VinitT commited on
Commit
d38d730
·
verified ·
1 Parent(s): df268de

Upload retrain embedding model

Browse files
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 384,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
README.md ADDED
@@ -0,0 +1,577 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ tags:
5
+ - sentence-transformers
6
+ - sentence-similarity
7
+ - feature-extraction
8
+ - dense
9
+ - generated_from_trainer
10
+ - dataset_size:556850
11
+ - loss:ContradictionMarginLoss
12
+ base_model: VinitT/Embeddings-Trivia
13
+ widget:
14
+ - source_sentence: Guy wearing sunglasses and blue shirt on skateboard in front of
15
+ a bright yellow building with palm trees.
16
+ sentences:
17
+ - Two people are standing by the street.
18
+ - A man rides a skateboard outside.
19
+ - The boys are inside laying down.
20
+ - source_sentence: In a park, a boy is bent to read the tree description, and a girl
21
+ is standing nearby waiting for him.
22
+ sentences:
23
+ - A boy and girl out in the park while looking at the scenery.
24
+ - Some girls are climbing.
25
+ - An illiterate boy standing up reading a tree description.
26
+ - source_sentence: A man in a blue shirt gesticulates as he speaks to a uniformed
27
+ official.
28
+ sentences:
29
+ - A man has a mouthfull of meatballs.
30
+ - A man is speaking with an official
31
+ - Women are working in a lab
32
+ - source_sentence: John left me, and a few minutes later I saw him from my window
33
+ walking slowly across the grass arm in arm with Cynthia Murdoch.
34
+ sentences:
35
+ - John left me to then walk with Cynthia Murdoch.
36
+ - A girl is wearing a crown while having a funny look on her face.
37
+ - John stayed and ignored Cynthia as she walked by.
38
+ - source_sentence: so he has overcome alcoholism at this point
39
+ sentences:
40
+ - A dog is holding a toy.
41
+ - He still is a heavy drinker and can't control it.
42
+ - He's gotten stronger and has overcome alcoholism.
43
+ datasets:
44
+ - sentence-transformers/all-nli
45
+ pipeline_tag: sentence-similarity
46
+ library_name: sentence-transformers
47
+ metrics:
48
+ - cosine_accuracy
49
+ model-index:
50
+ - name: SentenceTransformer based on VinitT/Embeddings-Trivia
51
+ results:
52
+ - task:
53
+ type: triplet
54
+ name: Triplet
55
+ dataset:
56
+ name: contra eval
57
+ type: contra_eval
58
+ metrics:
59
+ - type: cosine_accuracy
60
+ value: 0.949999988079071
61
+ name: Cosine Accuracy
62
+ ---
63
+
64
+ # SentenceTransformer based on VinitT/Embeddings-Trivia
65
+
66
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [VinitT/Embeddings-Trivia](https://huggingface.co/VinitT/Embeddings-Trivia) on the [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) dataset. It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
67
+
68
+ ## Model Details
69
+
70
+ ### Model Description
71
+ - **Model Type:** Sentence Transformer
72
+ - **Base model:** [VinitT/Embeddings-Trivia](https://huggingface.co/VinitT/Embeddings-Trivia) <!-- at revision f1c49cbecdbb76b4efd6ea97c91600816de94bb3 -->
73
+ - **Maximum Sequence Length:** 256 tokens
74
+ - **Output Dimensionality:** 384 dimensions
75
+ - **Similarity Function:** Cosine Similarity
76
+ - **Training Dataset:**
77
+ - [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli)
78
+ - **Language:** en
79
+ <!-- - **License:** Unknown -->
80
+
81
+ ### Model Sources
82
+
83
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
84
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
85
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
86
+
87
+ ### Full Model Architecture
88
+
89
+ ```
90
+ SentenceTransformer(
91
+ (0): Transformer({'max_seq_length': 256, 'do_lower_case': False, 'architecture': 'BertModel'})
92
+ (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
93
+ (2): Normalize()
94
+ )
95
+ ```
96
+
97
+ ## Usage
98
+
99
+ ### Direct Usage (Sentence Transformers)
100
+
101
+ First install the Sentence Transformers library:
102
+
103
+ ```bash
104
+ pip install -U sentence-transformers
105
+ ```
106
+
107
+ Then you can load this model and run inference.
