File size: 45,708 Bytes
9b076d1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
---
language:
- en
license: apache-2.0
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- dense
- generated_from_trainer
- dataset_size:441985
- loss:CachedMultipleNegativesRankingLoss
base_model: google/embeddinggemma-300m
widget:
- source_sentence: "title: \nSwords of Revealing Light  SDPLEN026  Common  1st Edition"
  sentences:
  - 'description

    Structure Deck Powercode Link is a Structure Deck in the YuGiOh Official Card
    Game and YuGiOh Trading Card Game It is the 35th Deck in the OCGs Structure Deck
    series following Structure Deck Cyberse Link It is the 42nd Deck in the TCGs Structure
    Deck series following Lair of Darkness Structure Deck'
  - 'description

    Features  Glowing in the dark effect These handmade luminous marbles are eyecathing
    and in line with the most party themes such as black light party neon theme party
    growth party Halloween party and so on so you can use them to decorate your party
    to create the party vibe you want  Play together These doted style glass marbles
    are funny toys for boys and girls They can play the related game on the garden
    yard lawn playground park and more It is a nice opportunity to develop their social
    skill and get away from electronic products  Specifications Material glass Diameter
    approx 063 inch Color as pictures shown  Package included 35 x Luminous marbles  Warm
    notices Manual measurement please allow slight errors on size The color might
    exist slightly difference due to different displayswhat you will get package comes
    with 35 pieces handmade glass marbles in 7 colors 5 pieces of each color with
    the same size enough quantity to meet your need You can also share them with your
    friends

    Size and muticolor doted style glass marbles are 063 inch in diameter featuring
    7 colors such as transparent pinkishorange yellow purple green sea blue and dark
    blue They are glowing at night and look very interesting then you can imagine
    yourself in the universe

    Quality material these muticolors luminous marbles are made from quality glass
    which are firm and sturdy smooth with a nice finish durable and highstrength designed
    for longtime use giving you nice play feeling

    How to make marbles glow brighter you can make the glow marbles sunbathe or illuminate
    them with ultraviolet light to make the marbles shine brighter and longer bringing
    you delightful mood Recommended for ages 5 years and up

    Multiple uses the luminous glass marbles can effectively improve boys and girls
    eye to hand coordination and enhance their fine motor skills They are also nice
    stocking stuffers Easter baskets birthday presents and home decoration For example
    you can use them to add a splash of color in your fish tanks and vases'
  - 'description

    Legendary Encounters A Predator Deck Building Game is really two games in one
    You can play as humans working together to escape and fight off the Predator Or
    your group can play as Predators competing to hunt humans and earn the most Honor
    You can mix both Legendary Encounters games in order to play as Predators hunting
    the Aliens You would play on the Aliens board pick a location from the Aliens
    game choose Aliens Objectives and more See Rulebook for details If youre feeling
    extra adventurous there are even more ways to mix the games such as playing as
    Predators and hunting a mix of Human Prey and Aliens Or you could mix in Marvel
    Legendary cards for even more craziness700 playable cards all with original art
    Game mat and rulebook included

    15 players

    3060 minute play time

    Cards included 35 Experience 25 Brute Strength 15 Role Avatars 10 Human 5 Predator
    15 Role Character Cards 10 Human 5 Predator 10 Commanders 8 Killer Instincts 224
    Character cards 16 characters with 14 cards each

    Cards Included 2 Locations 6 Objectives 66 Enemy cards 6 different minidecks 66
    Prey cards 6 different minidecks 24 Young Blood cards 24 Mercenary cards 40 Enemy
    Strikes 60 Prey Strikes 20 Traps 20 Gear 20 Tests 20 Challenges'
- source_sentence: "title: \nFunko POP Muppets VINYL Snowth"
  sentences:
  - 'description

    From the Manufacturer

    This Muppets Snowth Pop Vinyl figure will have you humming Do Dooo Do Doo doo
    too From Jim Hensons Muppets comes Snowth  the furry pink creature with horns
    and round yellow lips Snowths were performed simultaneously by Frank Oz with each
    puppet on each arm Snowth is adorable as ever as it stands 4 tall and comes in
    a stylized artful display boxCollect them all

