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, 0.6658, 0.1268],
#         [0.6658, 1.0000, 0.2392],
#         [0.1268, 0.2392, 1.0000]])

Evaluation

Metrics

Semantic Similarity

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

Metric	sts-dev	sts-test
pearson_cosine	0.8232	0.7945
spearman_cosine	0.8229	0.7977

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": [
          1024,
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          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": [
          1024,
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          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	13.1307	3.8711	0.8322	-
0.0574	1000	4.9118	2.2331	0.8644	-
0.0860	1500	3.9032	1.8578	0.8656	-
0.1147	2000	3.4689	1.6312	0.8714	-
0.1434	2500	3.1647	1.5546	0.8716	-
0.1721	3000	2.9685	1.4834	0.8801	-
0.2008	3500	2.8231	1.3738	0.8712	-
0.2294	4000	2.6908	1.3634	0.8718	-
0.2581	4500	2.6483	1.3831	0.8789	-
0.2868	5000	2.5545	1.3328	0.8756	-
0.3155	5500	2.428	1.2995	0.8737	-
0.3442	6000	2.4545	1.2703	0.8752	-
0.3729	6500	2.4895	1.2613	0.8704	-
0.4015	7000	2.406	1.2644	0.8632	-
0.4302	7500	2.2968	1.3179	0.8635	-
0.4589	8000	2.2072	1.3236	0.8724	-
0.4876	8500	2.3579	1.3357	0.8599	-
0.5163	9000	2.3808	1.2971	0.8592	-
0.5449	9500	2.2616	1.3413	0.8692	-
0.5736	10000	2.199	1.3170	0.8601	-
0.6023	10500	2.2254	1.3450	0.8590	-
0.6310	11000	2.16	1.3072	0.8606	-
0.6597	11500	2.1753	1.3070	0.8662	-
0.6883	12000	2.0891	1.2685	0.8687	-
0.7170	12500	2.1434	1.3496	0.8605	-
0.7457	13000	2.133	1.2944	0.8533	-
0.7744	13500	2.0775	1.3831	0.8528	-
0.8031	14000	2.0856	1.3325	0.8558	-
0.8318	14500	2.0905	1.3525	0.8713	-
0.8604	15000	2.0856	1.3079	0.8513	-
0.8891	15500	2.1206	1.3687	0.8509	-
0.9178	16000	2.0854	1.3752	0.8429	-
0.9465	16500	2.0765	1.4162	0.8423	-
0.9752	17000	2.0011	1.3374	0.8487	-
1.0038	17500	2.1728	1.4115	0.8504	-
1.0325	18000	1.8636	1.6978	0.8451	-
1.0612	18500	2.2661	1.3775	0.8501	-
1.0899	19000	1.9163	1.3913	0.8407	-
1.1186	19500	1.8524	1.3511	0.8495	-
1.1472	20000	1.9746	1.4419	0.8480	-
1.1759	20500	1.9949	1.4820	0.8406	-
1.2046	21000	2.0087	1.4877	0.8444	-
1.2333	21500	2.0073	1.3913	0.8475	-
1.2620	22000	1.9374	1.5215	0.8539	-
1.2907	22500	1.979	1.5082	0.8585	-
1.3193	23000	1.9629	1.4630	0.8454	-
1.3480	23500	2.0761	1.5292	0.8397	-
1.3767	24000	2.0052	1.5718	0.8464	-
1.4054	24500	2.0406	1.4504	0.8462	-
1.4341	25000	1.9931	1.6408	0.8470	-
1.4627	25500	2.0963	1.6715	0.8469	-
1.4914	26000	2.0744	1.7317	0.8344	-