108
+ ```python
109
+ from sentence_transformers import SentenceTransformer
110
+
111
+ # Download from the 🤗 Hub
112
+ model = SentenceTransformer("sentence_transformers_model_id")
113
+ # Run inference
114
+ sentences = [
115
+ 'so he has overcome alcoholism at this point',
116
+ "He's gotten stronger and has overcome alcoholism.",
117
+ "He still is a heavy drinker and can't control it.",
118
+ ]
119
+ embeddings = model.encode(sentences)
120
+ print(embeddings.shape)
121
+ # [3, 384]
122
+
123
+ # Get the similarity scores for the embeddings
124
+ similarities = model.similarity(embeddings, embeddings)
125
+ print(similarities)
126
+ # tensor([[1.0000, 0.7603, 0.0849],
127
+ # [0.7603, 1.0000, 0.0794],
128
+ # [0.0849, 0.0794, 1.0000]])
129
+ ```
130
+
131
+ <!--
132
+ ### Direct Usage (Transformers)
133
+
134
+ <details><summary>Click to see the direct usage in Transformers</summary>
135
+
136
+ </details>
137
+ -->
138
+
139
+ <!--
140
+ ### Downstream Usage (Sentence Transformers)
141
+
142
+ You can finetune this model on your own dataset.
143
+
144
+ <details><summary>Click to expand</summary>
145
+
146
+ </details>
147
+ -->
148
+
149
+ <!--
150
+ ### Out-of-Scope Use
151
+
152
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
153
+ -->
154
+
155
+ ## Evaluation
156
+
157
+ ### Metrics
158
+
159
+ #### Triplet
160
+
161
+ * Dataset: `contra_eval`
162
+ * Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
163
+
164
+ | Metric | Value |
165
+ |:--------------------|:---------|
166
+ | **cosine_accuracy** | **0.95** |
167
+
168
+ <!--
169
+ ## Bias, Risks and Limitations
170
+
171
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
172
+ -->
173
+
174
+ <!--
175
+ ### Recommendations
176
+
177
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
178
+ -->
179
+
180
+ ## Training Details
181
+
182
+ ### Training Dataset
183
+
184
+ #### all-nli
185
+
186
+ * Dataset: [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) at [d482672](https://huggingface.co/datasets/sentence-transformers/all-nli/tree/d482672c8e74ce18da116f430137434ba2e52fab)
187
+ * Size: 556,850 training samples
188
+ * Columns: <code>anchor</code>, <code>positive</code>, <code>negative</code>, and <code>label</code>
189
+ * Approximate statistics based on the first 1000 samples:
190
+ | | anchor | positive | negative | label |
191
+ |:--------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:-----------------------------|
192
+ | type | string | string | string | int |
193
+ | details | <ul><li>min: 5 tokens</li><li>mean: 19.16 tokens</li><li>max: 194 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 11.86 tokens</li><li>max: 32 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 12.23 tokens</li><li>max: 37 tokens</li></ul> | <ul><li>1: 100.00%</li></ul> |
194
+ * Samples:
195
+ | anchor | positive | negative | label |
196
+ |:----------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------|:---------------------------------------------------------------------|:---------------|
197
+ | <code>a young girl wearing blue smiles.</code> | <code>A little girl wears blue.</code> | <code>A little girl frowns as she wears an ugly burlap sack.</code> | <code>1</code> |
198
+ | <code>An old man wearing a tan jacket and blue pants standing on a sidewalk with a small suitcase.</code> | <code>A man wearing a jacket and jeans holds a suitcase.</code> | <code>A young woman sits on a bench holding her purse.</code> | <code>1</code> |
199
+ | <code>The people are inside.</code> | <code>Two people are dancing by a red couch.</code> | <code>People walk up and down the steps in front of a church.</code> | <code>1</code> |
200
+ * Loss: <code>custom_loss.ContradictionMarginLoss</code> with these parameters:
201
+ ```json
202
+ {
203
+ "margin_neutral": 0.2,
204
+ "margin_contradiction": 0.4
205
+ }
206
+ ```
207
+
208
+ ### Evaluation Dataset
209
+
210
+ #### all-nli
211
+
212
+ * Dataset: [all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) at [d482672](https://huggingface.co/datasets/sentence-transformers/all-nli/tree/d482672c8e74ce18da116f430137434ba2e52fab)
213
+ * Size: 1,000 evaluation samples
214
+ * Columns: <code>anchor</code>, <code>positive</code>, <code>negative</code>, and <code>label</code>
215
+ * Approximate statistics based on the first 1000 samples:
216
+ | | anchor | positive | negative | label |
217
+ |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:-----------------------------|
218
+ | type | string | string | string | int |
219
+ | details | <ul><li>min: 5 tokens</li><li>mean: 18.67 tokens</li><li>max: 86 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 11.92 tokens</li><li>max: 41 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 12.13 tokens</li><li>max: 40 tokens</li></ul> | <ul><li>1: 100.00%</li></ul> |
220
+ * Samples:
221
+ | anchor | positive | negative | label |
222
+ |:------------------------------------------------------|:---------------------------------------------------------------------------|:----------------------------------------------------|:---------------|
223
+ | <code>An older man riding a bike.</code> | <code>An elderly man is biking</code> | <code>an old man is sleeping</code> | <code>1</code> |
224
+ | <code>The man is on a skateboard.</code> | <code>A shirtless man is doing a skateboard trick over a bike rail.</code> | <code>A man performs a bike trick on a ramp.</code> | <code>1</code> |
225
+ | <code>The Episcopalians are all going to hell.</code> | <code>The Episcopalians will not be going to heaven.</code> | <code>All Episcopalians will go to heaven.</code> | <code>1</code> |
226
+ * Loss: <code>custom_loss.ContradictionMarginLoss</code> with these parameters:
227
+ ```json
228
+ {
229
+ "margin_neutral": 0.2,
230