    Displayable window box

    Muppets Classics'
  - 'description

    On September 20 1983 in Burbank CA the USPS released this setenant in an effort
    to break down he barriers and stereotypes which have been created One stamp simply
    shows the American Sign Language sign for I Love You while the other pictures
    a mother signing I Love You to her childADA

    Deaf

    Hearing Impaired'
  - 'description

    From Avatar Ty Lee as a stylized Pop vinyl from Funko Figure stands 3 34 inches
    and comes in a window display box Check out the other Avatar figures from Funko
    Collect them allImported

    Product TypeToys And Games

    Item Package Dimension35  L X45  W X625  H

    Item Package Weight025 Lbs

    Country Of Origin Viet Nam'
- source_sentence: "title: \nFuryu Accel World Kuroyukihime Figure"
  sentences:
  - 'description

    From the Manufacturer

    A rescue themed vehicle with a Dalmation RollaRound that both magically come to
    life when baby plays Push along the Firetruck to see the siren lights move up
    and down Put the Round into the the truck to see the Dogs head turn back and forth
    as he rollA rescue themed vehicle with a Dalmation RollaRound that both magically
    come to life when baby plays

    Push along the Firetruck to see the siren lights move up and down

    Put the Round into the the truck to see the Dogs head turn back and forth as he
    rolls along

    Part of RollARounds Collection

    Age Range 6 to 36 Months'
  - 'description

    Disney Toy Story Buzz Lightyear Little Lights with Hook for Samsung Galaxy S3
    Mini i8190'
  - 'description

    Accel World FuRyu 5 Kuroyukihime Sitting PVC Statue  KuroyukihimeBrand New Official
    Item

    Great for Collectors'
- source_sentence: "title: \nJellycat Pretty Patisserie Tarte Au Citron Food Plush"
  sentences:
  - 'description

    Features  Benefits  60pcs New Year latex balloons and foil balloons colored in
    gold and black printed with festive words are great atmosphere creator for your
    home and party  Made of durable latex or foil material wonderful for hanging on
    the wall ceiling window and other places you want  Specifications Size 12 inches
    18 inch Color black gold Material latex foil material  Package Includes 60 x New
    Year Balloon 4 x Foil BalloonLarge Quantity  Are you looking for New Year decorations
    Our cute balloons will be a great choice for you 60pcs latex balloons and 4pcs
    large foil balloons great combination for your home and party decoration

    Delicate Design  Taking black and gold as theme colors printed with festive words
    or designed with confetti the foil balloons are shaped in star or round classic
    and exquisite

    Good Quality  Made of good quality latex or foil material safe and durable the
    foil one is reusable you can fill them with air or helium

    Size  Each latex balloon measures 12 inch and the foil balloons measure 18 inch
    appropriate size for hanging on the wall ceiling window and other places you want

    Wide Application  Perfect to decorate your New Year party decoration Christmas
    school activities and other occasions suitable for both indoor and outdoor bar
    stage props etc'
  - 'description

    Finally A pet bear that wont gnaw on your head or bite your arms off This adorable
    bouncing head animal doll is ready for you to give him a great big bear hug Soft
    white fur makes this shaking head figure look great on any dashboard bookshelf
    desktop or countertop He loves the dashboard adhesive that travels with him because
    the sticky pad reminds him of honey Youll love it because it will securely attach
    him to your car dashboard but it wont leave any permanent residue With adorable
    beady eyes this bobblehead will make you the envy of all your friends  Perfect
    gift for any bobble head collector a trucker with an empty dashboard or children
    Kids love these cute little bobbing head toys One of many ridiculously awesome
    nodding head dash board therapists brought to you as part of Batty Bargains Legendary
    Bobbleheads collection Order yours todayNo live bears were harmed in the making
    of this bobblehead bear Fuzzy black fur made of velvet textured flocking

