Hugging Face's logo

Bheri
/
ithasa-mmbert-v3

Sentence Similarity
sentence-transformers
Safetensors
modernbert
feature-extraction
dense
Generated from Trainer
dataset_size:3749530
loss:MultipleNegativesRankingLoss
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
 Community
saikasyap commited on
Commit
ca55df8
·
verified ·
1 Parent(s): d3aa9ec

Initial commit

Browse files
Files changed (12) hide show
  1. .gitattributes +1 -0
  2. 1_Pooling/config.json +10 -0
  3. README.md +534 -0
  4. config.json +78 -0
  5. config_sentence_transformers.json +14 -0
  6. eval/translation_evaluation_eval-en-sa_results.csv +13 -0
  7. model.safetensors +3 -0
  8. modules.json +14 -0
  9. sentence_bert_config.json +4 -0
  10. tokenizer.json +3 -0
  11. tokenizer_config.json +24 -0
  12. training_args.bin +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 768,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
README.md ADDED
@@ -0,0 +1,534 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - sentence-similarity
5
+ - feature-extraction
6
+ - dense
7
+ - generated_from_trainer
8
+ - dataset_size:3749530
9
+ - loss:MultipleNegativesRankingLoss
10
+ base_model: jhu-clsp/mmBERT-base
11
+ widget:
12
+ - source_sentence: The false promise of sovereignty.
13
+ sentences:
14
+ - २ जिगीषोः प्रयाणे च अमरः।
15
+ - सत्त्वहेतु सुदृढकृत प्रतिज्ञा।
16
+ - उम्र क्रमशः 16 वर्ष व 15 वर्ष है।
17
+ - source_sentence: not of the world, even as I am not of the world."
18
+ sentences:
19
+ - १९.८॥ जगत्सर्वं सदा नास्ति चित्तमेव जगन्मयम् ।
20
+ - ५६३६-१ आस्तां मानः कथनं सखीषु वा मयि निवेद्यदुर्विनये ।
21
+ - (द) अन्तः वैयक्तिक सम्प्रेषण
22
+ - source_sentence: The Joys of May
23
+ sentences:
24
+ - ३ यात्रार्थं यष्टि र्वस्त्रपुटकं भक्ष्यं मुद्रा द्वितीयवस्त्रम्, एषां किमपि मा
25
+ गृह्लीत।
26
+ - भाग्य मुस्कान मई
27
+ - 19 अथ वायुः समुद्भूतॊ दावाग्निर अभवन महान
28
+ - source_sentence: plagiarized - 10.
29
+ sentences:
30
+ - अक्षरविनोदः - १० ।
31
+ - कवचेना वृतो नित्यं यत्र यत्रैव गच्छति . . ४३..
32
+ - खल त ते खेलं भवनकलहं सा न जहित ।
33
+ - source_sentence: 'attenuated vaccines:'
34
+ sentences:
35
+ - तृणैः कः वर्धते ?
36
+ - ६.५% दसादशे
37
+ - कम संवेदनशील टीकेः
38
+ pipeline_tag: sentence-similarity
39
+ library_name: sentence-transformers
40
+ metrics:
41
+ - src2trg_accuracy
42
+ - trg2src_accuracy
43
+ - mean_accuracy
44
+ model-index:
45
+ - name: SentenceTransformer based on jhu-clsp/mmBERT-base
46
+ results:
47
+ - task:
48
+ type: translation
49
+ name: Translation
50
+ dataset:
51
+ name: eval en sa
52
+ type: eval-en-sa
53
+ metrics:
54
+ - type: src2trg_accuracy
55
+ value: 0.616
56
+ name: Src2Trg Accuracy
57
+ - type: trg2src_accuracy
58
+ value: 0.604
59
+ name: Trg2Src Accuracy
60
+ - type: mean_accuracy
61
+ value: 0.61
62
+ name: Mean Accuracy
63
+ ---
64
+
65
+ # SentenceTransformer based on jhu-clsp/mmBERT-base
66
+
67
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [jhu-clsp/mmBERT-base](https://huggingface.co/jhu-clsp/mmBERT-base). 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.
68
+
69
+ ## Model Details
70
+
71
+ ### Model Description
72
+ - **Model Type:** Sentence Transformer
73
+ - **Base model:** [jhu-clsp/mmBERT-base](https://huggingface.co/jhu-clsp/mmBERT-base) <!-- at revision c5955035435e2bf121cde7f3c8863ef52ff35d82 -->
74
+ - **Maximum Sequence Length:** 128 tokens
75
+ - **Output Dimensionality:** 768 dimensions
76
+ - **Similarity Function:** Cosine Similarity
77
+ <!-- - **Training Dataset:** Unknown -->
78
+ <!-- - **Language:** Unknown -->
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': 128, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
92
+ (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})
93
+ )
94
+ ```
95
+
96
+ ## Usage
97
+
98
+ ### Direct Usage (Sentence Transformers)
99
+
100
+ First install the Sentence Transformers library:
101
+
102
+ ```bash
103
+ pip install -U sentence-transformers
104
+ ```
105
+
106
+ Then you can load this model and run inference.
107
+ ```python
108
+ from sentence_transformers import SentenceTransformer
109
+
110
+ # Download from the 🤗 Hub
111
+ model = SentenceTransformer("sentence_transformers_model_id")
112
+ # Run inference
113
+ sentences = [
114
+ 'attenuated vaccines:',
115
+ 'कम संवेदनशील टीकेः',
116
+ '६.५% दसादशे',
117
+ ]
118
+ embeddings = model.encode(sentences)
119
+ print(embeddings.shape)
120
+ # [3, 768]
121
+
122
+ # Get the similarity scores for the embeddings
123
+ similarities = model.similarity(embeddings, embeddings)
124
+ print(similarities)
125
+ # tensor([[1.0000, 0.3723, 0.1543],
126
+ # [0.3723, 1.0000, 0.2746],
127
+ # [0.1543, 0.2746, 1.0000]])
128
+ ```
129
+
130
+ <!--
131
+ ### Direct Usage (Transformers)
132
+
133
+ <details><summary>Click to see the direct usage in Transformers</summary>
134
+
135
+ </details>
136
+ -->
137
+
138
+ <!--
139
+ ### Downstream Usage (Sentence Transformers)
140
+
141
+ You can finetune this model on your own dataset.
142
+
143
+ <details><summary>Click to expand</summary>
144
+
145
+ </details>
146
+ -->
147
+
148
+ <!--
149
+ ### Out-of-Scope Use
150
+
151
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
152
+ -->
153
+
154
+ ## Evaluation
155
+
156
+ ### Metrics
157
+
158
+ #### Translation
159
+
160
+ * Dataset: `eval-en-sa`
161
+ * Evaluated with [<code>TranslationEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TranslationEvaluator)
162
+
163
+ | Metric | Value |
164
+ |:------------------|:---------|
165
+ | src2trg_accuracy | 0.616 |
166
+ | trg2src_accuracy | 0.604 |
167
+ | **mean_accuracy** | **0.61** |
168
+
169
+ <!--
170
+ ## Bias, Risks and Limitations
171
+
172
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
173
+ -->
174
+
175
+ <!--
176
+ ### Recommendations
177
+
178
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
179
+ -->
180
+
181
+ ## Training Details
182
+
183
+ ### Training Dataset
184
+
185
+ #### Unnamed Dataset
186
+
187
+ * Size: 3,749,530 training samples
188
+ * Columns: <code>sentence1</code> and <code>sentence2</code>
189
+ * Approximate statistics based on the first 1000 samples:
190
+ | | sentence1 | sentence2 |
191
+ |:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|
192
+ | type | string | string |
193
+ | details | <ul><li>min: 12 tokens</li><li>mean: 31.26 tokens</li><li>max: 88 tokens</li></ul> | <ul><li>min: 19 tokens</li><li>mean: 67.93 tokens</li><li>max: 128 tokens</li></ul> |
194
+ * Samples:
195
+ | sentence1 | sentence2 |
196
+ |:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
197
+ | <code>There was no Mughal tradition of primogeniture, the systematic passing of rule, upon an emperor's death, to his eldest son.<br></code> | <code>चक्रवर्तिनः मृत्योः अनन्तरं तस्य शासनस्य व्यवस्थितरूपेण सङ्क्रमणस्य, मुघलपरम्परायाः ज्येष्ठपुत्राधिकारपद्धतिः नासीत्।<br></code> |
198
+ | <code>The four sons of Shah Jahan all held governorships during their father's reign.<br></code> | <code>शाह्-जहाँ-नामकस्य चत्वारः पुत्राः, सर्वे पितुः शासनकाले शासकपदम् अधारयन्।<br></code> |
199
+ | <code>In this regard he discusses the correlation between social opportunities of education and health and how both of these complement economic and political freedoms as a healthy and well-educated person is better suited to make informed economic decisions and be involved in fruitful political demonstrations etc.<br></code> | <code>अस्मिन् विषये सः शिक्षणस्य स्वास्थ्यस्य च सामाजिकावकाशानाम् अन्योन्य-सम्बन्धस्य, तथा च एतद्द्वयम् अपि आर्थिक-राजनैतिक-स्वातन्त्र्ययोः कथं पूरकं भवतः इति च चर्चां करोति, यतोहि स्वस्था सुशिक्षिता च व्यक्तिः ज्ञानपूर्वम् आर्थिकविषयान् निर्णेतुं तथा फलप्रदेषु राजनैतिकेषु प्रतिपादनादिषु संलग्नः भवितुं च अधिकारी भवति इति।<br></code> |
200
+ * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
201
+ ```json
202
+ {
203
+ "scale": 20.0,
204
+ "similarity_fct": "cos_sim",
205
+ "gather_across_devices": false
206
+ }
207
+ ```
208
+
209
+ ### Evaluation Dataset
210
+
211
+ #### Unnamed Dataset
212
+
213
+ * Size: 1,000 evaluation samples
214
+ * Columns: <code>sentence1</code> and <code>sentence2</code>
215
+ * Approximate statistics based on the first 1000 samples:
216
+ | | sentence1 | sentence2 |
217
+ |:--------|:---------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
218
+ | type | string | string |
219
+ | details | <ul><li>min: 5 tokens</li><li>mean: 11.9 tokens</li><li>max: 67 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 23.13 tokens</li><li>max: 128 tokens</li></ul> |
220
+ * Samples:
221
+ | sentence1 | sentence2 |
222
+ |:------------------------------------------------------------------------------------------|:------------------------------------------------------------|
223
+ | <code>plus 2 tempered glass screen protectors:</code> | <code>6 पश्चात तापाभिसंतप्तॊ विदुर समार कर्शितः</code> |
224
+ | <code>"Take sadaqah (alms) from their wealth in order to purify them with it." (p.</code> | <code>अप्येकाङ्गेऽप्यधोवस्तुमिच्छामि च सुकुत्सिते" ॥</code> |
225
+ | <code>"Who could it possibly be?"</code> | <code>कश्च तासेः सम्भवति ?</code> |
226
+ * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
227
+ ```json
228
+ {
229
+ "scale": 20.0,
230
+ "similarity_fct": "cos_sim",
231
+ "gather_across_devices": false
232
+ }
233
+ ```
234
+
235
+ ### Training Hyperparameters
236
+ #### Non-Default Hyperparameters
237
+
238
+ - `per_device_train_batch_size`: 32
239
+ - `num_train_epochs`: 5
240
+ - `max_steps`: 12000
241
+ - `learning_rate`: 2e-05
242
+ - `warmup_steps`: 500
243
+ - `gradient_accumulation_steps`: 4
244
+ - `bf16`: True
245
+ - `eval_strategy`: steps
246
+ - `load_best_model_at_end`: True
247
+
248
+ #### All Hyperparameters
249
+ <details><summary>Click to expand</summary>
250
+
251
+ - `per_device_train_batch_size`: 32
252
+ - `num_train_epochs`: 5
253
+ - `max_steps`: 12000
254
+ - `learning_rate`: 2e-05
255
+ - `lr_scheduler_type`: linear
256
+ - `lr_scheduler_kwargs`: None
257
+ - `warmup_steps`: 500
258
+ - `optim`: adamw_torch_fused
259
+ - `optim_args`: None
260
+ - `weight_decay`: 0.0
261
+ - `adam_beta1`: 0.9
262
+ - `adam_beta2`: 0.999
263
+ - `adam_epsilon`: 1e-08
264
+ - `optim_target_modules`: None
265
+ - `gradient_accumulation_steps`: 4
266
+ - `average_tokens_across_devices`: True
267
+ - `max_grad_norm`: 1.0
268
+ - `label_smoothing_factor`: 0.0
269
+ - `bf16`: True
270
+ - `fp16`: False
271
+ - `bf16_full_eval`: False
272
+ - `fp16_full_eval`: False
273
+ - `tf32`: None
274
+ - `gradient_checkpointing`: False
275
+ - `gradient_checkpointing_kwargs`: None
276
+ - `torch_compile`: False
277
+ - `torch_compile_backend`: None
278
+ - `torch_compile_mode`: None
279
+ - `use_liger_kernel`: False
280
+ - `liger_kernel_config`: None
281
+ - `use_cache`: False
282
+ - `neftune_noise_alpha`: None
283
+ - `torch_empty_cache_steps`: None
284
+ - `auto_find_batch_size`: False
285
+ - `log_on_each_node`: True
286
+ - `logging_nan_inf_filter`: True
287
+ - `include_num_input_tokens_seen`: no
288
+ - `log_level`: passive
289
+ - `log_level_replica`: warning
290
+ - `disable_tqdm`: False
291
+ - `project`: huggingface
292
+ - `trackio_space_id`: trackio
293
+ - `eval_strategy`: steps
294
+ - `per_device_eval_batch_size`: 8
295
+ - `prediction_loss_only`: True
296
+ - `eval_on_start`: False
297
+ - `eval_do_concat_batches`: True
298
+ - `eval_use_gather_object`: False
299
+ - `eval_accumulation_steps`: None
300
+ - `include_for_metrics`: []
301
+ - `batch_eval_metrics`: False
302
+ - `save_only_model`: False
303
+ - `save_on_each_node`: False
304
+ - `enable_jit_checkpoint`: False
305
+ - `push_to_hub`: False
306
+ - `hub_private_repo`: None
307
+ - `hub_model_id`: None
308
+ - `hub_strategy`: every_save
309
+ - `hub_always_push`: False
310
+ - `hub_revision`: None
311
+ - `load_best_model_at_end`: True
312
+ - `ignore_data_skip`: False
313
+ - `restore_callback_states_from_checkpoint`: False
314
+ - `full_determinism`: False
315
+ - `seed`: 42
316
+ - `data_seed`: None
317
+ - `use_cpu`: False
318
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
319
+ - `parallelism_config`: None
320
+ - `dataloader_drop_last`: False
321
+ - `dataloader_num_workers`: 0
322
+ - `dataloader_pin_memory`: True
323
+ - `dataloader_persistent_workers`: False
324
+ - `dataloader_prefetch_factor`: None
325
+ - `remove_unused_columns`: True
326
+ - `label_names`: None
327
+ - `train_sampling_strategy`: random
328
+ - `length_column_name`: length
329
+ - `ddp_find_unused_parameters`: None
330
+ - `ddp_bucket_cap_mb`: None
331
+ - `ddp_broadcast_buffers`: False
332
+ - `ddp_backend`: None
333
+ - `ddp_timeout`: 1800
334
+ - `fsdp`: []
335
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
336
+ - `deepspeed`: None
337
+ - `debug`: []
338
+ - `skip_memory_metrics`: True
339
+ - `do_predict`: False
340
+ - `resume_from_checkpoint`: None
341
+ - `warmup_ratio`: None
342
+ - `local_rank`: -1
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 | eval-en-sa_mean_accuracy |
355
+ |:----------:|:--------:|:-------------:|:---------------:|:------------------------:|
356
+ | 0.0034 | 100 | 3.1353 | - | - |
357
+ | 0.0068 | 200 | 2.7273 | - | - |
358
+ | 0.0102 | 300 | 1.8263 | - | - |
359
+ | 0.0137 | 400 | 1.1810 | - | - |
360
+ | 0.0171 | 500 | 0.8952 | - | - |
361
+ | 0.0205 | 600 | 0.7068 | - | - |
362
+ | 0.0239 | 700 | 0.5979 | - | - |
363
+ | 0.0273 | 800 | 0.5412 | - | - |
364
+ | 0.0307 | 900 | 0.5255 | - | - |
365
+ | 0.0341 | 1000 | 0.4847 | 0.2013 | 0.5045 |
366
+ | 0.0376 | 1100 | 0.4752 | - | - |
367
+ | 0.0410 | 1200 | 0.4645 | - | - |
368
+ | 0.0444 | 1300 | 0.4173 | - | - |
369
+ | 0.0478 | 1400 | 0.4220 | - | - |
370
+ | 0.0512 | 1500 | 0.4163 | - | - |
371
+ | 0.0546 | 1600 | 0.3978 | - | - |
372
+ | 0.0580 | 1700 | 0.3895 | - | - |
373
+ | 0.0614 | 1800 | 0.3778 | - | - |
374
+ | 0.0649 | 1900 | 0.3904 | - | - |
375
+ | 0.0683 | 2000 | 0.3656 | 0.1436 | 0.563 |
376
+ | 0.0717 | 2100 | 0.3565 | - | - |
377
+ | 0.0751 | 2200 | 0.3526 | - | - |
378
+ | 0.0785 | 2300 | 0.3632 | - | - |
379
+ | 0.0819 | 2400 | 0.3468 | - | - |
380
+ | 0.0853 | 2500 | 0.3506 | - | - |
381
+ | 0.0888 | 2600 | 0.3505 | - | - |
382
+ | 0.0922 | 2700 | 0.3466 | - | - |
383
+ | 0.0956 | 2800 | 0.3422 | - | - |
384
+ | 0.0990 | 2900 | 0.3393 | - | - |
385
+ | 0.1024 | 3000 | 0.3345 | 0.1240 | 0.587 |
386
+ | 0.1058 | 3100 | 0.3238 | - | - |
387
+ | 0.1092 | 3200 | 0.3230 | - | - |
388
+ | 0.1127 | 3300 | 0.3281 | - | - |
389
+ | 0.1161 | 3400 | 0.3246 | - | - |
390
+ | 0.1195 | 3500 | 0.3111 | - | - |
391
+ | 0.1229 | 3600 | 0.3092 | - | - |
392
+ | 0.1263 | 3700 | 0.3187 | - | - |
393
+ | 0.1297 | 3800 | 0.3293 | - | - |
394
+ | 0.1331 | 3900 | 0.3246 | - | - |
395
+ | 0.1366 | 4000 | 0.3174 | 0.1165 | 0.598 |
396
+ | 0.1400 | 4100 | 0.3213 | - | - |
397
+ | 0.1434 | 4200 | 0.3167 | - | - |
398
+ | 0.1468 | 4300 | 0.3142 | - | - |
399
+ | 0.1502 | 4400 | 0.3070 | - | - |
400
+ | 0.1536 | 4500 | 0.3094 | - | - |
401
+ | 0.1570 | 4600 | 0.3084 | - | - |
402
+ | 0.1604 | 4700 | 0.3068 | - | - |
403
+ | 0.1639 | 4800 | 0.3060 | - | - |
404
+ | 0.1673 | 4900 | 0.3020 | - | - |
405
+ | 0.1707 | 5000 | 0.3072 | 0.1133 | 0.6045 |
406
+ | 0.1741 | 5100 | 0.3151 | - | - |
407
+ | 0.1775 | 5200 | 0.3121 | - | - |
408
+ | 0.1809 | 5300 | 0.3059 | - | - |
409
+ | 0.1843 | 5400 | 0.3069 | - | - |
410
+ | 0.1878 | 5500 | 0.3069 | - | - |
411
+ | 0.1912 | 5600 | 0.3134 | - | - |
412
+ | 0.1946 | 5700 | 0.3017 | - | - |
413
+ | 0.1980 | 5800 | 0.3088 | - | - |
414
+ | 0.2014 | 5900 | 0.3011 | - | - |
415
+ | 0.2048 | 6000 | 0.3075 | 0.1109 | 0.608 |
416
+ | 0.2082 | 6100 | 0.2957 | - | - |
417
+ | 0.2117 | 6200 | 0.3049 | - | - |
418
+ | 0.2151 | 6300 | 0.2994 | - | - |
419
+ | 0.2185 | 6400 | 0.2951 | - | - |
420
+ | 0.2219 | 6500 | 0.3116 | - | - |
421
+ | 0.2253 | 6600 | 0.3155 | - | - |
422
+ | 0.2287 | 6700 | 0.2938 | - | - |
423
+ | 0.2321 | 6800 | 0.2824 | - | - |
424
+ | 0.2355 | 6900 | 0.2973 | - | - |
425
+ | 0.2390 | 7000 | 0.3111 | 0.1100 | 0.6065 |
426
+ | 0.2424 | 7100 | 0.2973 | - | - |
427
+ | 0.2458 | 7200 | 0.2995 | - | - |
428
+ | 0.2492 | 7300 | 0.2962 | - | - |
429
+ | 0.2526 | 7400 | 0.2994 | - | - |
430
+ | 0.2560 | 7500 | 0.2964 | - | - |
431
+ | 0.2594 | 7600 | 0.2997 | - | - |
432
+ | 0.2629 | 7700 | 0.2932 | - | - |
433
+ | 0.2663 | 7800 | 0.2993 | - | - |
434
+ | 0.2697 | 7900 | 0.2987 | - | - |
435
+ | 0.2731 | 8000 | 0.2898 | 0.1084 | 0.6085 |
436
+ | 0.2765 | 8100 | 0.3007 | - | - |
437
+ | 0.2799 | 8200 | 0.2935 | - | - |
438
+ | 0.2833 | 8300 | 0.2885 | - | - |
439
+ | 0.2868 | 8400 | 0.3021 | - | - |
440
+ | 0.2902 | 8500 | 0.2958 | - | - |
441
+ | 0.2936 | 8600 | 0.3056 | - | - |
442
+ | 0.2970 | 8700 | 0.2908 | - | - |
443
+ | 0.3004 | 8800 | 0.3096 | - | - |
444
+ | 0.3038 | 8900 | 0.2924 | - | - |
445
+ | **0.3072** | **9000** | **0.3019** | **0.1077** | **0.607** |
446
+ | 0.3107 | 9100 | 0.2985 | - | - |
447
+ | 0.3141 | 9200 | 0.2906 | - | - |
448
+ | 0.3175 | 9300 | 0.2961 | - | - |
449
+ | 0.3209 | 9400 | 0.3044 | - | - |
450
+ | 0.3243 | 9500 | 0.3005 | - | - |
451
+ | 0.3277 | 9600 | 0.2943 | - | - |
452
+ | 0.3311 | 9700 | 0.2948 | - | - |
453
+ | 0.3345 | 9800 | 0.3046 | - | - |
454
+ | 0.3380 | 9900 | 0.2948 | - | - |
455
+ | 0.3414 | 10000 | 0.3060 | 0.1083 | 0.608 |
456
+ | 0.3448 | 10100 | 0.2906 | - | - |
457
+ | 0.3482 | 10200 | 0.2958 | - | - |
458
+ | 0.3516 | 10300 | 0.2919 | - | - |
459
+ | 0.3550 | 10400 | 0.3041 | - | - |
460
+ | 0.3584 | 10500 | 0.3055 | - | - |
461
+ | 0.3619 | 10600 | 0.2975 | - | - |
462
+ | 0.3653 | 10700 | 0.2984 | - | - |
463
+ | 0.3687 | 10800 | 0.2883 | - | - |
464
+ | 0.3721 | 10900 | 0.2949 | - | - |
465
+ | 0.3755 | 11000 | 0.2987 | 0.1083 | 0.6085 |
466
+ | 0.3789 | 11100 | 0.2938 | - | - |
467
+ | 0.3823 | 11200 | 0.2942 | - | - |
468
+ | 0.3858 | 11300 | 0.2879 | - | - |
469
+ | 0.3892 | 11400 | 0.2909 | - | - |
470
+ | 0.3926 | 11500 | 0.2899 | - | - |
471
+ | 0.3960 | 11600 | 0.2921 | - | - |
472
+ | 0.3994 | 11700 | 0.2944 | - | - |
473
+ | 0.4028 | 11800 | 0.2985 | - | - |
474
+ | 0.4062 | 11900 | 0.3027 | - | - |
475
+ | 0.4097 | 12000 | 0.2988 | 0.1082 | 0.61 |
476
+
477
+ * The bold row denotes the saved checkpoint.
478
+ </details>
479
+
480
+ ### Framework Versions
481
+ - Python: 3.10.18
482
+ - Sentence Transformers: 5.2.3
483
+ - Transformers: 5.2.0
484
+ - PyTorch: 2.8.0+cu128
485
+ - Accelerate: 1.12.0
486
+ - Datasets: 3.3.2
487
+ - Tokenizers: 0.22.1
488
+
489
+ ## Citation
490
+
491
+ ### BibTeX
492
+
493
+ #### Sentence Transformers
494
+ ```bibtex
495
+ @inproceedings{reimers-2019-sentence-bert,
496
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
497
+ author = "Reimers, Nils and Gurevych, Iryna",
498
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
499
+ month = "11",
500
+ year = "2019",
501
+ publisher = "Association for Computational Linguistics",
502
+ url = "https://arxiv.org/abs/1908.10084",
503
+ }
504
+ ```
505
+
506
+ #### MultipleNegativesRankingLoss
507
+ ```bibtex
508
+ @misc{henderson2017efficient,
509
+ title={Efficient Natural Language Response Suggestion for Smart Reply},
510
+ author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
511
+ year={2017},
512
+ eprint={1705.00652},
513
+ archivePrefix={arXiv},
514
+ primaryClass={cs.CL}
515
+ }
516
+ ```
517
+
518
+ <!--
519
+ ## Glossary
520
+
521
+ *Clearly define terms in order to be accessible across audiences.*
522
+ -->
523
+
524
+ <!--
525
+ ## Model Card Authors
526
+
527
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
528
+ -->
529
+
530
+ <!--
531
+ ## Model Card Contact
532
+
533
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
534
+ -->
config.json ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "ModernBertModel"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 2,
8
+ "classifier_activation": "gelu",
9
+ "classifier_bias": false,
10
+ "classifier_dropout": 0.0,
11
+ "classifier_pooling": "mean",
12
+ "cls_token_id": 1,
13
+ "decoder_bias": true,
14
+ "deterministic_flash_attn": false,
15
+ "dtype": "bfloat16",
16
+ "embedding_dropout": 0.0,
17
+ "eos_token_id": 1,
18
+ "global_attn_every_n_layers": 3,
19
+ "gradient_checkpointing": false,
20
+ "hidden_activation": "gelu",
21
+ "hidden_size": 768,
22
+ "initializer_cutoff_factor": 2.0,
23
+ "initializer_range": 0.02,
24
+ "intermediate_size": 1152,
25
+ "layer_norm_eps": 1e-05,
26
+ "layer_types": [
27
+ "full_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "full_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "full_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "full_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "full_attention",
40
+ "sliding_attention",
41
+ "sliding_attention",
42
+ "full_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "full_attention",
46
+ "sliding_attention",
47
+ "sliding_attention",
48
+ "full_attention"
49
+ ],
50
+ "local_attention": 128,
51
+ "mask_token_id": 4,
52
+ "max_position_embeddings": 8192,
53
+ "mlp_bias": false,
54
+ "mlp_dropout": 0.0,
55
+ "model_type": "modernbert",
56
+ "norm_bias": false,
57
+ "norm_eps": 1e-05,
58
+ "num_attention_heads": 12,
59
+ "num_hidden_layers": 22,
60
+ "pad_token_id": 0,
61
+ "position_embedding_type": "sans_pos",
62
+ "rope_parameters": {
63
+ "full_attention": {
64
+ "rope_theta": 160000,
65
+ "rope_type": "default"
66
+ },
67
+ "sliding_attention": {
68
+ "rope_theta": 160000,
69
+ "rope_type": "default"
70
+ }
71
+ },
72
+ "sep_token_id": 1,
73
+ "sparse_pred_ignore_index": -100,
74
+ "sparse_prediction": false,
75
+ "tie_word_embeddings": true,
76
+ "transformers_version": "5.2.0",
77
+ "vocab_size": 256000
78
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "SentenceTransformer",
3
+ "__version__": {
4
+ "sentence_transformers": "5.2.3",
5
+ "transformers": "5.2.0",
6
+ "pytorch": "2.8.0+cu128"
7
+ },
8
+ "prompts": {
9
+ "query": "",
10
+ "document": ""
11
+ },
12
+ "default_prompt_name": null,
13
+ "similarity_fn_name": "cosine"
14
+ }
eval/translation_evaluation_eval-en-sa_results.csv ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ epoch,steps,src2trg,trg2src
2
+ 0.034137557287088324,1000,0.512,0.497
3
+ 0.06827511457417665,2000,0.568,0.558
4
+ 0.10241267186126497,3000,0.593,0.581
5
+ 0.1365502291483533,4000,0.607,0.589
6
+ 0.1706877864354416,5000,0.614,0.595
7
+ 0.20482534372252995,6000,0.616,0.6
8
+ 0.23896290100961826,7000,0.613,0.6
9
+ 0.2731004582967066,8000,0.615,0.602
10
+ 0.3072380155837949,9000,0.612,0.602
11
+ 0.3413755728708832,10000,0.616,0.6
12
+ 0.37551313015797155,11000,0.618,0.599
13
+ 0.4096506874450599,12000,0.616,0.604
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e353f38ef2aace23d6bcfc994fd215c251f8c86bee247b0311a61501c66a8b79
3
+ size 613892480
modules.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ }
14
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 128,
3
+ "do_lower_case": false
4
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f231fae363627485dbad73e8b95f601378deede800416098c14c81a0deafa813
3
+ size 34363441
tokenizer_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "bos_token": "<bos>",
4
+ "clean_up_tokenization_spaces": false,
5
+ "cls_token": "<bos>",
6
+ "eos_token": "<eos>",
7
+ "extra_special_tokens": [
8
+ "<start_of_turn>",
9
+ "<end_of_turn>"
10
+ ],
11
+ "is_local": false,
12
+ "mask_token": "<mask>",
13
+ "model_input_names": [
14
+ "input_ids",
15
+ "attention_mask"
16
+ ],
17
+ "model_max_length": 8192,
18
+ "pad_token": "<pad>",
19
+ "padding_side": "right",
20
+ "sep_token": "<eos>",
21
+ "spaces_between_special_tokens": false,
22
+ "tokenizer_class": "TokenizersBackend",
23
+ "unk_token": "<unk>"
24
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:969c222c2b4019f08afe7f32e0c7734d6632ad5ef36d467bb62a1a1b82e36873
3
+ size 5521