+ "margin_contradiction": 0.4
231
+ }
232
+ ```
233
+
234
+ ### Training Hyperparameters
235
+ #### Non-Default Hyperparameters
236
+
237
+ - `eval_strategy`: steps
238
+ - `per_device_train_batch_size`: 64
239
+ - `per_device_eval_batch_size`: 64
240
+ - `learning_rate`: 2e-05
241
+ - `weight_decay`: 0.01
242
+ - `num_train_epochs`: 1
243
+ - `warmup_ratio`: 0.1
244
+ - `warmup_steps`: 0.1
245
+ - `fp16`: True
246
+ - `load_best_model_at_end`: True
247
+
248
+ #### All Hyperparameters
249
+ <details><summary>Click to expand</summary>
250
+
251
+ - `do_predict`: False
252
+ - `eval_strategy`: steps
253
+ - `prediction_loss_only`: True
254
+ - `per_device_train_batch_size`: 64
255
+ - `per_device_eval_batch_size`: 64
256
+ - `gradient_accumulation_steps`: 1
257
+ - `eval_accumulation_steps`: None
258
+ - `torch_empty_cache_steps`: None
259
+ - `learning_rate`: 2e-05
260
+ - `weight_decay`: 0.01
261
+ - `adam_beta1`: 0.9
262
+ - `adam_beta2`: 0.999
263
+ - `adam_epsilon`: 1e-08
264
+ - `max_grad_norm`: 1.0
265
+ - `num_train_epochs`: 1
266
+ - `max_steps`: -1
267
+ - `lr_scheduler_type`: linear
268
+ - `lr_scheduler_kwargs`: None
269
+ - `warmup_ratio`: 0.1
270
+ - `warmup_steps`: 0.1
271
+ - `log_level`: passive
272
+ - `log_level_replica`: warning
273
+ - `log_on_each_node`: True
274
+ - `logging_nan_inf_filter`: True
275
+ - `enable_jit_checkpoint`: False
276
+ - `save_on_each_node`: False
277
+ - `save_only_model`: False
278
+ - `restore_callback_states_from_checkpoint`: False
279
+ - `use_cpu`: False
280
+ - `seed`: 42
281
+ - `data_seed`: None
282
+ - `bf16`: False
283
+ - `fp16`: True
284
+ - `bf16_full_eval`: False
285
+ - `fp16_full_eval`: False
286
+ - `tf32`: None
287
+ - `local_rank`: -1
288
+ - `ddp_backend`: None
289
+ - `debug`: []
290
+ - `dataloader_drop_last`: False
291
+ - `dataloader_num_workers`: 0
292
+ - `dataloader_prefetch_factor`: None
293
+ - `disable_tqdm`: False
294
+ - `remove_unused_columns`: True
295
+ - `label_names`: None
296
+ - `load_best_model_at_end`: True
297
+ - `ignore_data_skip`: False
298
+ - `fsdp`: []
299
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
300
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
301
+ - `parallelism_config`: None
302
+ - `deepspeed`: None
303
+ - `label_smoothing_factor`: 0.0
304
+ - `optim`: adamw_torch_fused
305
+ - `optim_args`: None
306
+ - `group_by_length`: False
307
+ - `length_column_name`: length
308
+ - `project`: huggingface
309
+ - `trackio_space_id`: trackio
310
+ - `ddp_find_unused_parameters`: None
311
+ - `ddp_bucket_cap_mb`: None
312
+ - `ddp_broadcast_buffers`: False
313
+ - `dataloader_pin_memory`: True
314
+ - `dataloader_persistent_workers`: False
315
+ - `skip_memory_metrics`: True
316
+ - `push_to_hub`: False
317
+ - `resume_from_checkpoint`: None
318
+ - `hub_model_id`: None
319
+ - `hub_strategy`: every_save
320
+ - `hub_private_repo`: None