    Big strong paws and great listening skills make for a phenomenal dash board therapist
    When your not strong lean on this bobbing head bear Its bouncing head will help
    you though the tough times

    A beautiful toy dash board doll great for any road trip One of many dashboard
    figures available as part of Batty Bargains Legendary Bobbleheads Collection

    Adjustable bobble head moves easily and smoothly making it look like it is nodding
    or shaking its head Guaranteed to bob at a wide range of angles Just loosen or
    tighten the weighted bolt on the back of its bobbing head

    Easily mounts on almost any car or vehicle dashboard in seconds with included
    adhesive pad Easily removed without leaving any permanent residue'
  - 'description

    The Jellycat brand was established in London in 1999 to create quirky original
    and innovative soft toys for all ages Jellycat offers the best selection of soft
    plush stuffed animals and toys in the cutest and most luxurious of fabrics and
    textures With unmatched quality find the perfect stuffed animal with the coolest
    designs for babies kids and adults alike Plush toys from Jellycat come in a variety
    of sizes to cuddle including mini small medium large huge and really big They
    also feature a variety of collections and themes to adorn any nursery or childrens
    room for both boys and girls making a Jellycat the perfect gift Best of all your
    Jellycat stuffed animal will provide countless hours of soft hugs and memories
    for years to comePlush measures 4 x 4 x 2 inches

    Suitable for all ages

    Made of 95 Polyester 5 Spandex

    Spot clean only

    Designed by Jellycat in London UK'
- source_sentence: "title: \nFairytale and Historic Minifigure Set"
  sentences:
  - 'description

    Features 1 Cartoon Dinosaur Plush Doll is suitable for being Ornament Gift Pillow
    and Rag Toy 2 Made of quality soft plush and being fully filled with pp cotton
    Not afraid of squeezing 3 Not easy to deform with good resilience It has tightly
    threaded side which features not easy to produce cotton 4 Perfect to cuddle with
    in bed or to keep you company when you nap on the couch Great gift for kids girls
    and anyone you love 5 For the cleaning of such short plush doll clean the surface
    with cold water and place it in the sun after washing to ensure that the short
    plush is fluffy Notice Actual color may be slightly different from the image due
    to different display and light effect Please allow 13cm deviation due to manual
    measurement Description Material plush PP cotton Size 3560cm Decoration form hanging
    placing ColorStyle As the picture shown'
  - 'description

    Toysmiths mission is to supply quality toys and gifts while delivering superior
    customer service to retailers We offer products in many key categories including
    active play science  discovery arts  crafts impulse  novelty toys and nostalgic
    retro classics Since our inception in 1981 our owners Bill and Nancy Smith have
    worked very diligently to grow from a small family business shipping from their
    garage to a large office complex and distribution center that stocks and ships
    over 1600 products to more than 6000 accounts However our philosophy remains the
    same maintaining a strong family work environment while offering the best selection
    and quality of products together with the highest standards of service to all
    our customersEducational toys that help children learn

    Made using safe and high quality materials

    Toys for all age groups'
  - 'description

    The LEGO Education 779349 or 9349 227piece fairytale and historic minifigure set
    includes elements to build 22 multicultural male and female minifigures representing
    characters from fairytales and history Children four years and older can explore
    roleplaying while creating and reenacting fairytales and stories developing imagination
    and imitative skills Themes include pirates a witch and wizard a king and queen
    knights mine workers a mermaid and merman and more and accessories such as wands
    hats flowers spiders and snakes The set is compatible with any LEGO Education
    system supports a group of five students and comes with a decorated box for storage

    Since 1980 LEGO Education has delivered handson curriculumbased resources for
    teachers and students worldwide LEGO Education believes a handson mindson approach
    helps students actively take ownership of the learning process and develop 21stcentury
    skills such as creative thinking and problem solving through reallife engaging
    experiencesExploring differences to real life makebelieve and historic characters

    Storytelling through characters and things they do

    Developing fantasy imagination and imitative skills

    Featuring a variety of minifigures that enable children to create and act out
    their favourite fairytales and stories

