Hugging Face's logo

imageomics
/
trait2vec

Sentence Similarity
sentence-transformers
Safetensors
English
mpnet
ontology
nlp
biology
animals
fish
embedding
trait
feature-extraction
loss:CoSENTLoss
Eval Results
Model card Files
xet
Community
2
jjgarciac commited on
Commit
4fd4743
·
verified ·
1 Parent(s): 823c96a

Initial Trait2Vec model trained with 80% of the Phenoscape trait pairs.

Browse files
Files changed (13) hide show
  1. 1_Pooling/config.json +10 -0
  2. 2_Dense/config.json +6 -0
  3. 2_Dense/model.safetensors +3 -0
  4. README.md +578 -3
  5. config.json +23 -0
  6. config_sentence_transformers.json +14 -0
  7. model.safetensors +3 -0
  8. modules.json +20 -0
  9. sentence_bert_config.json +4 -0
  10. special_tokens_map.json +51 -0
  11. tokenizer.json +0 -0
  12. tokenizer_config.json +73 -0
  13. vocab.txt +0 -0
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 768,
3
+ "pooling_mode_cls_token": true,
4
+ "pooling_mode_mean_tokens": false,
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
+ }
2_Dense/config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "in_features": 768,
3
+ "out_features": 256,
4
+ "bias": true,
5
+ "activation_function": "torch.nn.modules.activation.Tanh"
6
+ }
2_Dense/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a633e13bd53c1896b2b55a07023ecfbde01591e22be6c38e8602e931e9190853
3
+ size 787616
README.md CHANGED
@@ -1,3 +1,578 @@
1
- ---
2
- license: mit
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - sentence-similarity
5
+ - feature-extraction
6
+ - generated_from_trainer
7
+ - dataset_size:438516
8
+ - loss:CoSENTLoss
9
+ base_model: sentence-transformers/all-mpnet-base-v2
10
+ widget:
11
+ - source_sentence: 'Ventral humeral ridge: or not'
12
+ sentences:
13
+ - 'If metasternum ossified, shape: long, narrow and tapering markedly anteriorly
14
+ to posteriorly, length up to 3.5 times maximum width'
15
+ - 'Astragalus, dorsolateral margin:: overlaps the anterior and posterior portions
16
+ of the calcaneum equally'
17
+ - 'Ulna size: does not apply'
18
+ - source_sentence: 'Form of distal portion of anteroventral process of ectopterygoid:
19
+ varyingly falcate'
20
+ sentences:
21
+ - 'Middle and distal radials in dorsal and anal fins: absent'
22
+ - 'Degree of development of primitively medial portion of fourth upper pharyngeal
23
+ tooth-plate: fourth upper pharyngeal tooth-plate covers ventral, posterior, dorsal
24
+ and sometimes anterior surfaces of fourth infrapharyngobranchial'
25
+ - 'Shape of pharyngeal apophysis (basioccipital): forked anteriorly'
26
+ - source_sentence: 'Form of distal portion of anteroventral process of ectopterygoid:
27
+ varyingly falcate'
28
+ sentences:
29
+ - 'parhypural: present'
30
+ - 'Epural: heavy'
31
+ - 'First infraorbital: short'
32
+ - source_sentence: 'Form of distal portion of anteroventral process of ectopterygoid:
33
+ varyingly falcate'
34
+ sentences:
35
+ - 'Dentary and angular: touch'
36
+ - 'Urohyal and first basibranchial: firmly attached'
37
+ - 'Supraneural 3-4 (nonadditive): absent'
38
+ - source_sentence: 'Form of distal portion of anteroventral process of ectopterygoid:
39
+ varyingly falcate'
40
+ sentences:
41
+ - 'Ventral diverging lamellae of mesethmoid: lamellae reduced or absent'
42
+ - 'Ventral ridge of the coracoid with a posterior process: absent'
43
+ - 'carpals: fully or partially ossified'
44
+ pipeline_tag: sentence-similarity
45
+ library_name: sentence-transformers
46
+ metrics:
47
+ - pearson_cosine
48
+ - spearman_cosine
49
+ model-index:
50
+ - name: SentenceTransformer based on sentence-transformers/all-mpnet-base-v2
51
+ results:
52
+ - task:
53
+ type: semantic-similarity
54
+ name: Semantic Similarity
55
+ dataset:
56
+ name: pheno dev
57
+ type: pheno-dev
58
+ metrics:
59
+ - type: pearson_cosine
60
+ value: 0.6082332469417436
61
+ name: Pearson Cosine
62
+ - type: spearman_cosine
63
+ value: 0.6250387873495056
64
+ name: Spearman Cosine
65
+ - task:
66
+ type: semantic-similarity
67
+ name: Semantic Similarity
68
+ dataset:
69
+ name: pheno test
70
+ type: pheno-test
71
+ metrics:
72
+ - type: pearson_cosine
73
+ value: 0.6822053314599665
74
+ name: Pearson Cosine
75
+ - type: spearman_cosine
76
+ value: 0.705688010939619
77
+ name: Spearman Cosine
78
+ ---
79
+
80
+ # SentenceTransformer based on sentence-transformers/all-mpnet-base-v2
81
+
82
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/all-mpnet-base-v2](https://huggingface.co/sentence-transformers/all-mpnet-base-v2). It maps sentences & paragraphs to a 256-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
83
+
84
+ ## Model Details
85
+
86
+ ### Model Description
87
+ - **Model Type:** Sentence Transformer
88
+ - **Base model:** [sentence-transformers/all-mpnet-base-v2](https://huggingface.co/sentence-transformers/all-mpnet-base-v2) <!-- at revision 9a3225965996d404b775526de6dbfe85d3368642 -->
89
+ - **Maximum Sequence Length:** 256 tokens
90
+ - **Output Dimensionality:** 256 dimensions
91
+ - **Similarity Function:** Cosine Similarity
92
+ <!-- - **Training Dataset:** Unknown -->
93
+ <!-- - **Language:** Unknown -->
94
+ <!-- - **License:** Unknown -->
95
+
96
+ ### Model Sources
97
+
98
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
99
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
100
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
101
+
102
+ ### Full Model Architecture
103
+
104
+ ```
105
+ SentenceTransformer(
106
+ (0): Transformer({'max_seq_length': 256, 'do_lower_case': False}) with Transformer model: MPNetModel
107
+ (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
108
+ (2): Dense({'in_features': 768, 'out_features': 256, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
109
+ )
110
+ ```
111
+
112
+ ## Usage
113
+
114
+ ### Direct Usage (Sentence Transformers)
115
+
116
+ First install the Sentence Transformers library:
117
+
118
+ ```bash
119
+ pip install -U sentence-transformers
120
+ ```
121
+
122
+ Then you can load this model and run inference.
123
+ ```python
124
+ from sentence_transformers import SentenceTransformer
125
+
126
+ # Download from the 🤗 Hub
127
+ model = SentenceTransformer("imageomics/trait2vec")
128
+ # Run inference
129
+ sentences = [
130
+ 'Form of distal portion of anteroventral process of ectopterygoid: varyingly falcate',
131
+ 'Ventral ridge of the coracoid with a posterior process: absent',
132
+ 'carpals: fully or partially ossified',
133
+ ]
134
+ embeddings = model.encode(sentences)
135
+ print(embeddings.shape)
136
+ # [3, 256]
137
+
138
+ # Get the similarity scores for the embeddings
139
+ similarities = model.similarity(embeddings, embeddings)
140
+ print(similarities.shape)
141
+ # [3, 3]
142
+ ```
143
+
144
+ <!--
145
+ ### Direct Usage (Transformers)
146
+
147
+ <details><summary>Click to see the direct usage in Transformers</summary>
148
+
149
+ </details>
150
+ -->
151
+
152
+ <!--
153
+ ### Downstream Usage (Sentence Transformers)
154
+
155
+ You can finetune this model on your own dataset.
156
+
157
+ <details><summary>Click to expand</summary>
158
+
159
+ </details>
160
+ -->
161
+
162
+ <!--
163
+ ### Out-of-Scope Use
164
+
165
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
166
+ -->
167
+
168
+ ## Evaluation
169
+
170
+ ### Metrics
171
+
172
+ #### Semantic Similarity
173
+
174
+ * Datasets: `pheno-dev` and `pheno-test`
175
+ * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
176
+
177
+ | Metric | pheno-dev | pheno-test |
178
+ |:--------------------|:----------|:-----------|
179
+ | pearson_cosine | 0.6082 | 0.6822 |
180
+ | **spearman_cosine** | **0.625** | **0.7057** |
181
+
182
+ <!--
183
+ ## Bias, Risks and Limitations
184
+
185
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
186
+ -->
187
+
188
+ <!--
189
+ ### Recommendations
190
+
191
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
192
+ -->
193
+
194
+ ## Training Details
195
+
196
+ ### Training Dataset
197
+
198
+ #### Unnamed Dataset
199
+
200
+
201
+ * Size: 438,516 training samples
202
+ * Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
203
+ * Approximate statistics based on the first 1000 samples:
204
+ | | sentence1 | sentence2 | score |
205
+ |:--------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:---------------------------------------------------------------|
206
+ | type | string | string | float |
207
+ | details | <ul><li>min: 9 tokens</li><li>mean: 42.84 tokens</li><li>max: 164 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 22.8 tokens</li><li>max: 164 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 0.1</li><li>max: 0.61</li></ul> |
208
+ * Samples:
209
+ | sentence1 | sentence2 | score |
210
+ |:-----------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------|
211
+ | <code>Gill raker shape between ceratobranchial 1 and ceratobranchials 2--4: Homomorphic</code> | <code>Extent of development of inferior lamella of lateral ethmoid: inferior lamella absent</code> | <code>0.014706667500582846</code> |
212
+ | <code>Gill raker shape between ceratobranchial 1 and ceratobranchials 2--4: Homomorphic</code> | <code>Shape of anal-fin pterygiophore tips: tips of pterygiophores shaped like an arrow-head; axial series of pterygiophores providing the ventral margin of the anal-fin base a scalloped appearance</code> | <code>0.030538703023734296</code> |
213
+ | <code>Gill raker shape between ceratobranchial 1 and ceratobranchials 2--4: Homomorphic</code> | <code>Suprapreopercle: present</code> | <code>0.3385057414877959</code> |
214
+ * Loss: [<code>CoSENTLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
215
+ ```json
216
+ {
217
+ "scale": 20.0,
218
+ "similarity_fct": "pairwise_cos_sim"
219
+ }
220
+ ```
221
+
222
+ ### Evaluation Dataset
223
+
224
+ #### Unnamed Dataset
225
+
226
+
227
+ * Size: 111,628 evaluation samples
228
+ * Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>score</code>
229
+ * Approximate statistics based on the first 1000 samples:
230
+ | | sentence1 | sentence2 | score |
231
+ |:--------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:---------------------------------------------------------------|
232
+ | type | string | string | float |
233
+ | details | <ul><li>min: 9 tokens</li><li>mean: 17.19 tokens</li><li>max: 24 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 21.97 tokens</li><li>max: 143 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 0.1</li><li>max: 0.86</li></ul> |
234
+ * Samples:
235
+ | sentence1 | sentence2 | score |
236
+ |:-------------------------------------------|:------------------------------------------------------------------------------------------------------------------|:----------------------------------|
237
+ | <code>Ventral humeral ridge: or not</code> | <code>Metacarpals, Metacarpal I, presence: absent</code> | <code>0.05558851078197206</code> |
238
+ | <code>Ventral humeral ridge: or not</code> | <code>Metapterygoid–quadrate fenestra: absent</code> | <code>0.004860625129173212</code> |
239
+ | <code>Ventral humeral ridge: or not</code> | <code>Dorsal and ventral borders of the maxillary articular process: straight or slightly curved ventrally</code> | <code>0.10380567059620477</code> |
240
+ * Loss: [<code>CoSENTLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cosentloss) with these parameters:
241
+ ```json
242
+ {
243
+ "scale": 20.0,
244
+ "similarity_fct": "pairwise_cos_sim"
245
+ }
246
+ ```
247
+
248
+ ### Training Hyperparameters
249
+ #### Non-Default Hyperparameters
250
+
251
+ - `eval_strategy`: steps
252
+ - `per_device_train_batch_size`: 64
253
+ - `per_device_eval_batch_size`: 64
254
+ - `learning_rate`: 2e-05
255
+ - `num_train_epochs`: 10
256
+ - `warmup_ratio`: 1e-06
257
+
258
+ #### All Hyperparameters
259
+ <details><summary>Click to expand</summary>
260
+
261
+ - `overwrite_output_dir`: False
262
+ - `do_predict`: False
263
+ - `eval_strategy`: steps
264
+ - `prediction_loss_only`: True
265
+ - `per_device_train_batch_size`: 64
266
+ - `per_device_eval_batch_size`: 64
267
+ - `per_gpu_train_batch_size`: None
268
+ - `per_gpu_eval_batch_size`: None
269
+ - `gradient_accumulation_steps`: 1
270
+ - `eval_accumulation_steps`: None
271
+ - `torch_empty_cache_steps`: None
272
+ - `learning_rate`: 2e-05
273
+ - `weight_decay`: 0.0
274
+ - `adam_beta1`: 0.9
275
+ - `adam_beta2`: 0.999
276
+ - `adam_epsilon`: 1e-08
277
+ - `max_grad_norm`: 1.0
278
+ - `num_train_epochs`: 10
279
+ - `max_steps`: -1
280
+ - `lr_scheduler_type`: linear
281
+ - `lr_scheduler_kwargs`: {}
282
+ - `warmup_ratio`: 1e-06
283
+ - `warmup_steps`: 0
284
+ - `log_level`: passive
285
+ - `log_level_replica`: warning
286
+ - `log_on_each_node`: True
287
+ - `logging_nan_inf_filter`: True
288
+ - `save_safetensors`: True
289
+ - `save_on_each_node`: False
290
+ - `save_only_model`: False
291
+ - `restore_callback_states_from_checkpoint`: False
292
+ - `no_cuda`: False
293
+ - `use_cpu`: False
294
+ - `use_mps_device`: False
295
+ - `seed`: 42
296
+ - `data_seed`: None
297
+ - `jit_mode_eval`: False
298
+ - `use_ipex`: False
299
+ - `bf16`: False
300
+ - `fp16`: False
301
+ - `fp16_opt_level`: O1
302
+ - `half_precision_backend`: auto
303
+ - `bf16_full_eval`: False
304
+ - `fp16_full_eval`: False
305
+ - `tf32`: None
306
+ - `local_rank`: 0
307
+ - `ddp_backend`: None
308
+ - `tpu_num_cores`: None
309
+ - `tpu_metrics_debug`: False
310
+ - `debug`: []
311
+ - `dataloader_drop_last`: False
312
+ - `dataloader_num_workers`: 0
313
+ - `dataloader_prefetch_factor`: None
314
+ - `past_index`: -1
315
+ - `disable_tqdm`: False
316
+ - `remove_unused_columns`: True
317
+ - `label_names`: None
318
+ - `load_best_model_at_end`: False
319
+ - `ignore_data_skip`: False
320
+ - `fsdp`: []
321
+ - `fsdp_min_num_params`: 0
322
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
323
+ - `fsdp_transformer_layer_cls_to_wrap`: None
324
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
325
+ - `deepspeed`: None
326
+ - `label_smoothing_factor`: 0.0
327
+ - `optim`: adamw_torch
328
+ - `optim_args`: None
329
+ - `adafactor`: False
330
+ - `group_by_length`: False
331
+ - `length_column_name`: length
332
+ - `ddp_find_unused_parameters`: None
333
+ - `ddp_bucket_cap_mb`: None
334
+ - `ddp_broadcast_buffers`: False
335
+ - `dataloader_pin_memory`: True
336
+ - `dataloader_persistent_workers`: False
337
+ - `skip_memory_metrics`: True
338
+ - `use_legacy_prediction_loop`: False
339
+ - `push_to_hub`: False
340
+ - `resume_from_checkpoint`: None
341
+ - `hub_model_id`: None
342
+ - `hub_strategy`: every_save
343
+ - `hub_private_repo`: None
344
+ - `hub_always_push`: False
345
+ - `gradient_checkpointing`: False
346
+ - `gradient_checkpointing_kwargs`: None
347
+ - `include_inputs_for_metrics`: False
348
+ - `include_for_metrics`: []
349
+ - `eval_do_concat_batches`: True
350
+ - `fp16_backend`: auto
351
+ - `push_to_hub_model_id`: None
352
+ - `push_to_hub_organization`: None
353
+ - `mp_parameters`:
354
+ - `auto_find_batch_size`: False
355
+ - `full_determinism`: False
356
+ - `torchdynamo`: None
357
+ - `ray_scope`: last
358
+ - `ddp_timeout`: 1800
359
+ - `torch_compile`: False
360
+ - `torch_compile_backend`: None
361
+ - `torch_compile_mode`: None
362
+ - `dispatch_batches`: None
363
+ - `split_batches`: None
364
+ - `include_tokens_per_second`: False
365
+ - `include_num_input_tokens_seen`: False
366
+ - `neftune_noise_alpha`: None
367
+ - `optim_target_modules`: None
368
+ - `batch_eval_metrics`: False
369
+ - `eval_on_start`: False
370
+ - `use_liger_kernel`: False
371
+ - `eval_use_gather_object`: False
372
+ - `average_tokens_across_devices`: False
373
+ - `prompts`: None
374
+ - `batch_sampler`: batch_sampler
375
+ - `multi_dataset_batch_sampler`: proportional
376
+
377
+ </details>
378
+
379
+ ### Training Logs
380
+ <details><summary>Click to expand</summary>
381
+
382
+ | Epoch | Step | Training Loss | Validation Loss | pheno-dev_spearman_cosine | pheno-test_spearman_cosine |
383
+ |:------:|:-----:|:-------------:|:---------------:|:-------------------------:|:--------------------------:|
384
+ | 0.0730 | 500 | 7.3492 | - | - | - |
385
+ | 0.1459 | 1000 | 6.9718 | - | - | - |
386
+ | 0.2189 | 1500 | 6.7986 | - | - | - |
387
+ | 0.2919 | 2000 | 6.7157 | 8.8773 | 0.6305 | - |
388
+ | 0.3649 | 2500 | 6.6327 | - | - | - |
389
+ | 0.4378 | 3000 | 6.5661 | - | - | - |
390
+ | 0.5108 | 3500 | 6.5309 | - | - | - |
391
+ | 0.5838 | 4000 | 6.4737 | 10.0841 | 0.6116 | - |
392
+ | 0.6567 | 4500 | 6.4516 | - | - | - |
393
+ | 0.7297 | 5000 | 6.4235 | - | - | - |
394
+ | 0.8027 | 5500 | 6.3908 | - | - | - |
395
+ | 0.8757 | 6000 | 6.3602 | 10.8098 | 0.6071 | - |
396
+ | 0.9486 | 6500 | 6.3315 | - | - | - |
397
+ | 1.0216 | 7000 | 6.3236 | - | - | - |
398
+ | 1.0946 | 7500 | 6.2753 | - | - | - |
399
+ | 1.1675 | 8000 | 6.2845 | 11.9185 | 0.6263 | - |
400
+ | 1.2405 | 8500 | 6.254 | - | - | - |
401
+ | 1.3135 | 9000 | 6.2351 | - | - | - |
402
+ | 1.3865 | 9500 | 6.2017 | - | - | - |
403
+ | 1.4594 | 10000 | 6.2138 | 12.3766 | 0.6161 | - |
404
+ | 1.5324 | 10500 | 6.2066 | - | - | - |
405
+ | 1.6054 | 11000 | 6.1834 | - | - | - |
406
+ | 1.6783 | 11500 | 6.1937 | - | - | - |
407
+ | 1.7513 | 12000 | 6.1661 | 12.9426 | 0.6113 | - |
408
+ | 1.8243 | 12500 | 6.1362 | - | - | - |
409
+ | 1.8973 | 13000 | 6.1065 | - | - | - |
410
+ | 1.9702 | 13500 | 6.1371 | - | - | - |
411
+ | 2.0432 | 14000 | 6.0983 | 13.5966 | 0.6156 | - |
412
+ | 2.1162 | 14500 | 6.0978 | - | - | - |
413
+ | 2.1891 | 15000 | 6.0767 | - | - | - |
414
+ | 2.2621 | 15500 | 6.066 | - | - | - |
415
+ | 2.3351 | 16000 | 6.0739 | 13.9316 | 0.6260 | - |
416
+ | 2.4081 | 16500 | 6.0635 | - | - | - |
417
+ | 2.4810 | 17000 | 6.0616 | - | - | - |
418
+ | 2.5540 | 17500 | 6.0219 | - | - | - |
419
+ | 2.6270 | 18000 | 6.0129 | 14.3098 | 0.6158 | - |
420
+ | 2.6999 | 18500 | 6.0414 | - | - | - |
421
+ | 2.7729 | 19000 | 6.0317 | - | - | - |
422
+ | 2.8459 | 19500 | 6.0158 | - | - | - |
423
+ | 2.9189 | 20000 | 6.0078 | 14.6487 | 0.6188 | - |
424
+ | 2.9918 | 20500 | 6.0295 | - | - | - |
425
+ | 3.0648 | 21000 | 5.9664 | - | - | - |
426
+ | 3.1378 | 21500 | 5.9682 | - | - | - |
427
+ | 3.2107 | 22000 | 5.9755 | 15.2314 | 0.6202 | - |
428
+ | 3.2837 | 22500 | 5.9608 | - | - | - |
429
+ | 3.3567 | 23000 | 5.9469 | - | - | - |
430
+ | 3.4297 | 23500 | 5.9673 | - | - | - |
431
+ | 3.5026 | 24000 | 5.9496 | 15.4385 | 0.6237 | - |
432
+ | 3.5756 | 24500 | 5.9148 | - | - | - |
433
+ | 3.6486 | 25000 | 5.9568 | - | - | - |
434
+ | 3.7215 | 25500 | 5.9135 | - | - | - |
435
+ | 3.7945 | 26000 | 5.9363 | 15.3029 | 0.6217 | - |
436
+ | 3.8675 | 26500 | 5.9096 | - | - | - |
437
+ | 3.9405 | 27000 | 5.9171 | - | - | - |
438
+ | 4.0134 | 27500 | 5.8955 | - | - | - |
439
+ | 4.0864 | 28000 | 5.861 | 15.3221 | 0.6265 | - |
440
+ | 4.1594 | 28500 | 5.8726 | - | - | - |
441
+ | 4.2323 | 29000 | 5.8835 | - | - | - |
442
+ | 4.3053 | 29500 | 5.8823 | - | - | - |
443
+ | 4.3783 | 30000 | 5.8702 | 15.7276 | 0.6266 | - |
444
+ | 4.4513 | 30500 | 5.8721 | - | - | - |
445
+ | 4.5242 | 31000 | 5.8988 | - | - | - |
446
+ | 4.5972 | 31500 | 5.8671 | - | - | - |
447
+ | 4.6702 | 32000 | 5.8705 | 15.9223 | 0.6212 | - |
448
+ | 4.7431 | 32500 | 5.8905 | - | - | - |
449
+ | 4.8161 | 33000 | 5.8634 | - | - | - |
450
+ | 4.8891 | 33500 | 5.8637 | - | - | - |
451
+ | 4.9621 | 34000 | 5.8385 | 16.1225 | 0.6045 | - |
452
+ | 5.0350 | 34500 | 5.8583 | - | - | - |
453
+ | 5.1080 | 35000 | 5.821 | - | - | - |
454
+ | 5.1810 | 35500 | 5.8219 | - | - | - |
455
+ | 5.2539 | 36000 | 5.8367 | 15.6937 | 0.6240 | - |
456
+ | 5.3269 | 36500 | 5.8245 | - | - | - |
457
+ | 5.3999 | 37000 | 5.8161 | - | - | - |
458
+ | 5.4729 | 37500 | 5.8138 | - | - | - |
459
+ | 5.5458 | 38000 | 5.815 | 15.7507 | 0.6279 | - |
460
+ | 5.6188 | 38500 | 5.8238 | - | - | - |
461
+ | 5.6918 | 39000 | 5.8235 | - | - | - |
462
+ | 5.7647 | 39500 | 5.8407 | - | - | - |
463
+ | 5.8377 | 40000 | 5.8258 | 15.8875 | 0.6213 | - |
464
+ | 5.9107 | 40500 | 5.7941 | - | - | - |
465
+ | 5.9837 | 41000 | 5.8301 | - | - | - |
466
+ | 6.0566 | 41500 | 5.7734 | - | - | - |
467
+ | 6.1296 | 42000 | 5.7759 | 16.0155 | 0.6212 | - |
468
+ | 6.2026 | 42500 | 5.7951 | - | - | - |
469
+ | 6.2755 | 43000 | 5.8023 | - | - | - |
470
+ | 6.3485 | 43500 | 5.7848 | - | - | - |
471
+ | 6.4215 | 44000 | 5.7774 | 16.0796 | 0.6152 | - |
472
+ | 6.4945 | 44500 | 5.7719 | - | - | - |
473
+ | 6.5674 | 45000 | 5.7822 | - | - | - |
474
+ | 6.6404 | 45500 | 5.7734 | - | - | - |
475
+ | 6.7134 | 46000 | 5.7856 | 16.2461 | 0.6142 | - |
476
+ | 6.7863 | 46500 | 5.7949 | - | - | - |
477
+ | 6.8593 | 47000 | 5.8346 | - | - | - |
478
+ | 6.9323 | 47500 | 5.7606 | - | - | - |
479
+ | 7.0053 | 48000 | 5.7839 | 16.0556 | 0.6249 | - |
480
+ | 7.0782 | 48500 | 5.7581 | - | - | - |
481
+ | 7.1512 | 49000 | 5.7472 | - | - | - |
482
+ | 7.2242 | 49500 | 5.7443 | - | - | - |
483
+ | 7.2971 | 50000 | 5.7481 | 16.1126 | 0.6248 | - |
484
+ | 7.3701 | 50500 | 5.7487 | - | - | - |
485
+ | 7.4431 | 51000 | 5.7443 | - | - | - |
486
+ | 7.5161 | 51500 | 5.76 | - | - | - |
487
+ | 7.5890 | 52000 | 5.7353 | 16.0932 | 0.6312 | - |
488
+ | 7.6620 | 52500 | 5.7632 | - | - | - |
489
+ | 7.7350 | 53000 | 5.7788 | - | - | - |
490
+ | 7.8079 | 53500 | 5.758 | - | - | - |
491
+ | 7.8809 | 54000 | 5.7324 | 16.1470 | 0.6247 | - |
492
+ | 7.9539 | 54500 | 5.7425 | - | - | - |
493
+ | 8.0269 | 55000 | 5.7416 | - | - | - |
494
+ | 8.0998 | 55500 | 5.7696 | - | - | - |
495
+ | 8.1728 | 56000 | 5.7493 | 16.2547 | 0.6313 | - |
496
+ | 8.2458 | 56500 | 5.7348 | - | - | - |
497
+ | 8.3187 | 57000 | 5.7173 | - | - | - |
498
+ | 8.3917 | 57500 | 5.7215 | - | - | - |
499
+ | 8.4647 | 58000 | 5.7163 | 16.3313 | 0.6237 | - |
500
+ | 8.5377 | 58500 | 5.722 | - | - | - |
501
+ | 8.6106 | 59000 | 5.7292 | - | - | - |
502
+ | 8.6836 | 59500 | 5.7295 | - | - | - |
503
+ | 8.7566 | 60000 | 5.7267 | 16.3434 | 0.6261 | - |
504
+ | 8.8295 | 60500 | 5.7207 | - | - | - |
505
+ | 8.9025 | 61000 | 5.7252 | - | - | - |
506
+ | 8.9755 | 61500 | 5.7061 | - | - | - |
507
+ | 9.0485 | 62000 | 5.7113 | 16.2999 | 0.6279 | - |
508
+ | 9.1214 | 62500 | 5.695 | - | - | - |
509
+ | 9.1944 | 63000 | 5.7152 | - | - | - |
510
+ | 9.2674 | 63500 | 5.7045 | - | - | - |
511
+ | 9.3403 | 64000 | 5.6907 | 16.2782 | 0.6264 | - |
512
+ | 9.4133 | 64500 | 5.7185 | - | - | - |
513
+ | 9.4863 | 65000 | 5.6903 | - | - | - |
514
+ | 9.5593 | 65500 | 5.705 | - | - | - |
515
+ | 9.6322 | 66000 | 5.7165 | 16.3625 | 0.6249 | - |
516
+ | 9.7052 | 66500 | 5.7027 | - | - | - |
517
+ | 9.7782 | 67000 | 5.7048 | - | - | - |
518
+ | 9.8511 | 67500 | 5.728 | - | - | - |
519
+ | 9.9241 | 68000 | 5.7111 | 16.3087 | 0.6250 | - |
520
+ | 9.9971 | 68500 | 5.7144 | - | - | - |
521
+ | 10.0 | 68520 | - | - | - | 0.7057 |
522
+
523
+ </details>
524
+
525
+ ### Framework Versions
526
+ - Python: 3.10.16
527
+ - Sentence Transformers: 3.3.1
528
+ - Transformers: 4.48.1
529
+ - PyTorch: 2.5.1.post303
530
+ - Accelerate: 1.3.0
531
+ - Datasets: 2.14.4
532
+ - Tokenizers: 0.21.0
533
+
534
+ ## Citation
535
+
536
+ ### BibTeX
537
+
538
+ #### Sentence Transformers
539
+ ```bibtex
540
+ @inproceedings{reimers-2019-sentence-bert,
541
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
542
+ author = "Reimers, Nils and Gurevych, Iryna",
543
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
544
+ month = "11",
545
+ year = "2019",
546
+ publisher = "Association for Computational Linguistics",
547
+ url = "https://arxiv.org/abs/1908.10084",
548
+ }
549
+ ```
550
+
551
+ #### CoSENTLoss
552
+ ```bibtex
553
+ @online{kexuefm-8847,
554
+ title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
555
+ author={Su Jianlin},
556
+ year={2022},
557
+ month={Jan},
558
+ url={https://kexue.fm/archives/8847},
559
+ }
560
+ ```
561
+
562
+ <!--
563
+ ## Glossary
564
+
565
+ *Clearly define terms in order to be accessible across audiences.*
566
+ -->
567
+
568
+ <!--
569
+ ## Model Card Authors
570
+
571
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
572
+ -->
573
+
574
+ <!--
575
+ ## Model Card Contact
576
+
577
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
578
+ -->
config.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "MPNetModel"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "dtype": "float32",
8
+ "eos_token_id": 2,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-05,
15
+ "max_position_embeddings": 514,
16
+ "model_type": "mpnet",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "pad_token_id": 1,
20
+ "relative_attention_num_buckets": 32,
21
+ "transformers_version": "4.57.1",
22
+ "vocab_size": 30527
23
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "5.1.1",
4
+ "transformers": "4.57.1",
5
+ "pytorch": "2.8.0+cu126"
6
+ },
7
+ "prompts": {
8
+ "query": "",
9
+ "document": ""
10
+ },
11
+ "default_prompt_name": null,
12
+ "similarity_fn_name": "cosine",
13
+ "model_type": "SentenceTransformer"
14
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86fbf085e0ca76d528543fa7f538b718833b7138671186746ccaf6c5f20d5b73
3
+ size 437967672
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_Dense",
18
+ "type": "sentence_transformers.models.Dense"
19
+ }
20
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 256,
3
+ "do_lower_case": false
4
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "[UNK]",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,73 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "104": {
36
+ "content": "[UNK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ },
43
+ "30526": {
44
+ "content": "<mask>",
45
+ "lstrip": true,
46
+ "normalized": false,
47
+ "rstrip": false,
48
+ "single_word": false,
49
+ "special": true
50
+ }
51
+ },
52
+ "bos_token": "<s>",
53
+ "clean_up_tokenization_spaces": false,
54
+ "cls_token": "<s>",
55
+ "do_lower_case": true,
56
+ "eos_token": "</s>",
57
+ "extra_special_tokens": {},
58
+ "mask_token": "<mask>",
59
+ "max_length": 128,
60
+ "model_max_length": 256,
61
+ "pad_to_multiple_of": null,
62
+ "pad_token": "<pad>",
63
+ "pad_token_type_id": 0,
64
+ "padding_side": "right",
65
+ "sep_token": "</s>",
66
+ "stride": 0,
67
+ "strip_accents": null,
68
+ "tokenize_chinese_chars": true,
69
+ "tokenizer_class": "MPNetTokenizer",
70
+ "truncation_side": "right",
71
+ "truncation_strategy": "longest_first",
72
+ "unk_token": "[UNK]"
73
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff