+ value: 0.998226953532341
959
+ name: Corpus Sparsity Ratio
960
+ - task:
961
+ type: sparse-information-retrieval
962
+ name: Sparse Information Retrieval
963
+ dataset:
964
+ name: NanoSCIDOCS
965
+ type: NanoSCIDOCS
966
+ metrics:
967
+ - type: dot_accuracy@1
968
+ value: 0.44
969
+ name: Dot Accuracy@1
970
+ - type: dot_accuracy@3
971
+ value: 0.62
972
+ name: Dot Accuracy@3
973
+ - type: dot_accuracy@5
974
+ value: 0.66
975
+ name: Dot Accuracy@5
976
+ - type: dot_accuracy@10
977
+ value: 0.76
978
+ name: Dot Accuracy@10
979
+ - type: dot_precision@1
980
+ value: 0.44
981
+ name: Dot Precision@1
982
+ - type: dot_precision@3
983
+ value: 0.2733333333333334
984
+ name: Dot Precision@3
985
+ - type: dot_precision@5
986
+ value: 0.22
987
+ name: Dot Precision@5
988
+ - type: dot_precision@10
989
+ value: 0.16399999999999998
990
+ name: Dot Precision@10
991
+ - type: dot_recall@1
992
+ value: 0.09166666666666666
993
+ name: Dot Recall@1
994
+ - type: dot_recall@3
995
+ value: 0.16866666666666666
996
+ name: Dot Recall@3
997
+ - type: dot_recall@5
998
+ value: 0.22566666666666665
999
+ name: Dot Recall@5
1000
+ - type: dot_recall@10
1001
+ value: 0.3356666666666666
1002
+ name: Dot Recall@10
1003
+ - type: dot_ndcg@10
1004
+ value: 0.326903742587538
1005
+ name: Dot Ndcg@10
1006
+ - type: dot_mrr@10
1007
+ value: 0.5351904761904761
1008
+ name: Dot Mrr@10
1009
+ - type: dot_map@100
1010
+ value: 0.24963439583895705
1011
+ name: Dot Map@100
1012
+ - type: query_active_dims
1013
+ value: 86.16000366210938
1014
+ name: Query Active Dims
1015
+ - type: query_sparsity_ratio
1016
+ value: 0.9971771180243068
1017
+ name: Query Sparsity Ratio
1018
+ - type: corpus_active_dims
1019
+ value: 115.15058898925781
1020
+ name: Corpus Active Dims
1021
+ - type: corpus_sparsity_ratio
1022
+ value: 0.9962272921502766
1023
+ name: Corpus Sparsity Ratio
1024
+ - task:
1025
+ type: sparse-information-retrieval
1026
+ name: Sparse Information Retrieval
1027
+ dataset:
1028
+ name: NanoArguAna
1029
+ type: NanoArguAna
1030
+ metrics:
1031
+ - type: dot_accuracy@1
1032
+ value: 0.1
1033
+ name: Dot Accuracy@1
1034
+ - type: dot_accuracy@3
1035
+ value: 0.52
1036
+ name: Dot Accuracy@3
1037
+ - type: dot_accuracy@5
1038
+ value: 0.66
1039
+ name: Dot Accuracy@5
1040
+ - type: dot_accuracy@10
1041
+ value: 0.78
1042
+ name: Dot Accuracy@10
1043
+ - type: dot_precision@1
1044
+ value: 0.1
1045
+ name: Dot Precision@1
1046
+ - type: dot_precision@3
1047
+ value: 0.17333333333333337
1048
+ name: Dot Precision@3
1049
+ - type: dot_precision@5
1050
+ value: 0.132
1051
+ name: Dot Precision@5
1052
+ - type: dot_precision@10
1053
+ value: 0.07800000000000001
1054
+ name: Dot Precision@10
1055
+ - type: dot_recall@1
1056
+ value: 0.1
1057
+ name: Dot Recall@1
1058
+ - type: dot_recall@3
1059
+ value: 0.52
1060
+ name: Dot Recall@3
1061
+ - type: dot_recall@5
1062
+ value: 0.66
1063
+ name: Dot Recall@5
1064
+ - type: dot_recall@10
1065
+ value: 0.78
1066
+ name: Dot Recall@10
1067
+ - type: dot_ndcg@10
1068
+ value: 0.44166045098306866
1069
+ name: Dot Ndcg@10
1070
+ - type: dot_mrr@10
1071
+ value: 0.3330793650793652
1072
+ name: Dot Mrr@10
1073
+ - type: dot_map@100
1074
+ value: 0.33989533146591966
1075
+ name: Dot Map@100
1076
+ - type: query_active_dims
1077
+ value: 119.26000213623047
1078
+ name: Query Active Dims
1079
+ - type: query_sparsity_ratio
1080
+ value: 0.9960926544087468
1081
+ name: Query Sparsity Ratio
1082
+ - type: corpus_active_dims
1083
+ value: 117.85887145996094
1084
+ name: Corpus Active Dims
1085
+ - type: corpus_sparsity_ratio
1086
+ value: 0.9961385600072092
1087
+ name: Corpus Sparsity Ratio
1088
+ - task:
1089
+ type: sparse-information-retrieval
1090
+ name: Sparse Information Retrieval
1091
+ dataset:
1092
+ name: NanoSciFact
1093
+ type: NanoSciFact
1094
+ metrics:
1095
+ - type: dot_accuracy@1
1096
+ value: 0.44
1097
+ name: Dot Accuracy@1
1098
+ - type: dot_accuracy@3
1099
+ value: 0.54
1100
+ name: Dot Accuracy@3
1101
+ - type: dot_accuracy@5
1102
+ value: 0.64
1103
+ name: Dot Accuracy@5
1104
+ - type: dot_accuracy@10
1105
+ value: 0.7
1106
+ name: Dot Accuracy@10
1107
+ - type: dot_precision@1
1108
+ value: 0.44
1109
+ name: Dot Precision@1
1110
+ - type: dot_precision@3
1111
+ value: 0.2
1112
+ name: Dot Precision@3
1113
+ - type: dot_precision@5
1114
+ value: 0.14400000000000002
1115
+ name: Dot Precision@5
1116
+ - type: dot_precision@10
1117
+ value: 0.07800000000000001
1118
+ name: Dot Precision@10
1119
+ - type: dot_recall@1
1120
+ value: 0.405
1121
+ name: Dot Recall@1
1122
+ - type: dot_recall@3
1123
+ value: 0.525
1124
+ name: Dot Recall@3
1125
+ - type: dot_recall@5
1126
+ value: 0.63
1127
+ name: Dot Recall@5
1128
+ - type: dot_recall@10
1129
+ value: 0.68
1130
+ name: Dot Recall@10
1131
+ - type: dot_ndcg@10
