RonanMcGovern commited on
Commit
195c123
1 Parent(s): c536d3e

Add new SentenceTransformer model.

Browse files
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 384,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
README.md ADDED
@@ -0,0 +1,531 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: cross-encoder/ms-marco-MiniLM-L-6-v2
3
+ datasets: []
4
+ language: []
5
+ library_name: sentence-transformers
6
+ pipeline_tag: sentence-similarity
7
+ tags:
8
+ - sentence-transformers
9
+ - sentence-similarity
10
+ - feature-extraction
11
+ - generated_from_trainer
12
+ - dataset_size:211
13
+ - loss:MultipleNegativesRankingLoss
14
+ widget:
15
+ - source_sentence: What happens if a player in possession enters the defending team's
16
+ seven-metre zone?
17
+ sentences:
18
+ - 10. 8 if a touch is made in the in - goal area before the ball is grounded, the
19
+ player in possession is to perform a rollball seven ( 7 ) metres from the team
20
+ ’ s attacking try line, provided it is not the sixth touch and the player is not
21
+ half. 10. 9 if a player in possession is touched while on or behind their defending
22
+ try line, the touch counts and once the referee sets the mark seven ( 7 ) metres
23
+ directly forward of the contact point from the defending team ’ s try line, a
24
+ rollball is performed. 10. 10 if a player in possession intentionally makes a
25
+ touch on an offside defender who is making every effort to retire and remain out
26
+ of play, the touch counts. fit playing rules - 5th edition copyright © touch football
27
+ australia 2020 9 10. 11 if a touch is made on a player in possession while the
28
+ player is juggling the ball in an attempt to maintain control of it, the touch
29
+ counts if the attacking player following the touch retains possession.
30
+ - 9. 2 on the change of possession due to an intercept, the first touch will be
31
+ zero ( 0 ) touch. 9. 3 following the sixth touch or a loss of possession due to
32
+ any other means, the ball must be returned to the mark without delay. ruling =
33
+ a deliberate delay in the changeover procedure will result in a penalty awarded
34
+ to the non - offending team ten ( 10 ) metres forward of the mark for the change
35
+ of possession. 9. 4 if the ball is dropped or passed and goes to ground during
36
+ play, a change of possession results. ruling = the mark for the change of possession
37
+ is where the ball makes initial contact with the ground. 9. 5 if the ball, while
38
+ still under the control of the half, contacts the ground in the in - goal area,
39
+ possession is lost. ruling = play will restart with a rollball at the nearest
40
+ point on the seven ( 7 ) metre line. fit playing rules - 5th edition 8 copyright
41
+ © touch football australia 2020 9. 6 if a player mishandles the ball and even
42
+ if in an effort to gain control, the ball is accidentally knocked forward into
43
+ any other player, a change of possession results.
44
+ - fit playing rules - 5th edition copyright © touch football australia 2020 9 10.
45
+ 11 if a touch is made on a player in possession while the player is juggling the
46
+ ball in an attempt to maintain control of it, the touch counts if the attacking
47
+ player following the touch retains possession. 10. 12 if a player in possession
48
+ is touched and subsequently makes contact with either the sideline, a field marker
49
+ or the ground outside the field of play, the touch counts and play continues with
50
+ a rollball at the mark where the touch occurred. 10. 13 when a player from the
51
+ defending team enters its defensive seven metre zone, the defending team must
52
+ move forward at a reasonable pace until a touch is imminent or made. ruling =
53
+ a penalty to the attacking team at the point of the infringement. 10. 14 when
54
+ a player in possession enters the defending teams ’ seven metre zone the defending
55
+ team is not obliged to move forward but cannot retire back towards their try line
56
+ until a touch is imminent or made. ruling = a penalty to the attacking team at
57
+ the seven ( 7 ) metre line in line with the point of the infringement.
58
+ - source_sentence: What is the maximum number of touches a team can have before a
59
+ change of possession?
