CMU-Wav2Gloss/Gitksan-encoder-bsz128-e1k-bsz32-e1k

SentenceTransformer based on sentence-transformers/LaBSE

This is a sentence-transformers model finetuned from sentence-transformers/LaBSE. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: sentence-transformers/LaBSE
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 256, 'do_lower_case': False, 'architecture': 'BertModel'})
  (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})
  (2): Dense({'in_features': 768, 'out_features': 768, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
  (3): Normalize()
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
    "Word: a l o o h l| Context: I i   ' n i i   y a t s h l   p i p e - -   n i i   g y a ' a   w i l   s g i h l   p i p e   a l o o h l   h a ' n i i y o ' o x s x w .| Translation: And I hit the pipe-- I saw there was a pipe on the sink.",
    'Morpheme: h l | Gloss: CN',
    'Morpheme: i i | Gloss: CCNJ',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.6834, 0.3528],
#         [0.6834, 1.0000, 0.4257],
#         [0.3528, 0.4257, 1.0000]])

Evaluation

Metrics

IR

• Dataset: validation
• Evaluated with main.IREvaluatorWithLogging

Metric	Value
cosine_accuracy@1	0.7584
cosine_accuracy@3	0.8959
cosine_accuracy@5	0.9331
cosine_accuracy@10	0.9591
cosine_precision@1	0.7584
cosine_precision@3	0.3656
cosine_precision@5	0.2335
cosine_precision@10	0.1223
cosine_recall@1	0.6247
cosine_recall@3	0.8255
cosine_recall@5	0.8721
cosine_recall@10	0.9062
cosine_ndcg@10	0.8269
cosine_mrr@10	0.8325
cosine_map@100	0.7885

Training Details

Training Dataset

Unnamed Dataset

• Size: 429 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 429 samples:

  	sentence_0	sentence_1	label
  type	string	string	int
  details	min: 40 tokens mean: 84.96 tokens max: 131 tokens	min: 11 tokens mean: 14.11 tokens max: 20 tokens	0: ~0.47% 1: ~0.47% 2: ~0.23% 3: ~0.47% 4: ~0.47% 5: ~0.23% 6: ~0.47% 7: ~0.23% 8: ~0.47% 9: ~0.23% 10: ~0.23% 11: ~0.23% 12: ~0.47% 13: ~0.47% 14: ~0.47% 15: ~0.23% 16: ~0.93% 17: ~0.23% 18: ~0.47% 19: ~0.47% 20: ~0.23% 21: ~0.23% 22: ~0.23% 23: ~0.47% 24: ~0.23% 25: ~0.23% 26: ~0.23% 27: ~0.47% 28: ~1.17% 29: ~0.47% 30: ~0.47% 31: ~0.47% 32: ~0.23% 33: ~0.23% 34: ~0.70% 35: ~0.23% 36: ~0.23% 37: ~0.23% 38: ~0.23% 39: ~0.70% 40: ~0.23% 41: ~0.70% 42: ~0.47% 43: ~0.23% 44: ~0.23% 45: ~0.23% 46: ~0.47% 47: ~0.23% 48: ~0.23% 49: ~0.47% 50: ~0.47% 51: ~0.23% 52: ~0.23% 53: ~0.23% 54: ~0.47% 55: ~0.47% 56: ~0.23% 57: ~0.23% 58: ~0.47% 59: ~0.23% 60: ~0.47% 61: ~0.23% 62: ~0.47% 63: ~0.47% 64: ~0.23% 65: ~0.23% 66: ~0.47% 67: ~0.47% 68: ~0.70% 69: ~0.47% 70: ~0.47% 71: ~0.23% 72: ~0.47% 73: ~0.47% 74: ~0.70% 75: ~0.23% 76: ~0.47% 77: ~0.70% 78: ~0.23% 79: ~0.70% 80: ~0.23% 81: ~0.23% 82: ~0.47% 83: ~0.23% 84: ~0.47% 85: ~0.47% 86: ~0.47% 87: ~0.47% 88: ~0.23% 89: ~0.23% 90: ~0.47% 91: ~0.23% 92: ~0.47% 93: ~0.23% 94: ~0.23% 95: ~0.47% 96: ~0.47% 97: ~0.23% 98: ~0.23% 99: ~0.23% 100: ~0.70% 101: ~0.47% 102: ~0.23% 103: ~0.47% 104: ~0.70% 105: ~0.23% 106: ~0.23% 107: ~0.23% 108: ~0.47% 109: ~0.23% 110: ~0.47% 111: ~0.23% 112: ~0.47% 113: ~0.23% 114: ~0.47% 115: ~0.23% 116: ~0.23% 117: ~0.23% 118: ~0.70% 119: ~0.47% 120: ~0.23% 121: ~0.23% 122: ~0.47% 123: ~0.70% 124: ~0.23% 125: ~0.47% 126: ~0.23% 127: ~0.23% 128: ~0.23% 129: ~0.47% 130: ~0.23% 131: ~0.70% 132: ~0.47% 133: ~0.23% 134: ~0.23% 135: ~0.47% 136: ~0.23% 137: ~0.23% 138: ~0.47% 139: ~0.23% 140: ~0.47% 141: ~0.23% 142: ~0.23% 143: ~0.47% 144: ~0.23% 145: ~0.70% 146: ~0.93% 147: ~0.47% 148: ~0.23% 149: ~0.47% 150: ~0.47% 151: ~0.47% 152: ~0.23% 153: ~0.47% 154: ~0.47% 155: ~0.23% 156: ~0.23% 157: ~0.47% 158: ~0.47% 159: ~0.23% 160: ~0.23% 161: ~0.70% 162: ~0.23% 163: ~0.23% 164: ~0.47% 165: ~0.47% 166: ~0.93% 167: ~0.23% 168: ~0.47% 169: ~0.70% 170: ~0.23% 171: ~0.23% 172: ~0.47% 173: ~0.23% 174: ~0.47% 175: ~0.70% 176: ~0.23% 177: ~0.23% 178: ~0.23% 179: ~0.47% 180: ~0.47% 181: ~0.47% 182: ~0.23% 183: ~0.23% 184: ~0.47% 185: ~0.23% 186: ~0.23% 187: ~0.70% 188: ~0.70% 189: ~0.23% 190: ~0.47% 191: ~0.23% 192: ~0.23% 193: ~0.70% 194: ~0.23% 195: ~0.23% 196: ~0.47% 197: ~0.23% 198: ~0.47% 199: ~0.47% 200: ~0.23% 201: ~0.23% 202: ~0.23% 203: ~0.47% 204: ~0.47% 205: ~0.23% 206: ~0.47% 207: ~0.23% 208: ~0.23% 209: ~0.47% 210: ~0.70% 211: ~0.47% 212: ~0.47% 213: ~0.47% 214: ~0.23% 215: ~0.23% 216: ~0.47% 217: ~0.47% 218: ~0.23% 219: ~0.23% 220: ~0.23% 221: ~0.23% 222: ~0.23% 223: ~0.70% 224: ~0.23% 225: ~0.47% 226: ~0.47% 227: ~0.23% 228: ~0.70% 229: ~0.47% 230: ~0.47% 231: ~0.23% 232: ~0.70% 233: ~0.70% 234: ~0.47% 235: ~0.23% 236: ~0.23% 237: ~0.23% 238: ~0.23% 239: ~0.47% 240: ~0.47% 241: ~0.23% 242: ~0.93% 243: ~0.47% 244: ~0.23% 245: ~0.70% 246: ~0.23% 247: ~0.70% 248: ~0.47% 249: ~0.23% 250: ~0.47% 251: ~0.23% 252: ~0.23% 253: ~0.23% 254: ~0.23% 255: ~0.47% 256: ~0.47% 257: ~0.70% 258: ~0.23% 259: ~0.23%

• Samples:

  sentence_0	sentence_1	label
  Word: h l a g ̲ o o k ̲| Context: I i h l a g ̲ o o k ̲ d i m h a ' w i ' y i i k y ' a a i s x w i ' y g ̲ o o h l w i l p x s e e k ̲ .| Translation: And before I went home I had a short pee in the bathroom.	Morpheme: g ̲ o o k ̲ | Gloss: first	90
  Word: x s a ' a k ̲ x w i ' y| Context: H l a a x s a ' a k ̲ x w i ' y ' n i i g ̲ a y o o t s ' i m i l t ' a a h l i h l j a b i ' y g ̲ o o h l t s ' i m w i l p x s e e k ̲ .| Translation: When I made it out, then I put what I had done (the rubble) back in the bathroom.	Morpheme: x s i | Gloss: out	228
  Word: n e e d i i| Context: I i ' n a k w h l ' w i h l w i l i ' m , g w i l a ' l h l g ̲ a n u u t x w , g ̲ a n w i h l n e e d i i l a x ̲ ' n i s x w i ' y g ̲ o o h l G i g e e n i x .| Translation: And we were away a long time, three weeks, and that's why I didn't hear from Gigeenix.	Morpheme: n e e | Gloss: NEG	67

• Loss: main.LossLogger

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 128
• per_device_eval_batch_size: 128
• num_train_epochs: 1000
• fp16: True
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 128
• per_device_eval_batch_size: 128
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1
• num_train_epochs: 1000
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• save_safetensors: True
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• no_cuda: False
• use_cpu: False
• use_mps_device: False
• seed: 42
• data_seed: None
• jit_mode_eval: False
• use_ipex: False
• bf16: False
• fp16: True
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: 0
• ddp_backend: None
• tpu_num_cores: None
• tpu_metrics_debug: False
• debug: []
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• past_index: -1
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: False
• ignore_data_skip: False
• fsdp: []
• fsdp_min_num_params: 0
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• fsdp_transformer_layer_cls_to_wrap: None
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• parallelism_config: None
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch_fused
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• skip_memory_metrics: True
• use_legacy_prediction_loop: False
• push_to_hub: False
• resume_from_checkpoint: None
• hub_model_id: None
• hub_strategy: every_save
• hub_private_repo: None
• hub_always_push: False
• hub_revision: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• eval_do_concat_batches: True
• fp16_backend: auto
• push_to_hub_model_id: None
• push_to_hub_organization: None
• mp_parameters:
• auto_find_batch_size: False
• full_determinism: False
• torchdynamo: None
• ray_scope: last
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• liger_kernel_config: None
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Click to expand

Epoch	Step	Training Loss	validation_cosine_ndcg@10
1.0	4	-	0.0647
2.0	8	-	0.0761
3.0	12	-	0.1236
4.0	16	-	0.2528
5.0	20	-	0.3902
6.0	24	-	0.4764
7.0	28	-	0.5325
8.0	32	-	0.6067
9.0	36	-	0.6709
10.0	40	-	0.7043
11.0	44	-	0.7018
12.0	48	-	0.6915
13.0	52	-	0.7073
14.0	56	-	0.7310
15.0	60	-	0.7335
16.0	64	-	0.7389
17.0	68	-	0.7586
18.0	72	-	0.7615
19.0	76	-	0.7586
20.0	80	-	0.7472
21.0	84	-	0.7588
22.0	88	-	0.7641
23.0	92	-	0.7740
24.0	96	-	0.7633
25.0	100	-	0.7721
26.0	104	-	0.7669
27.0	108	-	0.7728
28.0	112	-	0.7868
29.0	116	-	0.7735
30.0	120	-	0.7829
31.0	124	-	0.7937
32.0	128	-	0.7902
33.0	132	-	0.7656
34.0	136	-	0.7838
35.0	140	-	0.7821
36.0	144	-	0.7871
37.0	148	-	0.7869
38.0	152	-	0.7920
39.0	156	-	0.7905
40.0	160	-	0.7954
41.0	164	-	0.7966
42.0	168	-	0.7835
43.0	172	-	0.7800
44.0	176	-	0.8047
45.0	180	-	0.7990
46.0	184	-	0.7860
47.0	188	-	0.7891
48.0	192	-	0.7958
49.0	196	-	0.7813
50.0	200	-	0.7778
51.0	204	-	0.8001
52.0	208	-	0.7870
53.0	212	-	0.8027
54.0	216	-	0.7905
55.0	220	-	0.7827
56.0	224	-	0.8020
57.0	228	-	0.7919
58.0	232	-	0.7817
59.0	236	-	0.7994
60.0	240	-	0.8164
61.0	244	-	0.7788
62.0	248	-	0.7900
63.0	252	-	0.8173
64.0	256	-	0.7976
65.0	260	-	0.7905
66.0	264	-	0.7923
67.0	268	-	0.8071
68.0	272	-	0.7958
69.0	276	-	0.7871
70.0	280	-	0.8020
71.0	284	-	0.8103
72.0	288	-	0.8123
73.0	292	-	0.8118
74.0	296	-	0.7934
75.0	300	-	0.7882
76.0	304	-	0.8015
77.0	308	-	0.8201
78.0	312	-	0.8240
79.0	316	-	0.7994
80.0	320	-	0.8042
81.0	324	-	0.8114
82.0	328	-	0.8100
83.0	332	-	0.8041
84.0	336	-	0.8179
85.0	340	-	0.8197
86.0	344	-	0.7973
87.0	348	-	0.7985
88.0	352	-	0.8123
89.0	356	-	0.7997
90.0	360	-	0.8043
91.0	364	-	0.8057
92.0	368	-	0.7991
93.0	372	-	0.7983
94.0	376	-	0.8052
95.0	380	-	0.8026
96.0	384	-	0.8109
97.0	388	-	0.7929
98.0	392	-	0.8025
99.0	396	-	0.8218
100.0	400	-	0.8194
101.0	404	-	0.8023
102.0	408	-	0.8099
103.0	412	-	0.8110
104.0	416	-	0.8118
105.0	420	-	0.8004
106.0	424	-	0.8012
107.0	428	-	0.8070
108.0	432	-	0.8088
109.0	436	-	0.8073
110.0	440	-	0.8084
111.0	444	-	0.8038
112.0	448	-	0.8115
113.0	452	-	0.8169
114.0	456	-	0.8145
115.0	460	-	0.8020
116.0	464	-	0.7984
117.0	468	-	0.8077
118.0	472	-	0.8174
119.0	476	-	0.8200
120.0	480	-	0.8080
121.0	484	-	0.8093
122.0	488	-	0.8216
123.0	492	-	0.8240
124.0	496	-	0.8097
125.0	500	1.3195	0.8115
126.0	504	-	0.8176
127.0	508	-	0.8099
128.0	512	-	0.7977
129.0	516	-	0.7985
130.0	520	-	0.8015
131.0	524	-	0.8078
132.0	528	-	0.7985
133.0	532	-	0.8029
134.0	536	-	0.8087
135.0	540	-	0.8031
136.0	544	-	0.7999
137.0	548	-	0.8107
138.0	552	-	0.8110
139.0	556	-	0.7980
140.0	560	-	0.7977
141.0	564	-	0.8034
142.0	568	-	0.8053
143.0	572	-	0.7996
144.0	576	-	0.8014
145.0	580	-	0.8137
146.0	584	-	0.8221
147.0	588	-	0.8144
148.0	592	-	0.8020
149.0	596	-	0.7987
150.0	600	-	0.8029
151.0	604	-	0.8024
152.0	608	-	0.8045
153.0	612	-	0.8061
154.0	616	-	0.8035
155.0	620	-	0.8018
156.0	624	-	0.7992
157.0	628	-	0.8053
158.0	632	-	0.8134
159.0	636	-	0.8173
160.0	640	-	0.8118
161.0	644	-	0.8144
162.0	648	-	0.8145
163.0	652	-	0.8105
164.0	656	-	0.8011
165.0	660	-	0.8073
166.0	664	-	0.8111
167.0	668	-	0.8139
168.0	672	-	0.8030
169.0	676	-	0.8035
170.0	680	-	0.7993
171.0	684	-	0.8023
172.0	688	-	0.8081
173.0	692	-	0.8097
174.0	696	-	0.8060
175.0	700	-	0.8063
176.0	704	-	0.8114
177.0	708	-	0.8087
178.0	712	-	0.8090
179.0	716	-	0.8094
180.0	720	-	0.8071
181.0	724	-	0.8077
182.0	728	-	0.8108
183.0	732	-	0.8160
184.0	736	-	0.8099
185.0	740	-	0.8052
186.0	744	-	0.8103
187.0	748	-	0.8115
188.0	752	-	0.8105
189.0	756	-	0.8057
190.0	760	-	0.8157
191.0	764	-	0.8096
192.0	768	-	0.7998
193.0	772	-	0.8080
194.0	776	-	0.8207
195.0	780	-	0.8136
196.0	784	-	0.8029
197.0	788	-	0.8009
198.0	792	-	0.8150
199.0	796	-	0.8173
200.0	800	-	0.8070
201.0	804	-	0.8075
202.0	808	-	0.8164
203.0	812	-	0.8148
204.0	816	-	0.8077
205.0	820	-	0.8116
206.0	824	-	0.8148
207.0	828	-	0.8141
208.0	832	-	0.8085
209.0	836	-	0.8066
210.0	840	-	0.8154
211.0	844	-	0.8168
212.0	848	-	0.8132
213.0	852	-	0.8139
214.0	856	-	0.8200
215.0	860	-	0.8203
216.0	864	-	0.8100
217.0	868	-	0.8084
218.0	872	-	0.8115
219.0	876	-	0.8126
220.0	880	-	0.8126
221.0	884	-	0.8079
222.0	888	-	0.8101
223.0	892	-	0.8136
224.0	896	-	0.8124
225.0	900	-	0.8180
226.0	904	-	0.8173
227.0	908	-	0.8110
228.0	912	-	0.7991
229.0	916	-	0.8009
230.0	920	-	0.8096
231.0	924	-	0.8153
232.0	928	-	0.8177
233.0	932	-	0.8107
234.0	936	-	0.8066
235.0	940	-	0.8067
236.0	944	-	0.8198
237.0	948	-	0.8175
238.0	952	-	0.8077
239.0	956	-	0.8099
240.0	960	-	0.8073
241.0	964	-	0.8117
242.0	968	-	0.8148
243.0	972	-	0.8144
244.0	976	-	0.8050
245.0	980	-	0.8046
246.0	984	-	0.8107
247.0	988	-	0.8114
248.0	992	-	0.8065
249.0	996	-	0.8071
250.0	1000	1.081	0.8105
251.0	1004	-	0.8142
252.0	1008	-	0.8123
253.0	1012	-	0.8123
254.0	1016	-	0.8104
255.0	1020	-	0.8168
256.0	1024	-	0.8171
257.0	1028	-	0.8188
258.0	1032	-	0.8210
259.0	1036	-	0.8221
260.0	1040	-	0.8156
261.0	1044	-	0.8118
262.0	1048	-	0.8078
263.0	1052	-	0.8108
264.0	1056	-	0.8121
265.0	1060	-	0.8146
266.0	1064	-	0.8116
267.0	1068	-	0.8149
268.0	1072	-	0.8122
269.0	1076	-	0.8125
270.0	1080	-	0.8114
271.0	1084	-	0.8139
272.0	1088	-	0.8240
273.0	1092	-	0.8240
274.0	1096	-	0.8196
275.0	1100	-	0.8233
276.0	1104	-	0.8228
277.0	1108	-	0.8165
278.0	1112	-	0.8183
279.0	1116	-	0.8217
280.0	1120	-	0.8166
281.0	1124	-	0.8106
282.0	1128	-	0.8117
283.0	1132	-	0.8152
284.0	1136	-	0.8222
285.0	1140	-	0.8230
286.0	1144	-	0.8123
287.0	1148	-	0.8080
288.0	1152	-	0.8125
289.0	1156	-	0.8192
290.0	1160	-	0.8267
291.0	1164	-	0.8232
292.0	1168	-	0.8086
293.0	1172	-	0.8081
294.0	1176	-	0.8215
295.0	1180	-	0.8211
296.0	1184	-	0.8147
297.0	1188	-	0.8107
298.0	1192	-	0.8123
299.0	1196	-	0.8113
300.0	1200	-	0.8161
301.0	1204	-	0.8161
302.0	1208	-	0.8181
303.0	1212	-	0.8167
304.0	1216	-	0.8167
305.0	1220	-	0.8257
306.0	1224	-	0.8269

Framework Versions

• Python: 3.11.4
• Sentence Transformers: 5.1.1
• Transformers: 4.56.2
• PyTorch: 2.8.0+cu128
• Accelerate: 1.10.1
• Datasets: 4.1.1
• Tokenizers: 0.22.1

Citation

BibTeX

Sentence Transformers

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