    Ideal for 4 years old children or above'
datasets:
- guyhadad01/Amazon_2023_items_processed_filtered
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
- cosine_accuracy@1
- cosine_accuracy@3
- cosine_accuracy@5
- cosine_accuracy@10
- cosine_precision@1
- cosine_precision@3
- cosine_precision@5
- cosine_precision@10
- cosine_recall@1
- cosine_recall@3
- cosine_recall@5
- cosine_recall@10
- cosine_ndcg@10
- cosine_mrr@10
- cosine_map@100
model-index:
- name: EmbeddingGemma-300m fine-tuned on Amazon Toys & Games
  results:
  - task:
      type: information-retrieval
      name: Information Retrieval
    dataset:
      name: dev eval
      type: dev-eval
    metrics:
    - type: cosine_accuracy@1
      value: 0.308
      name: Cosine Accuracy@1
    - type: cosine_accuracy@3
      value: 0.437
      name: Cosine Accuracy@3
    - type: cosine_accuracy@5
      value: 0.507
      name: Cosine Accuracy@5
    - type: cosine_accuracy@10
      value: 0.585
      name: Cosine Accuracy@10
    - type: cosine_precision@1
      value: 0.308
      name: Cosine Precision@1
    - type: cosine_precision@3
      value: 0.1456666666666667
      name: Cosine Precision@3
    - type: cosine_precision@5
      value: 0.10139999999999998
      name: Cosine Precision@5
    - type: cosine_precision@10
      value: 0.0585
      name: Cosine Precision@10
    - type: cosine_recall@1
      value: 0.308
      name: Cosine Recall@1
    - type: cosine_recall@3
      value: 0.437
      name: Cosine Recall@3
    - type: cosine_recall@5
      value: 0.507
      name: Cosine Recall@5
    - type: cosine_recall@10
      value: 0.585
      name: Cosine Recall@10
    - type: cosine_ndcg@10
      value: 0.43779247661127096
      name: Cosine Ndcg@10
    - type: cosine_mrr@10
      value: 0.39167420634920597
      name: Cosine Mrr@10
    - type: cosine_map@100
      value: 0.3994706245247452
      name: Cosine Map@100
---

# EmbeddingGemma-300m fine-tuned on Amazon Toys & Games

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.

## Model Details

### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [google/embeddinggemma-300m](https://huggingface.co/google/embeddinggemma-300m) <!-- at revision 57c266a740f537b4dc058e1b0cda161fd15afa75 -->
- **Maximum Sequence Length:** 2048 tokens
- **Output Dimensionality:** 768 dimensions
- **Similarity Function:** Cosine Similarity
- **Training Dataset:**
    - [amazon_2023_items_processed_filtered](https://huggingface.co/datasets/guyhadad01/Amazon_2023_items_processed_filtered)
- **Language:** en
- **License:** apache-2.0

### Model Sources

- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)

### Full Model Architecture

```
SentenceTransformer(
  (0): Transformer({'max_seq_length': 2048, 'do_lower_case': False, 'architecture': 'Gemma3TextModel'})
  (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})
  (2): Dense({'in_features': 768, 'out_features': 3072, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (3): Dense({'in_features': 3072, 'out_features': 768, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (4): Normalize()
)
```

## Usage

### Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

```bash
pip install -U sentence-transformers
```

Then you can load this model and run inference.
```python
from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("rabaevn/EncodeRec")
# Run inference
queries = [
    "title: \nFairytale and Historic Minifigure Set",
]
documents = [
    'description\nThe LEGO Education 779349 or 9349 227piece fairytale and historic minifigure set includes elements to build 22 multicultural male and female minifigures representing characters from fairytales and history Children four years and older can explore roleplaying while creating and reenacting fairytales and stories developing imagination and imitative skills Themes include pirates a witch and wizard a king and queen knights mine workers a mermaid and merman and more and accessories such as wands hats flowers spiders and snakes The set is compatible with any LEGO Education system supports a group of five students and comes with a decorated box for storage\nSince 1980 LEGO Education has delivered handson curriculumbased resources for teachers and students worldwide LEGO Education believes a handson mindson approach helps students actively take ownership of the learning process and develop 21stcentury skills such as creative thinking and problem solving through reallife engaging experiencesExploring differences to real life makebelieve and historic characters\nStorytelling through characters and things they do\nDeveloping fantasy imagination and imitative skills\nFeaturing a variety of minifigures that enable children to create and act out their favourite fairytales and stories\nIdeal for 4 years old children or above',
    'description\nToysmiths mission is to supply quality toys and gifts while delivering superior customer service to retailers We offer products in many key categories including active play science  discovery arts  crafts impulse  novelty toys and nostalgic retro classics Since our inception in 1981 our owners Bill and Nancy Smith have worked very diligently to grow from a small family business shipping from their garage to a large office complex and distribution center that stocks and ships over 1600 products to more than 6000 accounts However our philosophy remains the same maintaining a strong family work environment while offering the best selection and quality of products together with the highest standards of service to all our customersEducational toys that help children learn\nMade using safe and high quality materials\nToys for all age groups',
    'description\nFeatures 1 Cartoon Dinosaur Plush Doll is suitable for being Ornament Gift Pillow and Rag Toy 2 Made of quality soft plush and being fully filled with pp cotton Not afraid of squeezing 3 Not easy to deform with good resilience It has tightly threaded side which features not easy to produce cotton 4 Perfect to cuddle with in bed or to keep you company when you nap on the couch Great gift for kids girls and anyone you love 5 For the cleaning of such short plush doll clean the surface with cold water and place it in the sun after washing to ensure that the short plush is fluffy Notice Actual color may be slightly different from the image due to different display and light effect Please allow 13cm deviation due to manual measurement Description Material plush PP cotton Size 3560cm Decoration form hanging placing ColorStyle As the picture shown',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 768] [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[0.4024, 0.3496, 0.2501]])
```

<!--
### Direct Usage (Transformers)

<details><summary>Click to see the direct usage in Transformers</summary>

</details>
-->

<!--
### Downstream Usage (Sentence Transformers)

You can finetune this model on your own dataset.

<details><summary>Click to expand</summary>

</details>
-->

<!--
### Out-of-Scope Use

*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->

## Evaluation

### Metrics

#### Information Retrieval

* Dataset: `dev-eval`
* Evaluated with [<code>InformationRetrievalEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator)

| Metric              | Value      |
|:--------------------|:-----------|
| cosine_accuracy@1   | 0.308      |
| cosine_accuracy@3   | 0.437      |
| cosine_accuracy@5   | 0.507      |
| cosine_accuracy@10  | 0.585      |
| cosine_precision@1  | 0.308      |
| cosine_precision@3  | 0.1457     |
| cosine_precision@5  | 0.1014     |
| cosine_precision@10 | 0.0585     |
| cosine_recall@1     | 0.308      |
| cosine_recall@3     | 0.437      |
| cosine_recall@5     | 0.507      |
| cosine_recall@10    | 0.585      |
| **cosine_ndcg@10**  | **0.4378** |
| cosine_mrr@10       | 0.3917     |
| cosine_map@100      | 0.3995     |

<!--
## Bias, Risks and Limitations

*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->

<!--
### Recommendations

*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->

## Training Details

### Training Dataset

#### amazon_2023_items_processed_filtered

* 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)
* Size: 441,985 training samples
* Columns: <code>question</code> and <code>passage_text</code>
* Approximate statistics based on the first 1000 samples:
  |         | question                                                                          | passage_text                                                                         |
  |:--------|:----------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
  | type    | string                                                                            | string                                                                               |
  | details | <ul><li>min: 9 tokens</li><li>mean: 25.35 tokens</li><li>max: 78 tokens</li></ul> | <ul><li>min: 17 tokens</li><li>mean: 194.73 tokens</li><li>max: 892 tokens</li></ul> |
* Samples:
  | question                                                                                                                                                                               | passage_text                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      |
  |:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
  | <code>title: <br>My Little Pony Twisty Twirly Hairstyles Applejack</code>                                                                                                              | <code>description<br>Style Applejack ponys hair for the best night ever Inspired by the My Little Pony Friendship is Magic television series Applejack is going to the Grand Galloping Gala at Canterlot Castle This Applejack figure is poseable with waxinfused yarn hair that can be styled up in barrettes or fancy braids Theres many style possibilities This pony figure comes with 10 accessories for making Applejack ponys beautiful yellow hair look gorgeous for the gala Also look for Twisty Twirly Hairstyles Rarity and Pinkie Pie figures and mix and match accessories for more hairstyling fun Each sold separately Subject to availability My Little Pony and all related characters are trademarks of HasbroInspired by the My Little Pony Friendship is Magic television series<br>Pretend to do Applejack ponys hair with clips and barrettes<br>Poseable waxinfused yarn hair can be styled up or down<br>Figure scale 3 inches<br>Includes pony figure and 10 accessories</code>                                                         |
  | <code>title: <br>TUANTUAN 4 Pcs 16 Scale Dollhouse Miniature Furniture Folding Chair Foldable Chair Model Folding Doll Chairs Decor Black Foldable Chair for Figure Accessories</code> | <code>description<br>MATERIALPlasticAlloyColour BlackSize Folding Size 173  85CmExpanding Size 8895153Cm GOOD QUALITYThese Mini Folding Chairs Are Made Of Plastic And Alloy Materials Which Are Environmentally Friendly And NonToxic Strong And Durable Stand Firm UNIQUE DESIGNMini Folding Chair Design Looks Like A Real Chair Small And Exquisite Realistic Appearance More Interesting To Match With Dolls WIDELY USEDThese Mini Folding Chairs Are 16 Scene Accessories Not For Real People Use And Are Exclusively For 12Inch Dolls Suitable For Doll House Accessories Miniature Furniture House Model Decoration It Can Also Be Used As A Mobile Phone Holder PACKAGE INCLUDE4 Pcs Dolls Folding ChairsMATERIALPlasticAlloyColour BlackSize Folding Size 173  85CmExpanding Size 8895153Cm<br>GOOD QUALITYThese Mini Folding Chairs Are Made Of Plastic And Alloy Materials Which Are Environmentally Friendly And NonToxic Strong And Durable Stand Firm<br>UNIQUE DESIGNMini Folding Chair Design Looks Like A Real Chair Small And Exquis...</code> |
  | <code>title: <br>Transformers Movie RD27 NEST Sky stalker japan import</code>                                                                                                          | <code>description<br>Manufactured by Takara Tomy Product name Transformers Revenge of the Fallen Scout class RD27 NEST Sky stalker  TRANSFORMERS REVENGE OF THE FALLEN SCUOT CLASS NEST SKYSTALKERb safety standards  b ST Mark<br>b target Gender  b boy<br>b Age  b from 5 years</code>                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         |
* Loss: [<code>CachedMultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cachedmultiplenegativesrankingloss) with these parameters:
  ```json
  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "mini_batch_size": 32,
      "gather_across_devices": false
  }
  ```

