guyhadad01 commited on
Commit
fbb52e5
·
verified ·
1 Parent(s): 6ccdddf

Training in progress, step 200, checkpoint

Browse files
.gitattributes CHANGED
@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
  tokenizer.json filter=lfs diff=lfs merge=lfs -text
 
 
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
  tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ last-checkpoint/tokenizer.json filter=lfs diff=lfs merge=lfs -text
last-checkpoint/1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 768,
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
+ }
last-checkpoint/2_Dense/config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "in_features": 768,
3
+ "out_features": 3072,
4
+ "bias": false,
5
+ "activation_function": "torch.nn.modules.linear.Identity"
6
+ }
last-checkpoint/2_Dense/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:245987c37984397149e2b7782bea1deeb6b41895e4c73137a0745ae1034fae74
3
+ size 4718680
last-checkpoint/3_Dense/config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "in_features": 3072,
3
+ "out_features": 768,
4
+ "bias": false,
5
+ "activation_function": "torch.nn.modules.linear.Identity"
6
+ }
last-checkpoint/3_Dense/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c06029654776794583e6462e88fb48f86fae257f300ef5195e19451cdd07501d
3
+ size 4718680
last-checkpoint/README.md ADDED
@@ -0,0 +1,744 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - sentence-similarity
5
+ - feature-extraction
6
+ - dense
7
+ - generated_from_trainer
8
+ - dataset_size:1007606
9
+ - loss:CachedMultipleNegativesRankingLoss
10
+ base_model: google/embeddinggemma-300m
11
+ widget:
12
+ - source_sentence: "title: \nPhi Tee Tail Retractable Golf Towel Set Portable Golf\
13
+ \ Ball Washer 12X12"
14
+ sentences:
15
+ - 'description
16
+
17
+ Want to hold up to 50 rounds of ammo conveniently These flip top boxes are available
18
+ in five sizes to accommodate any type of handgun ammo They are stackable for storage
19
+ feature an easytogrip textured surface and come with a load label
20
+
21
+ All MTM CaseGards are MADE IN USA in our factory in Dayton OhioFor 45 ACP 10mm
22
+ Auto 40 SW 41 Action Express 357 Sig 44 Auto Mag 45 Auto Rim 45 GAP 38 Casull
23
+ 400 CorBon 8mm Nambu
24
+
25
+ Scuffresistant textured surface
26
+
27
+ Maximum Overall Length  130
28
+
29
+ Stacking feet mechanical hinge and load labels
30
+
31
+ Snap lock latch 25 year guarantee Made in USA'
32
+ - 'description
33
+
34
+ Innovative Golf TowelBall Washer System for Every Golfer Cleverly solves the drawbacks
35
+ associated with golf towels Rare but if defective components are found free replacements
36
+ will be sent out upon request All of our products are HighQuality GUARANTEEDRetractable
37
+ Wearable Belt Loop dont forget or lose with 2 Quality Pocket Water Spray Pens
38
+
39
+ The Most Simple and Effective Portable Golf Ball Washer Set
40
+
41
+ Premium Microfiber Waffle Texture Towel Perfect 12X12 Size for Ball and Club
42
+ Head Cleaning
43
+
44
+ Easily Attached and Detached from the Durable Retractable Keyring for Frequent
45
+ Washing
46
+
47
+ All HighQuality Components Any Defects will be Replaced for Free'
48
+ - 'description
49
+
50
+ The Dura Outdoor Pickleballs also known as the Dura Fast 40 are a seamless plastic
51
+ whiffle ball specially designed for pickleball They like all outdoor pickleballs
52
+ have thick walls The purpose of using hard plastic for a pickleball is to increase
53
+ the life span of the ball in rough outdoor conditions The name of the ball is
54
+ short for durable and these balls live up to their name We are sure you will love
55
+ playing with them We certainly do So a little bit about the manufacturing process
56
+ To start the ball is made by injecting a single glob of hot plastic into a mold
57
+ The mold is closed and rotated to spread the plastic evenly around the walls of
58
+ the mold When the mold has stopped spinning it is opened and the new ball is removed
59
+ from the mold Sometimes there is a minor separation seam from excess plastic that
60
+ leaked out of the mold This excess plastic is sanded off Look inside the ball
61
+ and you will not see a seam We double dare you There are two sizes of holes drilled
62
+ into this ball The two sizes of holes and the hole configuration help make the
63
+ ball more aerodynamic The Dura Outdoor Pickleballs is a heavy ball weighing 092
64
+ ounces This weight helps keep the ball flying straight even in windy conditions
65
+ The weight also makes them a fast ball compared to the softer and lighter jugs
66
+ indoor pickleball The Dura are also a rather big the ball measures 29375 in diameter
67
+ The ball has an average bounce height of 32 promising lively response to every
68
+ hit Formerly called the Dura 56 the manufacturer changed the name several years
69
+ ago to the Dura Fast 40 We are happy with either of the names frankly For many
70
+ years this ball has been the official ball for the USA Pickleball Associations
71
+ National Pickleball Tournament so you know its got to be a great choice in the
72
+ pickleball world The Dura brand ball is the original pickleball although now there
73
+ are several companies making identical outdoor pickleballs such a'
74
+ - source_sentence: "title: \n47 NFL Atlanta Falcons Womens Trytop Cuff Knit Hat with\
75
+ \ Pom Red"
76
+ sentences:
77