321
+ - `hub_always_push`: False
322
+ - `hub_revision`: None
323
+ - `gradient_checkpointing`: False
324
+ - `gradient_checkpointing_kwargs`: None
325
+ - `include_for_metrics`: []
326
+ - `eval_do_concat_batches`: True
327
+ - `auto_find_batch_size`: False
328
+ - `full_determinism`: False
329
+ - `ddp_timeout`: 1800
330
+ - `torch_compile`: False
331
+ - `torch_compile_backend`: None
332
+ - `torch_compile_mode`: None
333
+ - `include_num_input_tokens_seen`: no
334
+ - `neftune_noise_alpha`: None
335
+ - `optim_target_modules`: None
336
+ - `batch_eval_metrics`: False
337
+ - `eval_on_start`: False
338
+ - `use_liger_kernel`: False
339
+ - `liger_kernel_config`: None
340
+ - `eval_use_gather_object`: False
341
+ - `average_tokens_across_devices`: True
342
+ - `use_cache`: False
343
+ - `prompts`: None
344
+ - `batch_sampler`: batch_sampler
345
+ - `multi_dataset_batch_sampler`: proportional
346
+ - `router_mapping`: {}
347
+ - `learning_rate_mapping`: {}
348
+
349
+ </details>
350
+
351
+ ### Training Logs
352
+ <details><summary>Click to expand</summary>
353
+
354
+ | Epoch | Step | Training Loss | Validation Loss | contra_eval_cosine_accuracy |
355
+ |:---------:|:--------:|:-------------:|:---------------:|:---------------------------:|
356
+ | 0.0001 | 1 | 0.2363 | - | - |
357
+ | 0.0057 | 50 | 0.1877 | - | - |
358
+ | 0.0115 | 100 | 0.1786 | - | - |
359
+ | 0.0172 | 150 | 0.1672 | - | - |
360
+ | 0.0230 | 200 | 0.1529 | - | - |
361
+ | 0.0287 | 250 | 0.1392 | - | - |
362
+ | 0.0345 | 300 | 0.1278 | - | - |
363
+ | 0.0402 | 350 | 0.1233 | - | - |
364
+ | 0.0460 | 400 | 0.1157 | - | - |
365
+ | 0.0517 | 450 | 0.1116 | - | - |
366
+ | 0.0575 | 500 | 0.1063 | 0.0983 | 0.9260 |
367
+ | 0.0632 | 550 | 0.1087 | - | - |
368
+ | 0.0690 | 600 | 0.1016 | - | - |
369
+ | 0.0747 | 650 | 0.1026 | - | - |
370
+ | 0.0805 | 700 | 0.0967 | - | - |
371
+ | 0.0862 | 750 | 0.0990 | - | - |
372
+ | 0.0919 | 800 | 0.0925 | - | - |
373
+ | 0.0977 | 850 | 0.0965 | - | - |
374
+ | 0.1034 | 900 | 0.0981 | - | - |
375
+ | 0.1092 | 950 | 0.0881 | - | - |
376
+ | 0.1149 | 1000 | 0.0920 | 0.0829 | 0.9410 |
377
+ | 0.1207 | 1050 | 0.0882 | - | - |
378
+ | 0.1264 | 1100 | 0.0839 | - | - |
379
+ | 0.1322 | 1150 | 0.0896 | - | - |
380
+ | 0.1379 | 1200 | 0.0858 | - | - |
381
+ | 0.1437 | 1250 | 0.0878 | - | - |
382
+ | 0.1494 | 1300 | 0.0857 | - | - |
383
+ | 0.1552 | 1350 | 0.0902 | - | - |
384
+ | 0.1609 | 1400 | 0.0793 | - | - |
385
+ | 0.1666 | 1450 | 0.0830 | - | - |
386
+ | 0.1724 | 1500 | 0.0827 | 0.0788 | 0.9380 |
387
+ | 0.1781 | 1550 | 0.0789 | - | - |
388
+ | 0.1839 | 1600 | 0.0834 | - | - |
389
+ | 0.1896 | 1650 | 0.0805 | - | - |
390
+ | 0.1954 | 1700 | 0.0795 | - | - |
391
+ | 0.2011 | 1750 | 0.0846 | - | - |
392
+ | 0.2069 | 1800 | 0.0822 | - | - |
393
+ | 0.2126 | 1850 | 0.0858 | - | - |
394
+ | 0.2184 | 1900 | 0.0785 | - | - |
395