60
+ sentences:
61
+ - touch count the progressive number of touches that each team has before a change
62
+ of possession, from zero ( 0 ) to six ( 6 ). try the result of any attacking player,
63
+ except the half, placing the ball on or over the team ’ s attacking try line before
64
+ being touched. try lines the lines separating the in - goal areas from the field
65
+ of play. see appendix 1. voluntary rollball the player in possession performs
66
+ a rollball before a touch is made with a defending player. wing the player outside
67
+ the link player. winner the team that scores the most tries during the match.
68
+ fit playing rules - 5th edition 4 copyright © touch football australia 2020 rules
69
+ of play mode of play the object of the game of touch is for each team to score
70
+ tries and to prevent the opposition from scoring. the ball may be passed, knocked
71
+ or handed between players of the attacking team who may in turn run or otherwise
72
+ move with the ball in an attempt to gain territorial advantage and to score tries.
73
+ defending players prevent the attacking team from gaining a territorial advantage
74
+ by touching the ball carrier.
75
+ - 4. 3. 1 identifying numbers must feature no more than two ( 2 ) digits. 4. 4 hats
76
+ or caps are permitted to be worn during a match provided they are safe and meet
77
+ any nta regulations. 4. 5 safe footwear must be worn with exceptions allowed for
78
+ game variants such as beach touch. 4. 6 light leather or synthetic boots with
79
+ soft moulded soles are permitted. 4. 6. 1 shoes with screw - in studs are not
80
+ to be worn by any player or referee. 4. 7 players are not to participate in any
81
+ match wearing any item of jewellery, chain, identification band / bracelet or
82
+ similar item that may prove dangerous. any jewellery or other items that cannot
83
+ be removed are to be taped to the satisfaction of the referee. 4. 8 long ( extend
84
+ beyond the finger flesh when viewed from the palm ) or sharp fingernails are not
85
+ allowed. 4. 9 referees and players may wear spectacles or sunglasses provided
86
+ they are safe and securely attached. 4. 10 referees and players may wear sport
87
+ monitoring equipment and medical supports such as knee or ankle braces provided,
88
+ at the sole discretion of competition ’ s controlling body, the items are not
89
+ dangerous.
90
+ - a player with both feet on or behind their defending try line. pass the act of
91
+ changing possession between individual attacking players by propelling the ball
92
+ laterally and / or backwards and may include a flick, knock or throw. perimeter
93
+ a border not less than five ( 5 ) metres from the boundary of the field of play.
94
+ see appendix 1. penalty the ruling by a referee to award a tap when a player or
95
+ team infringes the rules of the game. possession refers to the player or team
96
+ that has control of the ball. providing other rules do not apply, the team with
97
+ the ball is entitled to six ( 6 ) touches. referee the match official ( s ) appointed
98
+ to make rulings during the conduct of a match. rollball the act of bringing the
99
+ ball into play following a touch or a change of possession. ruck / rollball area
100
+ the area, not exceeding one ( 1 ) metre in distance, between the player performing
101
+ a rollball and the half. ruling the decision made by a referee as a result of
102
+ particular circumstance and may result in a play on, a tap penalty, a discipline
103
+ option, change of possession or a try. seven metre zone the area between the seven
104
+ ( 7 ) metre line and the try line.
105
+ - source_sentence: What is the definition of 'forward' in Touch Rugby?
106
+ sentences:
107
+ - end of play when the referee indicates completion of the match. exclusion when
108
+ a player is sent to the nearest sin bin area following three ( 3 ) penalties by
109
+ the defending team upon entering their seven metre zone. the player is counted
110
+ as a player on the field of play and cannot be replaced or interchanged. fit playing
111
+ rules - 5th edition copyright © touch football australia 2020 1 fit federation
112
+ of international touch field of play the playing area bounded by the sidelines
113
+ and dead ball lines, both of which are out of bounds. see appendix 1. forced interchange
114
+ when a player is required to undertake a compulsory interchange for an infringement
115
+ ruled more serious than a penalty but less serious than a permanent interchange,
116
+ sin bin or dismissal. forward a position or direction towards the dead ball line
117
+ beyond the team ’ s attacking try line. full time the expiration of the second
118
+ period of time allowed for play. half the player who takes possession following
119
+ a rollball. half time the break in play between the two halves of a match. imminent
120
+ about to occur, it is almost certain to occur. infringement the action of a player
121
+ contrary to the rules of the game.