+ - 'description
78
+
79
+ Show off your team spirit in style with officiallylicensed NFL team headwear apparel
80
+ from 47 Brand 47 provides the quality all true fans desire in their gear Known
81
+ for their vintage look and feel 47 has managed to also provide a new school spin
82
+ to this old school craze Their presentday success comes from never forgetting
83
+ their roots In 1947 twin brothers and Italian immigrants Arthur and Henry DAngelo
84
+ founded their company Twins Enterprises in Boston MA The DAngelos sold pennants
85
+ and other sports memorabilia on the streets around Fenway Park and through a combination
86
+ of hard work sound instincts and incredible passion the brothers were able to
87
+ grow their business from a single street cart to a premier sports lifestyle brand
88
+ that uniquely melds sport and style Now known as 47 Brand they produce a unique
89
+ mix of the finest headwear and apparel with an unparalleled attention to detail
90
+ which has helped established them as a premium global sportswear brand wellknown
91
+ by fans the world over 47 is proud to be an Officially Licensed partner with the
92
+ four key professional American sports leaguesMLB NFL NBA NHLas well as over 650
93
+ NCAA colleges universitiesImported
94
+
95
+ Support your favorite team in style comfort
96
+
97
+ Officially licensed product of the National Football League
98
+
99
+ 47 Brand produces only the finest sportswear for the fashionconscious fan
100
+
101
+ Widely known and loved by sports fans for its vintage look feel
102
+
103
+ Check out all new 47 caps knits tshirts hoodies pullovers socks scarves and more available
104
+ for all your favorite teams'
105
+ - 'description
106
+
107
+ 3 Pair Lenses Black Ice Blue Mirror Coating Titanium Mirror Coating
108
+
109
+ Features
110
+
111
+ 1 Made of premium quality which offers 100 UV Protection 100 Polarized 2 Precision
112
+ Cut and Guaranteed to fit with the original Oakley Eyepatch 2 Sunglasses frame
113
+ extremely seamless 3 Optional Mirror Coatings optimize usable light for specific
114
+ environment and activities 4 Easy to install 5 Reduces glare and enhances contrast
115
+ perfectly 6 Comes with fibre cleaning cloth and compatible size of hard protection
116
+ case
117
+
118
+ Not include the Frame
119
+
120
+ Disclaimer Our lenses are not affiliated with Oakley in any aspects
121
+
122
+ This Lenses Fit Oakley Model
123
+
124
+ Eyepatch 1 Not Fit
125
+
126
+ Eyepatch 2 Fit
127
+
128
+ Make sure the oakley model and lenses is the right one you need If you are not
129
+ sure please search the SKU which is marked on your frame
130
+
131
+ Fit SKU OO91363 Pair Lenses Black Ice Blue Mirror Coating Titanium Mirror Coating
132
+ This Lenses Only Fit Oakley Eyepatch 2 Sunglasses if you are not sure please search
133
+ the SKU which is marked on your frame OO9136
134
+
135
+ 100 Polarized and 100 UV Protection on all lenses
136
+
137
+ Suitable for cycling running driving and other outdoor sports and activities
138
+
139
+ Precision Cut and Guaranteed to Fit
140
+
141
+ Reduces glare and enhances contrast perfectly'
142
+ - 'description
143
+
144
+ NCAA Alabama Crimson Tide Neoway Cap MediumLarge100 Cotton
145
+
146
+ 39Thirty stretch fit 100 Cotton front with Spacer Mesh mids and rear of cap
147
+
148
+ Primary Team logo in raised embroidery on the front of the cap
149
+
150
+ Secondary Team logo in flat embroidery on back of cap
151
+
152
+ SmMed fits sizes 7 thru 7 38 MedLg fits sizes 7 14 thru 7 58'
153
+ - source_sentence: "title: \nDV8 Thug Bowling Ball"
154
+ sentences:
155
+ - 'description
156
+
157
+ Product name No Storage Swimming Lifesaving Drift Bag Size4829cm Material035mm
158
+ Ecofriendly PVC BuoyancyMax 10KG FeaturesEquipped with inflatable gas nozzlewithout
159
+ storage spaceIt can not put anything Note 11Befor you swimmingYou can blow air
160
+ through the air mouth and then ensure there is no leakage by submerging the buoy
161
+ 2 Do not let sharp objects damage the air bag after entering the water 3 Be careful
162
+ not to let the connection rope entangle the body 4 In case of emergency use in
163
+ the water keep calm and relaxed Hold fast and dont panic And quickly to a safe
164
+ area shallow water shore or call for help Package included 1Pcs Swim bouy without
165
+ storage spacePatron Saint of Swimming Safety With bright appearance eyecatching
166
+ in the water it is easy to be found in distress and can quickly use the buoyancy
167
+ of independent air bags for selfrescue and a short rest it is your guardian angel
168
+ of water safety
169
+
170
+ Easily Inflatable Storable and Easy to UseIt only takes 10 seconds to inflate
171
+ by blowing with the mouth very simple and convenient to use It has independent
172
+ storage space and can be used for personal belongings
173
+
174
+ UltraPortable and Easy to Store It is light and small after being deflated so
175
+ it is easy to store and carry Material is PVC its environmental protection safe
176
+ durable
177
+
178
+ As a Thoughtful Gift For relatives and friends who like swimming or water sports
179
+
180
+ Guarantee of water safety One of the necessary equipment for open water swimming
181
+ In the water can be used as a lifesaving sign Swimming float it can also remind
182
+ passing vessels'
183
+ - 'description
184
+
185
