sobamchan/roberta-large-mrl-768-512-256-128-64

SentenceTransformer based on FacebookAI/roberta-large

This is a sentence-transformers model finetuned from FacebookAI/roberta-large on the all-nli 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: FacebookAI/roberta-large
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 1024 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • all-nli
• Language: en

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': 512, 'do_lower_case': False, 'architecture': 'RobertaModel'})
  (1): Pooling({'word_embedding_dimension': 1024, '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})
)

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 = [
    'A construction worker peeking out of a manhole while his coworker sits on the sidewalk smiling.',
    'A worker is looking out of a manhole.',
    'The workers are both inside the manhole.',
]
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, 0.6579, 0.3481],
#         [0.6579, 1.0000, 0.5411],
#         [0.3481, 0.5411, 1.0000]])

Evaluation

Metrics

Semantic Similarity

• Datasets: sts-dev and sts-test
• Evaluated with EmbeddingSimilarityEvaluator

Metric	sts-dev	sts-test
pearson_cosine	0.7745	0.7441
spearman_cosine	0.7806	0.7525

Training Details

Training Dataset

all-nli

• Dataset: all-nli at d482672
• Size: 557,850 training samples
• Columns: anchor, positive, and negative
• Approximate statistics based on the first 1000 samples:

  	anchor	positive	negative
  type	string	string	string
  details	min: 7 tokens mean: 10.38 tokens max: 45 tokens	min: 6 tokens mean: 12.8 tokens max: 39 tokens	min: 6 tokens mean: 13.4 tokens max: 50 tokens

• Samples:

  anchor	positive	negative
  A person on a horse jumps over a broken down airplane.	A person is outdoors, on a horse.	A person is at a diner, ordering an omelette.
  Children smiling and waving at camera	There are children present	The kids are frowning
  A boy is jumping on skateboard in the middle of a red bridge.	The boy does a skateboarding trick.	The boy skates down the sidewalk.

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Evaluation Dataset

all-nli

• Dataset: all-nli at d482672
• Size: 6,584 evaluation samples
• Columns: anchor, positive, and negative
• Approximate statistics based on the first 1000 samples:

  	anchor	positive	negative
  type	string	string	string
  details	min: 6 tokens mean: 18.02 tokens max: 66 tokens	min: 5 tokens mean: 9.81 tokens max: 29 tokens	min: 5 tokens mean: 10.37 tokens max: 29 tokens

• Samples:

  anchor	positive	negative
  Two women are embracing while holding to go packages.	Two woman are holding packages.	The men are fighting outside a deli.
  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.	Two kids in numbered jerseys wash their hands.	Two kids in jackets walk to school.
  A man selling donuts to a customer during a world exhibition event held in the city of Angeles	A man selling donuts to a customer.	A woman drinks her coffee in a small cafe.

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• num_train_epochs: 15
• warmup_ratio: 0.1

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• 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: 1.0
• num_train_epochs: 15
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• 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: False
• 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: 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: 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: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Click to expand

