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

SentenceTransformer based on FacebookAI/roberta-base

This is a sentence-transformers model finetuned from FacebookAI/roberta-base on the all-nli dataset. 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: FacebookAI/roberta-base
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 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': 768, '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, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.6264, 0.0948],
#         [0.6264, 1.0000, 0.2493],
#         [0.0948, 0.2493, 1.0000]])

Evaluation

Metrics

Semantic Similarity

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

Metric	sts-dev	sts-test
pearson_cosine	0.8242	0.7899
spearman_cosine	0.8259	0.7921

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	-	-	0.6500	-
0.0287	500	12.9834	4.4446	0.8142	-
0.0574	1000	5.0558	2.6336	0.8433	-
0.0860	1500	4.0677	2.1055	0.8449	-
0.1147	2000	3.5554	1.8425	0.8488	-
0.1434	2500	3.1731	1.6827	0.8506	-
0.1721	3000	2.9998	1.6020	0.8574	-
0.2008	3500	2.8479	1.5138	0.8518	-
0.2294	4000	2.6391	1.4776	0.8531	-
0.2581	4500	2.6123	1.4301	0.8609	-
0.2868	5000	2.4732	1.3546	0.8591	-
0.3155	5500	2.3969	1.3039	0.8591	-
0.3442	6000	2.3997	1.3233	0.8656	-
0.3729	6500	2.3758	1.3031	0.8595	-
0.4015	7000	2.3403	1.2585	0.8578	-
0.4302	7500	2.2049	1.3044	0.8553	-
0.4589	8000	2.1295	1.2580	0.8525	-
0.4876	8500	2.2281	1.2584	0.8576	-
0.5163	9000	2.2291	1.2483	0.8470	-
0.5449	9500	2.1401	1.2757	0.8610	-
0.5736	10000	2.0297	1.2819	0.8525	-
0.6023	10500	2.0616	1.2209	0.8499	-
0.6310	11000	1.9647	1.2464	0.8547	-
0.6597	11500	2.0016	1.1847	0.8503	-
0.6883	12000	1.9446	1.2070	0.8525	-
0.7170	12500	1.9737	1.2414	0.8557	-
0.7457	13000	1.9648	1.1860	0.8508	-
0.7744	13500	1.9201	1.2226	0.8496	-
0.8031	14000	1.8554	1.2228	0.8484	-
0.8318	14500	1.8868	1.2439	0.8603	-
0.8604	15000	1.8792	1.1713	0.8495	-
0.8891	15500	1.8812	1.1805	0.8519	-
0.9178	16000	1.8516	1.2316	0.8514	-
0.9465	16500	1.8007	1.2269	0.8461	-
0.9752	17000	1.7603	1.2313	0.8490	-
1.0038	17500	1.7137	1.2067	0.8416	-
1.0325	18000	1.5488	1.1928	0.8384	-
1.0612	18500	1.5885	1.2342	0.8400	-
1.0899	19000	1.587	1.2192	0.8470	-
1.1186	19500	1.5847	1.2362	0.8412	-
1.1472	20000	1.6477	1.3148	0.8488	-
1.1759	20500	1.5791	1.2597	0.8412	-
1.2046	21000	1.6322	1.2991	0.8431	-
1.2333	21500	1.6587	1.2976	0.8355	-
1.2620	22000	1.5471	1.3057	0.8479	-
1.2907	22500	1.5954	1.3160	0.8394	-
1.3193	23000	1.5961	1.3306	0.8393	-
1.3480	23500	1.6175	1.3224	0.8338	-
1.3767	24000	1.58	1.3298	0.8423	-
1.4054	24500	1.5811	1.2604	0.8372	-
1.4341	25000	1.5488	1.3224	0.8380	-
1.4627	25500	1.5905	1.3510	0.8395	-
1.4914	26000	1.5759	1.3830	0.8355	-
1.5201	26500	1.5344	1.3678	0.8328	-
1.5488	27000	1.5502	1.3485	0.8268	-
1.5775	27500	1.6084	1.3426	0.8345	-
1.6061	28000	1.5494	1.3062	0.8434	-
1.6348	28500	1.5705	1.2855	0.8384	-
1.6635	29000	1.5043	1.3451	0.8387	-
1.6922	29500	1.5479	1.3092	0.8405	-
1.7209	30000	1.5569	1.2877	0.8352	-
1.7496	30500	1.4811	1.3088	0.8375	-
1.7782	31000	1.503	1.2647	0.8407	-
1.8069	31500	1.5416	1.3738	0.8445	-
1.8356	32000	1.4459	1.2674	0.8474	-
1.8643	32500	1.5038	1.2700	0.8440	-
1.8930	33000	1.4559	1.2964	0.8420	-
1.9216	33500	1.3829	1.3190	0.8395	-
1.9503	34000	1.4752	1.3236	0.8413	-
1.9790	34500	1.4366	1.2968	0.8402	-
2.0077	35000	1.3896	1.3126	0.8433	-
2.0364	35500	1.0903	1.3262	0.8322	-
2.0650	36000	1.1466	1.2708	0.8427	-
2.0937	36500	1.1498	1.2716	0.8434	-
2.1224	37000	1.1693	1.3225	0.8473	-
2.1511	37500	1.1437	1.3153	0.8346	-
2.1798	38000	1.1813	1.3388	0.8340	-
2.2085	38500	1.1034	1.2852	0.8428	-
2.2371	39000	1.0881	1.2978	0.8432	-
2.2658	39500	1.1366	1.2986	0.8397	-
2.2945	40000	1.1261	1.2811	0.8397	-
2.3232	40500	1.136	1.3165	0.8372	-
2.3519	41000	1.1146	1.2736	0.8423	-
2.3805	41500	1.1484	1.2822	0.8411	-
2.4092	42000	1.1052	1.2153	0.8443	-
2.4379	42500	1.0838	1.2890	0.8395	-
2.4666	43000	1.099	1.3082	0.8446	-
2.4953	43500	1.1038	1.3195	0.8352	-
2.5239	44000	1.0813	1.2516	0.8348	-
2.5526	44500	1.1078	1.3428	0.8381	-
2.5813	45000	1.1094	1.2394	0.8324	-
2.6100	45500	1.0594	1.2751	0.8369	-
2.6387	46000	1.0994	1.2607	0.8312	-
2.6674	46500	1.0946	1.2326	0.8308	-
2.6960	47000	1.0575	1.2957	0.8316	-
2.7247	47500	1.0936	1.3358	0.8234	-
2.7534	48000	1.0449	1.2434	0.8358	-
2.7821	48500	1.0078	1.3299	0.8292	-
2.8108	49000	1.0929	1.3079	0.8291	-
2.8394	49500	1.0517	1.2575	0.8338	-
2.8681	50000	1.0385	1.2555	0.8388	-
2.8968	50500	1.0458	1.2732	0.8368	-
2.9255	51000	1.0525	1.3205	0.8327	-
2.9542	51500	1.0272	1.2707	0.8428	-
2.9828	52000	0.9977	1.2854	0.8401	-
3.0115	52500	0.9116	1.3667	0.8371	-
3.0402	53000	0.7987	1.2610	0.8437	-
3.0689	53500	0.8297	1.3018	0.8428	-
3.0976	54000	0.8309	1.3320	0.8406	-
3.1263	54500	0.8354	1.2898	0.8353	-
3.1549	55000	0.8209	1.2937	0.8369	-
3.1836	55500	0.8346	1.2745	0.8413	-
3.2123	56000	0.8002	1.2647	0.8382	-
3.2410	56500	0.828	1.3219	0.8351	-
3.2697	57000	0.8146	1.2671	0.8421	-
3.2983	57500	0.7883	1.2576	0.8430	-
3.3270	58000	0.8037	1.2747	0.8402	-
3.3557	58500	0.8312	1.2527	0.8371	-
3.3844	59000	0.8239	1.3148	0.8344	-
3.4131	59500	0.8105	1.2951	0.8349	-
3.4417	60000	0.7824	1.3018	0.8330	-
3.4704	60500	0.7864	1.2749	0.8338	-
3.4991	61000	0.7713	1.2763	0.8368	-
3.5278	61500	0.7846	1.3069	0.8343	-
3.5565	62000	0.8137	1.2746	0.8372	-
3.5852	62500	0.8086	1.2858	0.8326	-
3.6138	63000	0.7581	1.3138	0.8373	-
3.6425	63500	0.8743	1.2905	0.8375	-
3.6712	64000	0.79	1.2679	0.8425	-
3.6999	64500	0.8028	1.3027	0.8392	-
3.7286	65000	0.7923	1.2678	0.8388	-
3.7572	65500	0.8628	1.2883	0.8362	-
3.7859	66000	0.8007	1.2476	0.8356	-
3.8146	66500	0.8163	1.2941	0.8327	-
3.8433	67000	0.8147	1.2695	0.8328	-
3.8720	67500	0.806	1.2737	0.8393	-
3.9006	68000	0.8017	1.2550	0.8380	-
3.9293	68500	0.7735	1.2912	0.8357	-
3.9580	69000	0.7936	1.2877	0.8403	-
3.9867	69500	0.7962	1.2525	0.8387	-
4.0154	70000	0.6995	1.2901	0.8382	-
4.0441	70500	0.6376	1.3132	0.8381	-
4.0727	71000	0.6268	1.3263	0.8401	-
4.1014	71500	0.6339	1.3005	0.8321	-
4.1301	72000	0.5927	1.3181	0.8355	-
4.1588	72500	0.674	1.3045	0.8361	-
4.1875	73000	0.6086	1.2831	0.8296	-
4.2161	73500	0.6055	1.2399	0.8369	-
4.2448	74000	0.6089	1.2837	0.8354	-
4.2735	74500	0.6201	1.2798	0.8370	-
4.3022	75000	0.6112	1.2372	0.8404	-
4.3309	75500	0.6228	1.3101	0.8372	-
4.3595	76000	0.6411	1.2884	0.8371	-
4.3882	76500	0.6705	1.2782	0.8376	-
4.4169	77000	0.6208	1.2618	0.8346	-
4.4456	77500	0.6243	1.2773	0.8417	-
4.4743	78000	0.6598	1.2746	0.8423	-
4.5030	78500	0.6265	1.3135	0.8350	-
4.5316	79000	0.6285	1.2945	0.8316	-
4.5603	79500	0.6514	1.2947	0.8373	-
4.5890	80000	0.6222	1.2861	0.8358	-
4.6177	80500	0.6499	1.3138	0.8352	-
4.6464	81000	0.6528	1.2633	0.8347	-
4.6750	81500	0.647	1.2300	0.8353	-
4.7037	82000	0.6339	1.3199	0.8304	-
4.7324	82500	0.6241	1.2745	0.8379	-
4.7611	83000	0.6196	1.2750	0.8333	-
4.7898	83500	0.6415	1.2656	0.8319	-
4.8184	84000	0.6285	1.2675	0.8349	-
4.8471	84500	0.6704	1.2692	0.8424	-
4.8758	85000	0.6346	1.3031	0.8366	-
4.9045	85500	0.6332	1.2819	0.8327	-
4.9332	86000	0.6201	1.2538	0.8396	-
4.9619	86500	0.5993	1.2576	0.8335	-
4.9905	87000	0.6233	1.2494	0.8403	-
5.0192	87500	0.5449	1.2462	0.8431	-
5.0479	88000	0.522	1.3205	0.8367	-
5.0766	88500	0.5046	1.2828	0.8404	-
5.1053	89000	0.5002	1.2980	0.8343	-
5.1339	89500	0.5024	1.2879	0.8413	-
5.1626	90000	0.4996	1.3013	0.8345	-
5.1913	90500	0.5184	1.2622	0.8326	-
5.2200	91000	0.5001	1.2429	0.8405	-
5.2487	91500	0.4899	1.2539	0.8314	-
5.2773	92000	0.5163	1.2123	0.8359	-
5.3060	92500	0.5124	1.2529	0.8321	-
5.3347	93000	0.5399	1.2868	0.8359	-
5.3634	93500	0.5095	1.2217	0.8355	-
5.3921	94000	0.5154	1.2764	0.8393	-
5.4208	94500	0.4904	1.2653	0.8386	-
5.4494	95000	0.4924	1.2685	0.8374	-
5.4781	95500	0.5201	1.2494	0.8280	-
5.5068	96000	0.5172	1.2479	0.8401	-
5.5355	96500	0.5385	1.3059	0.8389	-
5.5642	97000	0.5029	1.2922	0.8371	-
5.5928	97500	0.5158	1.2882	0.8394	-
5.6215	98000	0.5084	1.3059	0.8298	-
5.6502	98500	0.4991	1.2331	0.8354	-
5.6789	99000	0.5211	1.3025	0.8335	-
5.7076	99500	0.5106	1.2624	0.8374	-
5.7362	100000	0.5329	1.2238	0.8335	-
5.7649	100500	0.4982	1.2890	0.8362	-
5.7936	101000	0.4856	1.3022	0.8347	-
5.8223	101500	0.5195	1.2774	0.8434	-
5.8510	102000	0.5012	1.2318	0.8409	-
5.8797	102500	0.5086	1.3152	0.8296	-
5.9083	103000	0.5422	1.2187	0.8322	-
5.9370	103500	0.4863	1.2511	0.8358	-
5.9657	104000	0.536	1.2410	0.8303	-
5.9944	104500	0.5082	1.2399	0.8369	-
6.0231	105000	0.4542	1.3084	0.8309	-
6.0517	105500	0.4102	1.2852	0.8337	-
6.0804	106000	0.3911	1.2805	0.8333	-
6.1091	106500	0.4227	1.2853	0.8342	-
6.1378	107000	0.3923	1.3140	0.8347	-
6.1665	107500	0.4027	1.2836	0.8337	-
6.1951	108000	0.4268	1.2788	0.8335	-
6.2238	108500	0.4337	1.3135	0.8289	-
6.2525	109000	0.4152	1.2783	0.8375	-
6.2812	109500	0.4386	1.3863	0.8171	-
6.3099	110000	0.4269	1.2786	0.8344	-
6.3386	110500	0.4512	1.2746	0.8272	-
6.3672	111000	0.3879	1.3043	0.8382	-
6.3959	111500	0.3996	1.3095	0.8346	-
6.4246	112000	0.3942	1.3081	0.8339	-
6.4533	112500	0.413	1.2650	0.8392	-
6.4820	113000	0.4243	1.2872	0.8348	-
6.5106	113500	0.403	1.2970	0.8380	-
6.5393	114000	0.4151	1.2704	0.8383	-
6.5680	114500	0.4037	1.2972	0.8331	-
6.5967	115000	0.4269	1.3380	0.8376	-
6.6254	115500	0.4233	1.2927	0.8336	-
6.6540	116000	0.4289	1.2784	0.8337	-
6.6827	116500	0.4373	1.2634	0.8396	-
6.7114	117000	0.3996	1.2934	0.8356	-
6.7401	117500	0.4076	1.2793	0.8363	-
6.7688	118000	0.4091	1.3162	0.8300	-
6.7975	118500	0.4002	1.3168	0.8379	-
6.8261	119000	0.4128	1.2875	0.8385	-
6.8548	119500	0.3872	1.3245	0.8353	-
6.8835	120000	0.4091	1.2763	0.8406	-
6.9122	120500	0.3955	1.3050	0.8390	-
6.9409	121000	0.4223	1.2819	0.8340	-
6.9695	121500	0.4226	1.2982	0.8310	-
6.9982	122000	0.3973	1.2659	0.8354	-
7.0269	122500	0.3507	1.3166	0.8325	-
7.0556	123000	0.3665	1.2739	0.8373	-
7.0843	123500	0.3285	1.3148	0.8344	-
7.1129	124000	0.3441	1.2813	0.8390	-
7.1416	124500	0.3334	1.2939	0.8378	-
7.1703	125000	0.3616	1.2835	0.8372	-
7.1990	125500	0.3424	1.3163	0.8334	-
7.2277	126000	0.3729	1.3179	0.8333	-
7.2564	126500	0.3418	1.2671	0.8385	-
7.2850	127000	0.3425	1.2964	0.8338	-
7.3137	127500	0.3348	1.3023	0.8355	-
7.3424	128000	0.3261	1.3049	0.8408	-
7.3711	128500	0.3775	1.2912	0.8388	-
7.3998	129000	0.3512	1.3120	0.8340	-
7.4284	129500	0.3487	1.2879	0.8355	-
7.4571	130000	0.3346	1.2994	0.8298	-
7.4858	130500	0.3685	1.3090	0.8311	-
7.5145	131000	0.3604	1.2841	0.8314	-
7.5432	131500	0.3525	1.2841	0.8274	-
7.5718	132000	0.3406	1.2683	0.8310	-
7.6005	132500	0.3464	1.2790	0.8304	-
7.6292	133000	0.3559	1.2381	0.8327	-
7.6579	133500	0.3463	1.2864	0.8343	-
7.6866	134000	0.3748	1.2516	0.8307	-
7.7153	134500	0.3575	1.2559	0.8286	-
7.7439	135000	0.3318	1.2782	0.8254	-
7.7726	135500	0.3589	1.2826	0.8265	-
7.8013	136000	0.3473	1.2572	0.8337	-
7.8300	136500	0.3284	1.2599	0.8322	-
7.8587	137000	0.3334	1.3010	0.8307	-
7.8873	137500	0.3253	1.2655	0.8330	-
7.9160	138000	0.3402	1.2753	0.8343	-
7.9447	138500	0.345	1.2521	0.8338	-
7.9734	139000	0.3209	1.2731	0.8327	-
8.0021	139500	0.3481	1.2801	0.8296	-
8.0307	140000	0.2829	1.3438	0.8363	-
8.0594	140500	0.2889	1.3169	0.8290	-
8.0881	141000	0.2933	1.3344	0.8306	-
8.1168	141500	0.3098	1.2897	0.8317	-
8.1455	142000	0.2846	1.3296	0.8342	-
8.1742	142500	0.2738	1.3388	0.8254	-
8.2028	143000	0.2741	1.3294	0.8266	-
8.2315	143500	0.3133	1.3106	0.8303	-
8.2602	144000	0.3043	1.3315	0.8284	-
8.2889	144500	0.3257	1.2867	0.8319	-
8.3176	145000	0.299	1.2985	0.8293	-
8.3462	145500	0.2837	1.3443	0.8301	-
8.3749	146000	0.2886	1.3092	0.8316	-
8.4036	146500	0.2837	1.3053	0.8309	-
8.4323	147000	0.2898	1.3065	0.8285	-
8.4610	147500	0.2967	1.2968	0.8301	-
8.4896	148000	0.3062	1.3315	0.8288	-
8.5183	148500	0.3099	1.2954	0.8310	-
8.5470	149000	0.2906	1.2878	0.8282	-
8.5757	149500	0.3114	1.3483	0.8259	-
8.6044	150000	0.2784	1.2967	0.8318	-
8.6331	150500	0.2977	1.3032	0.8331	-
8.6617	151000	0.3194	1.3009	0.8336	-
8.6904	151500	0.292	1.2783	0.8331	-
8.7191	152000	0.2909	1.3243	0.8357	-
8.7478	152500	0.2867	1.3135	0.8316	-
8.7765	153000	0.2762	1.3196	0.8362	-
8.8051	153500	0.2973	1.3327	0.8344	-
8.8338	154000	0.2976	1.3533	0.8332	-
8.8625	154500	0.281	1.2960	0.8356	-
8.8912	155000	0.2885	1.2833	0.8347	-
8.9199	155500	0.286	1.2873	0.8367	-
8.9485	156000	0.2851	1.3146	0.8308	-
8.9772	156500	0.2999	1.3225	0.8313	-
9.0059	157000	0.2812	1.3481	0.8290	-
9.0346	157500	0.2415	1.3387	0.8276	-
9.0633	158000	0.2282	1.3657	0.8310	-
9.0920	158500	0.2315	1.3422	0.8332	-
9.1206	159000	0.2271	1.3270	0.8338	-
9.1493	159500	0.2578	1.3409	0.8346	-
9.1780	160000	0.2534	1.3682	0.8263	-
9.2067	160500	0.252	1.3480	0.8292	-
9.2354	161000	0.2426	1.3118	0.8324	-
9.2640	161500	0.241	1.3244	0.8342	-
9.2927	162000	0.2641	1.3550	0.8300	-
9.3214	162500	0.235	1.3633	0.8281	-
9.3501	163000	0.2376	1.3247	0.8295	-
9.3788	163500	0.2533	1.2885	0.8280	-
9.4074	164000	0.2496	1.2880	0.8262	-
9.4361	164500	0.2542	1.3147	0.8303	-
9.4648	165000	0.2398	1.3021	0.8283	-
9.4935	165500	0.255	1.3547	0.8331	-
9.5222	166000	0.2566	1.3392	0.8322	-
9.5509	166500	0.253	1.3471	0.8311	-
9.5795	167000	0.2521	1.3255	0.8320	-
9.6082	167500	0.2431	1.3392	0.8286	-
9.6369	168000	0.2455	1.3146	0.8325	-
9.6656	168500	0.2379	1.3177	0.8359	-
9.6943	169000	0.2446	1.3157	0.8338	-
9.7229	169500	0.244	1.2792	0.8323	-
9.7516	170000	0.243	1.2883	0.8327	-
9.7803	170500	0.2607	1.2789	0.8342	-
9.8090	171000	0.2457	1.3022	0.8286	-
9.8377	171500	0.2556	1.3212	0.8319	-
9.8663	172000	0.2483	1.2960	0.8326	-
9.8950	172500	0.2337	1.2957	0.8349	-
9.9237	173000	0.2279	1.3591	0.8322	-
9.9524	173500	0.2537	1.3259	0.8320	-
9.9811	174000	0.2501	1.3193	0.8283	-
10.0098	174500	0.2342	1.3031	0.8306	-
10.0384	175000	0.2107	1.3039	0.8320	-
10.0671	175500	0.2144	1.3254	0.8299	-
10.0958	176000	0.2195	1.3310	0.8289	-
10.1245	176500	0.2224	1.3331	0.8283	-
10.1532	177000	0.2314	1.3470	0.8290	-
10.1818	177500	0.2074	1.3448	0.8295	-
10.2105	178000	0.2144	1.3502	0.8289	-
10.2392	178500	0.2165	1.3407	0.8321	-
10.2679	179000	0.2225	1.3336	0.8300	-
10.2966	179500	0.2049	1.3322	0.8325	-
10.3252	180000	0.235	1.3099	0.8291	-
10.3539	180500	0.223	1.3654	0.8285	-
10.3826	181000	0.2486	1.3403	0.8318	-
10.4113	181500	0.2111	1.3213	0.8294	-
10.4400	182000	0.2272	1.3535	0.8311	-
10.4687	182500	0.2128	1.3455	0.8282	-
10.4973	183000	0.2087	1.3252	0.8326	-
10.5260	183500	0.2134	1.3360	0.8328	-
10.5547	184000	0.1955	1.3268	0.8345	-
10.5834	184500	0.2061	1.3499	0.8300	-
10.6121	185000	0.2089	1.3325	0.8294	-
10.6407	185500	0.2152	1.3630	0.8318	-
10.6694	186000	0.2086	1.3266	0.8285	-
10.6981	186500	0.2163	1.3130	0.8299	-
10.7268	187000	0.2218	1.3009	0.8317	-
10.7555	187500	0.1945	1.3064	0.8337	-
10.7841	188000	0.2323	1.3216	0.8319	-
10.8128	188500	0.1993	1.3334	0.8330	-
10.8415	189000	0.2302	1.3199	0.8315	-
10.8702	189500	0.2095	1.3527	0.8309	-
10.8989	190000	0.2019	1.3195	0.8335	-
10.9276	190500	0.2179	1.3258	0.8289	-
10.9562	191000	0.2264	1.2852	0.8345	-
10.9849	191500	0.2311	1.2949	0.8336	-
11.0136	192000	0.1846	1.3296	0.8317	-
11.0423	192500	0.1773	1.3000	0.8316	-
11.0710	193000	0.206	1.3370	0.8324	-
11.0996	193500	0.1861	1.3470	0.8267	-
11.1283	194000	0.2117	1.3355	0.8300	-
11.1570	194500	0.1756	1.3456	0.8265	-
11.1857	195000	0.2077	1.3432	0.8272	-
11.2144	195500	0.1997	1.3428	0.8280	-
11.2430	196000	0.198	1.3004	0.8309	-
11.2717	196500	0.1841	1.3437	0.8272	-
11.3004	197000	0.1927	1.3217	0.8326	-
11.3291	197500	0.1905	1.3424	0.8270	-
11.3578	198000	0.2011	1.3316	0.8251	-
11.3865	198500	0.1901	1.3385	0.8279	-
11.4151	199000	0.2003	1.3366	0.8264	-
11.4438	199500	0.2099	1.3282	0.8320	-
11.4725	200000	0.1838	1.3151	0.8317	-
11.5012	200500	0.1867	1.3451	0.8309	-
11.5299	201000	0.1887	1.3606	0.8273	-
11.5585	201500	0.2096	1.3450	0.8281	-
11.5872	202000	0.2017	1.3495	0.8288	-
11.6159	202500	0.1911	1.3665	0.8293	-
11.6446	203000	0.1752	1.3612	0.8276	-
11.6733	203500	0.1843	1.3630	0.8307	-
11.7019	204000	0.1771	1.3637	0.8300	-
11.7306	204500	0.1876	1.3336	0.8301	-
11.7593	205000	0.1908	1.3561	0.8291	-
11.7880	205500	0.154	1.3526	0.8284	-
11.8167	206000	0.1891	1.3541	0.8277	-
11.8454	206500	0.1888	1.3914	0.8276	-
11.8740	207000	0.1755	1.3729	0.8305	-
11.9027	207500	0.1747	1.3535	0.8332	-
11.9314	208000	0.1974	1.3643	0.8310	-
11.9601	208500	0.1912	1.3955	0.8251	-
11.9888	209000	0.1655	1.3693	0.8313	-
12.0174	209500	0.17	1.3893	0.8306	-
12.0461	210000	0.1523	1.3973	0.8298	-
12.0748	210500	0.1811	1.3727	0.8303	-
12.1035	211000	0.1631	1.4143	0.8299	-
12.1322	211500	0.1686	1.3605	0.8290	-
12.1608	212000	0.1644	1.3652	0.8295	-
12.1895	212500	0.1777	1.3657	0.8259	-
12.2182	213000	0.1618	1.3921	0.8272	-
12.2469	213500	0.1514	1.3664	0.8273	-
12.2756	214000	0.1651	1.3821	0.8249	-
12.3043	214500	0.1696	1.3777	0.8262	-
12.3329	215000	0.1638	1.3712	0.8265	-
12.3616	215500	0.1824	1.3719	0.8271	-
12.3903	216000	0.1646	1.3850	0.8271	-
12.4190	216500	0.1667	1.3704	0.8298	-
12.4477	217000	0.1764	1.3355	0.8307	-
12.4763	217500	0.1703	1.3312	0.8292	-
12.5050	218000	0.166	1.3544	0.8284	-
12.5337	218500	0.1659	1.3639	0.8275	-
12.5624	219000	0.1617	1.3830	0.8277	-
12.5911	219500	0.1557	1.3394	0.8292	-
12.6197	220000	0.1819	1.3540	0.8274	-
12.6484	220500	0.1611	1.3919	0.8254	-
12.6771	221000	0.1824	1.3366	0.8304	-
12.7058	221500	0.1621	1.3581	0.8277	-
12.7345	222000	0.1716	1.3638	0.8302	-
12.7632	222500	0.1604	1.3631	0.8294	-
12.7918	223000	0.1772	1.3577	0.8274	-
12.8205	223500	0.1592	1.3439	0.8290	-
12.8492	224000	0.1619	1.3722	0.8298	-
12.8779	224500	0.1561	1.3631	0.8294	-
12.9066	225000	0.1681	1.3768	0.8301	-
12.9352	225500	0.1666	1.3711	0.8257	-
12.9639	226000	0.1599	1.3684	0.8271	-
12.9926	226500	0.1702	1.3657	0.8260	-
13.0213	227000	0.1487	1.3755	0.8268	-
13.0500	227500	0.1428	1.3684	0.8280	-
13.0786	228000	0.1549	1.4036	0.8271	-
13.1073	228500	0.1623	1.3891	0.8281	-
13.1360	229000	0.1487	1.3758	0.8289	-
13.1647	229500	0.1619	1.3753	0.8283	-
13.1934	230000	0.1567	1.4021	0.8275	-
13.2221	230500	0.1622	1.3753	0.8288	-
13.2507	231000	0.1478	1.3688	0.8295	-
13.2794	231500	0.1482	1.3960	0.8279	-
13.3081	232000	0.1504	1.3679	0.8285	-
13.3368	232500	0.1651	1.3879	0.8270	-
13.3655	233000	0.1392	1.3828	0.8282	-
13.3941	233500	0.1652	1.3710	0.8271	-
13.4228	234000	0.1356	1.3715	0.8272	-
13.4515	234500	0.1515	1.3681	0.8289	-
13.4802	235000	0.1443	1.3815	0.8295	-
13.5089	235500	0.1438	1.3946	0.8280	-
13.5375	236000	0.1419	1.3664	0.8292	-
13.5662	236500	0.1494	1.3847	0.8285	-
13.5949	237000	0.1517	1.3948	0.8269	-
13.6236	237500	0.1522	1.3967	0.8276	-
13.6523	238000	0.152	1.3933	0.8285	-
13.6809	238500	0.1453	1.3879	0.8275	-
13.7096	239000	0.1514	1.4199	0.8263	-
13.7383	239500	0.1457	1.3954	0.8244	-
13.7670	240000	0.1556	1.3833	0.8261	-
13.7957	240500	0.149	1.3795	0.8262	-
13.8244	241000	0.1466	1.3849	0.8272	-
13.8530	241500	0.1417	1.3761	0.8283	-
13.8817	242000	0.149	1.3752	0.8256	-
13.9104	242500	0.1377	1.3687	0.8276	-
13.9391	243000	0.154	1.3717	0.8275	-
13.9678	243500	0.1419	1.3658	0.8286	-
13.9964	244000	0.1406	1.3816	0.8284	-
14.0251	244500	0.1365	1.3771	0.8281	-
14.0538	245000	0.1486	1.3772	0.8286	-
14.0825	245500	0.1344	1.3811	0.8293	-
14.1112	246000	0.1419	1.3766	0.8288	-
14.1398	246500	0.1399	1.3688	0.8275	-
14.1685	247000	0.1226	1.3883	0.8267	-
14.1972	247500	0.1373	1.3843	0.8257	-
14.2259	248000	0.152	1.3978	0.8269	-
14.2546	248500	0.1377	1.3878	0.8260	-
14.2833	249000	0.1455	1.3833	0.8258	-
14.3119	249500	0.1683	1.3903	0.8262	-
14.3406	250000	0.1248	1.4027	0.8259	-
14.3693	250500	0.1371	1.3947	0.8260	-
14.3980	251000	0.1419	1.3897	0.8265	-
14.4267	251500	0.1431	1.3956	0.8262	-
14.4553	252000	0.1485	1.3880	0.8271	-
14.4840	252500	0.1235	1.3851	0.8270	-
14.5127	253000	0.1119	1.3891	0.8268	-
14.5414	253500	0.1297	1.3973	0.8269	-
14.5701	254000	0.1388	1.3949	0.8261	-
14.5987	254500	0.1234	1.3931	0.8263	-
14.6274	255000	0.1392	1.3906	0.8270	-
14.6561	255500	0.1403	1.3913	0.8269	-
14.6848	256000	0.141	1.3999	0.8263	-
14.7135	256500	0.1356	1.3971	0.8264	-
14.7422	257000	0.1352	1.3998	0.8258	-
14.7708	257500	0.1495	1.3989	0.8256	-
14.7995	258000	0.1315	1.3983	0.8257	-
14.8282	258500	0.1097	1.4002	0.8259	-
14.8569	259000	0.125	1.3997	0.8259	-
14.8856	259500	0.1424	1.4013	0.8259	-
14.9142	260000	0.1393	1.3994	0.8259	-
14.9429	260500	0.1441	1.3972	0.8259	-
14.9716	261000	0.1391	1.3988	0.8259	-
-1	-1	-	-	-	0.7921

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}
}