+ | 0.2241 | 1950 | 0.0777 | - | - |
396
+ | 0.2299 | 2000 | 0.0746 | 0.0721 | 0.9460 |
397
+ | 0.2356 | 2050 | 0.0798 | - | - |
398
+ | 0.2414 | 2100 | 0.0798 | - | - |
399
+ | 0.2471 | 2150 | 0.0794 | - | - |
400
+ | 0.2528 | 2200 | 0.0769 | - | - |
401
+ | 0.2586 | 2250 | 0.0805 | - | - |
402
+ | 0.2643 | 2300 | 0.0782 | - | - |
403
+ | 0.2701 | 2350 | 0.0776 | - | - |
404
+ | 0.2758 | 2400 | 0.0776 | - | - |
405
+ | 0.2816 | 2450 | 0.0733 | - | - |
406
+ | 0.2873 | 2500 | 0.0750 | 0.0718 | 0.9440 |
407
+ | 0.2931 | 2550 | 0.0764 | - | - |
408
+ | 0.2988 | 2600 | 0.0775 | - | - |
409
+ | 0.3046 | 2650 | 0.0767 | - | - |
410
+ | 0.3103 | 2700 | 0.0766 | - | - |
411
+ | 0.3161 | 2750 | 0.0755 | - | - |
412
+ | 0.3218 | 2800 | 0.0752 | - | - |
413
+ | 0.3275 | 2850 | 0.0717 | - | - |
414
+ | 0.3333 | 2900 | 0.0714 | - | - |
415
+ | 0.3390 | 2950 | 0.0726 | - | - |
416
+ | 0.3448 | 3000 | 0.0751 | 0.0695 | 0.9470 |
417
+ | 0.3505 | 3050 | 0.0730 | - | - |
418
+ | 0.3563 | 3100 | 0.0733 | - | - |
419
+ | 0.3620 | 3150 | 0.0738 | - | - |
420
+ | 0.3678 | 3200 | 0.0701 | - | - |
421
+ | 0.3735 | 3250 | 0.0723 | - | - |
422
+ | 0.3793 | 3300 | 0.0759 | - | - |
423
+ | 0.3850 | 3350 | 0.0675 | - | - |
424
+ | 0.3908 | 3400 | 0.0696 | - | - |
425
+ | 0.3965 | 3450 | 0.0707 | - | - |
426
+ | 0.4023 | 3500 | 0.0705 | 0.0669 | 0.9440 |
427
+ | 0.4080 | 3550 | 0.0702 | - | - |
428
+ | 0.4137 | 3600 | 0.0716 | - | - |
429
+ | 0.4195 | 3650 | 0.0697 | - | - |
430
+ | 0.4252 | 3700 | 0.0721 | - | - |
431
+ | 0.4310 | 3750 | 0.0723 | - | - |
432
+ | 0.4367 | 3800 | 0.0741 | - | - |
433
+ | 0.4425 | 3850 | 0.0702 | - | - |
434
+ | 0.4482 | 3900 | 0.0653 | - | - |
435
+ | 0.4540 | 3950 | 0.0704 | - | - |
436
+ | 0.4597 | 4000 | 0.0718 | 0.0652 | 0.9450 |
437
+ | 0.4655 | 4050 | 0.0683 | - | - |
438
+ | 0.4712 | 4100 | 0.0719 | - | - |
439
+ | 0.4770 | 4150 | 0.0674 | - | - |
440
+ | 0.4827 | 4200 | 0.0659 | - | - |
441
+ | 0.4884 | 4250 | 0.0735 | - | - |
442
+ | 0.4942 | 4300 | 0.0737 | - | - |
443
+ | 0.4999 | 4350 | 0.0707 | - | - |
444
+ | 0.5057 | 4400 | 0.0690 | - | - |
445
+ | 0.5114 | 4450 | 0.0707 | - | - |
446
+ | 0.5172 | 4500 | 0.0696 | 0.0637 | 0.9470 |
447
+ | 0.5229 | 4550 | 0.0686 | - | - |
448
+ | 0.5287 | 4600 | 0.0710 | - | - |
449
+ | 0.5344 | 4650 | 0.0681 | - | - |
450
+ | 0.5402 | 4700 | 0.0667 | - | - |
451
+ | 0.5459 | 4750 | 0.0673 | - | - |
452
+ | 0.5517 | 4800 | 0.0618 | - | - |
453
+ | 0.5574 | 4850 | 0.0715 | - | - |
454
+ | 0.5632 | 4900 | 0.0703 | - | - |
455
+ | 0.5689 | 4950 | 0.0675 | - | - |
456
+ | 0.5746 | 5000 | 0.0715 | 0.0638 | 0.9500 |
457
+ | 0.5804 | 5050 | 0.0681 | - | - |
458
+ | 0.5861 | 5100 | 0.0628 | - | - |
459
+ | 0.5919 | 5150 | 0.0654 | - | - |
460
+ | 0.5976 | 5200 | 0.0662 | - | - |
461
+ | 0.6034 | 5250 | 0.0626 | - | - |
462
+ | 0.6091 | 5300 | 0.0660 | - | - |
463
+ | 0.6149 | 5350 | 0.0652 | - | - |
464