Epoch	Step	Training Loss	Validation Loss	sts-dev_spearman_cosine	sts-test_spearman_cosine
-1	-1	-	-	0.5730	-
0.0287	500	10.2711	3.3325	0.8185	-
0.0574	1000	4.1395	1.8744	0.8648	-
0.0860	1500	3.2579	1.5506	0.8684	-
0.1147	2000	2.9091	1.3770	0.8712	-
0.1434	2500	2.6568	1.3085	0.8713	-
0.1721	3000	2.4983	1.2535	0.8814	-
0.2008	3500	2.3645	1.1715	0.8724	-
0.2294	4000	2.2675	1.1691	0.8724	-
0.2581	4500	2.2197	1.1681	0.8793	-
0.2868	5000	2.1473	1.1175	0.8780	-
0.3155	5500	2.0275	1.0656	0.8779	-
0.3442	6000	2.0927	1.0879	0.8791	-
0.3729	6500	2.1037	1.0664	0.8702	-
0.4015	7000	2.0268	1.0565	0.8714	-
0.4302	7500	1.9499	1.1268	0.8660	-
0.4589	8000	1.8692	1.0884	0.8712	-
0.4876	8500	1.9749	1.1014	0.8681	-
0.5163	9000	1.9716	1.0833	0.8555	-
0.5449	9500	1.9051	1.1391	0.8690	-
0.5736	10000	1.8493	1.1056	0.8708	-
0.6023	10500	2.0184	1.0989	0.8686	-
0.6310	11000	1.7868	1.1040	0.8620	-
0.6597	11500	1.8045	1.0677	0.8643	-
0.6883	12000	1.7946	1.0693	0.8624	-
0.7170	12500	1.8075	1.1100	0.8651	-
0.7457	13000	1.7923	1.1331	0.8577	-
0.7744	13500	1.7866	1.1204	0.8552	-
0.8031	14000	1.7495	1.1104	0.8542	-
0.8318	14500	1.7729	1.1599	0.8647	-
0.8604	15000	1.7413	1.0973	0.8587	-
0.8891	15500	1.7937	1.1443	0.8572	-
0.9178	16000	1.7489	1.1553	0.8566	-
0.9465	16500	1.7695	1.1249	0.8518	-
0.9752	17000	1.6964	1.1616	0.8564	-
1.0038	17500	1.7009	1.2108	0.8419	-
1.0325	18000	1.496	1.1526	0.8572	-
1.0612	18500	1.5363	1.2081	0.8393	-
1.0899	19000	1.5324	1.2091	0.8421	-
1.1186	19500	1.532	1.2015	0.8478	-
1.1472	20000	1.6217	1.1716	0.8537	-
1.1759	20500	1.6181	1.2310	0.8553	-
1.2046	21000	1.6387	1.2562	0.8513	-
1.2333	21500	1.738	1.2336	0.8432	-
1.2620	22000	1.6215	1.2923	0.8507	-
1.2907	22500	1.6923	1.2786	0.8431	-
1.3193	23000	1.6976	1.2908	0.8410	-
1.3480	23500	1.7603	1.3793	0.8438	-
1.3767	24000	1.7386	1.3162	0.8399	-
1.4054	24500	1.6719	1.3176	0.8375	-
1.4341	25000	1.6746	1.4208	0.8358	-
1.4627	25500	1.7317	1.4188	0.8323	-
1.4914	26000	1.7185	1.3929	0.8395	-
1.5201	26500	1.6974	1.4667	0.8359	-
1.5488	27000	1.7197	1.4712	0.8325	-
1.5775	27500	1.8755	1.4054	0.8395	-
1.6061	28000	1.7886	1.4135	0.8417	-
1.6348	28500	1.8037	1.4044	0.8387	-
1.6635	29000	1.7862	1.4599	0.8344	-
1.6922	29500	1.7436	1.3778	0.8413	-
1.7209	30000	1.7524	1.3536	0.8354	-
1.7496	30500	1.7209	1.4169	0.8272	-
1.7782	31000	2.1612	13.8218	0.5311	-
1.8069	31500	2.1815	1.5036	0.8357	-
1.8356	32000	1.7241	1.4602	0.8326	-
1.8643	32500	1.6982	1.4521	0.8369	-
1.8930	33000	1.7243	1.4545	0.8428	-
1.9216	33500	1.7885	1.6161	0.8385	-
1.9503	34000	1.8334	1.5186	0.8347	-
1.9790	34500	1.8216	1.4084	0.8409	-
2.0077	35000	1.6731	1.4777	0.8364	-
2.0364	35500	1.4519	1.5688	0.8307	-
2.0650	36000	1.6391	1.4630	0.8377	-
2.0937	36500	1.5565	1.5380	0.8343	-
2.1224	37000	1.5275	1.4737	0.8244	-
2.1511	37500	1.4889	1.5225	0.8269	-
2.1798	38000	1.5439	1.4909	0.8378	-
2.2085	38500	1.4539	1.4877	0.8348	-
2.2371	39000	1.4442	1.4533	0.8321	-
2.2658	39500	1.5136	1.4661	0.8285	-
2.2945	40000	1.439	1.4510	0.8277	-
2.3232	40500	1.4663	1.4299	0.8380	-
2.3519	41000	1.4443	1.4788	0.8320	-
2.3805	41500	1.5818	1.4900	0.8261	-
2.4092	42000	1.4851	1.4582	0.8354	-
2.4379	42500	1.4569	1.4838	0.8309	-
2.4666	43000	1.4194	1.5216	0.8163	-
2.4953	43500	1.4702	1.4703	0.8196	-
2.5239	44000	1.5365	1.4784	0.8214	-
2.5526	44500	1.5114	1.4578	0.8217	-
2.5813	45000	1.5356	1.4729	0.8115	-
2.6100	45500	1.4716	1.4299	0.8309	-
2.6387	46000	1.4557	1.4769	0.8227	-
2.6674	46500	1.4415	1.5078	0.8177	-
2.6960	47000	1.4528	1.4463	0.8257	-
2.7247	47500	1.4631	1.5136	0.8232	-
2.7534	48000	1.5284	1.4869	0.8206	-
2.7821	48500	1.442	1.4322	0.8273	-
2.8108	49000	1.4305	1.4383	0.8292	-
2.8394	49500	1.4571	1.4309	0.8201	-
2.8681	50000	1.4359	1.4952	0.8178	-
2.8968	50500	1.4558	1.4832	0.8233	-
2.9255	51000	1.4631	1.5744	0.8227	-
2.9542	51500	1.4368	1.4477	0.8279	-
2.9828	52000	1.4329	1.8528	0.8233	-
3.0115	52500	1.4	1.5122	0.8238	-
3.0402	53000	1.2241	1.4225	0.8280	-
3.0689	53500	1.2561	1.5065	0.8204	-
3.0976	54000	1.3201	1.6042	0.8190	-
3.1263	54500	1.311	1.5385	0.8235	-
3.1549	55000	1.2471	1.4826	0.8256	-
3.1836	55500	1.2535	1.4369	0.8259	-
3.2123	56000	1.2698	1.6402	0.8045	-
3.2410	56500	1.2355	1.4863	0.8220	-
3.2697	57000	1.2081	1.4576	0.8195	-
3.2983	57500	1.1963	1.4918	0.8119	-
3.3270	58000	1.2593	1.4623	0.8246	-
3.3557	58500	1.282	1.4623	0.8090	-
3.3844	59000	1.2635	1.5247	0.8166	-
3.4131	59500	1.2815	1.5402	0.8202	-
3.4417	60000	1.1852	1.5276	0.8279	-
3.4704	60500	1.2092	1.3838	0.8203	-
3.4991	61000	1.216	1.4860	0.8148	-
3.5278	61500	1.2038	1.5535	0.8265	-
3.5565	62000	1.2619	1.4893	0.8213	-
3.5852	62500	1.2023	1.5940	0.8192	-
3.6138	63000	1.2061	1.5166	0.8166	-
3.6425	63500	1.3908	1.5104	0.8243	-
3.6712	64000	1.2893	1.8377	0.8200	-
3.6999	64500	1.2521	1.6505	0.8215	-
3.7286	65000	1.2866	1.5029	0.8145	-
3.7572	65500	1.4913	1.5370	0.8217	-
3.7859	66000	1.3785	1.5048	0.8168	-
3.8146	66500	1.3013	1.6031	0.8086	-
3.8433	67000	1.3738	1.6297	0.8115	-
3.8720	67500	1.2946	1.5696	0.8228	-
3.9006	68000	1.3555	1.5255	0.8089	-
3.9293	68500	1.2593	1.5023	0.8200	-
3.9580	69000	1.2875	1.5658	0.8173	-
3.9867	69500	1.2582	1.4963	0.8145	-
4.0154	70000	1.1927	1.5811	0.8084	-
4.0441	70500	1.103	1.7786	0.8005	-
4.0727	71000	1.1006	1.5336	0.8172	-
4.1014	71500	1.0872	1.6159	0.8271	-
4.1301	72000	1.1406	1.5320	0.8102	-
4.1588	72500	1.1652	1.5708	0.8213	-
4.1875	73000	1.1123	1.6201	0.8140	-
4.2161	73500	1.0834	1.5985	0.8213	-
4.2448	74000	1.0813	1.5889	0.8106	-
4.2735	74500	1.0598	1.5266	0.8152	-
4.3022	75000	1.0794	1.5154	0.8234	-
4.3309	75500	1.1016	1.6363	0.8189	-
4.3595	76000	1.1203	1.5820	0.8245	-
4.3882	76500	1.1166	1.5379	0.8197	-
4.4169	77000	1.1056	1.5454	0.8109	-
4.4456	77500	1.0499	1.4709	0.8128	-
4.4743	78000	1.0752	1.5489	0.8111	-
4.5030	78500	1.1039	1.5323	0.8238	-
4.5316	79000	1.0726	1.4388	0.8175	-
4.5603	79500	1.0873	1.5391	0.8165	-
4.5890	80000	1.1028	1.4887	0.8118	-
4.6177	80500	1.122	1.4914	0.8160	-
4.6464	81000	1.0842	1.5051	0.8167	-
4.6750	81500	1.0631	1.5653	0.8132	-
4.7037	82000	1.0724	1.5228	0.8120	-
4.7324	82500	1.0515	1.5087	0.8110	-
4.7611	83000	1.0537	1.5241	0.8120	-
4.7898	83500	1.0941	1.5083	0.8153	-
4.8184	84000	1.0812	1.5091	0.8001	-
4.8471	84500	1.0707	1.4898	0.8109	-
4.8758	85000	1.0467	1.4924	0.8184	-
4.9045	85500	1.0737	1.4708	0.8133	-
4.9332	86000	1.047	1.5657	0.8165	-
4.9619	86500	1.0175	1.5067	0.8086	-
4.9905	87000	1.0771	1.4804	0.8080	-
5.0192	87500	0.9551	1.5010	0.8059	-
5.0479	88000	0.9157	1.4884	0.8077	-
5.0766	88500	0.8326	1.4966	0.8090	-
5.1053	89000	0.8485	1.5071	0.8102	-
5.1339	89500	0.8998	1.5126	0.8049	-
5.1626	90000	0.9012	1.4982	0.8144	-
5.1913	90500	0.9354	1.4888	0.8066	-
5.2200	91000	0.9198	1.5137	0.8022	-
5.2487	91500	0.9074	1.4852	0.7977	-
5.2773	92000	0.9429	1.5120	0.8027	-
5.3060	92500	0.8891	1.5665	0.8002	-
5.3347	93000	1.054	1.5305	0.7982	-
5.3634	93500	0.9508	1.4608	0.8079	-
5.3921	94000	0.9275	1.5325	0.8086	-
5.4208	94500	0.9123	1.5041	0.8056	-
5.4494	95000	0.9666	1.5362	0.8005	-
5.4781	95500	0.9468	1.4727	0.7992	-
5.5068	96000	0.9501	1.4381	0.8078	-
5.5355	96500	0.9527	1.5504	0.8014	-
5.5642	97000	0.8989	1.4986	0.8082	-
5.5928	97500	0.9034	1.5549	0.8021	-
5.6215	98000	0.8865	1.6116	0.8084	-
5.6502	98500	1.5304	1.8928	0.7981	-
5.6789	99000	0.9919	1.4798	0.8050	-
5.7076	99500	0.9651	1.5517	0.8031	-
5.7362	100000	0.9372	1.5297	0.8010	-
5.7649	100500	0.9263	1.5323	0.8049	-
5.7936	101000	0.9242	1.5694	0.8080	-
5.8223	101500	1.0484	1.4544	0.8042	-
5.8510	102000	0.9351	1.5167	0.8114	-
5.8797	102500	0.9757	1.4759	0.7895	-
5.9083	103000	1.0185	1.4510	0.8017	-
5.9370	103500	0.8798	1.5102	0.8097	-
5.9657	104000	0.9518	1.4321	0.7969	-
5.9944	104500	0.978	1.5526	0.7946	-
6.0231	105000	0.818	1.5133	0.7992	-
6.0517	105500	0.7463	1.4665	0.8081	-
6.0804	106000	0.7313	1.5566	0.8032	-
6.1091	106500	0.7363	1.5826	0.8086	-
6.1378	107000	0.7184	1.5082	0.8091	-
6.1665	107500	0.7543	1.5200	0.8047	-
6.1951	108000	0.7492	1.4958	0.8050	-
6.2238	108500	0.7868	1.5206	0.8073	-
6.2525	109000	0.7714	1.5073	0.7990	-
6.2812	109500	0.7931	1.5886	0.8007	-
6.3099	110000	0.7768	1.4684	0.7970	-
6.3386	110500	0.7972	1.4479	0.8049	-
6.3672	111000	0.7286	1.4873	0.8109	-
6.3959	111500	0.7462	1.5758	0.8115	-
6.4246	112000	0.737	1.4757	0.8024	-
6.4533	112500	0.7437	1.4728	0.7962	-
6.4820	113000	0.7644	1.4875	0.7970	-
6.5106	113500	0.7563	1.5626	0.7925	-
6.5393	114000	0.7704	1.4859	0.8042	-
6.5680	114500	0.7455	1.5227	0.8017	-
6.5967	115000	0.7916	1.5085	0.8057	-
6.6254	115500	0.7804	1.4348	0.8029	-
6.6540	116000	0.797	1.4953	0.8025	-
6.6827	116500	0.7731	1.4998	0.8045	-
6.7114	117000	0.7324	1.5095	0.7936	-
6.7401	117500	0.7371	1.5010	0.8037	-
6.7688	118000	0.7596	1.5101	0.8008	-
6.7975	118500	0.7763	1.5442	0.8030	-
6.8261	119000	0.7941	1.4985	0.7879	-
6.8548	119500	0.7408	1.5652	0.7827	-
6.8835	120000	0.7568	1.5540	0.7862	-
6.9122	120500	0.7537	1.5316	0.7979	-
6.9409	121000	0.7741	1.5125	0.7983	-
6.9695	121500	0.7369	1.5109	0.7948	-
6.9982	122000	0.7617	1.4832	0.7926	-
7.0269	122500	0.652	1.4793	0.7923	-
7.0556	123000	0.6824	1.5213	0.7927	-
7.0843	123500	0.6285	1.5025	0.7954	-
7.1129	124000	0.6691	1.5328	0.7946	-
7.1416	124500	0.6422	1.6047	0.7942	-
7.1703	125000	0.6618	1.5424	0.7916	-
7.1990	125500	0.6601	1.5324	0.7961	-
7.2277	126000	0.67	1.5564	0.7914	-
7.2564	126500	0.6223	1.5353	0.7952	-
7.2850	127000	0.6344	1.5982	0.7944	-
7.3137	127500	0.6362	1.5258	0.8059	-
7.3424	128000	0.6254	1.5656	0.7936	-
7.3711	128500	0.6672	1.5142	0.7951	-
7.3998	129000	0.6411	1.6154	0.7938	-
7.4284	129500	0.6307	1.4875	0.7995	-
7.4571	130000	0.6369	1.5370	0.7943	-
7.4858	130500	0.6474	1.5084	0.7862	-
7.5145	131000	0.6475	1.4667	0.7995	-
7.5432	131500	0.6361	1.5099	0.7919	-
7.5718	132000	0.6047	1.5205	0.7861	-
7.6005	132500	0.6435	1.4809	0.7939	-
7.6292	133000	0.6277	1.4976	0.7934	-
7.6579	133500	0.6307	1.5139	0.7998	-
7.6866	134000	0.7114	1.5235	0.7971	-
7.7153	134500	0.728	1.4953	0.7977	-
7.7439	135000	0.6374	1.4649	0.7986	-
7.7726	135500	0.6409	1.4866	0.7964	-
7.8013	136000	0.6364	1.4959	0.7905	-
7.8300	136500	0.6309	1.4888	0.7968	-
7.8587	137000	0.6181	1.5176	0.7916	-
7.8873	137500	0.6002	1.5199	0.7885	-
7.9160	138000	0.6326	1.4986	0.7915	-
7.9447	138500	0.6285	1.5224	0.7886	-
7.9734	139000	0.6019	1.4330	0.7936	-
8.0021	139500	0.6534	1.5217	0.7820	-
8.0307	140000	0.4934	1.4483	0.7987	-
8.0594	140500	0.5176	1.4290	0.7938	-
8.0881	141000	0.5203	1.4782	0.7987	-
8.1168	141500	0.5333	1.5755	0.7809	-
8.1455	142000	0.5262	1.4008	0.7894	-
8.1742	142500	0.488	1.4576	0.7896	-
8.2028	143000	0.4995	1.4221	0.7850	-
8.2315	143500	0.5438	1.4670	0.7884	-
8.2602	144000	0.5358	1.5256	0.7909	-
8.2889	144500	0.5379	1.4966	0.7977	-
8.3176	145000	0.5281	1.4566	0.7987	-
8.3462	145500	0.5059	1.4206	0.7982	-
8.3749	146000	0.4993	1.4853	0.7994	-
8.4036	146500	0.5245	1.4286	0.8003	-
8.4323	147000	0.5277	1.4129	0.7958	-
8.4610	147500	0.5263	1.4098	0.8035	-
8.4896	148000	0.5452	1.5002	0.7957	-
8.5183	148500	0.5334	1.4246	0.8046	-
8.5470	149000	0.5358	1.4566	0.7905	-
8.5757	149500	0.55	1.4229	0.7950	-
8.6044	150000	0.5362	1.4068	0.7901	-
8.6331	150500	0.5321	1.3924	0.7931	-
8.6617	151000	0.5459	1.4455	0.7896	-
8.6904	151500	0.5243	1.4604	0.7969	-
8.7191	152000	0.5067	1.4185	0.7902	-
8.7478	152500	0.5085	1.4642	0.7919	-
8.7765	153000	0.4881	1.5082	0.7899	-
8.8051	153500	0.5203	1.5120	0.7919	-
8.8338	154000	0.5077	1.5209	0.7880	-
8.8625	154500	0.5083	1.4354	0.7938	-
8.8912	155000	0.5085	1.4376	0.7867	-
8.9199	155500	0.4906	1.4027	0.7922	-
8.9485	156000	0.5449	1.5246	0.7859	-
8.9772	156500	0.551	1.4236	0.7955	-
9.0059	157000	0.4974	1.4605	0.7932	-
9.0346	157500	0.4159	1.4128	0.7868	-
9.0633	158000	0.4002	1.3823	0.7921	-
9.0920	158500	0.3814	1.4415	0.7854	-
9.1206	159000	0.3694	1.4016	0.7885	-
9.1493	159500	0.4349	1.3871	0.7975	-
9.1780	160000	0.4258	1.3926	0.7919	-
9.2067	160500	0.4262	1.4316	0.7858	-
9.2354	161000	0.428	1.4839	0.7870	-
9.2640	161500	0.4186	1.4305	0.7917	-
9.2927	162000	0.4306	1.4962	0.7936	-
9.3214	162500	0.3977	1.4436	0.7971	-
9.3501	163000	0.3994	1.4661	0.7910	-
9.3788	163500	0.4227	1.5016	0.7952	-
9.4074	164000	0.4076	1.4551	0.7983	-
9.4361	164500	0.4095	1.4699	0.7889	-
9.4648	165000	0.4054	1.4351	0.7933	-
9.4935	165500	0.455	1.4284	0.7927	-
9.5222	166000	0.42	1.4683	0.7909	-
9.5509	166500	0.4418	1.4145	0.7965	-
9.5795	167000	0.4179	1.4001	0.7947	-
9.6082	167500	0.4094	1.4471	0.7875	-
9.6369	168000	0.4176	1.4185	0.7909	-
9.6656	168500	0.3991	1.3788	0.7894	-
9.6943	169000	0.4019	1.3530	0.7942	-
9.7229	169500	0.4141	1.4194	0.7848	-
9.7516	170000	0.4037	1.4556	0.7842	-
9.7803	170500	0.4298	1.3902	0.7847	-
9.8090	171000	0.4257	1.4237	0.7879	-
9.8377	171500	0.3989	1.4484	0.7916	-
9.8663	172000	0.4164	1.4447	0.7922	-
9.8950	172500	0.4184	1.4129	0.7876	-
9.9237	173000	0.3927	1.4382	0.7961	-
9.9524	173500	0.4355	1.4626	0.7885	-
9.9811	174000	0.4317	1.4384	0.7838	-
10.0098	174500	0.3654	1.4397	0.7868	-
10.0384	175000	0.3319	1.4772	0.7933	-
10.0671	175500	0.3218	1.4553	0.7862	-
10.0958	176000	0.3553	1.4422	0.7880	-
10.1245	176500	0.3572	1.4375	0.7865	-
10.1532	177000	0.3611	1.4634	0.7915	-
10.1818	177500	0.3511	1.4557	0.7875	-
10.2105	178000	0.3475	1.4579	0.7864	-
10.2392	178500	0.3456	1.5179	0.7839	-
10.2679	179000	0.3511	1.4576	0.7844	-
10.2966	179500	0.3336	1.4965	0.7879	-
10.3252	180000	0.3647	1.4619	0.7827	-
10.3539	180500	0.3451	1.4585	0.7871	-
10.3826	181000	0.3585	1.4688	0.7847	-
10.4113	181500	0.3432	1.4332	0.7922	-
10.4400	182000	0.3789	1.4236	0.7920	-
10.4687	182500	0.3313	1.3813	0.7794	-
10.4973	183000	0.3356	1.4369	0.7881	-
10.5260	183500	0.3187	1.4230	0.7889	-
10.5547	184000	0.3255	1.4207	0.7923	-
10.5834	184500	0.3252	1.4159	0.7924	-
10.6121	185000	0.3389	1.4502	0.7879	-
10.6407	185500	0.3407	1.4985	0.7945	-
10.6694	186000	0.3349	1.4637	0.7928	-
10.6981	186500	0.3459	1.4799	0.7922	-
10.7268	187000	0.3352	1.4447	0.7911	-
10.7555	187500	0.3188	1.4034	0.7908	-
10.7841	188000	0.3354	1.4559	0.7917	-
10.8128	188500	0.3087	1.4330	0.7901	-
10.8415	189000	0.3573	1.4262	0.7884	-
10.8702	189500	0.337	1.4397	0.7861	-
10.8989	190000	0.3284	1.4719	0.7875	-
10.9276	190500	0.3452	1.4200	0.7864	-
10.9562	191000	0.3407	1.4429	0.7909	-
10.9849	191500	0.3514	1.4511	0.7935	-
11.0136	192000	0.2932	1.4430	0.7905	-
11.0423	192500	0.2593	1.4349	0.7891	-
11.0710	193000	0.294	1.4235	0.7839	-
11.0996	193500	0.2742	1.4248	0.7853	-
11.1283	194000	0.2969	1.4282	0.7837	-
11.1570	194500	0.2549	1.4321	0.7812	-
11.1857	195000	0.3029	1.4152	0.7821	-
11.2144	195500	0.2948	1.4362	0.7815	-
11.2430	196000	0.2888	1.4211	0.7850	-
11.2717	196500	0.286	1.4898	0.7857	-
11.3004	197000	0.2875	1.4354	0.7878	-
11.3291	197500	0.2876	1.4378	0.7900	-
11.3578	198000	0.3074	1.4171	0.7861	-
11.3865	198500	0.2934	1.4582	0.7856	-
11.4151	199000	0.3017	1.4243	0.7853	-
11.4438	199500	0.2987	1.4444	0.7855	-
11.4725	200000	0.2801	1.4089	0.7869	-
11.5012	200500	0.2891	1.4545	0.7839	-
11.5299	201000	0.275	1.4979	0.7790	-
11.5585	201500	0.3127	1.3802	0.7888	-
11.5872	202000	0.2953	1.4369	0.7850	-
11.6159	202500	0.284	1.4590	0.7869	-
11.6446	203000	0.259	1.4573	0.7823	-
11.6733	203500	0.2787	1.4293	0.7840	-
11.7019	204000	0.2791	1.4671	0.7814	-
11.7306	204500	0.2942	1.4423	0.7885	-
11.7593	205000	0.2788	1.4622	0.7874	-
11.7880	205500	0.2614	1.4742	0.7846	-
11.8167	206000	0.2809	1.4380	0.7799	-
11.8454	206500	0.2933	1.4385	0.7859	-
11.8740	207000	0.2623	1.4415	0.7833	-
11.9027	207500	0.2494	1.4490	0.7823	-
11.9314	208000	0.2904	1.4445	0.7827	-
11.9601	208500	0.2737	1.4070	0.7764	-
11.9888	209000	0.262	1.4465	0.7811	-
12.0174	209500	0.2418	1.4411	0.7794	-
12.0461	210000	0.2315	1.4468	0.7847	-
12.0748	210500	0.2614	1.4399	0.7867	-
12.1035	211000	0.2256	1.4226	0.7837	-
12.1322	211500	0.2487	1.4494	0.7851	-
12.1608	212000	0.2273	1.4707	0.7821	-
12.1895	212500	0.2408	1.4935	0.7838	-
12.2182	213000	0.2299	1.4478	0.7842	-
12.2469	213500	0.2207	1.4369	0.7854	-
12.2756	214000	0.2234	1.4400	0.7833	-
12.3043	214500	0.2356	1.4631	0.7851	-
12.3329	215000	0.2256	1.4520	0.7827	-
12.3616	215500	0.2349	1.4680	0.7821	-
12.3903	216000	0.2253	1.4628	0.7852	-
12.4190	216500	0.2275	1.4912	0.7851	-
12.4477	217000	0.2395	1.4483	0.7848	-
12.4763	217500	0.2419	1.4556	0.7850	-
12.5050	218000	0.2391	1.4339	0.7841	-
12.5337	218500	0.2493	1.4502	0.7834	-
12.5624	219000	0.2197	1.4511	0.7815	-
12.5911	219500	0.2255	1.4263	0.7816	-
12.6197	220000	0.2428	1.4315	0.7784	-
12.6484	220500	0.2252	1.4596	0.7812	-
12.6771	221000	0.2408	1.4551	0.7852	-
12.7058	221500	0.2491	1.4858	0.7743	-
12.7345	222000	0.2378	1.4834	0.7832	-
12.7632	222500	0.227	1.4535	0.7828	-
12.7918	223000	0.2403	1.4725	0.7811	-
12.8205	223500	0.2211	1.4726	0.7789	-
12.8492	224000	0.2296	1.4557	0.7793	-
12.8779	224500	0.2289	1.4179	0.7819	-
12.9066	225000	0.2342	1.4385	0.7797	-
12.9352	225500	0.2378	1.4202	0.7771	-
12.9639	226000	0.2146	1.4290	0.7817	-
12.9926	226500	0.2375	1.4171	0.7776	-
13.0213	227000	0.1939	1.4448	0.7786	-
13.0500	227500	0.1943	1.4572	0.7795	-
13.0786	228000	0.2	1.4622	0.7833	-
13.1073	228500	0.2084	1.4683	0.7815	-
13.1360	229000	0.1927	1.4814	0.7777	-
13.1647	229500	0.2167	1.4574	0.7806	-
13.1934	230000	0.2003	1.4806	0.7810	-
13.2221	230500	0.2175	1.4794	0.7805	-
13.2507	231000	0.2051	1.4450	0.7820	-
13.2794	231500	0.2003	1.4658	0.7833	-
13.3081	232000	0.2025	1.4491	0.7841	-
13.3368	232500	0.2132	1.4462	0.7809	-
13.3655	233000	0.2028	1.4458	0.7817	-
13.3941	233500	0.2056	1.4331	0.7814	-
13.4228	234000	0.1834	1.4571	0.7790	-
13.4515	234500	0.2007	1.4393	0.7809	-
13.4802	235000	0.1882	1.4566	0.7813	-
13.5089	235500	0.1941	1.4503	0.7807	-
13.5375	236000	0.1993	1.4622	0.7782	-
13.5662	236500	0.1994	1.4631	0.7783	-
13.5949	237000	0.206	1.4430	0.7776	-
13.6236	237500	0.1969	1.4665	0.7810	-
13.6523	238000	0.2053	1.4888	0.7754	-
13.6809	238500	0.2034	1.4593	0.7761	-
13.7096	239000	0.1983	1.4838	0.7776	-
13.7383	239500	0.1945	1.4714	0.7773	-
13.7670	240000	0.2055	1.4640	0.7779	-
13.7957	240500	0.2024	1.4754	0.7787	-
13.8244	241000	0.1959	1.4552	0.7766	-
13.8530	241500	0.187	1.4456	0.7753	-
13.8817	242000	0.1906	1.4514	0.7739	-
13.9104	242500	0.1928	1.4691	0.7771	-
13.9391	243000	0.2021	1.4537	0.7779	-
13.9678	243500	0.1855	1.4683	0.7816	-
13.9964	244000	0.1997	1.4667	0.7802	-
14.0251	244500	0.1714	1.4906	0.7799	-
14.0538	245000	0.1878	1.4786	0.7811	-
14.0825	245500	0.1796	1.4974	0.7794	-
14.1112	246000	0.1826	1.4833	0.7796	-
14.1398	246500	0.1731	1.4995	0.7788	-
14.1685	247000	0.167	1.4896	0.7795	-
14.1972	247500	0.1871	1.4724	0.7797	-
14.2259	248000	0.1934	1.4777	0.7812	-
14.2546	248500	0.1764	1.4755	0.7822	-
14.2833	249000	0.1866	1.4718	0.7812	-
14.3119	249500	0.2047	1.4668	0.7817	-
14.3406	250000	0.1643	1.4811	0.7817	-
14.3693	250500	0.1715	1.4833	0.7790	-
14.3980	251000	0.1757	1.4786	0.7803	-
14.4267	251500	0.1844	1.4803	0.7807	-
14.4553	252000	0.1721	1.4953	0.7808	-
14.4840	252500	0.1549	1.4872	0.7810	-
14.5127	253000	0.1599	1.4582	0.7824	-
14.5414	253500	0.1691	1.4735	0.7813	-
14.5701	254000	0.1737	1.4741	0.7814	-
14.5987	254500	0.1612	1.4754	0.7810	-
14.6274	255000	0.1773	1.4656	0.7821	-
14.6561	255500	0.1758	1.4690	0.7814	-
14.6848	256000	0.1791	1.4730	0.7814	-
14.7135	256500	0.1848	1.4745	0.7810	-
14.7422	257000	0.1665	1.4855	0.7808	-
14.7708	257500	0.1827	1.4692	0.7809	-
14.7995	258000	0.1725	1.4647	0.7812	-
14.8282	258500	0.1535	1.4680	0.7812	-
14.8569	259000	0.1645	1.4720	0.7810	-
14.8856	259500	0.1704	1.4748	0.7807	-
14.9142	260000	0.1747	1.4699	0.7806	-
14.9429	260500	0.1893	1.4670	0.7807	-
14.9716	261000	0.1754	1.4679	0.7806	-
-1	-1	-	-	-	0.7525

Framework Versions

• Python: 3.13.0
• Sentence Transformers: 5.1.2
• Transformers: 4.57.1
• PyTorch: 2.9.1+cu128
• Accelerate: 1.11.0
• Datasets: 4.4.1
• 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",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

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}
}