+ Featuring the new Class 13F Hybrid Reactive coverstock wrapped around the Thug
186
+ core the Thug skids easily through the front and midlane flipping on the backend
187
+ for the most breakpoint potential of any DV8 ever on medium to oily lane conditions
188
+ The DV8 Thug can be drilled using the standard drilling techniques developed for
189
+ bowling balls with asymmetric cores Core Type Thug Low RG Coverstock Class 13F
190
+ Reactive Finishing Steps 500 Siaair Micro Pad Royal Compound Weights 1216 Asymmetrical
191
+ Diff 0015 at 15 lbs RG Max 2557 at 15 lbs RG Min 2505 at 15 lbs RG Differential
192
+ 0052 at 15 lbs To reduce oil absorption and remove dirt from the surface of the
193
+ ball clean your ball with a cleaner designed for reactive bowling balls after
194
+ each session Look We dont screw around DV8 bowling balls have been manufactured
195
+ to the highest standards of workmanship and material We warrant that theyll be
196
+ free of defects in materials and workmanship for a period of two years from the
197
+ date of purchase We agree to repair or replace the ball you bought if at any time
198
+ during the warranty period its found to be defective in material or workmanshipFeaturing
199
+ the new Class 13F Hybrid Reactive coverstock wrapped around the Thug core the
200
+ Thug skids easily through the front and midlane flipping on the backend for the
201
+ most breakpoint potential of any DV8 ever on medium to oily lane conditions
202
+
203
+ The DV8 Thug can be drilled using the standard drilling techniques developed for
204
+ bowling balls with asymmetric cores
205
+
206
+ Core Type Thug Low RG Coverstock Class 13F Reactive Finishing Steps 500 Siaair
207
+ Micro Pad Royal Compound Weights 1216 Asymmetrical Diff'
208
+ - 'description
209
+
210
+ If you havent tried it yet ultralight fishing can be one of the most enjoyable
211
+ ways to fish The Shakespeare Micro Series rods have a variety of actions and lengths
212
+ to fit any species and are perfect for fishing with lighter lures and linesMicro
213
+ Series Spinning Rod 7
214
+
215
+ Micro Series Spinning Rod 7
216
+
217
+ Micro Series Spinning Rod 7
218
+
219
+ Micro Series Spinning Rod 7
220
+
221
+ Micro Series Spinning Rod 7
222
+
223
+ Graphite composite rods with perfect light and ultra light actions
224
+
225
+ Full cork handles
226
+
227
+ Conventional reel seat with cushioned hoods
228
+
229
+ Stainless Steel guides with stainless steel inserts'
230
+ - source_sentence: "title: \nAltus Athletic 4Inch Premium Padded Lifting Belt Large"
231
+ sentences:
232
+ - 'description
233
+
234
+ Premium padded leather weight lifting belt is competition approved 4 inch size
235
+ and offers added support to the core muscle groups during heavy exerciseTraditional
236
+ Leather Belt
237
+
238
+ 4 in Wide Belt Saddle Stitched
239
+
240
+ TwoTongue Steel Buckle With Oval Belt Holes
241
+
242
+ Padded Suede Lining'
243
+ - 'description
244
+
245
+ NYX Lightning combines cutting edge performance and classic stylingNylon frame
246
+
247
+ Nylon lens
248
+
249
+ NonPolarized
250
+
251
+ Amber lens filters blue light and increases contrast
252
+
253
+ Italian made frame has smooth clean lines
254
+
255
+ 8 base Lens optimizes sun protection
256
+
257
+ TR90 Memory Nylon Frame and Polycarbonate Lens
258
+
259
+ Complete with carrying case and microfiber pouch'
260
+ - 'description
261
+
262
+ The Yoga Mat is made with high quality imported TPE materialhealthgreen soft odorlessit
263
+ comes with carry stripgift packingwe focus on the quality onlyit cant be compared
264
+ with the price for the normal exercise mat pilatespadif you want the best yoga
265
+ mateours are your best choice Remember No NBR PVC Only TPE Cleanng method Do
266
+ not use a washing machine or dryer Method 1 Please wipe the mat with cloth depped
267
+ in suds or laundry powder solution Then rinse with water and wipe it off from
268
+ mat with towel or cloth Method 2 Please soak yoga mat in warm water with some
269
+ vinegar for 30 minlay out the whole mat in bath tube with some warm water the
270
+ water must cover the mat and put some200ml white vinegar with itThen wipe the
271
+ water off from mat with towel or cloth72 x 24Inches14 thicky yoga mats ensure
272
+ comfortable for youeasy carry for travel
273
+
274
+ Health Fitness Non toxic Odorless Phthalate FreeThis pilates mat is safe for stretching
275
+ and toning workouts
276
+
277
+ TPE material lake blue rosered grassgreen violetgreen black bluegrey blackgreen
278
+ 8 color for choiceblackgreen and bluegray are doublesidememory foam protects your
279
+ knees and joints which allow you to grip the floor for balance
280
+
281
+ Ribbed surface on one side with a smooth surface on the other nonslip surface
282
+ grips the floor to prevent injuries
283
+
284
+ Elegant easytoclean durable and long lasting material Features an integrated carry
285
+ strap'
286
+ - source_sentence: "title: \niGreely Trolling Motor Plug and Receptacle Male Plug\
287
+ \ Female Panel Mount Socket Waterproof IP67 2Pin Connector 12V 24V 36V 48V Plug\
288
+ \ for Trolling Motor Marine Boat RV Solar Panel"
289
+ sentences:
290
+ - 'description
291
+
292
+ Miniature padded halter made of 34 inch webbing Features padded nose and crown
293
+ Snap at throat with adjustable nose and crownItem Package Dimensions 58 L X 188
294
+ W X 244 H Cm
295
+
296
+ Product Type Sporting Goods