122
+ - fit playing rules - 5th edition copyright © touch football australia 2020 9 10.
123
+ 11 if a touch is made on a player in possession while the player is juggling the
124
+ ball in an attempt to maintain control of it, the touch counts if the attacking
125
+ player following the touch retains possession. 10. 12 if a player in possession
126
+ is touched and subsequently makes contact with either the sideline, a field marker
127
+ or the ground outside the field of play, the touch counts and play continues with
128
+ a rollball at the mark where the touch occurred. 10. 13 when a player from the
129
+ defending team enters its defensive seven metre zone, the defending team must
130
+ move forward at a reasonable pace until a touch is imminent or made. ruling =
131
+ a penalty to the attacking team at the point of the infringement. 10. 14 when
132
+ a player in possession enters the defending teams ’ seven metre zone the defending
133
+ team is not obliged to move forward but cannot retire back towards their try line
134
+ until a touch is imminent or made. ruling = a penalty to the attacking team at
135
+ the seven ( 7 ) metre line in line with the point of the infringement.
136
+ - 10. 8 if a touch is made in the in - goal area before the ball is grounded, the
137
+ player in possession is to perform a rollball seven ( 7 ) metres from the team
138
+ ’ s attacking try line, provided it is not the sixth touch and the player is not
139
+ half. 10. 9 if a player in possession is touched while on or behind their defending
140
+ try line, the touch counts and once the referee sets the mark seven ( 7 ) metres
141
+ directly forward of the contact point from the defending team ’ s try line, a
142
+ rollball is performed. 10. 10 if a player in possession intentionally makes a
143
+ touch on an offside defender who is making every effort to retire and remain out
144
+ of play, the touch counts. fit playing rules - 5th edition copyright © touch football
145
+ australia 2020 9 10. 11 if a touch is made on a player in possession while the
146
+ player is juggling the ball in an attempt to maintain control of it, the touch
147
+ counts if the attacking player following the touch retains possession.
148
+ - source_sentence: What happens if neither team is leading at the end of the two-minute
149
+ period of extra time?
150
+ sentences:
151
+ - infringement the action of a player contrary to the rules of the game. in - goal
152
+ area the area in the field of play bounded by the sidelines, the try lines and
153
+ the dead ball lines. there are two ( 2 ), one ( 1 ) at each end of the field of
154
+ play. see appendix 1. interchange the act of an on - field player leaving the
155
+ field of play to be replaced by an off - field player entering the field of play.
156
+ interchange area a marked rectangle for each team on opposite sides of the field
157
+ of play usually measuring 20 metres long by no more than five ( 5 ) metres wide,
158
+ extending ten ( 10 ) metres either side of the halfway line and not less than
159
+ one ( 1 ) metre from the sideline. it is the area in which all off - field players
160
+ must remain until an interchange is initiated. see appendix 1. kick strike or
161
+ propel forcibly with the foot, a blow or forceful thrust with the foot to the
162
+ ball. a tap to commence or recommence play or a penalty tap is not defined as
163
+ a kick. line markings markings of the field of play. see appendix 1. link the
164
+ player beside the wing player.
165
+ - touch count the progressive number of touches that each team has before a change
166
+ of possession, from zero ( 0 ) to six ( 6 ). try the result of any attacking player,
167
+ except the half, placing the ball on or over the team ’ s attacking try line before
168
+ being touched. try lines the lines separating the in - goal areas from the field
169
+ of play. see appendix 1. voluntary rollball the player in possession performs
170
+ a rollball before a touch is made with a defending player. wing the player outside
171
+ the link player. winner the team that scores the most tries during the match.