+ | 0.6206 | 5400 | 0.0687 | - | - |
465
+ | 0.6264 | 5450 | 0.0677 | - | - |
466
+ | 0.6321 | 5500 | 0.0683 | 0.0631 | 0.9530 |
467
+ | 0.6379 | 5550 | 0.0666 | - | - |
468
+ | 0.6436 | 5600 | 0.0663 | - | - |
469
+ | 0.6494 | 5650 | 0.0637 | - | - |
470
+ | 0.6551 | 5700 | 0.0687 | - | - |
471
+ | 0.6608 | 5750 | 0.0620 | - | - |
472
+ | 0.6666 | 5800 | 0.0664 | - | - |
473
+ | 0.6723 | 5850 | 0.0666 | - | - |
474
+ | 0.6781 | 5900 | 0.0632 | - | - |
475
+ | 0.6838 | 5950 | 0.0676 | - | - |
476
+ | 0.6896 | 6000 | 0.0638 | 0.0634 | 0.9530 |
477
+ | 0.6953 | 6050 | 0.0655 | - | - |
478
+ | 0.7011 | 6100 | 0.0651 | - | - |
479
+ | 0.7068 | 6150 | 0.0675 | - | - |
480
+ | 0.7126 | 6200 | 0.0685 | - | - |
481
+ | 0.7183 | 6250 | 0.0647 | - | - |
482
+ | 0.7241 | 6300 | 0.0609 | - | - |
483
+ | 0.7298 | 6350 | 0.0643 | - | - |
484
+ | 0.7355 | 6400 | 0.0628 | - | - |
485
+ | 0.7413 | 6450 | 0.0627 | - | - |
486
+ | **0.747** | **6500** | **0.0639** | **0.0621** | **0.954** |
487
+ | 0.7528 | 6550 | 0.0658 | - | - |
488
+ | 0.7585 | 6600 | 0.0667 | - | - |
489
+ | 0.7643 | 6650 | 0.0632 | - | - |
490
+ | 0.7700 | 6700 | 0.0616 | - | - |
491
+ | 0.7758 | 6750 | 0.0666 | - | - |
492
+ | 0.7815 | 6800 | 0.0634 | - | - |
493
+ | 0.7873 | 6850 | 0.0647 | - | - |
494
+ | 0.7930 | 6900 | 0.0644 | - | - |
495
+ | 0.7988 | 6950 | 0.0617 | - | - |
496
+ | 0.8045 | 7000 | 0.0677 | 0.0626 | 0.9510 |
497
+ | 0.8103 | 7050 | 0.0616 | - | - |
498
+ | 0.8160 | 7100 | 0.0633 | - | - |
499
+ | 0.8217 | 7150 | 0.0645 | - | - |
500
+ | 0.8275 | 7200 | 0.0656 | - | - |
501
+ | 0.8332 | 7250 | 0.0597 | - | - |
502
+ | 0.8390 | 7300 | 0.0670 | - | - |
503
+ | 0.8447 | 7350 | 0.0638 | - | - |
504
+ | 0.8505 | 7400 | 0.0641 | - | - |
505
+ | 0.8562 | 7450 | 0.0660 | - | - |
506
+ | 0.8620 | 7500 | 0.0687 | 0.0618 | 0.9490 |
507
+ | 0.8677 | 7550 | 0.0654 | - | - |
508
+ | 0.8735 | 7600 | 0.0633 | - | - |
509
+ | 0.8792 | 7650 | 0.0660 | - | - |
510
+ | 0.8850 | 7700 | 0.0674 | - | - |
511
+ | 0.8907 | 7750 | 0.0681 | - | - |
512
+ | 0.8964 | 7800 | 0.0601 | - | - |
513
+ | 0.9022 | 7850 | 0.0612 | - | - |
514
+ | 0.9079 | 7900 | 0.0626 | - | - |
515
+ | 0.9137 | 7950 | 0.0641 | - | - |
516
+ | 0.9194 | 8000 | 0.0633 | 0.0619 | 0.9470 |
517
+ | 0.9252 | 8050 | 0.0637 | - | - |
518
+ | 0.9309 | 8100 | 0.0630 | - | - |
519
+ | 0.9367 | 8150 | 0.0646 | - | - |
520
+ | 0.9424 | 8200 | 0.0648 | - | - |
521
+ | 0.9482 | 8250 | 0.0647 | - | - |
522
+ | 0.9539 | 8300 | 0.0601 | - | - |
523
+ | 0.9597 | 8350 | 0.0600 | - | - |
524
+ | 0.9654 | 8400 | 0.0668 | - | - |
525
+ | 0.9712 | 8450 | 0.0640 | - | - |
526
+ | 0.9769 | 8500 | 0.0579 | 0.0618 | 0.9500 |
527
+ | 0.9826 | 8550 | 0.0645 | - | - |
528
+ | 0.9884 | 8600 | 0.0614 | - | - |
529
+ | 0.9941 | 8650 | 0.0642 | - | - |
530
+ | 0.9999 | 8700 | 0.0652 | - | - |
531
+
532
+ * The bold row denotes the saved checkpoint.
533
+ </details>
534
+
535
+ ### Framework Versions
536
+ - Python: 3.12.12
537
+ - Sentence Transformers: 5.2.2
538
+ - Transformers: 5.0.0
539
+ - PyTorch: 2.9.0+cu128
540
+ - Accelerate: 1.12.0
541
+ - Datasets: 4.0.0
542
+ - Tokenizers: 0.22.2
543
+
544
+ ## Citation
545
+
546
+ ### BibTeX
547
+
548
+ #### Sentence Transformers
549
+ ```bibtex
550
+ @inproceedings{reimers-2019-sentence-bert,
551
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
552
+ author = "Reimers, Nils and Gurevych, Iryna",
553
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
554
+ month = "11",
555
+ year = "2019",
556
+ publisher = "Association for Computational Linguistics",
557
+ url = "https://arxiv.org/abs/1908.10084",
558
+ }
559
+ ```
560
+
561
+ <!--
562
+ ## Glossary
563
+
564
+ *Clearly define terms in order to be accessible across audiences.*
565
+ -->
566
+
567
+ <!--
568
+ ## Model Card Authors
569
+
570
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
571
+ -->
572
+
573
+ <!--
574
+ ## Model Card Contact
575
+
576
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
577
+ -->
config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "dtype": "float32",
10
+ "eos_token_id": null,
11
+ "gradient_checkpointing": false,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 384,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 1536,
17
+ "is_decoder": false,
18
+ "layer_norm_eps": 1e-12,
19
+ "max_position_embeddings": 512,
20
+ "model_type": "bert",
21
+ "num_attention_heads": 12,
22
+ "num_hidden_layers": 6,
23
+ "pad_token_id": 0,
24
+ "position_embedding_type": "absolute",
25
+ "tie_word_embeddings": true,
26
+ "transformers_version": "5.0.0",
27
+ "type_vocab_size": 2,
28
+ "use_cache": true,
29
+ "vocab_size": 30522
30
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "5.2.2",
4
+ "transformers": "5.0.0",
5
+ "pytorch": "2.9.0+cu128"
6
+ },
7
+ "model_type": "SentenceTransformer",
8
+ "prompts": {
9
+ "query": "",
10
+ "document": ""
11
+ },
12
+ "default_prompt_name": null,
13
+ "similarity_fn_name": "cosine"
14
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01142cda08807a89c1ef4f32f31a2d6eb707e05c2ea160ebc9402023b54a16db
3
+ size 90864176
modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_Normalize",
18
+ "type": "sentence_transformers.models.Normalize"
19
+ }
20
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 256,
3
+ "do_lower_case": false
4
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "cls_token": "[CLS]",
4
+ "do_basic_tokenize": true,
5
+ "do_lower_case": true,
6
+ "is_local": false,
7
+ "mask_token": "[MASK]",
8
+ "model_max_length": 256,
9
+ "never_split": null,
10
+ "pad_token": "[PAD]",
11
+ "sep_token": "[SEP]",
12
+ "strip_accents": null,
13
+ "tokenize_chinese_chars": true,
14
+ "tokenizer_class": "BertTokenizer",
15
+ "unk_token": "[UNK]"
16
+ }