297
+
298
+ Item Package Weight 0272 kgs
299
+
300
+ Country Of Origin China'
301
+ - 'description
302
+
303
+ Note The terminals are small its a little difficult to solder 12 or 10 gauge cable
304
+ but its easy for 14 gauge
305
+
306
+ Product Parameter
307
+
308
+ Wire specification25M5M1412AWG
309
+
310
+ Line diameter range812
311
+
312
+ Diameternumber274
313
+
314
+ Rated currentA20A
315
+
316
+ Operation voltageACVRMS500V
317
+
318
+ Contact resistance1m
319
+
320
+ Temperature rise20A per contact at 3025A4030A50
321
+
322
+ Withstand voltage ACV1minute2500V
323
+
324
+ Insulation resistance500M
325
+
326
+ Protection levelIP65IP67
327
+
328
+ Mating cycle500 times
329
+
330
+ Working temperature4080
331
+
332
+ Connection modeQuick plug connection
333
+
334
+ Wiring modeWelding Crimping
335
+
336
+ Material
337
+
338
+ Insulation componentsHighperformance engineering plastics
339
+
340
+ HardwareZinc alloyNickel Plating
341
+
342
+ PINCopper alloy gold plated
343
+
344
+ Waterproof circleSilica gel
345
+
346
+ Flame retardant gradeUL94V02 pin power industrial circular connector is used for
347
+ power and signal perfectly fits Furrion solar port that many RVs come with eg
348
+ Grand Design Forrest River
349
+
350
+ Rated current 20A operation voltage 500VAC outdoor waterproof IP67Goldplated contacts
351
+ high strength corrosion resistance good electrical conductivity effective response
352
+ to current temperature changes
353
+
354
+ Protected against solids liquids provide a secure and reliable connection in
355
+ the harsh environment
356
+
357
+ PBT material of the protective housing with characters of fire safety compression
358
+ antiexplosion and antideformation
359
+
360
+ Widely used in data acquisition systems inclinometer computer automation measurement
361
+ and control systems mechnical equipment audiovideo communications automotive and
362
+ other industries
363
+
364
+ Trolling Motor Plug and Receptacle Male Plug Female Panel Mount Socket Waterproof
365
+ IP67 2Pin Connector 12V 24V 36V 48V Plug for Trolling Motor Marine Boat RV Solar
366
+ Panel
367
+
368
+ Note The terminals are small its a little difficult to solder 12 or 10 gauge cable
369
+ but its easy for 14 gauge'
370
+ - 'description
371
+
372
+ Full leather construction perforated palm and fingers for breathability an optifeel
373
+ closure and bright colors make Callaways Opticolor glove the best choice for performance
374
+ and fashion
375
+
376
+ Opti Feel Leather
377
+
378
+ Premium Feel Fit and Comfort
379
+
380
+ Premium Feel Fit and Comfort
381
+
382
+ Perforations on Palm Top of Hand and Fingers
383
+
384
+ Moisture Reduction Increased Breathability
385
+
386
+ Moisture Reduction Increased Breathability
387
+
388
+ Opti Fit Adjustable Closure
389
+
390
+ Thin Light and Secure Fit
391
+
392
+ Thin Light and Secure FitLeather
393
+
394
+ LeftWorn on Left Hand
395
+
396
+ Opti feel plus leather premium feel fit and comfort
397
+
398
+ Perforations on palm fingers and thumb moisture reduction and increased breathability
399
+
400
+ Opti fit adjustable closure thin light and secure fit'
401
+ datasets:
402
+ - guyhadad01/Amazon_2023_items_processed_filtered
403
+ pipeline_tag: sentence-similarity
404
+ library_name: sentence-transformers
405
+ ---
406
+
407
+ # SentenceTransformer based on google/embeddinggemma-300m
408
+
409
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [google/embeddinggemma-300m](https://huggingface.co/google/embeddinggemma-300m) on the [amazon_2023_items_processed_filtered](https://huggingface.co/datasets/guyhadad01/Amazon_2023_items_processed_filtered) dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
410
+
411
+ ## Model Details
412
+
413
+ ### Model Description
414
+ - **Model Type:** Sentence Transformer
415
+ - **Base model:** [google/embeddinggemma-300m](https://huggingface.co/google/embeddinggemma-300m) <!-- at revision 57c266a740f537b4dc058e1b0cda161fd15afa75 -->
416
+ - **Maximum Sequence Length:** 512 tokens
417
+ - **Output Dimensionality:** 768 dimensions
418
+ - **Similarity Function:** Cosine Similarity
419
+ - **Training Dataset:**
420
+ - [amazon_2023_items_processed_filtered](https://huggingface.co/datasets/guyhadad01/Amazon_2023_items_processed_filtered)
421
+ <!-- - **Language:** Unknown -->
422
+ <!-- - **License:** Unknown -->
423
+
424
+ ### Model Sources
425
+
426
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
427
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
428
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
429
+
430
+ ### Full Model Architecture
431
+
432
+ ```
433
+ SentenceTransformer(
434
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False, 'architecture': 'Gemma3TextModel'})
435
+ (1): Pooling({'word_embedding_dimension': 768, '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})
436
+ (2): Dense({'in_features': 768, 'out_features': 3072, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
437
+ (3): Dense({'in_features': 3072, 'out_features': 768, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
438
+ (4): Normalize()
439
+ )
440
+ ```
441
+
442
+ ## Usage
443
+
444
+ ### Direct Usage (Sentence Transformers)
445
+
446
+ First install the Sentence Transformers library:
447
+
448
+ ```bash
449
+ pip install -U sentence-transformers
450
+ ```
451
+
452
+ Then you can load this model and run inference.