172
+ fit playing rules - 5th edition 4 copyright © touch football australia 2020 rules
173
+ of play mode of play the object of the game of touch is for each team to score
174
+ tries and to prevent the opposition from scoring. the ball may be passed, knocked
175
+ or handed between players of the attacking team who may in turn run or otherwise
176
+ move with the ball in an attempt to gain territorial advantage and to score tries.
177
+ defending players prevent the attacking team from gaining a territorial advantage
178
+ by touching the ball carrier.
179
+ - 24. 1. 2 the drop - off commences with a tap from the centre of the halfway line
180
+ by the team that did not commence the match with possession. 24. 1. 3 the drop
181
+ - off will commence with a two ( 2 ) minute period of extra time. 24. 1. 4 should
182
+ a team be leading at the expiration of the two ( 2 ) minute period of extra time
183
+ then that team will be declared the winner and match complete. 24. 1. 5 should
184
+ neither team be leading at the expiration of two ( 2 ) minutes, a signal is given
185
+ and the match will pause at the next touch or dead ball. each team will then remove
186
+ another player from the field of play. 24. 1. 6 the match will recommence immediately
187
+ after the players have left the field at the same place where it paused ( i. e.
188
+ the team retains possession at the designated number of touches, or at change
189
+ of possession due to some infringement or the sixth touch ) and the match will
190
+ continue until a try is scored. 24. 1. 7 there is no time off during the drop
191
+ - off and the clock does not stop at the two ( 2 ) minute interval.
192
+ - source_sentence: What happens if a team is leading at the end of the two-minute
193
+ period of extra time?
194
+ sentences:
195
+ - 24. 1. 2 the drop - off commences with a tap from the centre of the halfway line
196
+ by the team that did not commence the match with possession. 24. 1. 3 the drop
197
+ - off will commence with a two ( 2 ) minute period of extra time. 24. 1. 4 should
198
+ a team be leading at the expiration of the two ( 2 ) minute period of extra time
199
+ then that team will be declared the winner and match complete. 24. 1. 5 should
200
+ neither team be leading at the expiration of two ( 2 ) minutes, a signal is given
201
+ and the match will pause at the next touch or dead ball. each team will then remove
202
+ another player from the field of play. 24. 1. 6 the match will recommence immediately
203
+ after the players have left the field at the same place where it paused ( i. e.
204
+ the team retains possession at the designated number of touches, or at change
205
+ of possession due to some infringement or the sixth touch ) and the match will
206
+ continue until a try is scored. 24. 1. 7 there is no time off during the drop
207
+ - off and the clock does not stop at the two ( 2 ) minute interval.
208
+ - 7. 7 the tap to commence or recommence play must be performed without delay. ruling
209
+ = a penalty to the non - offending team at the centre of the halfway line. 8 match
210
+ duration 8. 1 a match is 40 minutes in duration, consisting of two ( 2 ) x 20
211
+ minute halves with a half time break. 8. 1. 1 there is no time off for injury
212
+ during a match. 8. 2 local competition and tournament conditions may vary the
213
+ duration of a match. 8. 3 when time expires, play is to continue until the next
214
+ touch or dead ball and end of play is signaled by the referee. 8. 3. 1 should
215
+ a penalty be awarded during this period, the penalty is to be taken. 8. 4 if a
216
+ match is abandoned in any circumstances other than those referred to in clause
217
+ 24. 1. 6 the nta or nta competition provider in its sole discretion shall determine
218
+ the result of the match. 9 possession 9. 1 the team with the ball is entitled
219
+ to six ( 6 ) touches prior to a change of possession. 9. 2 on the change of possession
220
+ due to an intercept, the first touch will be zero ( 0 ) touch.