### Evaluation Dataset

#### amazon_2023_items_processed_filtered

* 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)
* Size: 110,497 evaluation samples
* Columns: <code>question</code> and <code>passage_text</code>
* Approximate statistics based on the first 1000 samples:
  |         | question                                                                         | passage_text                                                                         |
  |:--------|:---------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
  | type    | string                                                                           | string                                                                               |
  | details | <ul><li>min: 7 tokens</li><li>mean: 25.1 tokens</li><li>max: 93 tokens</li></ul> | <ul><li>min: 15 tokens</li><li>mean: 189.31 tokens</li><li>max: 948 tokens</li></ul> |
* Samples:
  | question                                                                                                                                            | passage_text                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         |
  |:----------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
  | <code>title: <br>Reaper Miniatures Tengu Warrior 03774 Dark Heaven Legends RPG DD Mini Figure</code>                                                | <code>description<br>Tengu WarriorBy Artist Derek SchubertDark Heaven Legends25mm Heroic Scale Fantasy MiniaturesIntegral builtin basesUnpainted metal models that may require assembly If assembly is needed glue or putty is required not includedHuge selection of characters and monsters for fantasy roleplayers miniatures painters and wargamersDark Heaven Legends is the premier 25mm Heroic Scale fantasy line for miniature painters roleplayers and wargamers Over the last thirteen years the Dark Heaven line has produced over 1300 fantasy miniatures designed and crafted by the top miniatures sculptors in the worldFound in Reaper Miniatures Category Dark Heaven LegendsFound in Reaper Miniatures Category Dark Heaven Legends<br>Unpainted Metal Miniature<br>Figure designed by artist Derek Schubert</code>                                                                                                                                                                                                                                |
  | <code>title: <br>Marvel Legends  Black Widow Action Figure</code>                                                                                   | <code>description<br>Scale 6 Inch Format Action Figures Packaging Clamshell Manufacturer Toy Biz Natalia Romanova aka Natasha Romanoff is the Black Widow as beautiful as Russian spy as their has ever been and as deadly as her namesake Working with everyone from SHIELD to the Avengers to the Marvel Knights she is widely known as a valuable asset to any crimefighting team In this 6inch Marvel Legends her beauty and power are perfectly captured in the sculpt of her skintight black outfit We receive both US and Canadian Cases You will get either English bilingual or trilingual carded figures based on availability Please understand this before ordering Package condition may vary due to size and weight We do not guarantee mint on cardMarvel Legends 6 Inch Action Figure Man Thing Series  Black Widow<br>Marvel<br>Toybiz<br>6 Inch</code>                                                                                                                                                                                             |
  | <code>title: <br>Cabbage Patch Kids Cutie Dash The Deer 9  Collectible Adoptable Baby Doll Toy  Officially Licensed  Gift for Girls and Boys</code> | <code>description<br>This Holiday complete your collection of Cabbage Patch Cuties by adopting the Dash Reindeer doll Each Cabbage Patch Cutie features a snuggly onesie with adjustable hood and can really suck its thumb Cutie dolls come with the traditional signature baby powder scent that Cabbage Patch Kid fans know and love Take the Oath of Adoption Cabbage Patch Kid Cuties are numbered for collectibility and make a great toy gift for boys and girls who love Cabbage Patch Kids Officially licensed Cabbage Patch Kids merchandise 3 Pack Each Measures approximately 9 tall Comes in sealed polybag packaging with official Cabbage Patch tagThis Holiday complete your collection of Cabbage Patch Cuties by adopting Dash the Reindeer baby doll<br>Its a Cabbage Christmas Each Cabbage Patch Cutie features a snuggly onesie with adjustable hood and can really suck its thumb<br>Cutie baby dolls come with the traditional signature baby powder scent that Cabbage Patch Kid fans know and love<br>Take the Oath of Adoption ...</code> |
* Loss: [<code>CachedMultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cachedmultiplenegativesrankingloss) with these parameters:
  ```json
  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "mini_batch_size": 32,
      "gather_across_devices": false
  }
  ```

### Training Hyperparameters
#### Non-Default Hyperparameters

- `eval_strategy`: steps
- `per_device_train_batch_size`: 2
- `per_device_eval_batch_size`: 1
- `gradient_accumulation_steps`: 32
- `torch_empty_cache_steps`: 50
- `learning_rate`: 2e-05
- `num_train_epochs`: 1
- `warmup_ratio`: 0.1
- `fp16`: True
- `dataloader_num_workers`: 4
- `push_to_hub`: True
- `hub_model_id`: rabaevn/EncodeRec
- `hub_strategy`: checkpoint
- `gradient_checkpointing`: True
- `batch_sampler`: no_duplicates

