dkqjrm/bge-m3-embedding-augmented-olive-lora-sentence

SentenceTransformer based on unsloth/bge-m3

This is a sentence-transformers model finetuned from unsloth/bge-m3 on the augmented-olive-product-sentence dataset. It maps sentences & paragraphs to a 1024-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: unsloth/bge-m3
• Maximum Sequence Length: 8192 tokens
• Output Dimensionality: 1024 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • augmented-olive-product-sentence

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': 8192, 'do_lower_case': False, 'architecture': 'PeftModelForFeatureExtraction'})
  (1): Pooling({'word_embedding_dimension': 1024, '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): 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("dkqjrm/bge-m3-embedding-augmented-olive-lora-sentence")
# Run inference
sentences = [
    '유리아쥬 [3중장벽강화 미스트] 유리아쥬 오떼르말 50ml(N)',
    'ユリアージュミスト',
    'ギャツビー ワックス',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000,  0.6145, -0.0901],
#         [ 0.6145,  1.0000,  0.0137],
#         [-0.0901,  0.0137,  1.0000]])

Training Details

Training Dataset

augmented-olive-product-sentence

• Dataset: augmented-olive-product-sentence at b513dad
• Size: 3,682,898 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 14 tokens mean: 20.32 tokens max: 41 tokens	min: 3 tokens mean: 8.76 tokens max: 35 tokens

• Samples:

  anchor	positive
  베르사체 베르사체 브라이트 크리스탈 50ml 택1	베르사체 브라이트 크리스탈 50ml 1
  베르사체 베르사체 브라이트 크리스탈 50ml 택1	베르사체 브라이트
  베르사체 베르사체 브라이트 크리스탈 50ml 택1	베르사체 크리스탈

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "gather_across_devices": false
  }
  

Evaluation Dataset

augmented-olive-product-sentence

• Dataset: augmented-olive-product-sentence at b513dad
• Size: 193,728 evaluation samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 12 tokens mean: 21.91 tokens max: 31 tokens	min: 3 tokens mean: 9.19 tokens max: 28 tokens

• Samples:

  anchor	positive
  랑방 랑방 루머 2 로즈 50ml	랑방 루머 2 로즈 50ml
  랑방 랑방 루머 2 로즈 50ml	랑방 향수
  랑방 랑방 루머 2 로즈 50ml	랑방 루머

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "gather_across_devices": false
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 16
• gradient_accumulation_steps: 16
• learning_rate: 3e-05
• num_train_epochs: 2
• lr_scheduler_type: cosine
• warmup_ratio: 0.1
• fp16: True
• push_to_hub: True
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 8
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 16
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 3e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 2
• max_steps: -1
• lr_scheduler_type: cosine
• lr_scheduler_kwargs: None
• warmup_ratio: 0.1
• 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
• 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
• project: huggingface
• trackio_space_id: trackio
• 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: True
• 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: no
• 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: True
• prompts: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Click to expand

Epoch	Step	Training Loss	Validation Loss
0.0035	50	1.0635	-
0.0070	100	0.9772	-
0.0104	150	0.87	-
0.0139	200	0.7313	-
0.0174	250	0.5824	-
0.0209	300	0.4781	-
0.0243	350	0.4168	-
0.0278	400	0.386	-
0.0313	450	0.3552	-
0.0348	500	0.3387	-
0.0382	550	0.321	-
0.0417	600	0.3036	-
0.0452	650	0.2955	-
0.0487	700	0.2779	-
0.0521	750	0.2647	-
0.0556	800	0.2498	-
0.0591	850	0.2489	-
0.0626	900	0.2434	-
0.0660	950	0.2238	-
0.0695	1000	0.2243	-
0.0730	1050	0.2119	-
0.0765	1100	0.2211	-
0.0799	1150	0.1966	-
0.0834	1200	0.1942	-
0.0869	1250	0.1928	-
0.0904	1300	0.1849	-
0.0938	1350	0.1823	-
0.0973	1400	0.1743	-
0.1008	1450	0.1677	-
0.1043	1500	0.1684	-
0.1077	1550	0.1686	-
0.1112	1600	0.1603	-
0.1147	1650	0.1685	-
0.1182	1700	0.1537	-
0.1216	1750	0.1539	-
0.1251	1800	0.15	-
0.1286	1850	0.1477	-
0.1321	1900	0.1516	-
0.1355	1950	0.1466	-
0.1390	2000	0.1434	-
0.1425	2050	0.1361	-
0.1460	2100	0.1438	-
0.1494	2150	0.1359	-
0.1529	2200	0.1409	-
0.1564	2250	0.1329	-
0.1599	2300	0.1376	-
0.1633	2350	0.1331	-
0.1668	2400	0.1265	-
0.1703	2450	0.1177	-
0.1738	2500	0.1273	-
0.1773	2550	0.1217	-
0.1807	2600	0.1205	-
0.1842	2650	0.1176	-
0.1877	2700	0.1208	-
0.1912	2750	0.1155	-
0.1946	2800	0.1185	-
0.1981	2850	0.1169	-
0.2016	2900	0.1179	-
0.2051	2950	0.114	-
0.2085	3000	0.1148	0.0420
0.2120	3050	0.1074	-
0.2155	3100	0.1108	-
0.2190	3150	0.1084	-
0.2224	3200	0.1132	-
0.2259	3250	0.1082	-
0.2294	3300	0.1065	-
0.2329	3350	0.1042	-
0.2363	3400	0.1052	-
0.2398	3450	0.1054	-
0.2433	3500	0.1006	-
0.2468	3550	0.0967	-
0.2502	3600	0.1005	-
0.2537	3650	0.0985	-
0.2572	3700	0.0997	-
0.2607	3750	0.0963	-
0.2641	3800	0.096	-
0.2676	3850	0.0928	-
0.2711	3900	0.0903	-
0.2746	3950	0.0925	-
0.2780	4000	0.0946	-
0.2815	4050	0.0981	-
0.2850	4100	0.0866	-
0.2885	4150	0.0889	-
0.2919	4200	0.0899	-
0.2954	4250	0.0958	-
0.2989	4300	0.0888	-
0.3024	4350	0.0876	-
0.3058	4400	0.0859	-
0.3093	4450	0.0857	-
0.3128	4500	0.0868	-
0.3163	4550	0.0841	-
0.3197	4600	0.0853	-
0.3232	4650	0.0808	-
0.3267	4700	0.0812	-
0.3302	4750	0.0797	-
0.3336	4800	0.0834	-
0.3371	4850	0.0791	-
0.3406	4900	0.0799	-
0.3441	4950	0.0754	-
0.3476	5000	0.0814	-
0.3510	5050	0.08	-
0.3545	5100	0.0735	-
0.3580	5150	0.0791	-
0.3615	5200	0.077	-
0.3649	5250	0.0745	-
0.3684	5300	0.0738	-
0.3719	5350	0.0786	-
0.3754	5400	0.0762	-
0.3788	5450	0.0736	-
0.3823	5500	0.0793	-
0.3858	5550	0.0726	-
0.3893	5600	0.0728	-
0.3927	5650	0.0748	-
0.3962	5700	0.0734	-
0.3997	5750	0.0698	-
0.4032	5800	0.073	-
0.4066	5850	0.0719	-
0.4101	5900	0.0735	-
0.4136	5950	0.0671	-
0.4171	6000	0.0689	0.0275
0.4205	6050	0.0713	-
0.4240	6100	0.0707	-
0.4275	6150	0.0631	-
0.4310	6200	0.0691	-
0.4344	6250	0.065	-
0.4379	6300	0.0681	-
0.4414	6350	0.0695	-
0.4449	6400	0.0678	-
0.4483	6450	0.0648	-
0.4518	6500	0.0662	-
0.4553	6550	0.0691	-
0.4588	6600	0.0689	-
0.4622	6650	0.0685	-
0.4657	6700	0.0709	-
0.4692	6750	0.0652	-
0.4727	6800	0.0655	-
0.4761	6850	0.065	-
0.4796	6900	0.0682	-
0.4831	6950	0.0681	-
0.4866	7000	0.0635	-
0.4900	7050	0.0641	-
0.4935	7100	0.0636	-
0.4970	7150	0.0657	-
0.5005	7200	0.0627	-
0.5039	7250	0.0663	-
0.5074	7300	0.0638	-
0.5109	7350	0.06	-
0.5144	7400	0.06	-
0.5179	7450	0.0612	-
0.5213	7500	0.06	-
0.5248	7550	0.0588	-
0.5283	7600	0.0612	-
0.5318	7650	0.0606	-
0.5352	7700	0.0627	-
0.5387	7750	0.0612	-
0.5422	7800	0.0624	-
0.5457	7850	0.059	-
0.5491	7900	0.0617	-
0.5526	7950	0.0573	-
0.5561	8000	0.0583	-
0.5596	8050	0.0577	-
0.5630	8100	0.0577	-
0.5665	8150	0.0628	-
0.5700	8200	0.058	-
0.5735	8250	0.06	-
0.5769	8300	0.0593	-
0.5804	8350	0.0594	-
0.5839	8400	0.056	-
0.5874	8450	0.0543	-
0.5908	8500	0.0568	-
0.5943	8550	0.0516	-
0.5978	8600	0.0566	-
0.6013	8650	0.0583	-
0.6047	8700	0.0581	-
0.6082	8750	0.0566	-
0.6117	8800	0.0535	-
0.6152	8850	0.0571	-
0.6186	8900	0.055	-
0.6221	8950	0.0528	-
0.6256	9000	0.0531	0.0217
0.6291	9050	0.0549	-
0.6325	9100	0.0528	-
0.6360	9150	0.0579	-
0.6395	9200	0.053	-
0.6430	9250	0.052	-
0.6464	9300	0.056	-
0.6499	9350	0.0605	-
0.6534	9400	0.0542	-
0.6569	9450	0.0516	-
0.6603	9500	0.0541	-
0.6638	9550	0.054	-
0.6673	9600	0.0518	-
0.6708	9650	0.0517	-
0.6742	9700	0.0507	-
0.6777	9750	0.0526	-
0.6812	9800	0.0492	-
0.6847	9850	0.0543	-
0.6882	9900	0.0503	-
0.6916	9950	0.0515	-
0.6951	10000	0.0516	-
0.6986	10050	0.0499	-
0.7021	10100	0.0544	-
0.7055	10150	0.0497	-
0.7090	10200	0.0564	-
0.7125	10250	0.0533	-
0.7160	10300	0.0502	-
0.7194	10350	0.0528	-
0.7229	10400	0.0507	-
0.7264	10450	0.0518	-
0.7299	10500	0.0493	-
0.7333	10550	0.0518	-
0.7368	10600	0.0524	-
0.7403	10650	0.0515	-
0.7438	10700	0.0504	-
0.7472	10750	0.0509	-
0.7507	10800	0.0495	-
0.7542	10850	0.0532	-
0.7577	10900	0.048	-
0.7611	10950	0.0511	-
0.7646	11000	0.0511	-
0.7681	11050	0.0465	-
0.7716	11100	0.0447	-
0.7750	11150	0.0477	-
0.7785	11200	0.0497	-
0.7820	11250	0.0488	-
0.7855	11300	0.0469	-
0.7889	11350	0.0502	-
0.7924	11400	0.0478	-
0.7959	11450	0.0479	-
0.7994	11500	0.049	-
0.8028	11550	0.0452	-
0.8063	11600	0.0484	-
0.8098	11650	0.047	-
0.8133	11700	0.0464	-
0.8167	11750	0.0436	-
0.8202	11800	0.0452	-
0.8237	11850	0.0475	-
0.8272	11900	0.0477	-
0.8306	11950	0.047	-
0.8341	12000	0.0444	0.0197
0.8376	12050	0.043	-
0.8411	12100	0.0478	-
0.8445	12150	0.0443	-
0.8480	12200	0.0463	-
0.8515	12250	0.0456	-
0.8550	12300	0.0449	-
0.8585	12350	0.0489	-
0.8619	12400	0.0451	-
0.8654	12450	0.044	-
0.8689	12500	0.0453	-
0.8724	12550	0.0434	-
0.8758	12600	0.045	-
0.8793	12650	0.0452	-
0.8828	12700	0.0423	-
0.8863	12750	0.0446	-
0.8897	12800	0.045	-
0.8932	12850	0.0466	-
0.8967	12900	0.0448	-
0.9002	12950	0.0475	-
0.9036	13000	0.0443	-
0.9071	13050	0.0457	-
0.9106	13100	0.0463	-
0.9141	13150	0.043	-
0.9175	13200	0.0435	-
0.9210	13250	0.0425	-
0.9245	13300	0.0451	-
0.9280	13350	0.0447	-
0.9314	13400	0.043	-
0.9349	13450	0.0431	-
0.9384	13500	0.0454	-
0.9419	13550	0.0484	-
0.9453	13600	0.0453	-
0.9488	13650	0.0444	-
0.9523	13700	0.0438	-
0.9558	13750	0.0415	-
0.9592	13800	0.0438	-
0.9627	13850	0.044	-
0.9662	13900	0.0433	-
0.9697	13950	0.0439	-
0.9731	14000	0.0428	-
0.9766	14050	0.0423	-
0.9801	14100	0.0419	-
0.9836	14150	0.0443	-
0.9870	14200	0.0406	-
0.9905	14250	0.0422	-
0.9940	14300	0.0414	-
0.9975	14350	0.0438	-
1.0009	14400	0.042	-
1.0044	14450	0.0404	-
1.0079	14500	0.0429	-
1.0113	14550	0.0395	-
1.0148	14600	0.0402	-
1.0183	14650	0.0403	-
1.0218	14700	0.0413	-
1.0252	14750	0.0399	-
1.0287	14800	0.0426	-
1.0322	14850	0.0384	-
1.0357	14900	0.0387	-
1.0391	14950	0.0383	-
1.0426	15000	0.0436	0.0183
1.0461	15050	0.039	-
1.0496	15100	0.0415	-
1.0530	15150	0.0394	-
1.0565	15200	0.0375	-
1.0600	15250	0.0399	-
1.0635	15300	0.0379	-
1.0669	15350	0.0413	-
1.0704	15400	0.0373	-
1.0739	15450	0.0411	-
1.0774	15500	0.0449	-
1.0808	15550	0.0392	-
1.0843	15600	0.0389	-
1.0878	15650	0.0387	-
1.0913	15700	0.0394	-
1.0947	15750	0.0383	-
1.0982	15800	0.0435	-
1.1017	15850	0.0382	-
1.1052	15900	0.0429	-
1.1086	15950	0.0366	-
1.1121	16000	0.0404	-
1.1156	16050	0.0431	-
1.1191	16100	0.0382	-
1.1225	16150	0.0376	-
1.1260	16200	0.0385	-
1.1295	16250	0.0406	-
1.1330	16300	0.0387	-
1.1364	16350	0.0376	-
1.1399	16400	0.0375	-
1.1434	16450	0.0412	-
1.1469	16500	0.0406	-
1.1504	16550	0.0383	-
1.1538	16600	0.039	-
1.1573	16650	0.0381	-
1.1608	16700	0.039	-
1.1643	16750	0.04	-
1.1677	16800	0.0385	-
1.1712	16850	0.0377	-
1.1747	16900	0.0374	-
1.1782	16950	0.0394	-
1.1816	17000	0.0383	-
1.1851	17050	0.0384	-
1.1886	17100	0.0392	-
1.1921	17150	0.0386	-
1.1955	17200	0.0368	-
1.1990	17250	0.037	-
1.2025	17300	0.035	-
1.2060	17350	0.038	-
1.2094	17400	0.0354	-
1.2129	17450	0.0385	-
1.2164	17500	0.0388	-
1.2199	17550	0.0424	-
1.2233	17600	0.0435	-
1.2268	17650	0.036	-
1.2303	17700	0.0381	-
1.2338	17750	0.0358	-
1.2372	17800	0.0369	-
1.2407	17850	0.0385	-
1.2442	17900	0.0368	-
1.2477	17950	0.0355	-
1.2511	18000	0.0419	0.0166
1.2546	18050	0.0369	-
1.2581	18100	0.0362	-
1.2616	18150	0.0365	-
1.2650	18200	0.0369	-
1.2685	18250	0.0382	-
1.2720	18300	0.0394	-
1.2755	18350	0.0371	-
1.2789	18400	0.0358	-
1.2824	18450	0.0376	-
1.2859	18500	0.0362	-
1.2894	18550	0.0368	-
1.2928	18600	0.0371	-
1.2963	18650	0.0374	-
1.2998	18700	0.0378	-
1.3033	18750	0.0372	-
1.3067	18800	0.0382	-
1.3102	18850	0.037	-
1.3137	18900	0.0366	-
1.3172	18950	0.0369	-
1.3207	19000	0.0347	-
1.3241	19050	0.0379	-
1.3276	19100	0.0369	-
1.3311	19150	0.0364	-
1.3346	19200	0.0356	-
1.3380	19250	0.0361	-
1.3415	19300	0.0392	-
1.3450	19350	0.035	-
1.3485	19400	0.0349	-
1.3519	19450	0.0359	-
1.3554	19500	0.0373	-
1.3589	19550	0.0386	-
1.3624	19600	0.0353	-
1.3658	19650	0.0359	-
1.3693	19700	0.0382	-
1.3728	19750	0.0379	-
1.3763	19800	0.0353	-
1.3797	19850	0.0367	-
1.3832	19900	0.0346	-
1.3867	19950	0.0336	-
1.3902	20000	0.0341	-
1.3936	20050	0.0388	-
1.3971	20100	0.0329	-
1.4006	20150	0.0334	-
1.4041	20200	0.0354	-
1.4075	20250	0.0338	-
1.4110	20300	0.0344	-
1.4145	20350	0.0362	-
1.4180	20400	0.0357	-
1.4214	20450	0.0372	-
1.4249	20500	0.0329	-
1.4284	20550	0.0389	-
1.4319	20600	0.0359	-
1.4353	20650	0.0377	-
1.4388	20700	0.0343	-
1.4423	20750	0.0381	-
1.4458	20800	0.0351	-
1.4492	20850	0.0381	-
1.4527	20900	0.0377	-
1.4562	20950	0.0384	-
1.4597	21000	0.0332	0.0164
1.4631	21050	0.0377	-
1.4666	21100	0.0351	-
1.4701	21150	0.037	-
1.4736	21200	0.0327	-
1.4770	21250	0.0364	-
1.4805	21300	0.0369	-
1.4840	21350	0.0344	-
1.4875	21400	0.0352	-
1.4910	21450	0.0371	-
1.4944	21500	0.0364	-
1.4979	21550	0.0332	-
1.5014	21600	0.0367	-
1.5049	21650	0.0355	-
1.5083	21700	0.0364	-
1.5118	21750	0.0361	-
1.5153	21800	0.035	-
1.5188	21850	0.036	-
1.5222	21900	0.0313	-
1.5257	21950	0.0334	-
1.5292	22000	0.0319	-
1.5327	22050	0.0358	-
1.5361	22100	0.0323	-
1.5396	22150	0.032	-
1.5431	22200	0.034	-
1.5466	22250	0.0344	-
1.5500	22300	0.0382	-
1.5535	22350	0.0353	-
1.5570	22400	0.0335	-
1.5605	22450	0.0318	-
1.5639	22500	0.0335	-
1.5674	22550	0.0346	-
1.5709	22600	0.0357	-
1.5744	22650	0.0342	-
1.5778	22700	0.0358	-
1.5813	22750	0.0338	-
1.5848	22800	0.0334	-
1.5883	22850	0.0374	-
1.5917	22900	0.0335	-
1.5952	22950	0.0321	-
1.5987	23000	0.0351	-
1.6022	23050	0.0322	-
1.6056	23100	0.0319	-
1.6091	23150	0.0357	-
1.6126	23200	0.0341	-
1.6161	23250	0.0319	-
1.6195	23300	0.0356	-
1.6230	23350	0.0333	-
1.6265	23400	0.0348	-
1.6300	23450	0.0341	-
1.6334	23500	0.0339	-
1.6369	23550	0.0362	-
1.6404	23600	0.033	-
1.6439	23650	0.0334	-
1.6473	23700	0.0351	-
1.6508	23750	0.0339	-
1.6543	23800	0.0324	-
1.6578	23850	0.0342	-
1.6613	23900	0.0316	-
1.6647	23950	0.0373	-
1.6682	24000	0.0336	0.0157
1.6717	24050	0.0334	-
1.6752	24100	0.035	-
1.6786	24150	0.0354	-
1.6821	24200	0.0348	-
1.6856	24250	0.0347	-
1.6891	24300	0.0338	-
1.6925	24350	0.0363	-
1.6960	24400	0.0321	-
1.6995	24450	0.0326	-
1.7030	24500	0.0365	-
1.7064	24550	0.0335	-
1.7099	24600	0.0363	-
1.7134	24650	0.032	-
1.7169	24700	0.0323	-
1.7203	24750	0.0332	-
1.7238	24800	0.034	-
1.7273	24850	0.0354	-
1.7308	24900	0.0335	-
1.7342	24950	0.0342	-
1.7377	25000	0.0355	-
1.7412	25050	0.0371	-
1.7447	25100	0.0349	-
1.7481	25150	0.0325	-
1.7516	25200	0.0334	-
1.7551	25250	0.0342	-
1.7586	25300	0.0327	-
1.7620	25350	0.0324	-
1.7655	25400	0.0323	-
1.7690	25450	0.0329	-
1.7725	25500	0.0329	-
1.7759	25550	0.0316	-
1.7794	25600	0.0331	-
1.7829	25650	0.0347	-
1.7864	25700	0.0321	-
1.7898	25750	0.0343	-
1.7933	25800	0.0324	-
1.7968	25850	0.0325	-
1.8003	25900	0.0348	-
1.8037	25950	0.0336	-
1.8072	26000	0.0345	-
1.8107	26050	0.0347	-
1.8142	26100	0.0329	-
1.8176	26150	0.0346	-
1.8211	26200	0.0343	-
1.8246	26250	0.033	-
1.8281	26300	0.0343	-
1.8316	26350	0.0354	-
1.8350	26400	0.0357	-
1.8385	26450	0.0339	-
1.8420	26500	0.0338	-
1.8455	26550	0.0344	-
1.8489	26600	0.0328	-
1.8524	26650	0.0333	-
1.8559	26700	0.0316	-
1.8594	26750	0.0332	-
1.8628	26800	0.0355	-
1.8663	26850	0.0325	-
1.8698	26900	0.0347	-
1.8733	26950	0.0366	-
1.8767	27000	0.0317	0.0155
1.8802	27050	0.0346	-
1.8837	27100	0.0313	-
1.8872	27150	0.0365	-
1.8906	27200	0.0313	-
1.8941	27250	0.0315	-
1.8976	27300	0.0336	-
1.9011	27350	0.0326	-
1.9045	27400	0.033	-
1.9080	27450	0.0343	-
1.9115	27500	0.0346	-
1.9150	27550	0.0353	-
1.9184	27600	0.0322	-
1.9219	27650	0.0351	-
1.9254	27700	0.0337	-
1.9289	27750	0.0334	-
1.9323	27800	0.035	-
1.9358	27850	0.0314	-
1.9393	27900	0.0331	-
1.9428	27950	0.0333	-
1.9462	28000	0.0358	-
1.9497	28050	0.0366	-
1.9532	28100	0.035	-
1.9567	28150	0.037	-
1.9601	28200	0.035	-
1.9636	28250	0.0311	-
1.9671	28300	0.0333	-
1.9706	28350	0.0326	-
1.9740	28400	0.0323	-
1.9775	28450	0.0331	-
1.9810	28500	0.0336	-
1.9845	28550	0.0309	-
1.9879	28600	0.0354	-
1.9914	28650	0.0339	-
1.9949	28700	0.033	-
1.9984	28750	0.0323	-

Framework Versions

• Python: 3.12.12
• Sentence Transformers: 5.2.1
• Transformers: 4.57.6
• PyTorch: 2.10.0+cu128
• Accelerate: 1.12.0
• Datasets: 4.3.0
• Tokenizers: 0.22.2

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",
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    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},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}