221
+ - '12. 6 if a player from the defending team unintentionally makes contact with
222
+ the ball in flight and the ball is retrieved by an attacking player, play and
223
+ the touch count continues. 12. 7 a player from the attacking team cannot pass
224
+ the ball into a defending player intentionally seeking a rebound or a restart
225
+ of the touch count. ruling = a penalty to the defending team at the point of the
226
+ pass. 13 the rollball 13. 1 the attacking player is to position on the mark, face
227
+ the opponent ’ s try line, make a genuine attempt to stand parallel to the sidelines,
228
+ place the ball on the ground between the feet in a controlled manner and : 13.
229
+ 1. 1 step forward over the ball ; or 13. 1. 2 roll the ball back between the feet
230
+ no more than one ( 1 ) metre ; or 13. 1. 3 pass a foot over the ball. ruling =
231
+ a change of possession to the defending team at the point of the infringement.
232
+ 13. 2 a player must perform the rollball on the mark. ruling = a penalty to the
233
+ defending team at the point of the infringement. 13. 3 a player must not perform
234
+ a voluntary rollball.'
235
+ ---
236
+
237
+ # SentenceTransformer based on cross-encoder/ms-marco-MiniLM-L-6-v2
238
+
239
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [cross-encoder/ms-marco-MiniLM-L-6-v2](https://huggingface.co/cross-encoder/ms-marco-MiniLM-L-6-v2). It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
240
+
241
+ ## Model Details
242
+
243
+ ### Model Description
244
+ - **Model Type:** Sentence Transformer
245
+ - **Base model:** [cross-encoder/ms-marco-MiniLM-L-6-v2](https://huggingface.co/cross-encoder/ms-marco-MiniLM-L-6-v2) <!-- at revision b2cfda50a1a9fc7919e7444afbb52610d268af92 -->
246
+ - **Maximum Sequence Length:** 512 tokens
247
+ - **Output Dimensionality:** 384 tokens
248
+ - **Similarity Function:** Cosine Similarity
249
+ <!-- - **Training Dataset:** Unknown -->
250
+ <!-- - **Language:** Unknown -->
251
+ <!-- - **License:** Unknown -->
252
+
253
+ ### Model Sources
254
+
255
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
256
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
257
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
258
+
259
+ ### Full Model Architecture
260
+
261
+ ```
262
+ SentenceTransformer(
263
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel
264
+ (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
265
+ )
266
+ ```
267
+
268
+ ## Usage
269
+
270
+ ### Direct Usage (Sentence Transformers)
271
+
272
+ First install the Sentence Transformers library:
273
+
274
+ ```bash
275
+ pip install -U sentence-transformers
276
+ ```
277
+
278
+ Then you can load this model and run inference.