1132
+ value: 0.5538495558550187
1133
+ name: Dot Ndcg@10
1134
+ - type: dot_mrr@10
1135
+ value: 0.5216904761904761
1136
+ name: Dot Mrr@10
1137
+ - type: dot_map@100
1138
+ value: 0.5176733402786808
1139
+ name: Dot Map@100
1140
+ - type: query_active_dims
1141
+ value: 111.37999725341797
1142
+ name: Query Active Dims
1143
+ - type: query_sparsity_ratio
1144
+ value: 0.9963508290002812
1145
+ name: Query Sparsity Ratio
1146
+ - type: corpus_active_dims
1147
+ value: 109.96676635742188
1148
+ name: Corpus Active Dims
1149
+ - type: corpus_sparsity_ratio
1150
+ value: 0.9963971310413007
1151
+ name: Corpus Sparsity Ratio
1152
+ - task:
1153
+ type: sparse-information-retrieval
1154
+ name: Sparse Information Retrieval
1155
+ dataset:
1156
+ name: NanoTouche2020
1157
+ type: NanoTouche2020
1158
+ metrics:
1159
+ - type: dot_accuracy@1
1160
+ value: 0.5714285714285714
1161
+ name: Dot Accuracy@1
1162
+ - type: dot_accuracy@3
1163
+ value: 0.7959183673469388
1164
+ name: Dot Accuracy@3
1165
+ - type: dot_accuracy@5
1166
+ value: 0.8367346938775511
1167
+ name: Dot Accuracy@5
1168
+ - type: dot_accuracy@10
1169
+ value: 0.9795918367346939
1170
+ name: Dot Accuracy@10
1171
+ - type: dot_precision@1
1172
+ value: 0.5714285714285714
1173
+ name: Dot Precision@1
1174
+ - type: dot_precision@3
1175
+ value: 0.5578231292517006
1176
+ name: Dot Precision@3
1177
+ - type: dot_precision@5
1178
+ value: 0.4734693877551021
1179
+ name: Dot Precision@5
1180
+ - type: dot_precision@10
1181
+ value: 0.3938775510204081
1182
+ name: Dot Precision@10
1183
+ - type: dot_recall@1
1184
+ value: 0.03883194419290252
1185
+ name: Dot Recall@1
1186
+ - type: dot_recall@3
1187
+ value: 0.10835972457173955
1188
+ name: Dot Recall@3
1189
+ - type: dot_recall@5
1190
+ value: 0.15843173631430588
1191
+ name: Dot Recall@5
1192
+ - type: dot_recall@10
1193
+ value: 0.25572265913299697
1194
+ name: Dot Recall@10
1195
+ - type: dot_ndcg@10
1196
+ value: 0.4521113986238841
1197
+ name: Dot Ndcg@10
1198
+ - type: dot_mrr@10
1199
+ value: 0.7052883057985098
1200
+ name: Dot Mrr@10
1201
+ - type: dot_map@100
1202
+ value: 0.33689523249457515
1203
+ name: Dot Map@100
1204
+ - type: query_active_dims
1205
+ value: 51.163265228271484
1206
+ name: Query Active Dims
1207
+ - type: query_sparsity_ratio
1208
+ value: 0.9983237250105409
1209
+ name: Query Sparsity Ratio
1210
+ - type: corpus_active_dims
1211
+ value: 122.40800476074219
1212
+ name: Corpus Active Dims
1213
+ - type: corpus_sparsity_ratio
1214
+ value: 0.9959895156031472
1215
+ name: Corpus Sparsity Ratio
1216
+ ---
1217
+
1218
+ # splade-distilbert-base-uncased trained on Natural Questions
1219
+
1220
+ This is a [SPLADE Sparse Encoder](https://www.sbert.net/docs/sparse_encoder/usage/usage.html) model finetuned from [distilbert/distilbert-base-uncased](https://huggingface.co/distilbert/distilbert-base-uncased) on the [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions) dataset using the [sentence-transformers](https://www.SBERT.net) library. It maps sentences & paragraphs to a 30522-dimensional sparse vector space and can be used for semantic search and sparse retrieval.
1221
+ ## Model Details
1222
+
1223
+ ### Model Description
1224
+ - **Model Type:** SPLADE Sparse Encoder
1225
+ - **Base model:** [distilbert/distilbert-base-uncased](https://huggingface.co/distilbert/distilbert-base-uncased) <!-- at revision 12040accade4e8a0f71eabdb258fecc2e7e948be -->
1226
+ - **Maximum Sequence Length:** 256 tokens
1227
+ - **Output Dimensionality:** 30522 dimensions
1228
+ - **Similarity Function:** Dot Product
1229
+ - **Training Dataset:**
1230
+ - [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions)
1231
+ - **Language:** en
1232
+ - **License:** apache-2.0
1233
+
1234
+ ### Model Sources
1235
+
1236
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
1237
+ - **Documentation:** [Sparse Encoder Documentation](https://www.sbert.net/docs/sparse_encoder/usage/usage.html)
1238
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
1239
+ - **Hugging Face:** [Sparse Encoders on Hugging Face](https://huggingface.co/models?library=sentence-transformers&other=sparse-encoder)
1240
+
1241
+ ### Full Model Architecture
1242
+
1243
+ ```
1244
+ SparseEncoder(
1245
+ (0): MLMTransformer({'max_seq_length': 256, 'do_lower_case': False}) with MLMTransformer model: DistilBertForMaskedLM
1246
+ (1): SpladePooling({'pooling_strategy': 'max', 'activation_function': 'relu', 'word_embedding_dimension': 30522})
1247
+ )
1248
+ ```
1249
+
1250
+ ## Usage
1251
+
1252
+ ### Direct Usage (Sentence Transformers)
1253
+
1254
+ First install the Sentence Transformers library:
1255
+
1256
+ ```bash
1257
+ pip install -U sentence-transformers
1258
+ ```
1259
+
1260
+ Then you can load this model and run inference.