1.5201	26500	2.0192	1.7196	0.8301	-
1.5488	27000	2.072	1.8939	0.8272	-
1.5775	27500	2.1233	1.6756	0.8381	-
1.6061	28000	2.1309	1.6725	0.8399	-
1.6348	28500	2.1307	1.7151	0.8337	-
1.6635	29000	2.0551	1.6822	0.8348	-
1.6922	29500	2.0623	1.6315	0.8463	-
1.7209	30000	2.1413	1.6249	0.8420	-
1.7496	30500	2.0578	1.8499	0.8327	-
1.7782	31000	2.1535	1.6687	0.8427	-
1.8069	31500	2.1894	1.6501	0.8369	-
1.8356	32000	1.979	1.7413	0.8341	-
1.8643	32500	2.1038	1.7312	0.8345	-
1.8930	33000	2.0746	1.7323	0.8313	-
1.9216	33500	2.2393	1.9724	0.8354	-
1.9503	34000	2.171	1.7513	0.8395	-
1.9790	34500	2.0399	1.7751	0.8345	-
2.0077	35000	2.0002	1.7431	0.8411	-
2.0364	35500	1.6843	1.7284	0.8365	-
2.0650	36000	1.772	1.8173	0.8394	-
2.0937	36500	1.7372	1.8277	0.8381	-
2.1224	37000	1.8665	1.7897	0.8392	-
2.1511	37500	1.8157	1.8601	0.8326	-
2.1798	38000	1.8641	1.6849	0.8388	-
2.2085	38500	1.7293	1.6760	0.8388	-
2.2371	39000	1.7038	1.6455	0.8348	-
2.2658	39500	1.8139	1.7665	0.8205	-
2.2945	40000	1.7791	1.7799	0.8302	-
2.3232	40500	2.8435	1.9061	0.8196	-
2.3519	41000	1.9696	1.8294	0.8344	-
2.3805	41500	1.9685	1.9805	0.8169	-
2.4092	42000	1.7893	1.8200	0.8283	-
2.4379	42500	1.74	1.7132	0.8366	-
2.4666	43000	1.7877	1.7723	0.8433	-
2.4953	43500	1.8317	1.6720	0.8367	-
2.5239	44000	1.7922	1.7199	0.8249	-
2.5526	44500	1.7841	1.7628	0.8300	-
2.5813	45000	1.8367	1.8752	0.8328	-
2.6100	45500	1.773	1.8062	0.8367	-
2.6387	46000	1.8124	1.8124	0.8347	-
2.6674	46500	1.7595	1.7697	0.8340	-
2.6960	47000	1.7422	1.8300	0.8231	-
2.7247	47500	1.8007	1.7629	0.8303	-
2.7534	48000	1.7744	1.7752	0.8287	-
2.7821	48500	1.6891	1.6854	0.8341	-
2.8108	49000	1.7044	1.8094	0.8213	-
2.8394	49500	1.6808	1.6874	0.8243	-
2.8681	50000	1.6774	1.7517	0.8229	-
2.8968	50500	1.684	1.7038	0.8314	-
2.9255	51000	1.7204	1.7657	0.8268	-
2.9542	51500	1.6877	1.7660	0.8306	-
2.9828	52000	1.8228	1.7883	0.8241	-
3.0115	52500	1.5882	1.7963	0.8278	-
3.0402	53000	1.4159	1.8232	0.8262	-
3.0689	53500	1.4347	1.8152	0.8246	-
3.0976	54000	1.5007	1.8113	0.8216	-
3.1263	54500	1.5196	1.7677	0.8250	-
3.1549	55000	1.4994	1.7585	0.8302	-
3.1836	55500	1.5854	1.7113	0.8310	-
3.2123	56000	1.4578	1.8058	0.8238	-
3.2410	56500	1.525	1.7659	0.8255	-
3.2697	57000	1.4602	1.8074	0.8262	-
3.2983	57500	2.1095	1.7176	0.8306	-
3.3270	58000	1.4814	1.8732	0.8312	-
3.3557	58500	1.6221	1.7636	0.8297	-
3.3844	59000	1.4695	1.7405	0.8232	-
3.4131	59500	1.5805	1.7804	0.8376	-
3.4417	60000	1.4774	1.7737	0.8289	-
3.4704	60500	1.4614	1.7717	0.8369	-
3.4991	61000	1.5027	1.8026	0.8312	-
3.5278	61500	1.4788	1.8565	0.8187	-
3.5565	62000	1.5613	1.8170	0.8268	-
3.5852	62500	1.529	1.9040	0.8255	-
3.6138	63000	1.5549	2.0451	0.8263	-
3.6425	63500	1.5604	1.8017	0.8323	-
3.6712	64000	1.4462	1.8130	0.8291	-
3.6999	64500	1.5074	1.8368	0.8236	-
3.7286	65000	1.4982	1.8025	0.8251	-
3.7572	65500	1.5496	1.7867	0.8235	-
3.7859	66000	1.5688	1.8279	0.8200	-
3.8146	66500	1.4988	1.8225	0.8315	-
3.8433	67000	1.5178	1.7781	0.8211	-
3.8720	67500	1.4558	1.8183	0.8279	-
3.9006	68000	1.52	1.7993	0.8219	-
3.9293	68500	1.4339	1.7622	0.8211	-
3.9580	69000	1.4377	1.8122	0.8131	-
3.9867	69500	1.5208	1.8728	0.8200	-
4.0154	70000	1.3749	1.7789	0.8285	-
4.0441	70500	1.3293	1.9063	0.8301	-
4.0727	71000	1.2638	2.1195	0.8174	-
4.1014	71500	1.454	1.8508	0.8265	-
4.1301	72000	1.3227	1.8074	0.8169	-
4.1588	72500	1.3982	1.9820	0.8127	-
4.1875	73000	1.3168	1.9202	0.8223	-
4.2161	73500	1.2791	1.7751	0.8279	-
4.2448	74000	1.2821	1.7690	0.8259	-
4.2735	74500	1.2799	1.8505	0.8202	-
4.3022	75000	1.2453	1.7584	0.8334	-
4.3309	75500	1.2421	1.7923	0.8297	-
4.3595	76000	1.332	1.8744	0.8216	-
4.3882	76500	1.3413	1.8049	0.8292	-
4.4169	77000	1.3342	1.7446	0.8271	-
4.4456	77500	1.2565	1.7859	0.8229	-
4.4743	78000	1.2976	2.0875	0.8284	-
4.5030	78500	1.2861	1.8081	0.8260	-
4.5316	79000	1.2982	1.7828	0.8267	-
4.5603	79500	1.3014	1.7792	0.8200	-
4.5890	80000	1.2867	1.8072	0.8251	-
4.6177	80500	1.3247	1.7776	0.8256	-
4.6464	81000	1.3646	1.7684	0.8210	-
4.6750	81500	1.4309	1.8437	0.8186	-
4.7037	82000	1.3742	1.9158	0.8157	-
4.7324	82500	4.1451	1.8141	0.8249	-
4.7611	83000	1.3416	1.7796	0.8277	-
4.7898	83500	1.3342	1.7990	0.8238	-
4.8184	84000	1.3027	1.9050	0.8180	-
4.8471	84500	1.3237	1.7734	0.8227	-
4.8758	85000	1.2319	1.8478	0.8243	-
4.9045	85500	1.279	1.7974	0.8255	-
4.9332	86000	1.2646	1.7283	0.8305	-
4.9619	86500	1.1886	1.9569	0.8212	-
4.9905	87000	1.2567	1.7428	0.8295	-
5.0192	87500	1.1228	1.8055	0.8306	-
5.0479	88000	1.0618	1.7539	0.8274	-
5.0766	88500	1.0226	1.8684	0.8298	-
5.1053	89000	1.0808	1.7666	0.8208	-
5.1339	89500	1.044	1.7659	0.8211	-
5.1626	90000	1.0438	1.7997	0.8260	-
5.1913	90500	1.1137	1.8361	0.8175	-
5.2200	91000	1.0646	1.8627	0.8264	-
5.2487	91500	1.0429	1.8203	0.8254	-
5.2773	92000	1.108	1.7993	0.8255	-
5.3060	92500	1.031	1.8521	0.8239	-
5.3347	93000	1.1251	1.8060	0.8244	-
5.3634	93500	1.1004	1.8634	0.8247	-
5.3921	94000	1.1488	1.8411	0.8288	-
5.4208	94500	1.0505	1.7374	0.8253	-
5.4494	95000	1.1083	1.8251	0.8238	-
5.4781	95500	1.1011	1.7409	0.8328	-
5.5068	96000	1.1082	1.6637	0.8354	-
5.5355	96500	1.105	1.8248	0.8252	-
5.5642	97000	1.1057	1.7803	0.8261	-
5.5928	97500	1.0842	1.8545	0.8195	-
5.6215	98000	1.0576	1.8561	0.8216	-
5.6502	98500	1.0779	1.7923	0.8272	-
5.6789	99000	1.0408	1.7872	0.8314	-
5.7076	99500	1.1003	1.7327	0.8320	-
5.7362	100000	1.123	1.8170	0.8191	-
5.7649	100500	1.0396	1.6897	0.8258	-
5.7936	101000	1.0426	1.7272	0.8331	-
5.8223	101500	1.0729	1.7506	0.8244	-
5.8510	102000	1.0641	1.7722	0.8306	-
5.8797	102500	1.0518	1.6785	0.8273	-
5.9083	103000	1.0955	1.9886	0.8288	-
5.9370	103500	1.21	2.0539	0.8261	-
5.9657	104000	3.942	1.7589	0.8341	-
5.9944	104500	1.1229	1.8700	0.8185	-
6.0231	105000	0.9885	1.8072	0.8278	-
6.0517	105500	0.9292	1.8001	0.8230	-
6.0804	106000	0.8982	1.8051	0.8317	-
6.1091	106500	0.8904	1.7529	0.8260	-
6.1378	107000	0.8534	1.7874	0.8190	-
6.1665	107500	0.9079	1.7166	0.8289	-
6.1951	108000	0.9005	1.7859	0.8209	-
6.2238	108500	0.9184	1.7757	0.8215	-
6.2525	109000	0.9333	1.7849	0.8261	-
6.2812	109500	0.9627	1.8212	0.8209	-
6.3099	110000	0.9174	1.7716	0.8239	-
6.3386	110500	0.9259	1.8290	0.8278	-
6.3672	111000	0.8882	1.7430	0.8272	-
6.3959	111500	0.8686	1.8061	0.8245	-
6.4246	112000	0.9222	1.8112	0.8221	-
6.4533	112500	0.9037	1.8119	0.8211	-
6.4820	113000	0.8855	1.8029	0.8118	-
6.5106	113500	0.9046	1.8553	0.8245	-
6.5393	114000	0.9272	1.7863	0.8176	-
6.5680	114500	0.931	1.8363	0.8161	-
6.5967	115000	1.0015	1.9976	0.8130	-
6.6254	115500	1.0549	1.8178	0.8212	-
6.6540	116000	0.9827	1.7530	0.8265	-
6.6827	116500	0.9652	1.8149	0.8206	-
6.7114	117000	0.9022	1.8423	0.8259	-
6.7401	117500	0.9249	1.7947	0.8176	-
6.7688	118000	0.8837	1.8191	0.8204	-
6.7975	118500	0.9227	1.7489	0.8259	-
6.8261	119000	0.925	1.7993	0.8160	-
6.8548	119500	0.9141	1.8146	0.8196	-
6.8835	120000	0.8956	1.7155	0.8253	-
6.9122	120500	0.8889	1.7959	0.8347	-
6.9409	121000	0.9551	1.7828	0.8286	-
6.9695	121500	0.8899	1.7918	0.8205	-
6.9982	122000	0.9091	1.7596	0.8236	-
7.0269	122500	0.7642	1.8320	0.8206	-
7.0556	123000	0.9046	1.9385	0.8222	-
7.0843	123500	0.766	1.9060	0.8259	-
7.1129	124000	0.789	1.8289	0.8264	-
7.1416	124500	0.7838	1.8628	0.8281	-
7.1703	125000	0.778	1.7788	0.8154	-
7.1990	125500	0.7838	1.8501	0.8157	-
7.2277	126000	0.8169	1.7758	0.8273	-
7.2564	126500	0.7384	1.9127	0.8194	-
7.2850	127000	0.7868	1.9334	0.8216	-
7.3137	127500	0.7678	1.7547	0.8274	-
7.3424	128000	0.7264	1.7854	0.8275	-
7.3711	128500	0.7925	1.8186	0.8318	-
7.3998	129000	0.7739	1.8462	0.8284	-
7.4284	129500	0.7604	1.8237	0.8277	-
7.4571	130000	0.7492	1.8005	0.8266	-
7.4858	130500	0.7634	1.8099	0.8191	-
7.5145	131000	0.724	1.7840	0.8328	-
7.5432	131500	0.7595	1.8725	0.8219	-
7.5718	132000	0.7308	1.8795	0.8154	-
7.6005	132500	0.7312	1.7966	0.8291	-
7.6292	133000	0.743	1.7927	0.8278	-
7.6579	133500	0.7586	1.7952	0.8294	-
7.6866	134000	0.8312	1.7386	0.8213	-
7.7153	134500	0.7633	1.7157	0.8260	-
7.7439	135000	0.7448	1.8266	0.8261	-
7.7726	135500	0.818	1.7873	0.8275	-
7.8013	136000	0.8235	1.7485	0.8198	-
7.8300	136500	0.7899	1.8871	0.8176	-
7.8587	137000	0.8828	1.9689	0.8184	-
7.8873	137500	0.7736	1.7805	0.8213	-
7.9160	138000	0.7228	1.8282	0.8248	-
7.9447	138500	0.7677	1.7306	0.8244	-
7.9734	139000	0.7351	1.8036	0.8220	-
8.0021	139500	0.7599	1.7727	0.8148	-
8.0307	140000	0.6264	1.7673	0.8177	-
8.0594	140500	0.6125	1.7771	0.8227	-
8.0881	141000	0.6353	1.7675	0.8195	-
8.1168	141500	0.6346	1.7946	0.8229	-
8.1455	142000	0.6101	1.7527	0.8280	-
8.1742	142500	0.5788	1.7372	0.8236	-
8.2028	143000	0.6028	1.7798	0.8248	-
8.2315	143500	0.649	1.7616	0.8198	-
8.2602	144000	0.6672	1.7052	0.8319	-
8.2889	144500	0.665	1.8043	0.8249	-
8.3176	145000	0.619	1.8087	0.8207	-
8.3462	145500	0.6151	1.7635	0.8305	-
8.3749	146000	0.6022	1.7403	0.8313	-
8.4036	146500	0.6258	1.7289	0.8250	-
8.4323	147000	0.6407	1.7225	0.8277	-
8.4610	147500	0.6372	1.7056	0.8284	-
8.4896	148000	0.6761	1.7248	0.8212	-
8.5183	148500	0.6568	1.7226	0.8265	-
8.5470	149000	0.6383	1.6703	0.8281	-
8.5757	149500	0.624	1.7020	0.8245	-
8.6044	150000	0.6188	1.7051	0.8293	-
8.6331	150500	0.6376	1.6799	0.8298	-
8.6617	151000	0.6795	1.7103	0.8247	-
8.6904	151500	0.6274	1.6895	0.8208	-
8.7191	152000	0.6165	1.7270	0.8241	-
8.7478	152500	0.6016	1.7310	0.8217	-
8.7765	153000	0.5853	1.7136	0.8252	-
8.8051	153500	0.666	1.7093	0.8288	-
8.8338	154000	0.61	1.7469	0.8250	-
8.8625	154500	0.6542	1.7309	0.8237	-
8.8912	155000	0.6038	1.6728	0.8213	-
8.9199	155500	0.6195	1.6677	0.8189	-
8.9485	156000	0.646	1.7323	0.8253	-
8.9772	156500	0.6538	1.6865	0.8238	-
9.0059	157000	0.592	1.7343	0.8209	-
9.0346	157500	0.5138	1.7442	0.8233	-
9.0633	158000	0.4933	1.7031	0.8232	-
9.0920	158500	0.4745	1.7306	0.8272	-
9.1206	159000	0.4669	1.7311	0.8289	-
9.1493	159500	0.5194	1.6786	0.8285	-
9.1780	160000	0.536	1.7298	0.8257	-
9.2067	160500	0.4942	1.7287	0.8260	-
9.2354	161000	0.5187	1.6976	0.8235	-
9.2640	161500	0.4831	1.6702	0.8305	-
9.2927	162000	0.5253	1.7145	0.8242	-
9.3214	162500	0.4667	1.6928	0.8245	-
9.3501	163000	0.5022	1.6803	0.8252	-
9.3788	163500	0.5203	1.7851	0.8236	-
9.4074	164000	0.4864	1.6996	0.8217	-
9.4361	164500	0.5125	1.7387	0.8176	-
9.4648	165000	0.4808	1.6818	0.8287	-
9.4935	165500	0.5257	1.7030	0.8255	-
9.5222	166000	0.4963	1.7088	0.8237	-
9.5509	166500	0.5304	1.6953	0.8275	-
9.5795	167000	0.5243	1.6535	0.8236	-
9.6082	167500	0.5012	1.6995	0.8259	-
9.6369	168000	0.5155	1.6797	0.8267	-
9.6656	168500	0.511	1.6843	0.8258	-
9.6943	169000	0.4822	1.6736	0.8308	-
9.7229	169500	0.4908	1.6450	0.8233	-
9.7516	170000	0.5098	1.6952	0.8243	-
9.7803	170500	0.5232	1.7315	0.8263	-
9.8090	171000	0.5174	1.7310	0.8273	-
9.8377	171500	0.5064	1.6783	0.8290	-
9.8663	172000	0.5096	1.7544	0.8248	-
9.8950	172500	0.4885	1.6620	0.8270	-
9.9237	173000	0.4612	1.6874	0.8210	-
9.9524	173500	0.5025	1.7113	0.8221	-
9.9811	174000	0.5071	1.7020	0.8237	-
10.0098	174500	0.4593	1.7157	0.8234	-
10.0384	175000	0.3894	1.7493	0.8260	-
10.0671	175500	0.3875	1.7702	0.8223	-
10.0958	176000	0.4322	1.8000	0.8227	-
10.1245	176500	0.4227	1.7576	0.8276	-
10.1532	177000	0.4368	1.7613	0.8285	-
10.1818	177500	0.4236	1.7270	0.8299	-
10.2105	178000	0.4149	1.7487	0.8298	-
10.2392	178500	0.4108	1.7098	0.8274	-
10.2679	179000	0.4116	1.6843	0.8284	-
10.2966	179500	0.3987	1.6513	0.8284	-
10.3252	180000	0.4667	1.7409	0.8261	-
10.3539	180500	0.4278	1.7262	0.8249	-
10.3826	181000	0.4427	1.7291	0.8250	-
10.4113	181500	0.4157	1.6731	0.8270	-
10.4400	182000	0.4301	1.6889	0.8266	-
10.4687	182500	0.3917	1.7171	0.8221	-
10.4973	183000	0.3984	1.6740	0.8204	-
10.5260	183500	0.3972	1.6973	0.8226	-
10.5547	184000	0.3958	1.7018	0.8276	-
10.5834	184500	0.4144	1.7134	0.8218	-
10.6121	185000	0.3967	1.7309	0.8196	-
10.6407	185500	0.43	1.6597	0.8273	-
10.6694	186000	0.409	1.7010	0.8212	-
10.6981	186500	0.4209	1.6872	0.8224	-
10.7268	187000	0.4165	1.6793	0.8230	-
10.7555	187500	0.3769	1.6417	0.8254	-
10.7841	188000	0.4232	1.6933	0.8187	-
10.8128	188500	0.3872	1.6862	0.8225	-
10.8415	189000	0.4396	1.6578	0.8212	-
10.8702	189500	0.409	1.6828	0.8242	-
10.8989	190000	0.3897	1.6644	0.8239	-
10.9276	190500	0.4072	1.6723	0.8299	-
10.9562	191000	0.4102	1.7174	0.8263	-
10.9849	191500	0.4437	1.6481	0.8246	-
11.0136	192000	0.3587	1.7165	0.8224	-
11.0423	192500	0.3295	1.6652	0.8264	-
11.0710	193000	0.3662	1.7058	0.8245	-
11.0996	193500	0.3254	1.6834	0.8217	-
11.1283	194000	0.3416	1.6786	0.8236	-
11.1570	194500	0.3161	1.7102	0.8244	-
11.1857	195000	0.3641	1.7259	0.8240	-
11.2144	195500	0.3503	1.7683	0.8211	-
11.2430	196000	0.3574	1.7092	0.8207	-
11.2717	196500	0.3519	1.7105	0.8204	-
11.3004	197000	0.3439	1.6659	0.8255	-
11.3291	197500	0.3401	1.6938	0.8194	-
11.3578	198000	0.3542	1.6713	0.8204	-
11.3865	198500	0.3451	1.6958	0.8229	-
11.4151	199000	0.3548	1.6717	0.8213	-
11.4438	199500	0.3607	1.6450	0.8270	-
11.4725	200000	0.3242	1.7143	0.8214	-
11.5012	200500	0.3547	1.6688	0.8206	-
11.5299	201000	0.3443	1.6909	0.8218	-
11.5585	201500	0.3799	1.6252	0.8224	-
11.5872	202000	0.3599	1.6647	0.8211	-
11.6159	202500	0.3385	1.6586	0.8227	-
11.6446	203000	0.3176	1.6887	0.8225	-
11.6733	203500	0.3387	1.7232	0.8247	-
11.7019	204000	0.3399	1.6772	0.8265	-
11.7306	204500	0.3491	1.7123	0.8213	-
11.7593	205000	0.3416	1.6950	0.8233	-
11.7880	205500	0.3029	1.6988	0.8207	-
11.8167	206000	0.3348	1.6667	0.8259	-
11.8454	206500	0.3491	1.6693	0.8238	-
11.8740	207000	0.3096	1.6617	0.8236	-
11.9027	207500	0.2888	1.6873	0.8261	-
11.9314	208000	0.3492	1.6676	0.8253	-
11.9601	208500	0.344	1.6592	0.8254	-
11.9888	209000	0.2991	1.6427	0.8289	-
12.0174	209500	0.2895	1.6966	0.8244	-
12.0461	210000	0.2764	1.6716	0.8227	-
12.0748	210500	0.3001	1.6863	0.8220	-
12.1035	211000	0.2832	1.6749	0.8250	-
12.1322	211500	0.2937	1.6697	0.8267	-
12.1608	212000	0.2737	1.6615	0.8236	-
12.1895	212500	0.2909	1.7160	0.8206	-
12.2182	213000	0.2847	1.6509	0.8268	-
12.2469	213500	0.2711	1.6814	0.8233	-
12.2756	214000	0.2868	1.6701	0.8241	-
12.3043	214500	0.2898	1.6717	0.8223	-
12.3329	215000	0.2847	1.7059	0.8233	-
12.3616	215500	0.3015	1.6790	0.8240	-
12.3903	216000	0.2793	1.6922	0.8261	-
12.4190	216500	0.2803	1.7192	0.8230	-
12.4477	217000	0.2892	1.6702	0.8260	-
12.4763	217500	0.2903	1.6929	0.8237	-
12.5050	218000	0.295	1.6340	0.8264	-
12.5337	218500	0.293	1.6505	0.8270	-
12.5624	219000	0.2701	1.6945	0.8271	-
12.5911	219500	0.267	1.6784	0.8278	-
12.6197	220000	0.3009	1.6514	0.8269	-
12.6484	220500	0.266	1.6717	0.8261	-
12.6771	221000	0.3	1.6844	0.8280	-
12.7058	221500	0.3059	1.6771	0.8314	-
12.7345	222000	0.2901	1.6663	0.8319	-
12.7632	222500	0.279	1.6392	0.8314	-
12.7918	223000	0.2949	1.6556	0.8270	-
12.8205	223500	0.2616	1.6746	0.8265	-
12.8492	224000	0.2809	1.6477	0.8284	-
12.8779	224500	0.2609	1.6443	0.8281	-
12.9066	225000	0.2799	1.6440	0.8274	-
12.9352	225500	0.2869	1.6878	0.8258	-
12.9639	226000	0.253	1.6778	0.8246	-
12.9926	226500	0.2926	1.6454	0.8255	-
13.0213	227000	0.2348	1.6859	0.8242	-
13.0500	227500	0.2353	1.6554	0.8231	-
13.0786	228000	0.2488	1.6847	0.8226	-
13.1073	228500	0.259	1.6820	0.8255	-
13.1360	229000	0.2341	1.6892	0.8237	-
13.1647	229500	0.2603	1.7153	0.8228	-
13.1934	230000	0.2411	1.6844	0.8235	-
13.2221	230500	0.2626	1.6940	0.8240	-
13.2507	231000	0.241	1.6811	0.8247	-
13.2794	231500	0.2342	1.6801	0.8262	-
13.3081	232000	0.2334	1.6911	0.8261	-
13.3368	232500	0.2575	1.6722	0.8236	-
13.3655	233000	0.2329	1.6650	0.8244	-
13.3941	233500	0.2547	1.6775	0.8251	-
13.4228	234000	0.2234	1.6631	0.8239	-
13.4515	234500	0.2365	1.6691	0.8235	-
13.4802	235000	0.2268	1.7275	0.8231	-
13.5089	235500	0.2306	1.6805	0.8245	-
13.5375	236000	0.2388	1.6765	0.8258	-
13.5662	236500	0.2474	1.6769	0.8240	-
13.5949	237000	0.2499	1.7176	0.8228	-
13.6236	237500	0.2406	1.6807	0.8241	-
13.6523	238000	0.2481	1.7075	0.8234	-
13.6809	238500	0.2472	1.6630	0.8234	-
13.7096	239000	0.231	1.7123	0.8238	-
13.7383	239500	0.2294	1.6875	0.8243	-
13.7670	240000	0.2459	1.7007	0.8250	-
13.7957	240500	0.2512	1.6751	0.8278	-
13.8244	241000	0.2355	1.7079	0.8262	-
13.8530	241500	0.2265	1.7144	0.8263	-
13.8817	242000	0.2324	1.7026	0.8268	-
13.9104	242500	0.2299	1.6978	0.8273	-
13.9391	243000	0.2362	1.7243	0.8267	-
13.9678	243500	0.2315	1.6821	0.8290	-
13.9964	244000	0.2386	1.7134	0.8270	-
14.0251	244500	0.2062	1.6998	0.8269	-
14.0538	245000	0.219	1.7169	0.8249	-
14.0825	245500	0.2071	1.7173	0.8264	-
14.1112	246000	0.2178	1.7058	0.8257	-
14.1398	246500	0.2071	1.7181	0.8251	-
14.1685	247000	0.1918	1.7252	0.8243	-
14.1972	247500	0.2307	1.7096	0.8241	-
14.2259	248000	0.2288	1.7527	0.8235	-
14.2546	248500	0.2097	1.7030	0.8250	-
14.2833	249000	0.2275	1.7006	0.8249	-
14.3119	249500	0.2361	1.7337	0.8235	-
14.3406	250000	0.2023	1.7084	0.8234	-
14.3693	250500	0.2112	1.7090	0.8232	-
14.3980	251000	0.2193	1.7033	0.8241	-
14.4267	251500	0.2157	1.7041	0.8236	-
14.4553	252000	0.2059	1.7023	0.8236	-
14.4840	252500	0.194	1.7170	0.8240	-
14.5127	253000	0.1852	1.7050	0.8246	-
14.5414	253500	0.2043	1.7011	0.8246	-
14.5701	254000	0.2103	1.7024	0.8245	-
14.5987	254500	0.1906	1.7177	0.8242	-
14.6274	255000	0.2176	1.7233	0.8237	-
14.6561	255500	0.2065	1.7247	0.8231	-
14.6848	256000	0.222	1.7163	0.8236	-
14.7135	256500	0.2234	1.7166	0.8232	-
14.7422	257000	0.2093	1.7230	0.8223	-
14.7708	257500	0.2321	1.7148	0.8222	-
14.7995	258000	0.2046	1.7102	0.8225	-
14.8282	258500	0.1773	1.7073	0.8230	-
14.8569	259000	0.1961	1.7131	0.8231	-
14.8856	259500	0.2092	1.7097	0.8232	-
14.9142	260000	0.208	1.7093	0.8232	-
14.9429	260500	0.2159	1.7110	0.8230	-
14.9716	261000	0.2106	1.7138	0.8229	-
-1	-1	-	-	-	0.7977

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}
}