279
+ ```python
280
+ from sentence_transformers import SentenceTransformer
281
+
282
+ # Download from the 🤗 Hub
283
+ model = SentenceTransformer("Trelis/ms-marco-MiniLM-L-6-v2-2-constant-ep-MNRLpairs-2e-5-batch32-cuda-overlap")
284
+ # Run inference
285
+ sentences = [
286
+ 'What happens if a team is leading at the end of the two-minute period of extra time?',
287
+ '24. 1. 2 the drop - off commences with a tap from the centre of the halfway line by the team that did not commence the match with possession. 24. 1. 3 the drop - off will commence with a two ( 2 ) minute period of extra time. 24. 1. 4 should a team be leading at the expiration of the two ( 2 ) minute period of extra time then that team will be declared the winner and match complete. 24. 1. 5 should neither team be leading at the expiration of two ( 2 ) minutes, a signal is given and the match will pause at the next touch or dead ball. each team will then remove another player from the field of play. 24. 1. 6 the match will recommence immediately after the players have left the field at the same place where it paused ( i. e. the team retains possession at the designated number of touches, or at change of possession due to some infringement or the sixth touch ) and the match will continue until a try is scored. 24. 1. 7 there is no time off during the drop - off and the clock does not stop at the two ( 2 ) minute interval.',
288
+ '7. 7 the tap to commence or recommence play must be performed without delay. ruling = a penalty to the non - offending team at the centre of the halfway line. 8 match duration 8. 1 a match is 40 minutes in duration, consisting of two ( 2 ) x 20 minute halves with a half time break. 8. 1. 1 there is no time off for injury during a match. 8. 2 local competition and tournament conditions may vary the duration of a match. 8. 3 when time expires, play is to continue until the next touch or dead ball and end of play is signaled by the referee. 8. 3. 1 should a penalty be awarded during this period, the penalty is to be taken. 8. 4 if a match is abandoned in any circumstances other than those referred to in clause 24. 1. 6 the nta or nta competition provider in its sole discretion shall determine the result of the match. 9 possession 9. 1 the team with the ball is entitled to six ( 6 ) touches prior to a change of possession. 9. 2 on the change of possession due to an intercept, the first touch will be zero ( 0 ) touch.',
289
+ ]
290
+ embeddings = model.encode(sentences)
291
+ print(embeddings.shape)
292
+ # [3, 384]
293
+
294
+ # Get the similarity scores for the embeddings
295
+ similarities = model.similarity(embeddings, embeddings)
296
+ print(similarities.shape)
297
+ # [3, 3]
298
+ ```
299
+
300
+ <!--
301
+ ### Direct Usage (Transformers)
302
+
303
+ <details><summary>Click to see the direct usage in Transformers</summary>
304
+
305
+ </details>
306
+ -->
307
+
308
+ <!--
309
+ ### Downstream Usage (Sentence Transformers)
310
+
311
+ You can finetune this model on your own dataset.
312
+
313
+ <details><summary>Click to expand</summary>
314
+
315
+ </details>
316
+ -->
317
+
318
+ <!--
319
+ ### Out-of-Scope Use
320
+
321
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
322
+ -->
323
+
324
+ <!--
325
+ ## Bias, Risks and Limitations
326
+
327
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
328
+ -->
329
+
330
+ <!--
331
+ ### Recommendations
332
+
333
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
334
+ -->
335
+
336
+ ## Training Details
337
+
338
+ ### Training Hyperparameters
339
+ #### Non-Default Hyperparameters
340
+
341
+ - `eval_strategy`: steps
342
+ - `per_device_train_batch_size`: 32
343
+ - `per_device_eval_batch_size`: 32
344
+ - `learning_rate`: 2e-05