453
+ ```python
454
+ from sentence_transformers import SentenceTransformer
455
+
456
+ # Download from the 🤗 Hub
457
+ model = SentenceTransformer("guyhadad01/EncodeRec_300M_Sports")
458
+ # Run inference
459
+ queries = [
460
+ "title: \niGreely Trolling Motor Plug and Receptacle Male Plug Female Panel Mount Socket Waterproof IP67 2Pin Connector 12V 24V 36V 48V Plug for Trolling Motor Marine Boat RV Solar Panel",
461
+ ]
462
+ documents = [
463
+ 'description\nNote\xa0The terminals are small its a little difficult to\xa0solder 12 or 10 gauge cable but its easy for 14 gauge\nProduct Parameter\nWire specification25M5M1412AWG\nLine diameter range812\nDiameternumber274\nRated currentA20A\nOperation voltageACVRMS500V\nContact resistance1m\nTemperature rise20A per contact at 3025A4030A50\nWithstand voltage ACV1minute2500V\nInsulation resistance500M\nProtection levelIP65IP67\nMating cycle500 times\nWorking temperature4080\nConnection modeQuick plug connection\nWiring modeWelding Crimping\nMaterial\nInsulation componentsHighperformance engineering plastics\nHardwareZinc alloyNickel Plating\nPINCopper alloy gold plated\nWaterproof circleSilica gel\nFlame retardant gradeUL94V02 pin power industrial circular connector is used for power and signal perfectly fits Furrion solar port that many RVs come with eg Grand Design Forrest River\nRated current 20A operation voltage 500VAC outdoor waterproof IP67Goldplated contacts high strength corrosion resistance good electrical conductivity effective response to current temperature changes\nProtected against solids liquids provide a secure and reliable connection in the harsh environment\nPBT material of the protective housing with characters of fire safety compression antiexplosion and antideformation\nWidely used in data acquisition systems inclinometer computer automation measurement and control systems mechnical equipment audiovideo communications automotive and other industries\nTrolling Motor Plug and Receptacle Male Plug Female Panel Mount Socket Waterproof IP67 2Pin Connector 12V 24V 36V 48V Plug for Trolling Motor Marine Boat RV Solar Panel\nNote The terminals are small its a little difficult to solder 12 or 10 gauge cable but its easy for 14 gauge',
464
+ 'description\nFull leather construction perforated palm and fingers for breathability an optifeel closure and bright colors make Callaways Opticolor glove the best choice for performance and fashion\nOpti Feel Leather\nPremium Feel Fit and Comfort\nPremium Feel Fit and Comfort\nPerforations on Palm Top of Hand and Fingers\nMoisture Reduction Increased Breathability\nMoisture Reduction Increased Breathability\nOpti Fit Adjustable Closure\nThin Light and Secure Fit\nThin Light and Secure FitLeather\nLeftWorn on Left Hand\nOpti feel plus leather premium feel fit and comfort\nPerforations on palm fingers and thumb moisture reduction and increased breathability\nOpti fit adjustable closure thin light and secure fit',
465
+ 'description\nMiniature padded halter made of 34 inch webbing Features padded nose and crown Snap at throat with adjustable nose and crownItem Package Dimensions 58 L X 188 W X 244 H Cm\nProduct Type Sporting Goods\nItem Package Weight 0272 kgs\nCountry Of Origin China',
466
+ ]
467
+ query_embeddings = model.encode_query(queries)
468
+ document_embeddings = model.encode_document(documents)
469
+ print(query_embeddings.shape, document_embeddings.shape)
470
+ # [1, 768] [3, 768]
471
+
472
+ # Get the similarity scores for the embeddings
473
+ similarities = model.similarity(query_embeddings, document_embeddings)
474
+ print(similarities)
475
+ # tensor([[ 0.7186, -0.1245, -0.0664]])
476
+ ```
477
+
478
+ <!--
479
+ ### Direct Usage (Transformers)
480
+
481
+ <details><summary>Click to see the direct usage in Transformers</summary>
482
+
483
+ </details>
484
+ -->
485
+
486
+ <!--
487
+ ### Downstream Usage (Sentence Transformers)
488
+
489
+ You can finetune this model on your own dataset.
490
+
491
+ <details><summary>Click to expand</summary>
492
+
493
+ </details>
494
+ -->
495
+
496
+ <!--
497
+ ### Out-of-Scope Use
498
+
499
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
500
+ -->
501
+
502
+ <!--
503
+ ## Bias, Risks and Limitations
504
+
505
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
506
+ -->
507
+
508
+ <!--
509
+ ### Recommendations
510
+
511
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
512
+ -->
513
+
514
+ ## Training Details
515
+
516
+ ### Training Dataset
517
+
518
+ #### amazon_2023_items_processed_filtered
519
+
520
+ * Dataset: [amazon_2023_items_processed_filtered](https://huggingface.co/datasets/guyhadad01/Amazon_2023_items_processed_filtered) at [6b58dd1](https://huggingface.co/datasets/guyhadad01/Amazon_2023_items_processed_filtered/tree/6b58dd18854109aac31652e941c667725f6352f0)
521
+ * Size: 1,007,606 training samples
522
+ * Columns: <code>title</code> and <code>description</code>
523
+ * Approximate statistics based on the first 1000 samples:
524
+ | | title | description |
525
+ |:--------|:----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
526
+ | type | string | string |
527
+ | details | <ul><li>min: 8 tokens</li><li>mean: 22.72 tokens</li><li>max: 65 tokens</li></ul> | <ul><li>min: 18 tokens</li><li>mean: 188.1 tokens</li><li>max: 512 tokens</li></ul> |
528
+ * Samples:
529
+ | title | description |
530
+ |:------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
531
+ | <code>title: <br>SureGrip Zombie Wheels Low 59mm 4 Pack</code> | <code>description<br>All Zombie wheels are made in the USA Zombie wheels feature anodized aluminum hubs for maximum durability and precise feel while maintaining rock solid stability This allows our unique urethane compounds to deliver all your power to the floor Choose the Zombie combination that fits your skating style and surface Zombie Aluminum Core Designed in house and manufactured using state of the art machining technology What makes this different from other aluminum cores The Zombie core is machined from a solid billet of aluminum using the same manufacturing processes as we use to make our famous Power Trac racing plates This increases strength and allows us to machine the hub to a tighter tolerance than any other hub on the market Creating the perfect fit between bearing and inner core Cores are then treated with a special anodizing process to increase hub durability and urethane bonding This technology is unique to the Zombie wheel line Zombie Insane Urethane Specially formulated...</code> |
532
+ | <code>title: <br>NHL San Jose Sharks Team Logo Post Earrings</code> | <code>description<br>Complete your game day outfit with these cute earrings and showcase your team spirit all year round with these NHL Team Logo Post Earrings by Aminco These adorable earrings measure approximately 1inch by 1inch and is decorated with your favorite team colored logo The earrings feature post backings for carefree wear and are NHL officially licensed These would make a great gift for the loyal sports fan in your life100 Synthetic<br>Imported<br>Adorable Post Earrings Measures Approximately 1inch by 1inch<br>Decorated with Team Colored Logo<br>NHL Officially Licensed<br>Features Post Backings for Carefree Wear<br>Complete your Game Day Outfit with these Cute Earrings</code> |
533
+ | <code>title: <br>Team Golf Alamaba Crimson Tide Embroidered Towel from</code> | <code>description<br>Keep your clubs clean while supporting your favorite collegiate team with this officially licensed NCAA embroidered golf towel from Team Golf The trifold embroidered towel features a hook and grommet and measures 16in x 25inCotton<br>Trifold golf towel is embroidered with your favorite collegiate teams logo<br>Hook and grommet<br>Dimensions W 16 in H 25 in<br>Officially licensed</code> |
534
+ * Loss: [<code>CachedMultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cachedmultiplenegativesrankingloss) with these parameters:
535
+ ```json
536
+ {
537
+ "scale": 20.0,
538
+ "similarity_fct": "cos_sim",
539
+ "mini_batch_size": 64,
540
+ "gather_across_devices": false
541
+ }
542
+ ```
543
+
544
+ ### Training Hyperparameters
545
+ #### Non-Default Hyperparameters
546
+
547
+ - `per_device_train_batch_size`: 512
548
+ - `num_train_epochs`: 1
549
+ - `warmup_ratio`: 0.1
550
+ - `bf16`: True
551
+ - `push_to_hub`: True
552
+ - `hub_model_id`: guyhadad01/EncodeRec_300M_Sports
553
+ - `hub_strategy`: checkpoint
554
+ - `prompts`: {'question': 'task: search result | query: ', 'passage_text': 'title: none | text: '}
555
+
556
+ #### All Hyperparameters
557
+ <details><summary>Click to expand</summary>
558
+
559
+ - `overwrite_output_dir`: False
560
+ - `do_predict`: False
561
+ - `eval_strategy`: no
562
+ - `prediction_loss_only`: True
563
+ - `per_device_train_batch_size`: 512
564
+ - `per_device_eval_batch_size`: 8
565
+ - `per_gpu_train_batch_size`: None
566
+ - `per_gpu_eval_batch_size`: None
567
+ - `gradient_accumulation_steps`: 1
568
+ - `eval_accumulation_steps`: None
569
+ - `torch_empty_cache_steps`: None
570
+ - `learning_rate`: 5e-05
571
+ - `weight_decay`: 0.0
572
+ - `adam_beta1`: 0.9
573
+ - `adam_beta2`: 0.999
574
+ - `adam_epsilon`: 1e-08
575
+ - `max_grad_norm`: 1.0
576
+ - `num_train_epochs`: 1
577
+ - `max_steps`: -1
578
+ - `lr_scheduler_type`: linear
579
+ - `lr_scheduler_kwargs`: {}
580
+ - `warmup_ratio`: 0.1
581
+ - `warmup_steps`: 0
582
+ - `log_level`: passive
583
+ - `log_level_replica`: warning
584
+ - `log_on_each_node`: True
585
+ - `logging_nan_inf_filter`: True
586
+ - `save_safetensors`: True
587
+ - `save_on_each_node`: False
588
+ - `save_only_model`: False
589
+ - `restore_callback_states_from_checkpoint`: False
590
+ - `no_cuda`: False
591
+ - `use_cpu`: False
592
+ - `use_mps_device`: False
593
+ - `seed`: 42
594
+ - `data_seed`: None
595
+ - `jit_mode_eval`: False
596
+ - `bf16`: True
597
+ - `fp16`: False
598
+ - `fp16_opt_level`: O1
599
+ - `half_precision_backend`: auto
600
+ - `bf16_full_eval`: False
601
+ - `fp16_full_eval`: False
602
+ - `tf32`: None
603
+ - `local_rank`: 0
604
+ - `ddp_backend`: None
605
+ - `tpu_num_cores`: None
606
+ - `tpu_metrics_debug`: False
607
+ - `debug`: []
608
+ - `dataloader_drop_last`: False
609
+ - `dataloader_num_workers`: 0
610
+ - `dataloader_prefetch_factor`: None
611
+ - `past_index`: -1
612
+ - `disable_tqdm`: False
613
+ - `remove_unused_columns`: True
614
+ - `label_names`: None
615
+ - `load_best_model_at_end`: False
616
+ - `ignore_data_skip`: False
617
+ - `fsdp`: []
618
+ - `fsdp_min_num_params`: 0
619
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
620
+ - `fsdp_transformer_layer_cls_to_wrap`: None
621
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
622
+ - `parallelism_config`: None
623
+ - `deepspeed`: None
624
+ - `label_smoothing_factor`: 0.0
625
+ - `optim`: adamw_torch
626
+ - `optim_args`: None
627
+ - `adafactor`: False
628
+ - `group_by_length`: False
629
+ - `length_column_name`: length
630
+ - `project`: huggingface
631
+ - `trackio_space_id`: trackio
632
+ - `ddp_find_unused_parameters`: None
633
+ - `ddp_bucket_cap_mb`: None
634
+ - `ddp_broadcast_buffers`: False
635
+ - `dataloader_pin_memory`: True
636
+ - `dataloader_persistent_workers`: False
637
+ - `skip_memory_metrics`: True
638
+ - `use_legacy_prediction_loop`: False
639
+ - `push_to_hub`: True
640
+ - `resume_from_checkpoint`: None
641
+ - `hub_model_id`: guyhadad01/EncodeRec_300M_Sports
642
+ - `hub_strategy`: checkpoint
643
+ - `hub_private_repo`: None
644
+ - `hub_always_push`: False
645
+ - `hub_revision`: None
646
+ - `gradient_checkpointing`: False
647
+ - `gradient_checkpointing_kwargs`: None
648
+ - `include_inputs_for_metrics`: False
649
+ - `include_for_metrics`: []
650
+ - `eval_do_concat_batches`: True
651
+ - `fp16_backend`: auto
652
+ - `push_to_hub_model_id`: None
653
+ - `push_to_hub_organization`: None
654
+ - `mp_parameters`:
655
+ - `auto_find_batch_size`: False
656
+ - `full_determinism`: False
657
+ - `torchdynamo`: None
658
+ - `ray_scope`: last
659
+ - `ddp_timeout`: 1800
660
+ - `torch_compile`: False
661
+ - `torch_compile_backend`: None
662
+ - `torch_compile_mode`: None
663
+ - `include_tokens_per_second`: False
664
+ - `include_num_input_tokens_seen`: no
665
+ - `neftune_noise_alpha`: None
666
+ - `optim_target_modules`: None
667
+ - `batch_eval_metrics`: False
668
+ - `eval_on_start`: False
669
+ - `use_liger_kernel`: False
670
+ - `liger_kernel_config`: None
671
+ - `eval_use_gather_object`: False
672
+ - `average_tokens_across_devices`: True
673
+ - `prompts`: {'question': 'task: search result | query: ', 'passage_text': 'title: none | text: '}
674
+ - `batch_sampler`: batch_sampler
675
+ - `multi_dataset_batch_sampler`: proportional
676
+ - `router_mapping`: {}
677
+ - `learning_rate_mapping`: {}
678
+
679
+ </details>
680
+
681
+ ### Training Logs
682
+ | Epoch | Step | Training Loss |
683
+ |:------:|:----:|:-------------:|
684
+ | 0.0254 | 50 | 0.4625 |
685
+ | 0.0508 | 100 | 0.2331 |
686
+ | 0.0762 | 150 | 0.2031 |
687
+ | 0.1016 | 200 | 0.2042 |
688
+
689
+
690
+ ### Framework Versions
691
+ - Python: 3.12.11
692
+ - Sentence Transformers: 5.1.0
693
+ - Transformers: 4.57.0
694
+ - PyTorch: 2.7.1+cu126
695
+ - Accelerate: 1.10.0
696
+ - Datasets: 3.6.0
697
+ - Tokenizers: 0.22.1
698
+
699
+ ## Citation
700
+
701
+ ### BibTeX
702
+
703
+ #### Sentence Transformers
704
+ ```bibtex
705
+ @inproceedings{reimers-2019-sentence-bert,
706
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
707
+ author = "Reimers, Nils and Gurevych, Iryna",
708
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
709
+ month = "11",
710
+ year = "2019",
711
+ publisher = "Association for Computational Linguistics",
712
+ url = "https://arxiv.org/abs/1908.10084",
713
+ }
714
+ ```
715
+
716
+ #### CachedMultipleNegativesRankingLoss
717
+ ```bibtex
718
+ @misc{gao2021scaling,
719
+ title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
720
+ author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
721
+ year={2021},
722
+ eprint={2101.06983},
723
+ archivePrefix={arXiv},
724
+ primaryClass={cs.LG}
725
+ }
726
+ ```
727
+
728
+ <!--
729
+ ## Glossary
730
+
731
+ *Clearly define terms in order to be accessible across audiences.*
732
+ -->
733
+
734
+ <!--
735
+ ## Model Card Authors
736
+
737
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
738
+ -->
739
+
740
+ <!--
741
+ ## Model Card Contact
742
+
743
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
744
+ -->
last-checkpoint/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<image_soft_token>": 262144
3
+ }
last-checkpoint/config.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_sliding_window_pattern": 6,
3
+ "architectures": [
4
+ "Gemma3TextModel"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "attn_logit_softcapping": null,
9
+ "bos_token_id": 2,
10
+ "dtype": "bfloat16",
11
+ "eos_token_id": 1,
12
+ "final_logit_softcapping": null,
13
+ "head_dim": 256,
14
+ "hidden_activation": "gelu_pytorch_tanh",
15
+ "hidden_size": 768,
16
+ "initializer_range": 0.02,
17
+ "intermediate_size": 1152,
18
+ "layer_types": [
19
+ "sliding_attention",
20
+ "sliding_attention",
21
+ "sliding_attention",
22
+ "sliding_attention",
23
+ "sliding_attention",
24
+ "full_attention",
25
+ "sliding_attention",
26
+ "sliding_attention",
27
+ "sliding_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "full_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "sliding_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "full_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "sliding_attention",
40
+ "sliding_attention",
41
+ "sliding_attention",
42
+ "full_attention"
43
+ ],
44
+ "max_position_embeddings": 2048,
45
+ "model_type": "gemma3_text",
46
+ "num_attention_heads": 3,
47
+ "num_hidden_layers": 24,
48
+ "num_key_value_heads": 1,
49
+ "pad_token_id": 0,
50
+ "query_pre_attn_scalar": 256,
51
+ "rms_norm_eps": 1e-06,
52
+ "rope_local_base_freq": 10000.0,
53
+ "rope_scaling": null,
54
+ "rope_theta": 1000000.0,
55
+ "sliding_window": 257,
56
+ "transformers_version": "4.57.0",
57
+ "use_bidirectional_attention": true,
58
+ "use_cache": true,
59
+ "vocab_size": 262144
60
+ }
last-checkpoint/config_sentence_transformers.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "SentenceTransformer",
3
+ "__version__": {
4
+ "sentence_transformers": "5.1.0",
5
+ "transformers": "4.57.0",
6
+ "pytorch": "2.7.1+cu126"
7
+ },
8
+ "prompts": {
9
+ "query": "task: search result | query: ",
10
+ "document": "title: none | text: ",
11
+ "BitextMining": "task: search result | query: ",
12
+ "Clustering": "task: clustering | query: ",
13
+ "Classification": "task: classification | query: ",
14
+ "InstructionRetrieval": "task: code retrieval | query: ",
15
+ "MultilabelClassification": "task: classification | query: ",
16
+ "PairClassification": "task: sentence similarity | query: ",
17
+ "Reranking": "task: search result | query: ",
18
+ "Retrieval": "task: search result | query: ",
19
+ "Retrieval-query": "task: search result | query: ",
20
+ "Retrieval-document": "title: none | text: ",
21
+ "STS": "task: sentence similarity | query: ",
22
+ "Summarization": "task: summarization | query: "
23
+ },
24
+ "default_prompt_name": null,
25
+ "similarity_fn_name": "cosine"
26
+ }
last-checkpoint/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dcff2bc35d4b89f9920dbfc24c9784230f133e60b7efd8421b728caa3129765
3
+ size 605759848
last-checkpoint/modules.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Dense",
18
+ "type": "sentence_transformers.models.Dense"
19
+ },
20
+ {
21
+ "idx": 3,
22
+ "name": "3",
23
+ "path": "3_Dense",
24
+ "type": "sentence_transformers.models.Dense"
25
+ },
26
+ {
27
+ "idx": 4,
28
+ "name": "4",
29
+ "path": "4_Normalize",
30
+ "type": "sentence_transformers.models.Normalize"
31
+ }
32
+ ]
last-checkpoint/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89649a22f3916e197be45a082597d611800c3e9d7c9d75aba96246deab4411d1
3
+ size 1230592267
last-checkpoint/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c262bfa40b2df2776b7e218b544f3cfccc93f25ff6995c84b8366443b0e13522
3
+ size 14645
last-checkpoint/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1d4e739551319e957e6c472fd8c9bacf3bae41231a764211a8f19abc4bae6961
3
+ size 1465
last-checkpoint/sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": false
4
+ }
last-checkpoint/special_tokens_map.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "boi_token": "<start_of_image>",
3
+ "bos_token": {
4
+ "content": "<bos>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "eoi_token": "<end_of_image>",
11
+ "eos_token": {
12
+ "content": "<eos>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "image_token": "<image_soft_token>",
19
+ "pad_token": {
20
+ "content": "<pad>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false
25
+ },
26
+ "unk_token": {
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ }
33
+ }
last-checkpoint/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c79a190be01275b078b3574d02188abc5784e5651a101b20d826371ba8e897dc
3
+ size 33385261
last-checkpoint/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
3
+ size 4689074
last-checkpoint/tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff
 
last-checkpoint/trainer_state.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.1016260162601626,
6
+ "eval_steps": 500,
7
+ "global_step": 200,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.02540650406504065,
14
+ "grad_norm": 9.75,
15
+ "learning_rate": 1.2436548223350254e-05,
16
+ "loss": 0.4625,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.0508130081300813,
21
+ "grad_norm": 7.6875,
22
+ "learning_rate": 2.5126903553299492e-05,
23
+ "loss": 0.2331,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.07621951219512195,
28
+ "grad_norm": 6.78125,
29
+ "learning_rate": 3.7817258883248735e-05,
30
+ "loss": 0.2031,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.1016260162601626,
35
+ "grad_norm": 5.84375,
36
+ "learning_rate": 4.994353472614343e-05,
37
+ "loss": 0.2042,
38
+ "step": 200
39
+ }
40
+ ],
41
+ "logging_steps": 50,
42
+ "max_steps": 1968,
43
+ "num_input_tokens_seen": 0,
44
+ "num_train_epochs": 1,
45
+ "save_steps": 200,
46
+ "stateful_callbacks": {
47
+ "TrainerControl": {
48
+ "args": {
49
+ "should_epoch_stop": false,
50
+ "should_evaluate": false,
51
+ "should_log": false,
52
+ "should_save": true,
53
+ "should_training_stop": false
54
+ },
55
+ "attributes": {}
56
+ }
57
+ },
58
+ "total_flos": 0.0,
59
+ "train_batch_size": 512,
60
+ "trial_name": null,
61
+ "trial_params": null
62
+ }
last-checkpoint/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe13c102155fc21a5f15e266bc335ba3dbed5536c392ea69398c1eb5eef2d4b0
3
+ size 6289