#### All Hyperparameters
<details><summary>Click to expand</summary>

- `overwrite_output_dir`: False
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 2
- `per_device_eval_batch_size`: 1
- `per_gpu_train_batch_size`: None
- `per_gpu_eval_batch_size`: None
- `gradient_accumulation_steps`: 32
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: 50
- `learning_rate`: 2e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 1
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: {}
- `warmup_ratio`: 0.1
- `warmup_steps`: 0
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `save_safetensors`: True
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `no_cuda`: False
- `use_cpu`: False
- `use_mps_device`: False
- `seed`: 42
- `data_seed`: None
- `jit_mode_eval`: False
- `use_ipex`: False
- `bf16`: False
- `fp16`: True
- `fp16_opt_level`: O1
- `half_precision_backend`: auto
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: 0
- `ddp_backend`: None
- `tpu_num_cores`: None
- `tpu_metrics_debug`: False
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 4
- `dataloader_prefetch_factor`: None
- `past_index`: -1
- `disable_tqdm`: False
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: False
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_min_num_params`: 0
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `fsdp_transformer_layer_cls_to_wrap`: None
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch_fused
- `optim_args`: None
- `adafactor`: False
- `group_by_length`: False
- `length_column_name`: length
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `use_legacy_prediction_loop`: False
- `push_to_hub`: True
- `resume_from_checkpoint`: None
- `hub_model_id`: rabaevn/EncodeRec
- `hub_strategy`: checkpoint
- `hub_private_repo`: None
- `hub_always_push`: False
- `hub_revision`: None
- `gradient_checkpointing`: True
- `gradient_checkpointing_kwargs`: None
- `include_inputs_for_metrics`: False
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `fp16_backend`: auto
- `push_to_hub_model_id`: None
- `push_to_hub_organization`: None
- `mp_parameters`: 
- `auto_find_batch_size`: False
- `full_determinism`: False
- `torchdynamo`: None
- `ray_scope`: last
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `include_tokens_per_second`: False
- `include_num_input_tokens_seen`: False
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `liger_kernel_config`: None
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: False
- `prompts`: None
- `batch_sampler`: no_duplicates
- `multi_dataset_batch_sampler`: proportional
- `router_mapping`: {}
- `learning_rate_mapping`: {}

</details>

### Training Logs
| Epoch  | Step | Training Loss | Validation Loss | dev-eval_cosine_ndcg@10 |
|:------:|:----:|:-------------:|:---------------:|:-----------------------:|
| 0.0072 | 50   | 0.0428        | -               | -                       |
| 0.0145 | 100  | 0.0118        | -               | -                       |
| 0.0217 | 150  | 0.0087        | -               | -                       |
| 0.0290 | 200  | 0.0064        | -               | -                       |
| 0.0362 | 250  | 0.0069        | -               | -                       |
| 0.0434 | 300  | 0.0088        | -               | -                       |
| 0.0507 | 350  | 0.0055        | -               | -                       |
| 0.0579 | 400  | 0.0067        | -               | -                       |
| 0.0652 | 450  | 0.0098        | -               | -                       |
| 0.0724 | 500  | 0.0096        | -               | -                       |
| 0.0796 | 550  | 0.0104        | -               | -                       |
| 0.0869 | 600  | 0.0155        | -               | -                       |
| 0.0941 | 650  | 0.0109        | -               | -                       |
| 0.1014 | 700  | 0.0144        | -               | -                       |
| 0.1086 | 750  | 0.0109        | -               | -                       |
| 0.1158 | 800  | 0.0107        | 0.0             | 0.4378                  |


### Framework Versions
- Python: 3.10.18
- Sentence Transformers: 5.1.0
- Transformers: 4.55.2
- PyTorch: 2.8.0+cu126
- Accelerate: 1.10.0
- Datasets: 4.1.1
- Tokenizers: 0.21.4

## Citation

### BibTeX

#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}
```

#### CachedMultipleNegativesRankingLoss
```bibtex
@misc{gao2021scaling,
    title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
    author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
    year={2021},
    eprint={2101.06983},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}
```

<!--
## Glossary

*Clearly define terms in order to be accessible across audiences.*
-->

<!--
## Model Card Authors

*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->

<!--
## Model Card Contact

*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->