345
+ - `num_train_epochs`: 2
346
+ - `lr_scheduler_type`: constant
347
+ - `warmup_ratio`: 0.3
348
+ - `bf16`: True
349
+
350
+ #### All Hyperparameters
351
+ <details><summary>Click to expand</summary>
352
+
353
+ - `overwrite_output_dir`: False
354
+ - `do_predict`: False
355
+ - `eval_strategy`: steps
356
+ - `prediction_loss_only`: True
357
+ - `per_device_train_batch_size`: 32
358
+ - `per_device_eval_batch_size`: 32
359
+ - `per_gpu_train_batch_size`: None
360
+ - `per_gpu_eval_batch_size`: None
361
+ - `gradient_accumulation_steps`: 1
362
+ - `eval_accumulation_steps`: None
363
+ - `learning_rate`: 2e-05
364
+ - `weight_decay`: 0.0
365
+ - `adam_beta1`: 0.9
366
+ - `adam_beta2`: 0.999
367
+ - `adam_epsilon`: 1e-08
368
+ - `max_grad_norm`: 1.0
369
+ - `num_train_epochs`: 2
370
+ - `max_steps`: -1
371
+ - `lr_scheduler_type`: constant
372
+ - `lr_scheduler_kwargs`: {}
373
+ - `warmup_ratio`: 0.3
374
+ - `warmup_steps`: 0
375
+ - `log_level`: passive
376
+ - `log_level_replica`: warning
377
+ - `log_on_each_node`: True
378
+ - `logging_nan_inf_filter`: True
379
+ - `save_safetensors`: True
380
+ - `save_on_each_node`: False
381
+ - `save_only_model`: False
382
+ - `restore_callback_states_from_checkpoint`: False
383
+ - `no_cuda`: False
384
+ - `use_cpu`: False
385
+ - `use_mps_device`: False
386
+ - `seed`: 42
387
+ - `data_seed`: None
388
+ - `jit_mode_eval`: False
389
+ - `use_ipex`: False
390
+ - `bf16`: True
391
+ - `fp16`: False
392
+ - `fp16_opt_level`: O1
393
+ - `half_precision_backend`: auto
394
+ - `bf16_full_eval`: False
395
+ - `fp16_full_eval`: False
396
+ - `tf32`: None
397
+ - `local_rank`: 0
398
+ - `ddp_backend`: None
399
+ - `tpu_num_cores`: None
400
+ - `tpu_metrics_debug`: False
401
+ - `debug`: []
402
+ - `dataloader_drop_last`: False
403
+ - `dataloader_num_workers`: 0
404
+ - `dataloader_prefetch_factor`: None
405
+ - `past_index`: -1
406
+ - `disable_tqdm`: False
407
+ - `remove_unused_columns`: True
408
+ - `label_names`: None
409
+ - `load_best_model_at_end`: False
410
+ - `ignore_data_skip`: False
411
+ - `fsdp`: []
412
+ - `fsdp_min_num_params`: 0
413
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
414
+ - `fsdp_transformer_layer_cls_to_wrap`: None
415
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
416
+ - `deepspeed`: None
417
+ - `label_smoothing_factor`: 0.0
418
+ - `optim`: adamw_torch
419
+ - `optim_args`: None
420
+ - `adafactor`: False
421
+ - `group_by_length`: False
422
+ - `length_column_name`: length
423
+ - `ddp_find_unused_parameters`: None
424
+ - `ddp_bucket_cap_mb`: None
425
+ - `ddp_broadcast_buffers`: False
426
+ - `dataloader_pin_memory`: True
427
+ - `dataloader_persistent_workers`: False
428
+ - `skip_memory_metrics`: True
429
+ - `use_legacy_prediction_loop`: False
430
+ - `push_to_hub`: False
431
+ - `resume_from_checkpoint`: None
432
+ - `hub_model_id`: None
433
+ - `hub_strategy`: every_save
434
+ - `hub_private_repo`: False
435
+ - `hub_always_push`: False
436
+ - `gradient_checkpointing`: False
437
+ - `gradient_checkpointing_kwargs`: None
438
+ - `include_inputs_for_metrics`: False
439
+ - `eval_do_concat_batches`: True
440
+ - `fp16_backend`: auto
441
+ - `push_to_hub_model_id`: None
442
+ - `push_to_hub_organization`: None
443
+ - `mp_parameters`:
444
+ - `auto_find_batch_size`: False
445
+ - `full_determinism`: False
446
+ - `torchdynamo`: None
447
+ - `ray_scope`: last
448
+ - `ddp_timeout`: 1800
449
+ - `torch_compile`: False
450
+ - `torch_compile_backend`: None
451
+ - `torch_compile_mode`: None
452
+ - `dispatch_batches`: None
453
+ - `split_batches`: None
454
+ - `include_tokens_per_second`: False
455
+ - `include_num_input_tokens_seen`: False
456
+ - `neftune_noise_alpha`: None
457
+ - `optim_target_modules`: None
458
+ - `batch_eval_metrics`: False
459
+ - `eval_on_start`: False
460
+ - `batch_sampler`: batch_sampler
461
+ - `multi_dataset_batch_sampler`: proportional
462
+
463
+ </details>
464
+
465
+ ### Training Logs
466
+ | Epoch | Step | Training Loss | loss |
467
+ |:------:|:----:|:-------------:|:------:|
468
+ | 0.2857 | 2 | 3.7086 | 3.1734 |
469
+ | 0.5714 | 4 | 3.4714 | 3.0742 |
470
+ | 0.8571 | 6 | 3.41 | 3.0204 |
471
+ | 1.1429 | 8 | 3.042 | 2.9657 |
472
+ | 1.4286 | 10 | 3.3335 | 2.9125 |
473
+ | 1.7143 | 12 | 3.2224 | 2.8573 |
474
+ | 2.0 | 14 | 2.9969 | 2.8300 |
475
+
476
+
477
+ ### Framework Versions
478
+ - Python: 3.10.12
479
+ - Sentence Transformers: 3.0.1
480
+ - Transformers: 4.42.3
481
+ - PyTorch: 2.1.1+cu121
482
+ - Accelerate: 0.31.0
483
+ - Datasets: 2.17.1
484
+ - Tokenizers: 0.19.1
485
+
486
+ ## Citation
487
+
488
+ ### BibTeX
489
+
490
+ #### Sentence Transformers
491
+ ```bibtex
492
+ @inproceedings{reimers-2019-sentence-bert,
493
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
494
+ author = "Reimers, Nils and Gurevych, Iryna",
495
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
496
+ month = "11",
497
+ year = "2019",
498
+ publisher = "Association for Computational Linguistics",
499
+ url = "https://arxiv.org/abs/1908.10084",
500
+ }
501
+ ```
502
+
503
+ #### MultipleNegativesRankingLoss
504
+ ```bibtex
505
+ @misc{henderson2017efficient,
506
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
507
+ 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},
508
+ year={2017},
509
+ eprint={1705.00652},
510
+ archivePrefix={arXiv},
511
+ primaryClass={cs.CL}
512
+ }
513
+ ```
514
+
515
+ <!--
516
+ ## Glossary
517
+
518
+ *Clearly define terms in order to be accessible across audiences.*
519
+ -->
520
+
521
+ <!--
522
+ ## Model Card Authors
523
+
524
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
525
+ -->
526
+
527
+ <!--
528
+ ## Model Card Contact
529
+
530
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
531
+ -->
config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "cross-encoder/ms-marco-MiniLM-L-6-v2",
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "classifier_dropout": null,
8
+ "gradient_checkpointing": false,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 384,
12
+ "id2label": {
13
+ "0": "LABEL_0"
14
+ },
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 1536,
17
+ "label2id": {
18
+ "LABEL_0": 0
19
+ },
20
+ "layer_norm_eps": 1e-12,
21
+ "max_position_embeddings": 512,
22
+ "model_type": "bert",
23
+ "num_attention_heads": 12,
24
+ "num_hidden_layers": 6,
25
+ "pad_token_id": 0,
26
+ "position_embedding_type": "absolute",
27
+ "sbert_ce_default_activation_function": "torch.nn.modules.linear.Identity",
28
+ "torch_dtype": "float32",
29
+ "transformers_version": "4.42.3",
30
+ "type_vocab_size": 2,
31
+ "use_cache": true,
32
+ "vocab_size": 30522
33
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.0.1",
4
+ "transformers": "4.42.3",
5
+ "pytorch": "2.1.1+cu121"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null,
9
+ "similarity_fn_name": null
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:240bd4ab0edc659d30a168627e60f2bccf62f0f63b72dc22bb4c9a65fdeca782
3
+ size 90864192
modules.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ }
14
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
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 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "never_split": null,
51
+ "pad_token": "[PAD]",
52
+ "sep_token": "[SEP]",
53
+ "strip_accents": null,
54
+ "tokenize_chinese_chars": true,
55
+ "tokenizer_class": "BertTokenizer",
56
+ "unk_token": "[UNK]"
57
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff