manu commited on
Commit
5f71831
1 Parent(s): 0825023

Add new SentenceTransformer model.

Browse files
.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": 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 +1,524 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ library_name: sentence-transformers
5
+ tags:
6
+ - sentence-transformers
7
+ - sentence-similarity
8
+ - feature-extraction
9
+ - dataset_size:100K<n<1M
10
+ - loss:CachedMultipleNegativesRankingLoss
11
+ base_model: Unbabel/xlm-roberta-comet-small
12
+ metrics:
13
+ - cosine_accuracy
14
+ - dot_accuracy
15
+ - manhattan_accuracy
16
+ - euclidean_accuracy
17
+ - max_accuracy
18
+ widget:
19
+ - source_sentence: There's a dock
20
+ sentences:
21
+ - There is a door.
22
+ - the animal is running
23
+ - The woman is singing.
24
+ - source_sentence: The boy scowls
25
+ sentences:
26
+ - A boy is blowing bubbles.
27
+ - He is playing a song.
28
+ - They are driving cars.
29
+ - source_sentence: A bird flying.
30
+ sentences:
31
+ - A butterfly flys freely.
32
+ - A dog carries a bone.
33
+ - Two dogs are playing.
34
+ - source_sentence: A woman sings.
35
+ sentences:
36
+ - The woman is singing.
37
+ - The man is in a city.
38
+ - there is a man in a pool.
39
+ - source_sentence: a baby smiling
40
+ sentences:
41
+ - A baby is unhappy.
42
+ - The dog has big ears.
43
+ - They are driving cars.
44
+ pipeline_tag: sentence-similarity
45
+ model-index:
46
+ - name: SentenceTransformer based on Unbabel/xlm-roberta-comet-small
47
+ results:
48
+ - task:
49
+ type: triplet
50
+ name: Triplet
51
+ dataset:
52
+ name: all nli dev
53
+ type: all-nli-dev
54
+ metrics:
55
+ - type: cosine_accuracy
56
+ value: 0.849
57
+ name: Cosine Accuracy
58
+ - type: dot_accuracy
59
+ value: 0.163
60
+ name: Dot Accuracy
61
+ - type: manhattan_accuracy
62
+ value: 0.837
63
+ name: Manhattan Accuracy
64
+ - type: euclidean_accuracy
65
+ value: 0.841
66
+ name: Euclidean Accuracy
67
+ - type: max_accuracy
68
+ value: 0.849
69
+ name: Max Accuracy
70
+ - task:
71
+ type: triplet
72
+ name: Triplet
73
+ dataset:
74
+ name: all nli test
75
+ type: all-nli-test
76
+ metrics:
77
+ - type: cosine_accuracy
78
+ value: 0.839
79
+ name: Cosine Accuracy
80
+ - type: dot_accuracy
81
+ value: 0.15
82
+ name: Dot Accuracy
83
+ - type: manhattan_accuracy
84
+ value: 0.827
85
+ name: Manhattan Accuracy
86
+ - type: euclidean_accuracy
87
+ value: 0.827
88
+ name: Euclidean Accuracy
89
+ - type: max_accuracy
90
+ value: 0.839
91
+ name: Max Accuracy
92
+ ---
93
+
94
+ # SentenceTransformer based on Unbabel/xlm-roberta-comet-small
95
+
96
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Unbabel/xlm-roberta-comet-small](https://huggingface.co/Unbabel/xlm-roberta-comet-small) on the [sentence-transformers/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.
97
+
98
+ ## Model Details
99
+
100
+ ### Model Description
101
+ - **Model Type:** Sentence Transformer
102
+ - **Base model:** [Unbabel/xlm-roberta-comet-small](https://huggingface.co/Unbabel/xlm-roberta-comet-small) <!-- at revision df568a015df5cefbf2f449314b61ce9afb0cb593 -->
103
+ - **Maximum Sequence Length:** 512 tokens
104
+ - **Output Dimensionality:** 384 tokens
105
+ - **Similarity Function:** Cosine Similarity
106
+ - **Training Dataset:**
107
+ - [sentence-transformers/all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli)
108
+ - **Language:** en
109
+ <!-- - **License:** Unknown -->
110
+
111
+ ### Model Sources
112
+
113
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
114
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
115
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
116
+
117
+ ### Full Model Architecture
118
+
119
+ ```
120
+ SentenceTransformer(
121
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: XLMRobertaModel
122
+ (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})
123
+ )
124
+ ```
125
+
126
+ ## Usage
127
+
128
+ ### Direct Usage (Sentence Transformers)
129
+
130
+ First install the Sentence Transformers library:
131
+
132
+ ```bash
133
+ pip install -U sentence-transformers
134
+ ```
135
+
136
+ Then you can load this model and run inference.
137
+ ```python
138
+ from sentence_transformers import SentenceTransformer
139
+
140
+ # Download from the 🤗 Hub
141
+ model = SentenceTransformer("mics-nlp/xlm-roberta-small-all-nli-triplet")
142
+ # Run inference
143
+ sentences = [
144
+ 'a baby smiling',
145
+ 'A baby is unhappy.',
146
+ 'The dog has big ears.',
147
+ ]
148
+ embeddings = model.encode(sentences)
149
+ print(embeddings.shape)
150
+ # [3, 384]
151
+
152
+ # Get the similarity scores for the embeddings
153
+ similarities = model.similarity(embeddings, embeddings)
154
+ print(similarities.shape)
155
+ # [3, 3]
156
+ ```
157
+
158
+ <!--
159
+ ### Direct Usage (Transformers)
160
+
161
+ <details><summary>Click to see the direct usage in Transformers</summary>
162
+
163
+ </details>
164
+ -->
165
+
166
+ <!--
167
+ ### Downstream Usage (Sentence Transformers)
168
+
169
+ You can finetune this model on your own dataset.
170
+
171
+ <details><summary>Click to expand</summary>
172
+
173
+ </details>
174
+ -->
175
+
176
+ <!--
177
+ ### Out-of-Scope Use
178
+
179
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
180
+ -->
181
+
182
+ ## Evaluation
183
+
184
+ ### Metrics
185
+
186
+ #### Triplet
187
+ * Dataset: `all-nli-dev`
188
+ * Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
189
+
190
+ | Metric | Value |
191
+ |:-------------------|:----------|
192
+ | cosine_accuracy | 0.849 |
193
+ | dot_accuracy | 0.163 |
194
+ | manhattan_accuracy | 0.837 |
195
+ | euclidean_accuracy | 0.841 |
196
+ | **max_accuracy** | **0.849** |
197
+
198
+ #### Triplet
199
+ * Dataset: `all-nli-test`
200
+ * Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
201
+
202
+ | Metric | Value |
203
+ |:-------------------|:----------|
204
+ | cosine_accuracy | 0.839 |
205
+ | dot_accuracy | 0.15 |
206
+ | manhattan_accuracy | 0.827 |
207
+ | euclidean_accuracy | 0.827 |
208
+ | **max_accuracy** | **0.839** |
209
+
210
+ <!--
211
+ ## Bias, Risks and Limitations
212
+
213
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
214
+ -->
215
+
216
+ <!--
217
+ ### Recommendations
218
+
219
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
220
+ -->
221
+
222
+ ## Training Details
223
+
224
+ ### Training Dataset
225
+
226
+ #### sentence-transformers/all-nli
227
+
228
+ * Dataset: [sentence-transformers/all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) at [d482672](https://huggingface.co/datasets/sentence-transformers/all-nli/tree/d482672c8e74ce18da116f430137434ba2e52fab)
229
+ * Size: 100,000 training samples
230
+ * Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
231
+ * Approximate statistics based on the first 1000 samples:
232
+ | | anchor | positive | negative |
233
+ |:--------|:---------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
234
+ | type | string | string | string |
235
+ | details | <ul><li>min: 7 tokens</li><li>mean: 10.9 tokens</li><li>max: 52 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 13.62 tokens</li><li>max: 42 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 14.76 tokens</li><li>max: 55 tokens</li></ul> |
236
+ * Samples:
237
+ | anchor | positive | negative |
238
+ |:---------------------------------------------------------------------------|:-------------------------------------------------|:-----------------------------------------------------------|
239
+ | <code>A person on a horse jumps over a broken down airplane.</code> | <code>A person is outdoors, on a horse.</code> | <code>A person is at a diner, ordering an omelette.</code> |
240
+ | <code>Children smiling and waving at camera</code> | <code>There are children present</code> | <code>The kids are frowning</code> |
241
+ | <code>A boy is jumping on skateboard in the middle of a red bridge.</code> | <code>The boy does a skateboarding trick.</code> | <code>The boy skates down the sidewalk.</code> |
242
+ * Loss: [<code>CachedMultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cachedmultiplenegativesrankingloss) with these parameters:
243
+ ```json
244
+ {
245
+ "scale": 20.0,
246
+ "similarity_fct": "cos_sim"
247
+ }
248
+ ```
249
+
250
+ ### Evaluation Dataset
251
+
252
+ #### sentence-transformers/all-nli
253
+
254
+ * Dataset: [sentence-transformers/all-nli](https://huggingface.co/datasets/sentence-transformers/all-nli) at [d482672](https://huggingface.co/datasets/sentence-transformers/all-nli/tree/d482672c8e74ce18da116f430137434ba2e52fab)
255
+ * Size: 1,000 evaluation samples
256
+ * Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
257
+ * Approximate statistics based on the first 1000 samples:
258
+ | | anchor | positive | negative |
259
+ |:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
260
+ | type | string | string | string |
261
+ | details | <ul><li>min: 6 tokens</li><li>mean: 20.31 tokens</li><li>max: 83 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 10.71 tokens</li><li>max: 35 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 11.39 tokens</li><li>max: 32 tokens</li></ul> |
262
+ * Samples:
263
+ | anchor | positive | negative |
264
+ |:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------|:--------------------------------------------------------|
265
+ | <code>Two women are embracing while holding to go packages.</code> | <code>Two woman are holding packages.</code> | <code>The men are fighting outside a deli.</code> |
266
+ | <code>Two young children in blue jerseys, one with the number 9 and one with the number 2 are standing on wooden steps in a bathroom and washing their hands in a sink.</code> | <code>Two kids in numbered jerseys wash their hands.</code> | <code>Two kids in jackets walk to school.</code> |
267
+ | <code>A man selling donuts to a customer during a world exhibition event held in the city of Angeles</code> | <code>A man selling donuts to a customer.</code> | <code>A woman drinks her coffee in a small cafe.</code> |
268
+ * Loss: [<code>CachedMultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cachedmultiplenegativesrankingloss) with these parameters:
269
+ ```json
270
+ {
271
+ "scale": 20.0,
272
+ "similarity_fct": "cos_sim"
273
+ }
274
+ ```
275
+
276
+ ### Training Hyperparameters
277
+ #### Non-Default Hyperparameters
278
+
279
+ - `eval_strategy`: steps
280
+ - `per_device_train_batch_size`: 16
281
+ - `per_device_eval_batch_size`: 16
282
+ - `num_train_epochs`: 1
283
+ - `warmup_ratio`: 0.1
284
+ - `bf16`: True
285
+ - `batch_sampler`: no_duplicates
286
+
287
+ #### All Hyperparameters
288
+ <details><summary>Click to expand</summary>
289
+
290
+ - `overwrite_output_dir`: False
291
+ - `do_predict`: False
292
+ - `eval_strategy`: steps
293
+ - `prediction_loss_only`: True
294
+ - `per_device_train_batch_size`: 16
295
+ - `per_device_eval_batch_size`: 16
296
+ - `per_gpu_train_batch_size`: None
297
+ - `per_gpu_eval_batch_size`: None
298
+ - `gradient_accumulation_steps`: 1
299
+ - `eval_accumulation_steps`: None
300
+ - `learning_rate`: 5e-05
301
+ - `weight_decay`: 0.0
302
+ - `adam_beta1`: 0.9
303
+ - `adam_beta2`: 0.999
304
+ - `adam_epsilon`: 1e-08
305
+ - `max_grad_norm`: 1.0
306
+ - `num_train_epochs`: 1
307
+ - `max_steps`: -1
308
+ - `lr_scheduler_type`: linear
309
+ - `lr_scheduler_kwargs`: {}
310
+ - `warmup_ratio`: 0.1
311
+ - `warmup_steps`: 0
312
+ - `log_level`: passive
313
+ - `log_level_replica`: warning
314
+ - `log_on_each_node`: True
315
+ - `logging_nan_inf_filter`: True
316
+ - `save_safetensors`: True
317
+ - `save_on_each_node`: False
318
+ - `save_only_model`: False
319
+ - `restore_callback_states_from_checkpoint`: False
320
+ - `no_cuda`: False
321
+ - `use_cpu`: False
322
+ - `use_mps_device`: False
323
+ - `seed`: 42
324
+ - `data_seed`: None
325
+ - `jit_mode_eval`: False
326
+ - `use_ipex`: False
327
+ - `bf16`: True
328
+ - `fp16`: False
329
+ - `fp16_opt_level`: O1
330
+ - `half_precision_backend`: auto
331
+ - `bf16_full_eval`: False
332
+ - `fp16_full_eval`: False
333
+ - `tf32`: None
334
+ - `local_rank`: 0
335
+ - `ddp_backend`: None
336
+ - `tpu_num_cores`: None
337
+ - `tpu_metrics_debug`: False
338
+ - `debug`: []
339
+ - `dataloader_drop_last`: False
340
+ - `dataloader_num_workers`: 0
341
+ - `dataloader_prefetch_factor`: None
342
+ - `past_index`: -1
343
+ - `disable_tqdm`: False
344
+ - `remove_unused_columns`: True
345
+ - `label_names`: None
346
+ - `load_best_model_at_end`: False
347
+ - `ignore_data_skip`: False
348
+ - `fsdp`: []
349
+ - `fsdp_min_num_params`: 0
350
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
351
+ - `fsdp_transformer_layer_cls_to_wrap`: None
352
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
353
+ - `deepspeed`: None
354
+ - `label_smoothing_factor`: 0.0
355
+ - `optim`: adamw_torch
356
+ - `optim_args`: None
357
+ - `adafactor`: False
358
+ - `group_by_length`: False
359
+ - `length_column_name`: length
360
+ - `ddp_find_unused_parameters`: None
361
+ - `ddp_bucket_cap_mb`: None
362
+ - `ddp_broadcast_buffers`: False
363
+ - `dataloader_pin_memory`: True
364
+ - `dataloader_persistent_workers`: False
365
+ - `skip_memory_metrics`: True
366
+ - `use_legacy_prediction_loop`: False
367
+ - `push_to_hub`: False
368
+ - `resume_from_checkpoint`: None
369
+ - `hub_model_id`: None
370
+ - `hub_strategy`: every_save
371
+ - `hub_private_repo`: False
372
+ - `hub_always_push`: False
373
+ - `gradient_checkpointing`: False
374
+ - `gradient_checkpointing_kwargs`: None
375
+ - `include_inputs_for_metrics`: False
376
+ - `eval_do_concat_batches`: True
377
+ - `fp16_backend`: auto
378
+ - `push_to_hub_model_id`: None
379
+ - `push_to_hub_organization`: None
380
+ - `mp_parameters`:
381
+ - `auto_find_batch_size`: False
382
+ - `full_determinism`: False
383
+ - `torchdynamo`: None
384
+ - `ray_scope`: last
385
+ - `ddp_timeout`: 1800
386
+ - `torch_compile`: False
387
+ - `torch_compile_backend`: None
388
+ - `torch_compile_mode`: None
389
+ - `dispatch_batches`: None
390
+ - `split_batches`: None
391
+ - `include_tokens_per_second`: False
392
+ - `include_num_input_tokens_seen`: False
393
+ - `neftune_noise_alpha`: None
394
+ - `optim_target_modules`: None
395
+ - `batch_eval_metrics`: False
396
+ - `batch_sampler`: no_duplicates
397
+ - `multi_dataset_batch_sampler`: proportional
398
+
399
+ </details>
400
+
401
+ ### Training Logs
402
+ | Epoch | Step | Training Loss | loss | all-nli-dev_max_accuracy | all-nli-test_max_accuracy |
403
+ |:-----:|:----:|:-------------:|:------:|:------------------------:|:-------------------------:|
404
+ | 0 | 0 | - | - | 0.541 | - |
405
+ | 0.016 | 100 | 3.5308 | 3.1817 | 0.558 | - |
406
+ | 0.032 | 200 | 3.2784 | 3.0406 | 0.597 | - |
407
+ | 0.048 | 300 | 3.113 | 2.7572 | 0.635 | - |
408
+ | 0.064 | 400 | 2.8296 | 2.4646 | 0.68 | - |
409
+ | 0.08 | 500 | 2.631 | 2.3583 | 0.676 | - |
410
+ | 0.096 | 600 | 2.3247 | 2.1394 | 0.706 | - |
411
+ | 0.112 | 700 | 2.2211 | 2.0201 | 0.711 | - |
412
+ | 0.128 | 800 | 2.1263 | 1.9560 | 0.757 | - |
413
+ | 0.144 | 900 | 2.2105 | 1.9074 | 0.748 | - |
414
+ | 0.16 | 1000 | 2.0637 | 1.9289 | 0.728 | - |
415
+ | 0.176 | 1100 | 2.1772 | 1.8796 | 0.741 | - |
416
+ | 0.192 | 1200 | 2.1518 | 1.8346 | 0.761 | - |
417
+ | 0.208 | 1300 | 1.728 | 1.8213 | 0.765 | - |
418
+ | 0.224 | 1400 | 1.8101 | 1.6321 | 0.772 | - |
419
+ | 0.24 | 1500 | 1.7516 | 1.5669 | 0.793 | - |
420
+ | 0.256 | 1600 | 1.4988 | 1.5538 | 0.8 | - |
421
+ | 0.272 | 1700 | 1.6695 | 1.5462 | 0.803 | - |
422
+ | 0.288 | 1800 | 1.5971 | 1.5499 | 0.783 | - |
423
+ | 0.304 | 1900 | 1.5614 | 1.5047 | 0.788 | - |
424
+ | 0.32 | 2000 | 1.522 | 1.4957 | 0.794 | - |
425
+ | 0.336 | 2100 | 1.3624 | 1.4153 | 0.814 | - |
426
+ | 0.352 | 2200 | 1.4773 | 1.4169 | 0.809 | - |
427
+ | 0.368 | 2300 | 1.6066 | 1.3697 | 0.813 | - |
428
+ | 0.384 | 2400 | 1.5106 | 1.3203 | 0.819 | - |
429
+ | 0.4 | 2500 | 1.4783 | 1.3417 | 0.817 | - |
430
+ | 0.416 | 2600 | 1.3696 | 1.2650 | 0.824 | - |
431
+ | 0.432 | 2700 | 1.5115 | 1.2779 | 0.829 | - |
432
+ | 0.448 | 2800 | 1.4834 | 1.2668 | 0.834 | - |
433
+ | 0.464 | 2900 | 1.4823 | 1.2621 | 0.836 | - |
434
+ | 0.48 | 3000 | 1.4163 | 1.2465 | 0.837 | - |
435
+ | 0.496 | 3100 | 1.4232 | 1.2475 | 0.837 | - |
436
+ | 0.512 | 3200 | 1.2193 | 1.1975 | 0.838 | - |
437
+ | 0.528 | 3300 | 1.2569 | 1.1816 | 0.838 | - |
438
+ | 0.544 | 3400 | 1.2988 | 1.1936 | 0.839 | - |
439
+ | 0.56 | 3500 | 1.5068 | 1.2213 | 0.835 | - |
440
+ | 0.576 | 3600 | 1.3022 | 1.1799 | 0.842 | - |
441
+ | 0.592 | 3700 | 1.3823 | 1.1910 | 0.831 | - |
442
+ | 0.608 | 3800 | 1.4224 | 1.1786 | 0.834 | - |
443
+ | 0.624 | 3900 | 1.3765 | 1.1541 | 0.843 | - |
444
+ | 0.64 | 4000 | 1.4987 | 1.1365 | 0.844 | - |
445
+ | 0.656 | 4100 | 1.7525 | 1.1394 | 0.843 | - |
446
+ | 0.672 | 4200 | 1.6013 | 1.1178 | 0.841 | - |
447
+ | 0.688 | 4300 | 1.3326 | 1.0959 | 0.846 | - |
448
+ | 0.704 | 4400 | 1.355 | 1.0757 | 0.848 | - |
449
+ | 0.72 | 4500 | 1.2834 | 1.0681 | 0.846 | - |
450
+ | 0.736 | 4600 | 1.2939 | 1.0696 | 0.85 | - |
451
+ | 0.752 | 4700 | 1.4069 | 1.0645 | 0.848 | - |
452
+ | 0.768 | 4800 | 1.4503 | 1.0609 | 0.849 | - |
453
+ | 0.784 | 4900 | 1.2833 | 1.0587 | 0.847 | - |
454
+ | 0.8 | 5000 | 1.3321 | 1.0563 | 0.849 | - |
455
+ | 0.816 | 5100 | 1.3006 | 1.0539 | 0.847 | - |
456
+ | 0.832 | 5200 | 1.4332 | 1.0527 | 0.847 | - |
457
+ | 0.848 | 5300 | 1.3101 | 1.0505 | 0.848 | - |
458
+ | 0.864 | 5400 | 1.3658 | 1.0523 | 0.849 | - |
459
+ | 0.88 | 5500 | 1.353 | 1.0520 | 0.849 | - |
460
+ | 0.896 | 5600 | 1.2429 | 1.0521 | 0.848 | - |
461
+ | 0.912 | 5700 | 1.3512 | 1.0505 | 0.848 | - |
462
+ | 0.928 | 5800 | 1.2995 | 1.0501 | 0.848 | - |
463
+ | 0.944 | 5900 | 1.3514 | 1.0491 | 0.849 | - |
464
+ | 0.96 | 6000 | 1.3976 | 1.0490 | 0.848 | - |
465
+ | 0.976 | 6100 | 1.2112 | 1.0487 | 0.848 | - |
466
+ | 0.992 | 6200 | 0.0033 | 1.0492 | 0.849 | - |
467
+ | 1.0 | 6250 | - | - | - | 0.839 |
468
+
469
+
470
+ ### Framework Versions
471
+ - Python: 3.9.10
472
+ - Sentence Transformers: 3.0.0
473
+ - Transformers: 4.41.2
474
+ - PyTorch: 2.3.0+cu121
475
+ - Accelerate: 0.26.1
476
+ - Datasets: 2.16.1
477
+ - Tokenizers: 0.19.1
478
+
479
+ ## Citation
480
+
481
+ ### BibTeX
482
+
483
+ #### Sentence Transformers
484
+ ```bibtex
485
+ @inproceedings{reimers-2019-sentence-bert,
486
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
487
+ author = "Reimers, Nils and Gurevych, Iryna",
488
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
489
+ month = "11",
490
+ year = "2019",
491
+ publisher = "Association for Computational Linguistics",
492
+ url = "https://arxiv.org/abs/1908.10084",
493
+ }
494
+ ```
495
+
496
+ #### CachedMultipleNegativesRankingLoss
497
+ ```bibtex
498
+ @misc{gao2021scaling,
499
+ title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
500
+ author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
501
+ year={2021},
502
+ eprint={2101.06983},
503
+ archivePrefix={arXiv},
504
+ primaryClass={cs.LG}
505
+ }
506
+ ```
507
+
508
+ <!--
509
+ ## Glossary
510
+
511
+ *Clearly define terms in order to be accessible across audiences.*
512
+ -->
513
+
514
+ <!--
515
+ ## Model Card Authors
516
+
517
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
518
+ -->
519
+
520
+ <!--
521
+ ## Model Card Contact
522
+
523
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
524
+ -->
config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Unbabel/xlm-roberta-comet-small",
3
+ "architectures": [
4
+ "XLMRobertaModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "eos_token_id": 2,
10
+ "gradient_checkpointing": false,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 384,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 1536,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 514,
18
+ "model_type": "xlm-roberta",
19
+ "num_attention_heads": 12,
20
+ "num_hidden_layers": 6,
21
+ "pad_token_id": 1,
22
+ "position_embedding_type": "absolute",
23
+ "torch_dtype": "bfloat16",
24
+ "transformers_version": "4.41.2",
25
+ "type_vocab_size": 1,
26
+ "use_cache": true,
27
+ "vocab_size": 250002
28
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.0.0",
4
+ "transformers": "4.41.2",
5
+ "pytorch": "2.3.0+cu121"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null,
9
+ "similarity_fn_name": null
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6cf80df00cf2a36fd7d85a04e82cf9eeddb56a6485a8fef3be209600336e19db
3
+ size 213999464
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": 512,
3
+ "do_lower_case": false
4
+ }
sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
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
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:883b037111086fd4dfebbbc9b7cee11e1517b5e0c0514879478661440f137085
3
+ size 17082987
tokenizer_config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "mask_token": "<mask>",
49
+ "model_max_length": 512,
50
+ "pad_token": "<pad>",
51
+ "sep_token": "</s>",
52
+ "tokenizer_class": "XLMRobertaTokenizer",
53
+ "unk_token": "<unk>"
54
+ }