1261
+ ```python
1262
+ from sentence_transformers import SparseEncoder
1263
+
1264
+ # Download from the 🤗 Hub
1265
+ model = SparseEncoder("arthurbresnu/splade-distilbert-base-uncased-nq")
1266
+ # Run inference
1267
+ sentences = [
1268
+ 'is send in the clowns from a musical',
1269
+ 'Send In the Clowns "Send In the Clowns" is a song written by Stephen Sondheim for the 1973 musical A Little Night Music, an adaptation of Ingmar Bergman\'s film Smiles of a Summer Night. It is a ballad from Act Two, in which the character Desirée reflects on the ironies and disappointments of her life. Among other things, she looks back on an affair years earlier with the lawyer Fredrik, who was deeply in love with her but whose marriage proposals she had rejected. Meeting him after so long, she realizes she is in love with him and finally ready to marry him, but now it is he who rejects her: he is in an unconsummated marriage with a much younger woman. Desirée proposes marriage to rescue him from this situation, but he declines, citing his dedication to his bride. Reacting to his rejection, Desirée sings this song. The song is later reprised as a coda after Fredrik\'s young wife runs away with his son, and Fredrik is finally free to accept Desirée\'s offer.[1]',
1270
+ 'The Suite Life on Deck The Suite Life on Deck is an American sitcom that aired on Disney Channel from September 26, 2008 to May 6, 2011. It is a sequel/spin-off of the Disney Channel Original Series The Suite Life of Zack & Cody. The series follows twin brothers Zack and Cody Martin and hotel heiress London Tipton in a new setting, the SS Tipton, where they attend classes at "Seven Seas High School" and meet Bailey Pickett while Mr. Moseby manages the ship. The ship travels around the world to nations such as Italy, France, Greece, India, Sweden and the United Kingdom where the characters experience different cultures, adventures, and situations.[1]',
1271
+ ]
1272
+ embeddings = model.encode(sentences)
1273
+ print(embeddings.shape)
1274
+ # (3, 30522)
1275
+
1276
+ # Get the similarity scores for the embeddings
1277
+ similarities = model.similarity(embeddings, embeddings)
1278
+ print(similarities.shape)
1279
+ # [3, 3]
1280
+ ```
1281
+
1282
+ <!--
1283
+ ### Direct Usage (Transformers)
1284
+
1285
+ <details><summary>Click to see the direct usage in Transformers</summary>
1286
+
1287
+ </details>
1288
+ -->
1289
+
1290
+ <!--
1291
+ ### Downstream Usage (Sentence Transformers)
1292
+
1293
+ You can finetune this model on your own dataset.
1294
+
1295
+ <details><summary>Click to expand</summary>
1296
+
1297
+ </details>
1298
+ -->
1299
+
1300
+ <!--
1301
+ ### Out-of-Scope Use
1302
+
1303
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
1304
+ -->
1305
+
1306
+ ## Evaluation
1307
+
1308
+ ### Metrics
1309
+
1310
+ #### Sparse Information Retrieval
1311
+
1312
+ * Datasets: `NanoMSMARCO`, `NanoNFCorpus`, `NanoNQ`, `NanoClimateFEVER`, `NanoDBPedia`, `NanoFEVER`, `NanoFiQA2018`, `NanoHotpotQA`, `NanoMSMARCO`, `NanoNFCorpus`, `NanoNQ`, `NanoQuoraRetrieval`, `NanoSCIDOCS`, `NanoArguAna`, `NanoSciFact` and `NanoTouche2020`
1313
+ * Evaluated with [<code>SparseInformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseInformationRetrievalEvaluator)
1314
+
1315
+ | Metric | NanoMSMARCO | NanoNFCorpus | NanoNQ | NanoClimateFEVER | NanoDBPedia | NanoFEVER | NanoFiQA2018 | NanoHotpotQA | NanoQuoraRetrieval | NanoSCIDOCS | NanoArguAna | NanoSciFact | NanoTouche2020 |
1316
+ |:----------------------|:------------|:-------------|:-----------|:-----------------|:------------|:-----------|:-------------|:-------------|:-------------------|:------------|:------------|:------------|:---------------|
1317
+ | dot_accuracy@1 | 0.28 | 0.38 | 0.36 | 0.26 | 0.62 | 0.54 | 0.22 | 0.74 | 0.8 | 0.44 | 0.1 | 0.44 | 0.5714 |
1318
+ | dot_accuracy@3 | 0.5 | 0.46 | 0.58 | 0.42 | 0.82 | 0.72 | 0.42 | 0.9 | 0.92 | 0.62 | 0.52 | 0.54 | 0.7959 |
1319
+ | dot_accuracy@5 | 0.66 | 0.52 | 0.66 | 0.48 | 0.86 | 0.82 | 0.44 | 0.9 | 0.94 | 0.66 | 0.66 | 0.64 | 0.8367 |
1320
+ | dot_accuracy@10 | 0.74 | 0.62 | 0.76 | 0.62 | 0.92 | 0.92 | 0.54 | 0.98 | 0.98 | 0.76 | 0.78 | 0.7 | 0.9796 |
1321
+ | dot_precision@1 | 0.28 | 0.38 | 0.36 | 0.26 | 0.62 | 0.54 | 0.22 | 0.74 | 0.8 | 0.44 | 0.1 | 0.44 | 0.5714 |
1322
+ | dot_precision@3 | 0.1667 | 0.32 | 0.2 | 0.1533 | 0.5267 | 0.24 | 0.1667 | 0.42 | 0.36 | 0.2733 | 0.1733 | 0.2 | 0.5578 |
1323
+ | dot_precision@5 | 0.132 | 0.292 | 0.14 | 0.108 | 0.464 | 0.168 | 0.112 | 0.284 | 0.236 | 0.22 | 0.132 | 0.144 | 0.4735 |
1324
+ | dot_precision@10 | 0.074 | 0.236 | 0.08 | 0.078 | 0.432 | 0.096 | 0.072 | 0.164 | 0.13 | 0.164 | 0.078 | 0.078 | 0.3939 |
1325
+ | dot_recall@1 | 0.28 | 0.0242 | 0.34 | 0.125 | 0.0704 | 0.5267 | 0.1234 | 0.37 | 0.724 | 0.0917 | 0.1 | 0.405 | 0.0388 |
1326
+ | dot_recall@3 | 0.5 | 0.0531 | 0.56 | 0.2017 | 0.1332 | 0.6867 | 0.2904 | 0.63 | 0.8613 | 0.1687 | 0.52 | 0.525 | 0.1084 |
1327
+ | dot_recall@5 | 0.66 | 0.0727 | 0.64 | 0.24 | 0.1783 | 0.7667 | 0.3084 | 0.71 | 0.906 | 0.2257 | 0.66 | 0.63 | 0.1584 |
1328
+ | dot_recall@10 | 0.74 | 0.0959 | 0.72 | 0.3217 | 0.3024 | 0.8767 | 0.3804 | 0.82 | 0.9633 | 0.3357 | 0.78 | 0.68 | 0.2557 |
1329
+ | **dot_ndcg@10** | **0.4958** | **0.2785** | **0.5342** | **0.266** | **0.5285** | **0.6963** | **0.2843** | **0.7251** | **0.8827** | **0.3269** | **0.4417** | **0.5538** | **0.4521** |
1330
+ | dot_mrr@10 | 0.4185 | 0.4394 | 0.4837 | 0.3618 | 0.7312 | 0.6532 | 0.3184 | 0.8267 | 0.8674 | 0.5352 | 0.3331 | 0.5217 | 0.7053 |
1331
+ | dot_map@100 | 0.4294 | 0.1174 | 0.4799 | 0.208 | 0.3981 | 0.6384 | 0.2433 | 0.6315 | 0.8507 | 0.2496 | 0.3399 | 0.5177 | 0.3369 |
1332
+ | query_active_dims | 52.56 | 62.5 | 45.64 | 89.9 | 48.78 | 82.72 | 52.92 | 69.82 | 49.62 | 86.16 | 119.26 | 111.38 | 51.1633 |
1333
+ | query_sparsity_ratio | 0.9983 | 0.998 | 0.9985 | 0.9971 | 0.9984 | 0.9973 | 0.9983 | 0.9977 | 0.9984 | 0.9972 | 0.9961 | 0.9964 | 0.9983 |
1334
+ | corpus_active_dims | 106.134 | 126.2465 | 104.3785 | 107.8876 | 112.2791 | 121.6111 | 104.2389 | 134.8499 | 54.1169 | 115.1506 | 117.8589 | 109.9668 | 122.408 |
1335
+ | corpus_sparsity_ratio | 0.9965 | 0.9959 | 0.9966 | 0.9965 | 0.9963 | 0.996 | 0.9966 | 0.9956 | 0.9982 | 0.9962 | 0.9961 | 0.9964 | 0.996 |
1336
+
1337
+ #### Sparse Nano BEIR
1338
+
1339
+ * Dataset: `NanoBEIR_mean`
1340
+ * Evaluated with [<code>SparseNanoBEIREvaluator</code>](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseNanoBEIREvaluator) with these parameters:
1341
+ ```json
1342
+ {
1343
+ "dataset_names": [
1344
+ "msmarco",
1345
+ "nfcorpus",
1346
+ "nq"
1347
+ ]
1348
+ }
1349
+ ```
1350
+
1351
+ | Metric | Value |
1352
+ |:----------------------|:-----------|
1353
+ | dot_accuracy@1 | 0.34 |
1354
+ | dot_accuracy@3 | 0.5133 |
1355
+ | dot_accuracy@5 | 0.6133 |
1356
+ | dot_accuracy@10 | 0.7067 |
1357
+ | dot_precision@1 | 0.34 |
1358
+ | dot_precision@3 | 0.2289 |
1359
+ | dot_precision@5 | 0.188 |
1360
+ | dot_precision@10 | 0.13 |
1361
+ | dot_recall@1 | 0.2147 |
1362
+ | dot_recall@3 | 0.371 |
1363
+ | dot_recall@5 | 0.4576 |
1364
+ | dot_recall@10 | 0.5186 |
1365
+ | **dot_ndcg@10** | **0.4362** |
1366
+ | dot_mrr@10 | 0.4472 |
1367
+ | dot_map@100 | 0.3422 |
1368
+ | query_active_dims | 53.5667 |
1369
+ | query_sparsity_ratio | 0.9982 |
1370
+ | corpus_active_dims | 110.0135 |
1371
+ | corpus_sparsity_ratio | 0.9964 |
1372
+
1373
+ #### Sparse Nano BEIR
1374
+
1375
+ * Dataset: `NanoBEIR_mean`
1376
+ * Evaluated with [<code>SparseNanoBEIREvaluator</code>](https://sbert.net/docs/package_reference/sparse_encoder/evaluation.html#sentence_transformers.sparse_encoder.evaluation.SparseNanoBEIREvaluator) with these parameters:
1377
+ ```json
1378
+ {
1379
+ "dataset_names": [
1380
+ "climatefever",
1381
+ "dbpedia",
1382
+ "fever",
1383
+ "fiqa2018",
1384
+ "hotpotqa",
1385
+ "msmarco",
1386
+ "nfcorpus",
1387
+ "nq",
1388
+ "quoraretrieval",
1389
+ "scidocs",
1390
+ "arguana",
1391
+ "scifact",
1392
+ "touche2020"
1393
+ ]
1394
+ }
1395
+ ```
1396
+
1397
+ | Metric | Value |
1398
+ |:----------------------|:-----------|
1399
+ | dot_accuracy@1 | 0.4424 |
1400
+ | dot_accuracy@3 | 0.632 |
1401
+ | dot_accuracy@5 | 0.6982 |
1402
+ | dot_accuracy@10 | 0.7923 |
1403
+ | dot_precision@1 | 0.4424 |
1404
+ | dot_precision@3 | 0.2891 |
1405
+ | dot_precision@5 | 0.2235 |
1406
+ | dot_precision@10 | 0.1597 |
1407
+ | dot_recall@1 | 0.2476 |
1408
+ | dot_recall@3 | 0.403 |
1409
+ | dot_recall@5 | 0.4736 |
1410
+ | dot_recall@10 | 0.5594 |
1411
+ | **dot_ndcg@10** | **0.4974** |
1412
+ | dot_mrr@10 | 0.5535 |
1413
+ | dot_map@100 | 0.4185 |
1414
+ | query_active_dims | 70.9861 |
1415
+ | query_sparsity_ratio | 0.9977 |
1416
+ | corpus_active_dims | 109.8633 |
1417
+ | corpus_sparsity_ratio | 0.9964 |
1418
+
1419
+ <!--
1420
+ ## Bias, Risks and Limitations
1421
+
1422
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
1423
+ -->
1424
+
1425
+ <!--
1426
+ ### Recommendations
1427
+
1428
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
1429
+ -->
1430
+
1431
+ ## Training Details
1432
+
1433
+ ### Training Dataset
1434
+
1435
+ #### natural-questions
1436
+
1437
+ * Dataset: [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions) at [f9e894e](https://huggingface.co/datasets/sentence-transformers/natural-questions/tree/f9e894e1081e206e577b4eaa9ee6de2b06ae6f17)
1438
+ * Size: 99,000 training samples
1439
+ * Columns: <code>query</code> and <code>answer</code>
1440
+ * Approximate statistics based on the first 1000 samples:
1441
+ | | query | answer |
1442
+ |:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
1443
+ | type | string | string |
1444
+ | details | <ul><li>min: 10 tokens</li><li>mean: 11.71 tokens</li><li>max: 26 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 131.81 tokens</li><li>max: 450 tokens</li></ul> |
1445
+ * Samples:
1446
+ | query | answer |
1447
+ |:--------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
1448
+ | <code>who played the father in papa don't preach</code> | <code>Alex McArthur Alex McArthur (born March 6, 1957) is an American actor.</code> |
1449
+ | <code>where was the location of the battle of hastings</code> | <code>Battle of Hastings The Battle of Hastings[a] was fought on 14 October 1066 between the Norman-French army of William, the Duke of Normandy, and an English army under the Anglo-Saxon King Harold Godwinson, beginning the Norman conquest of England. It took place approximately 7 miles (11 kilometres) northwest of Hastings, close to the present-day town of Battle, East Sussex, and was a decisive Norman victory.</code> |
1450
+ | <code>how many puppies can a dog give birth to</code> | <code>Canine reproduction The largest litter size to date was set by a Neapolitan Mastiff in Manea, Cambridgeshire, UK on November 29, 2004; the litter was 24 puppies.[22]</code> |
1451
+ * Loss: [<code>SpladeLoss</code>](https://sbert.net/docs/package_reference/sparse_encoder/losses.html#spladeloss) with these parameters:
1452
+ ```json
1453
+ {
1454
+ "loss": "SparseMultipleNegativesRankingLoss(scale=1.0, similarity_fct='dot_score')",
1455
+ "lambda_corpus": 3e-05,
1456
+ "lambda_query": 5e-05
1457
+ }
1458
+ ```
1459
+
1460
+ ### Evaluation Dataset
1461
+
1462
+ #### natural-questions
1463
+
1464
+ * Dataset: [natural-questions](https://huggingface.co/datasets/sentence-transformers/natural-questions) at [f9e894e](https://huggingface.co/datasets/sentence-transformers/natural-questions/tree/f9e894e1081e206e577b4eaa9ee6de2b06ae6f17)
1465
+ * Size: 1,000 evaluation samples
1466
+ * Columns: <code>query</code> and <code>answer</code>
1467
+ * Approximate statistics based on the first 1000 samples:
1468
+ | | query | answer |
1469
+ |:--------|:-----------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
1470
+ | type | string | string |
1471
+ | details | <ul><li>min: 10 tokens</li><li>mean: 11.69 tokens</li><li>max: 23 tokens</li></ul> | <ul><li>min: 15 tokens</li><li>mean: 134.01 tokens</li><li>max: 512 tokens</li></ul> |
1472
+ * Samples:
1473
+ | query | answer |
1474
+ |:-------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
1475
+ | <code>where is the tiber river located in italy</code> | <code>Tiber The Tiber (/ˈtaɪbər/, Latin: Tiberis,[1] Italian: Tevere [ˈteːvere])[2] is the third-longest river in Italy, rising in the Apennine Mountains in Emilia-Romagna and flowing 406 kilometres (252 mi) through Tuscany, Umbria and Lazio, where it is joined by the river Aniene, to the Tyrrhenian Sea, between Ostia and Fiumicino.[3] It drains a basin estimated at 17,375 square kilometres (6,709 sq mi). The river has achieved lasting fame as the main watercourse of the city of Rome, founded on its eastern banks.</code> |
1476
+ | <code>what kind of car does jay gatsby drive</code> | <code>Jay Gatsby At the Buchanan home, Jordan Baker, Nick, Jay, and the Buchanans decide to visit New York City. Tom borrows Gatsby's yellow Rolls Royce to drive up to the city. On the way to New York City, Tom makes a detour at a gas station in "the Valley of Ashes", a run-down part of Long Island. The owner, George Wilson, shares his concern that his wife, Myrtle, may be having an affair. This unnerves Tom, who has been having an affair with Myrtle, and he leaves in a hurry.</code> |
1477
+ | <code>who sings if i can dream about you</code> | <code>I Can Dream About You "I Can Dream About You" is a song performed by American singer Dan Hartman on the soundtrack album of the film Streets of Fire. Released in 1984 as a single from the soundtrack, and included on Hartman's album I Can Dream About You, it reached number 6 on the Billboard Hot 100.[1]</code> |
1478
+ * Loss: [<code>SpladeLoss</code>](https://sbert.net/docs/package_reference/sparse_encoder/losses.html#spladeloss) with these parameters:
1479
+ ```json
1480
+ {
1481
+ "loss": "SparseMultipleNegativesRankingLoss(scale=1.0, similarity_fct='dot_score')",
1482
+ "lambda_corpus": 3e-05,
1483
+ "lambda_query": 5e-05
1484
+ }
1485
+ ```
1486
+
1487
+ ### Training Hyperparameters
1488
+ #### Non-Default Hyperparameters
1489
+
1490
+ - `eval_strategy`: steps
1491
+ - `per_device_train_batch_size`: 12
1492
+ - `per_device_eval_batch_size`: 12
1493
+ - `learning_rate`: 2e-05
1494
+ - `num_train_epochs`: 1
1495
+ - `bf16`: True
1496
+ - `load_best_model_at_end`: True
1497
+ - `batch_sampler`: no_duplicates
1498
+
1499
+ #### All Hyperparameters
1500
+ <details><summary>Click to expand</summary>
1501
+
1502
+ - `overwrite_output_dir`: False
1503
+ - `do_predict`: False
1504
+ - `eval_strategy`: steps
1505
+ - `prediction_loss_only`: True
1506
+ - `per_device_train_batch_size`: 12
1507
+ - `per_device_eval_batch_size`: 12
1508
+ - `per_gpu_train_batch_size`: None
1509
+ - `per_gpu_eval_batch_size`: None
1510
+ - `gradient_accumulation_steps`: 1
1511
+ - `eval_accumulation_steps`: None
1512
+ - `torch_empty_cache_steps`: None
1513
+ - `learning_rate`: 2e-05
1514
+ - `weight_decay`: 0.0
1515
+ - `adam_beta1`: 0.9
1516
+ - `adam_beta2`: 0.999
1517
+ - `adam_epsilon`: 1e-08
1518
+ - `max_grad_norm`: 1.0
1519
+ - `num_train_epochs`: 1
1520
+ - `max_steps`: -1
1521
+ - `lr_scheduler_type`: linear
1522
+ - `lr_scheduler_kwargs`: {}
1523
+ - `warmup_ratio`: 0.0
1524
+ - `warmup_steps`: 0
1525
+ - `log_level`: passive
1526
+ - `log_level_replica`: warning
1527
+ - `log_on_each_node`: True
1528
+ - `logging_nan_inf_filter`: True
1529
+ - `save_safetensors`: True
1530
+ - `save_on_each_node`: False
1531
+ - `save_only_model`: False
1532
+ - `restore_callback_states_from_checkpoint`: False
1533
+ - `no_cuda`: False
1534
+ - `use_cpu`: False
1535
+ - `use_mps_device`: False
1536
+ - `seed`: 42
1537
+ - `data_seed`: None
1538
+ - `jit_mode_eval`: False
1539
+ - `use_ipex`: False
1540
+ - `bf16`: True
1541
+ - `fp16`: False
1542
+ - `fp16_opt_level`: O1
1543
+ - `half_precision_backend`: auto
1544
+ - `bf16_full_eval`: False
1545
+ - `fp16_full_eval`: False
1546
+ - `tf32`: None
1547
+ - `local_rank`: 0
1548
+ - `ddp_backend`: None
1549
+ - `tpu_num_cores`: None
1550
+ - `tpu_metrics_debug`: False
1551
+ - `debug`: []
1552
+ - `dataloader_drop_last`: False
1553
+ - `dataloader_num_workers`: 0
1554
+ - `dataloader_prefetch_factor`: None
1555
+ - `past_index`: -1
1556
+ - `disable_tqdm`: False
1557
+ - `remove_unused_columns`: True
1558
+ - `label_names`: None
1559
+ - `load_best_model_at_end`: True
1560
+ - `ignore_data_skip`: False
1561
+ - `fsdp`: []
1562
+ - `fsdp_min_num_params`: 0
1563
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
1564
+ - `tp_size`: 0
1565
+ - `fsdp_transformer_layer_cls_to_wrap`: None
1566
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
1567
+ - `deepspeed`: None
1568
+ - `label_smoothing_factor`: 0.0
1569
+ - `optim`: adamw_torch
1570
+ - `optim_args`: None
1571
+ - `adafactor`: False
1572
+ - `group_by_length`: False
1573
+ - `length_column_name`: length
1574
+ - `ddp_find_unused_parameters`: None
1575
+ - `ddp_bucket_cap_mb`: None
1576
+ - `ddp_broadcast_buffers`: False
1577
+ - `dataloader_pin_memory`: True
1578
+ - `dataloader_persistent_workers`: False
1579
+ - `skip_memory_metrics`: True
1580
+ - `use_legacy_prediction_loop`: False
1581
+ - `push_to_hub`: False
1582
+ - `resume_from_checkpoint`: None
1583
+ - `hub_model_id`: None
1584
+ - `hub_strategy`: every_save
1585
+ - `hub_private_repo`: None
1586
+ - `hub_always_push`: False
1587
+ - `gradient_checkpointing`: False
1588
+ - `gradient_checkpointing_kwargs`: None
1589
+ - `include_inputs_for_metrics`: False
1590
+ - `include_for_metrics`: []
1591
+ - `eval_do_concat_batches`: True
1592
+ - `fp16_backend`: auto
1593
+ - `push_to_hub_model_id`: None
1594
+ - `push_to_hub_organization`: None
1595
+ - `mp_parameters`:
1596
+ - `auto_find_batch_size`: False
1597
+ - `full_determinism`: False
1598
+ - `torchdynamo`: None
1599
+ - `ray_scope`: last
1600
+ - `ddp_timeout`: 1800
1601
+ - `torch_compile`: False
1602
+ - `torch_compile_backend`: None
1603
+ - `torch_compile_mode`: None
1604
+ - `dispatch_batches`: None
1605
+ - `split_batches`: None
1606
+ - `include_tokens_per_second`: False
1607
+ - `include_num_input_tokens_seen`: False
1608
+ - `neftune_noise_alpha`: None
1609
+ - `optim_target_modules`: None
1610
+ - `batch_eval_metrics`: False
1611
+ - `eval_on_start`: False
1612
+ - `use_liger_kernel`: False
1613
+ - `eval_use_gather_object`: False
1614
+ - `average_tokens_across_devices`: False
1615
+ - `prompts`: None
1616
+ - `batch_sampler`: no_duplicates
1617
+ - `multi_dataset_batch_sampler`: proportional
1618
+
1619
+ </details>
1620
+
1621
+ ### Training Logs
1622
+ | Epoch | Step | Training Loss | Validation Loss | NanoMSMARCO_dot_ndcg@10 | NanoNFCorpus_dot_ndcg@10 | NanoNQ_dot_ndcg@10 | NanoBEIR_mean_dot_ndcg@10 | NanoClimateFEVER_dot_ndcg@10 | NanoDBPedia_dot_ndcg@10 | NanoFEVER_dot_ndcg@10 | NanoFiQA2018_dot_ndcg@10 | NanoHotpotQA_dot_ndcg@10 | NanoQuoraRetrieval_dot_ndcg@10 | NanoSCIDOCS_dot_ndcg@10 | NanoArguAna_dot_ndcg@10 | NanoSciFact_dot_ndcg@10 | NanoTouche2020_dot_ndcg@10 |
1623
+ |:-------:|:--------:|:-------------:|:---------------:|:-----------------------:|:------------------------:|:------------------:|:-------------------------:|:----------------------------:|:-----------------------:|:---------------------:|:------------------------:|:------------------------:|:------------------------------:|:-----------------------:|:-----------------------:|:-----------------------:|:--------------------------:|
1624
+ | 0.0242 | 200 | 6.3626 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1625
+ | 0.0485 | 400 | 0.0957 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1626
+ | 0.0727 | 600 | 0.0927 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1627
+ | 0.0970 | 800 | 0.0588 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1628
+ | 0.1212 | 1000 | 0.0408 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1629
+ | 0.1455 | 1200 | 0.0515 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1630
+ | 0.1697 | 1400 | 0.0517 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1631
+ | 0.1939 | 1600 | 0.0213 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1632
+ | 0.2 | 1650 | - | 0.0520 | 0.4929 | 0.2618 | 0.4572 | 0.4040 | - | - | - | - | - | - | - | - | - | - |
1633
+ | 0.2182 | 1800 | 0.019 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1634
+ | 0.2424 | 2000 | 0.0333 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1635
+ | 0.2667 | 2200 | 0.0282 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1636
+ | 0.2909 | 2400 | 0.0418 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1637
+ | 0.3152 | 2600 | 0.0386 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1638
+ | 0.3394 | 2800 | 0.0289 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1639
+ | 0.3636 | 3000 | 0.0242 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1640
+ | 0.3879 | 3200 | 0.0335 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1641
+ | 0.4 | 3300 | - | 0.0360 | 0.4715 | 0.2808 | 0.5340 | 0.4288 | - | - | - | - | - | - | - | - | - | - |
1642
+ | 0.4121 | 3400 | 0.0264 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1643
+ | 0.4364 | 3600 | 0.0331 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1644
+ | 0.4606 | 3800 | 0.0339 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1645
+ | 0.4848 | 4000 | 0.0225 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1646
+ | 0.5091 | 4200 | 0.0164 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1647
+ | 0.5333 | 4400 | 0.0247 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1648
+ | 0.5576 | 4600 | 0.0213 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1649
+ | 0.5818 | 4800 | 0.0187 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1650
+ | 0.6 | 4950 | - | 0.0217 | 0.4901 | 0.2930 | 0.5072 | 0.4301 | - | - | - | - | - | - | - | - | - | - |
1651
+ | 0.6061 | 5000 | 0.0153 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1652
+ | 0.6303 | 5200 | 0.0186 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1653
+ | 0.6545 | 5400 | 0.0096 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1654
+ | 0.6788 | 5600 | 0.0115 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1655
+ | 0.7030 | 5800 | 0.0255 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1656
+ | 0.7273 | 6000 | 0.0219 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1657
+ | 0.7515 | 6200 | 0.033 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1658
+ | 0.7758 | 6400 | 0.0199 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1659
+ | 0.8 | 6600 | 0.0175 | 0.0224 | 0.4700 | 0.2743 | 0.5136 | 0.4193 | - | - | - | - | - | - | - | - | - | - |
1660
+ | 0.8242 | 6800 | 0.0236 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1661
+ | 0.8485 | 7000 | 0.0145 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1662
+ | 0.8727 | 7200 | 0.0372 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1663
+ | 0.8970 | 7400 | 0.0107 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1664
+ | 0.9212 | 7600 | 0.0131 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1665
+ | 0.9455 | 7800 | 0.0127 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1666
+ | 0.9697 | 8000 | 0.0207 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1667
+ | 0.9939 | 8200 | 0.0217 | - | - | - | - | - | - | - | - | - | - | - | - | - | - | - |
1668
+ | **1.0** | **8250** | **-** | **0.0219** | **0.4958** | **0.2785** | **0.5342** | **0.4362** | **-** | **-** | **-** | **-** | **-** | **-** | **-** | **-** | **-** | **-** |
1669
+ | -1 | -1 | - | - | 0.4958 | 0.2785 | 0.5342 | 0.4974 | 0.2660 | 0.5285 | 0.6963 | 0.2843 | 0.7251 | 0.8827 | 0.3269 | 0.4417 | 0.5538 | 0.4521 |
1670
+
1671
+ * The bold row denotes the saved checkpoint.
1672
+
1673
+ ### Environmental Impact
1674
+ Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
1675
+ - **Energy Consumed**: 0.040 kWh
1676
+ - **Carbon Emitted**: 0.002 kg of CO2
1677
+ - **Hours Used**: 0.322 hours
1678
+
1679
+ ### Training Hardware
1680
+ - **On Cloud**: No
1681
+ - **GPU Model**: 1 x NVIDIA GeForce RTX 3070 Ti Laptop GPU
1682
+ - **CPU Model**: AMD Ryzen 9 6900HX with Radeon Graphics
1683
+ - **RAM Size**: 30.61 GB
1684
+
1685
+ ### Framework Versions
1686
+ - Python: 3.12.9
1687
+ - Sentence Transformers: 4.2.0.dev0
1688
+ - Transformers: 4.50.3
1689
+ - PyTorch: 2.6.0+cu124
1690
+ - Accelerate: 1.6.0
1691
+ - Datasets: 3.5.0
1692
+ - Tokenizers: 0.21.1
1693
+
1694
+ ## Citation
1695
+
1696
+ ### BibTeX
1697
+
1698
+ #### Sentence Transformers
1699
+ ```bibtex
1700
+ @inproceedings{reimers-2019-sentence-bert,
1701
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
1702
+ author = "Reimers, Nils and Gurevych, Iryna",
1703
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
1704
+ month = "11",
1705
+ year = "2019",
1706
+ publisher = "Association for Computational Linguistics",
1707
+ url = "https://arxiv.org/abs/1908.10084",
1708
+ }
1709
+ ```
1710
+
1711
+ #### SpladeLoss
1712
+ ```bibtex
1713
+ @misc{formal2022distillationhardnegativesampling,
1714
+ title={From Distillation to Hard Negative Sampling: Making Sparse Neural IR Models More Effective},
1715
+ author={Thibault Formal and Carlos Lassance and Benjamin Piwowarski and Stéphane Clinchant},
1716
+ year={2022},
1717
+ eprint={2205.04733},
1718
+ archivePrefix={arXiv},
1719
+ primaryClass={cs.IR},
1720
+ url={https://arxiv.org/abs/2205.04733},
1721
+ }
1722
+ ```
1723
+
1724
+ #### SparseMultipleNegativesRankingLoss
1725
+ ```bibtex
1726
+ @misc{henderson2017efficient,
1727
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
1728
+ author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
1729
+ year={2017},
1730
+ eprint={1705.00652},
1731
+ archivePrefix={arXiv},
1732
+ primaryClass={cs.CL}
1733
+ }
1734
+ ```
1735
+
1736
+ #### FlopsLoss
1737
+ ```bibtex
1738
+ @article{paria2020minimizing,
1739
+ title={Minimizing flops to learn efficient sparse representations},
1740
+ author={Paria, Biswajit and Yeh, Chih-Kuan and Yen, Ian EH and Xu, Ning and Ravikumar, Pradeep and P{'o}czos, Barnab{'a}s},
1741
+ journal={arXiv preprint arXiv:2004.05665},
1742
+ year={2020}
1743
+ }
1744
+ ```
1745
+
1746
+ <!--
1747
+ ## Glossary
1748
+
1749
+ *Clearly define terms in order to be accessible across audiences.*
1750
+ -->
1751
+
1752
+ <!--
1753
+ ## Model Card Authors
1754
+
1755
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
1756
+ -->
1757
+
1758
+ <!--
1759
+ ## Model Card Contact
1760
+
1761
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
1762
+ -->
config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "gelu",
3
+ "architectures": [
4
+ "DistilBertForMaskedLM"
5
+ ],
6
+ "attention_dropout": 0.1,
7
+ "dim": 768,
8
+ "dropout": 0.1,
9
+ "hidden_dim": 3072,
10
+ "initializer_range": 0.02,
11
+ "max_position_embeddings": 512,
12
+ "model_type": "distilbert",
13
+ "n_heads": 12,
14
+ "n_layers": 6,
15
+ "pad_token_id": 0,
16
+ "qa_dropout": 0.1,
17
+ "seq_classif_dropout": 0.2,
18
+ "sinusoidal_pos_embds": false,
19
+ "tie_weights_": true,
20
+ "torch_dtype": "float32",
21
+ "transformers_version": "4.50.3",
22
+ "vocab_size": 30522
23
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "SparseEncoder",
3
+ "__version__": {
4
+ "sentence_transformers": "4.2.0.dev0",
5
+ "transformers": "4.50.3",
6
+ "pytorch": "2.6.0+cu124"
7
+ },
8
+ "prompts": {},
9
+ "default_prompt_name": null,
10
+ "similarity_fn_name": "dot"
11
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71983673d3beaf042160ad432b7162cff1ac0fe7f0f75c6826334dcd18a88ff0
3
+ size 267954768
modules.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.sparse_encoder.models.MLMTransformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_SpladePooling",
12
+ "type": "sentence_transformers.sparse_encoder.models.SpladePooling"
13
+ }
14
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 256,
3
+ "do_lower_case": false
4
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "DistilBertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff