sobamchan/roberta-large-no-mrl

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, 0.8065, 0.5747],
#         [0.8065, 1.0000, 0.6756],
#         [0.5747, 0.6756, 1.0000]])

Evaluation

Metrics

Semantic Similarity

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

Metric	sts-dev	sts-test
pearson_cosine	0.7451	0.71
spearman_cosine	0.7649	0.7351

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
      ],
      "matryoshka_weights": [
          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
      ],
      "matryoshka_weights": [
          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	2.658	0.6755	0.8371	-
0.0574	1000	0.8483	0.3792	0.8641	-
0.0860	1500	0.6459	0.3036	0.8668	-
0.1147	2000	0.57	0.2605	0.8709	-
0.1434	2500	0.5211	0.2503	0.8685	-
0.1721	3000	0.4947	0.2367	0.8770	-
0.2008	3500	0.4651	0.2154	0.8688	-
0.2294	4000	0.4386	0.2154	0.8716	-
0.2581	4500	0.4351	0.2128	0.8766	-
0.2868	5000	0.4189	0.2061	0.8751	-
0.3155	5500	0.3989	0.2028	0.8744	-
0.3442	6000	0.4064	0.1998	0.8768	-
0.3729	6500	0.4056	0.2044	0.8718	-
0.4015	7000	0.3975	0.1953	0.8644	-
0.4302	7500	0.3799	0.2079	0.8629	-
0.4589	8000	0.3562	0.2059	0.8656	-
0.4876	8500	0.3789	0.1991	0.8637	-
0.5163	9000	0.3812	0.1941	0.8666	-
0.5449	9500	0.3697	0.2086	0.8655	-
0.5736	10000	0.3529	0.2041	0.8649	-
0.6023	10500	0.3591	0.2099	0.8606	-
0.6310	11000	0.3479	0.2068	0.8588	-
0.6597	11500	0.3532	0.1941	0.8615	-
0.6883	12000	0.3444	0.1949	0.8633	-
0.7170	12500	0.3574	0.2172	0.8560	-
0.7457	13000	0.3581	0.2073	0.8460	-
0.7744	13500	0.3501	0.2108	0.8592	-
0.8031	14000	0.3397	0.2087	0.8554	-
0.8318	14500	0.3468	0.2164	0.8655	-
0.8604	15000	0.3414	0.2055	0.8537	-
0.8891	15500	0.3441	0.2222	0.8549	-
0.9178	16000	0.3457	0.2153	0.8561	-
0.9465	16500	0.3432	0.2215	0.8532	-
0.9752	17000	0.3315	0.2237	0.8541	-
1.0038	17500	0.3272	0.2292	0.8451	-
1.0325	18000	0.3015	0.2257	0.8472	-
1.0612	18500	0.3041	0.2224	0.8370	-
1.0899	19000	0.3016	0.2207	0.8411	-
1.1186	19500	0.3113	0.2331	0.8464	-
1.1472	20000	0.3274	0.2427	0.8393	-
1.1759	20500	0.3215	0.2405	0.8395	-
1.2046	21000	0.3268	0.2332	0.8505	-
1.2333	21500	0.3324	0.2339	0.8351	-
1.2620	22000	0.3128	0.2450	0.8423	-
1.2907	22500	0.3247	0.2546	0.8457	-
1.3193	23000	0.3432	0.2366	0.8429	-
1.3480	23500	0.3324	0.2496	0.8414	-
1.3767	24000	0.3301	0.2424	0.8451	-
1.4054	24500	0.331	0.2472	0.8431	-
1.4341	25000	0.3273	0.2829	0.8421	-
1.4627	25500	0.3768	0.2615	0.8395	-
1.4914	26000	0.3409	0.2745	0.8323	-
1.5201	26500	0.3249	0.2575	0.8317	-
1.5488	27000	0.339	0.2651	0.8303	-
1.5775	27500	0.3873	0.2624	0.8406	-
1.6061	28000	0.3376	0.2623	0.8335	-
1.6348	28500	0.3497	0.2707	0.8336	-
1.6635	29000	0.3332	0.2694	0.8421	-
1.6922	29500	0.3439	0.2632	0.8398	-
1.7209	30000	0.3469	0.2748	0.8397	-
1.7496	30500	0.3408	0.3037	0.8194	-
1.7782	31000	0.3313	0.2587	0.8308	-
1.8069	31500	0.35	0.2783	0.8315	-
1.8356	32000	0.3273	0.2645	0.8191	-
1.8643	32500	0.3409	0.2507	0.8396	-
1.8930	33000	0.338	0.2742	0.8338	-
1.9216	33500	0.3164	0.2675	0.8289	-
1.9503	34000	0.3334	0.2672	0.8365	-
1.9790	34500	0.3275	0.2773	0.8411	-
2.0077	35000	0.3323	0.2758	0.8302	-
2.0364	35500	0.2837	0.2711	0.8254	-
2.0650	36000	0.341	0.2609	0.8347	-
2.0937	36500	0.2828	0.2615	0.8298	-
2.1224	37000	0.299	0.2707	0.8228	-
2.1511	37500	0.2901	0.2846	0.8156	-
2.1798	38000	0.3316	0.2579	0.8275	-
2.2085	38500	0.2837	0.2720	0.8220	-
2.2371	39000	0.2844	0.2937	0.8139	-
2.2658	39500	0.3028	0.2967	0.8260	-
2.2945	40000	0.2857	0.2785	0.8235	-
2.3232	40500	0.2975	0.2648	0.8339	-
2.3519	41000	0.2881	0.2818	0.8163	-
2.3805	41500	0.3047	0.2811	0.8219	-
2.4092	42000	0.315	0.2915	0.8067	-
2.4379	42500	0.3044	0.2871	0.8124	-
2.4666	43000	0.3454	0.4459	0.7998	-
2.4953	43500	0.3065	0.2800	0.8248	-
2.5239	44000	0.3011	0.3524	0.7959	-
2.5526	44500	0.2923	0.2935	0.8167	-
2.5813	45000	0.3105	0.2752	0.8165	-
2.6100	45500	0.3029	0.2990	0.8139	-
2.6387	46000	0.3102	0.3041	0.8111	-
2.6674	46500	0.2992	0.2826	0.8169	-
2.6960	47000	0.2954	0.2656	0.8226	-
2.7247	47500	0.2939	0.2861	0.8014	-
2.7534	48000	0.2871	0.2799	0.8076	-
2.7821	48500	0.2878	0.2694	0.8128	-
2.8108	49000	0.2879	0.2790	0.8168	-
2.8394	49500	0.2759	0.2907	0.8162	-
2.8681	50000	0.2824	0.2829	0.8149	-
2.8968	50500	0.2835	0.2980	0.8198	-
2.9255	51000	0.2914	0.2934	0.8030	-
2.9542	51500	0.3028	0.2898	0.8149	-
2.9828	52000	0.2744	0.2873	0.8210	-
3.0115	52500	0.2674	0.2872	0.8225	-
3.0402	53000	0.2319	0.2849	0.8136	-
3.0689	53500	0.2411	0.3113	0.8129	-
3.0976	54000	0.2564	0.2783	0.8207	-
3.1263	54500	0.2508	0.2751	0.8201	-
3.1549	55000	0.2318	0.2748	0.8236	-
3.1836	55500	0.2587	0.2945	0.8007	-
3.2123	56000	0.2697	0.2882	0.8217	-
3.2410	56500	0.2535	0.2917	0.8179	-
3.2697	57000	0.25	0.2752	0.8173	-
3.2983	57500	0.2299	0.2946	0.8070	-
3.3270	58000	0.2418	0.2832	0.8207	-
3.3557	58500	0.25	0.2761	0.8154	-
3.3844	59000	0.2422	0.2763	0.8173	-
3.4131	59500	0.2598	0.2772	0.8183	-
3.4417	60000	0.2353	0.2828	0.8199	-
3.4704	60500	0.2362	0.2827	0.8154	-
3.4991	61000	0.231	0.2869	0.8040	-
3.5278	61500	0.2326	0.2862	0.7984	-
3.5565	62000	0.2424	0.2769	0.8225	-
3.5852	62500	0.2492	0.2691	0.8112	-
3.6138	63000	0.2344	0.2680	0.8070	-
3.6425	63500	0.2579	0.2736	0.8196	-
3.6712	64000	0.2294	0.2861	0.8165	-
3.6999	64500	0.2403	0.2744	0.8140	-
3.7286	65000	0.2406	0.2680	0.8119	-
3.7572	65500	0.2529	0.2703	0.8179	-
3.7859	66000	0.2464	0.2803	0.8157	-
3.8146	66500	0.2489	0.2709	0.8069	-
3.8433	67000	0.2492	0.2638	0.8202	-
3.8720	67500	0.2401	0.2813	0.8123	-
3.9006	68000	0.2487	0.2720	0.8140	-
3.9293	68500	0.2289	0.2700	0.8145	-
3.9580	69000	0.2371	0.2672	0.8266	-
3.9867	69500	0.239	0.2721	0.8218	-
4.0154	70000	0.2241	0.2669	0.8280	-
4.0441	70500	0.1993	0.2769	0.8083	-
4.0727	71000	0.2028	0.2719	0.8072	-
4.1014	71500	0.2019	0.2827	0.8126	-
4.1301	72000	0.232	0.2704	0.8093	-
4.1588	72500	0.2145	0.2763	0.8154	-
4.1875	73000	0.2233	0.2855	0.8125	-
4.2161	73500	0.2029	0.2732	0.8142	-
4.2448	74000	0.2114	0.2788	0.8063	-
4.2735	74500	0.1968	0.2824	0.8078	-
4.3022	75000	0.2015	0.2691	0.8144	-
4.3309	75500	0.2052	0.2784	0.7987	-
4.3595	76000	0.2162	0.2695	0.8106	-
4.3882	76500	0.2234	0.2618	0.8123	-
4.4169	77000	0.2074	0.2775	0.8170	-
4.4456	77500	0.2086	0.2794	0.8073	-
4.4743	78000	0.2076	0.2836	0.8069	-
4.5030	78500	0.2186	0.2613	0.8182	-
4.5316	79000	0.1963	0.2713	0.8128	-
4.5603	79500	0.2058	0.2828	0.8184	-
4.5890	80000	0.2103	0.2716	0.8123	-
4.6177	80500	0.2319	0.2869	0.8031	-
4.6464	81000	0.2156	0.2841	0.7948	-
4.6750	81500	0.2017	0.2814	0.7971	-
4.7037	82000	0.2171	0.2996	0.8040	-
4.7324	82500	0.2174	0.2771	0.8061	-
4.7611	83000	0.2061	0.2679	0.7892	-
4.7898	83500	0.2168	0.2817	0.7896	-
4.8184	84000	0.2123	0.2868	0.7827	-
4.8471	84500	0.2125	0.2782	0.7980	-
4.8758	85000	0.2032	0.2857	0.8017	-
4.9045	85500	0.2177	0.3339	0.7665	-
4.9332	86000	0.2049	0.2761	0.7998	-
4.9619	86500	0.1928	0.2857	0.8029	-
4.9905	87000	0.2098	0.2788	0.7886	-
5.0192	87500	0.1869	0.2714	0.7951	-
5.0479	88000	0.1765	0.2783	0.7931	-
5.0766	88500	0.1707	0.2867	0.8087	-
5.1053	89000	0.1732	0.2722	0.8106	-
5.1339	89500	0.1778	0.2673	0.8034	-
5.1626	90000	0.1746	0.2945	0.8036	-
5.1913	90500	0.1809	0.2710	0.7987	-
5.2200	91000	0.219	0.2871	0.8113	-
5.2487	91500	0.177	0.2949	0.7935	-
5.2773	92000	0.1846	0.2761	0.8098	-
5.3060	92500	0.1728	0.2838	0.7929	-
5.3347	93000	0.1786	0.2849	0.7999	-
5.3634	93500	0.1783	0.2770	0.8003	-
5.3921	94000	0.1764	0.2868	0.8032	-
5.4208	94500	0.1741	0.2880	0.7912	-
5.4494	95000	0.1755	0.2798	0.8094	-
5.4781	95500	0.1845	0.2822	0.7921	-
5.5068	96000	0.1761	0.2856	0.7984	-
5.5355	96500	0.1803	0.2714	0.8006	-
5.5642	97000	0.1748	0.2938	0.7972	-
5.5928	97500	0.181	0.2818	0.7904	-
5.6215	98000	0.1723	0.2773	0.8043	-
5.6502	98500	0.176	0.2783	0.8099	-
5.6789	99000	0.1718	0.2694	0.7979	-
5.7076	99500	0.1724	0.2738	0.8002	-
5.7362	100000	0.1823	0.2781	0.7712	-
5.7649	100500	0.1684	0.2677	0.7971	-
5.7936	101000	0.1706	0.2727	0.7934	-
5.8223	101500	0.1742	0.2898	0.7976	-
5.8510	102000	0.1699	0.2746	0.7794	-
5.8797	102500	0.1801	0.2697	0.7909	-
5.9083	103000	0.1792	0.2774	0.7920	-
5.9370	103500	0.1719	0.2618	0.7981	-
5.9657	104000	0.1806	0.2657	0.7990	-
5.9944	104500	0.1767	0.2914	0.7838	-
6.0231	105000	0.1854	0.2886	0.7900	-
6.0517	105500	0.1449	0.2889	0.7756	-
6.0804	106000	0.1433	0.2772	0.7974	-
6.1091	106500	0.1429	0.2909	0.7976	-
6.1378	107000	0.1385	0.2763	0.7934	-
6.1665	107500	0.1452	0.2920	0.7954	-
6.1951	108000	0.1463	0.2715	0.7904	-
6.2238	108500	0.1488	0.2839	0.7982	-
6.2525	109000	0.1506	0.2741	0.8023	-
6.2812	109500	0.1524	0.2835	0.8007	-
6.3099	110000	0.1443	0.2720	0.7975	-
6.3386	110500	0.152	0.2882	0.7896	-
6.3672	111000	0.142	0.2759	0.8041	-
6.3959	111500	0.1431	0.2841	0.8054	-
6.4246	112000	0.1406	0.2857	0.7917	-
6.4533	112500	0.1478	0.3215	0.7817	-
6.4820	113000	0.1523	0.2796	0.7851	-
6.5106	113500	0.148	0.2736	0.7870	-
6.5393	114000	0.1481	0.2835	0.7993	-
6.5680	114500	0.1387	0.2844	0.7914	-
6.5967	115000	0.1475	0.2798	0.7981	-
6.6254	115500	0.1463	0.2739	0.7940	-
6.6540	116000	0.1491	0.2739	0.7987	-
6.6827	116500	0.1537	0.2708	0.7965	-
6.7114	117000	0.143	0.2685	0.8018	-
6.7401	117500	0.1481	0.2654	0.7902	-
6.7688	118000	0.1461	0.2741	0.7928	-
6.7975	118500	0.1489	0.2719	0.7965	-
6.8261	119000	0.1503	0.2852	0.7849	-
6.8548	119500	0.1435	0.2729	0.7983	-
6.8835	120000	0.1432	0.2703	0.7924	-
6.9122	120500	0.1481	0.2694	0.7889	-
6.9409	121000	0.1514	0.2735	0.7968	-
6.9695	121500	0.1424	0.2671	0.7914	-
6.9982	122000	0.143	0.2626	0.8006	-
7.0269	122500	0.1287	0.2754	0.7856	-
7.0556	123000	0.1269	0.2748	0.7850	-
7.0843	123500	0.1225	0.2821	0.7807	-
7.1129	124000	0.1223	0.2753	0.7781	-
7.1416	124500	0.1253	0.2688	0.7972	-
7.1703	125000	0.1214	0.2737	0.7905	-
7.1990	125500	0.1208	0.2689	0.7926	-
7.2277	126000	0.127	0.2754	0.7923	-
7.2564	126500	0.1152	0.2715	0.7867	-
7.2850	127000	0.1183	0.2766	0.7792	-
7.3137	127500	0.1195	0.2786	0.7850	-
7.3424	128000	0.1195	0.2885	0.7763	-
7.3711	128500	0.1332	0.2796	0.7868	-
7.3998	129000	0.1217	0.2838	0.7840	-
7.4284	129500	0.1191	0.2711	0.7819	-
7.4571	130000	0.1234	0.2752	0.7744	-
7.4858	130500	0.1297	0.2663	0.7802	-
7.5145	131000	0.1238	0.2643	0.7878	-
7.5432	131500	0.1196	0.2752	0.7809	-
7.5718	132000	0.1164	0.2744	0.7780	-
7.6005	132500	0.1208	0.2682	0.7722	-
7.6292	133000	0.1319	0.2774	0.7811	-
7.6579	133500	0.1208	0.2705	0.7921	-
7.6866	134000	0.1336	0.2681	0.7804	-
7.7153	134500	0.1226	0.3096	0.7763	-
7.7439	135000	0.1293	0.2724	0.7763	-
7.7726	135500	0.1309	0.2707	0.7718	-
7.8013	136000	0.1218	0.2636	0.7799	-
7.8300	136500	0.1253	0.2805	0.7719	-
7.8587	137000	0.1198	0.2619	0.7924	-
7.8873	137500	0.1195	0.2788	0.7822	-
7.9160	138000	0.1264	0.2795	0.7794	-
7.9447	138500	0.1186	0.2687	0.7811	-
7.9734	139000	0.1173	0.2743	0.7758	-
8.0021	139500	0.1216	0.2658	0.7735	-
8.0307	140000	0.1008	0.2725	0.7985	-
8.0594	140500	0.1026	0.2752	0.7897	-
8.0881	141000	0.1031	0.2743	0.7885	-
8.1168	141500	0.1019	0.2623	0.7881	-
8.1455	142000	0.1034	0.2590	0.7870	-
8.1742	142500	0.0986	0.2714	0.7872	-
8.2028	143000	0.0946	0.2729	0.7872	-
8.2315	143500	0.1018	0.2799	0.7842	-
8.2602	144000	0.1029	0.2796	0.7837	-
8.2889	144500	0.1031	0.2760	0.7832	-
8.3176	145000	0.0979	0.2751	0.7863	-
8.3462	145500	0.0958	0.2726	0.7899	-
8.3749	146000	0.0945	0.2709	0.7898	-
8.4036	146500	0.0982	0.2726	0.7944	-
8.4323	147000	0.1048	0.2639	0.7820	-
8.4610	147500	0.1006	0.2630	0.7787	-
8.4896	148000	0.1092	0.2716	0.7771	-
8.5183	148500	0.1024	0.2676	0.7903	-
8.5470	149000	0.1038	0.2619	0.7891	-
8.5757	149500	0.1032	0.2596	0.7960	-
8.6044	150000	0.1022	0.2660	0.7862	-
8.6331	150500	0.103	0.2767	0.7863	-
8.6617	151000	0.1138	0.2657	0.7781	-
8.6904	151500	0.1071	0.2607	0.7884	-
8.7191	152000	0.098	0.2567	0.7900	-
8.7478	152500	0.1019	0.2670	0.7854	-
8.7765	153000	0.0972	0.2647	0.7851	-
8.8051	153500	0.1089	0.2715	0.7759	-
8.8338	154000	0.102	0.2799	0.7817	-
8.8625	154500	0.102	0.2796	0.7808	-
8.8912	155000	0.1055	0.2691	0.7860	-
8.9199	155500	0.0989	0.2636	0.7843	-
8.9485	156000	0.0978	0.2671	0.7800	-
8.9772	156500	0.1085	0.2760	0.7889	-
9.0059	157000	0.102	0.2715	0.7824	-
9.0346	157500	0.0829	0.2759	0.7798	-
9.0633	158000	0.0811	0.2728	0.7897	-
9.0920	158500	0.0849	0.2641	0.7757	-
9.1206	159000	0.0795	0.2570	0.7824	-
9.1493	159500	0.0914	0.2624	0.7725	-
9.1780	160000	0.089	0.2677	0.7727	-
9.2067	160500	0.0874	0.2682	0.7760	-
9.2354	161000	0.0843	0.2690	0.7756	-
9.2640	161500	0.0805	0.2677	0.7732	-
9.2927	162000	0.09	0.2792	0.7723	-
9.3214	162500	0.0842	0.2727	0.7731	-
9.3501	163000	0.0861	0.2647	0.7737	-
9.3788	163500	0.0881	0.2748	0.7769	-
9.4074	164000	0.0838	0.2644	0.7864	-
9.4361	164500	0.0822	0.2609	0.7712	-
9.4648	165000	0.0826	0.2609	0.7787	-
9.4935	165500	0.0839	0.2688	0.7743	-
9.5222	166000	0.0863	0.2600	0.7781	-
9.5509	166500	0.0865	0.2663	0.7800	-
9.5795	167000	0.0816	0.2517	0.7738	-
9.6082	167500	0.0801	0.2597	0.7774	-
9.6369	168000	0.0849	0.2550	0.7764	-
9.6656	168500	0.0821	0.2629	0.7727	-
9.6943	169000	0.0845	0.2696	0.7737	-
9.7229	169500	0.0846	0.2647	0.7686	-
9.7516	170000	0.0818	0.2700	0.7729	-
9.7803	170500	0.0878	0.2620	0.7699	-
9.8090	171000	0.0808	0.2623	0.7672	-
9.8377	171500	0.0777	0.2622	0.7704	-
9.8663	172000	0.0835	0.2606	0.7734	-
9.8950	172500	0.0791	0.2568	0.7771	-
9.9237	173000	0.0739	0.2676	0.7747	-
9.9524	173500	0.0831	0.2566	0.7753	-
9.9811	174000	0.0857	0.2711	0.7637	-
10.0098	174500	0.0711	0.2718	0.7784	-
10.0384	175000	0.0638	0.2612	0.7787	-
10.0671	175500	0.0655	0.2647	0.7781	-
10.0958	176000	0.0702	0.2622	0.7713	-
10.1245	176500	0.0698	0.2672	0.7784	-
10.1532	177000	0.074	0.2678	0.7844	-
10.1818	177500	0.0672	0.2575	0.7830	-
10.2105	178000	0.0685	0.2667	0.7746	-
10.2392	178500	0.0662	0.2650	0.7719	-
10.2679	179000	0.0685	0.2647	0.7743	-
10.2966	179500	0.0666	0.2584	0.7787	-
10.3252	180000	0.073	0.2567	0.7730	-
10.3539	180500	0.0678	0.2665	0.7676	-
10.3826	181000	0.074	0.2621	0.7727	-
10.4113	181500	0.0698	0.2580	0.7798	-
10.4400	182000	0.0729	0.2529	0.7729	-
10.4687	182500	0.0645	0.2548	0.7714	-
10.4973	183000	0.0644	0.2599	0.7742	-
10.5260	183500	0.0638	0.2597	0.7754	-
10.5547	184000	0.0656	0.2606	0.7699	-
10.5834	184500	0.0645	0.2576	0.7776	-
10.6121	185000	0.0639	0.2600	0.7730	-
10.6407	185500	0.0668	0.2580	0.7742	-
10.6694	186000	0.0641	0.2571	0.7765	-
10.6981	186500	0.0689	0.2592	0.7708	-
10.7268	187000	0.067	0.2537	0.7672	-
10.7555	187500	0.0626	0.2549	0.7759	-
10.7841	188000	0.0704	0.2678	0.7751	-
10.8128	188500	0.0616	0.2692	0.7718	-
10.8415	189000	0.0717	0.2583	0.7750	-
10.8702	189500	0.0679	0.2594	0.7722	-
10.8989	190000	0.064	0.2616	0.7713	-
10.9276	190500	0.0695	0.2667	0.7829	-
10.9562	191000	0.0703	0.2619	0.7834	-
10.9849	191500	0.0715	0.2564	0.7813	-
11.0136	192000	0.0594	0.2624	0.7820	-
11.0423	192500	0.0526	0.2616	0.7830	-
11.0710	193000	0.0595	0.2636	0.7799	-
11.0996	193500	0.0537	0.2571	0.7875	-
11.1283	194000	0.0589	0.2617	0.7810	-
11.1570	194500	0.052	0.2632	0.7825	-
11.1857	195000	0.0607	0.2609	0.7829	-
11.2144	195500	0.057	0.2712	0.7757	-
11.2430	196000	0.0587	0.2672	0.7790	-
11.2717	196500	0.0593	0.2585	0.7731	-
11.3004	197000	0.0589	0.2721	0.7706	-
11.3291	197500	0.0556	0.2656	0.7706	-
11.3578	198000	0.0622	0.2584	0.7741	-
11.3865	198500	0.0572	0.2695	0.7750	-
11.4151	199000	0.0586	0.2649	0.7755	-
11.4438	199500	0.0595	0.2671	0.7767	-
11.4725	200000	0.0563	0.2630	0.7707	-
11.5012	200500	0.0574	0.2642	0.7674	-
11.5299	201000	0.0542	0.2644	0.7740	-
11.5585	201500	0.0613	0.2605	0.7694	-
11.5872	202000	0.0593	0.2604	0.7712	-
11.6159	202500	0.0556	0.2628	0.7699	-
11.6446	203000	0.0524	0.2631	0.7728	-
11.6733	203500	0.0602	0.2705	0.7622	-
11.7019	204000	0.0582	0.2631	0.7739	-
11.7306	204500	0.0579	0.2573	0.7721	-
11.7593	205000	0.057	0.2558	0.7774	-
11.7880	205500	0.051	0.2597	0.7757	-
11.8167	206000	0.0559	0.2515	0.7711	-
11.8454	206500	0.0543	0.2566	0.7720	-
11.8740	207000	0.0517	0.2554	0.7763	-
11.9027	207500	0.0493	0.2598	0.7722	-
11.9314	208000	0.0567	0.2592	0.7691	-
11.9601	208500	0.0559	0.2618	0.7755	-
11.9888	209000	0.0503	0.2615	0.7804	-
12.0174	209500	0.0499	0.2648	0.7812	-
12.0461	210000	0.047	0.2614	0.7819	-
12.0748	210500	0.0511	0.2632	0.7717	-
12.1035	211000	0.0464	0.2660	0.7693	-
12.1322	211500	0.0484	0.2658	0.7719	-
12.1608	212000	0.0465	0.2676	0.7756	-
12.1895	212500	0.0478	0.2689	0.7696	-
12.2182	213000	0.0467	0.2564	0.7684	-
12.2469	213500	0.0435	0.2606	0.7674	-
12.2756	214000	0.048	0.2602	0.7701	-
12.3043	214500	0.0471	0.2641	0.7687	-
12.3329	215000	0.0473	0.2557	0.7683	-
12.3616	215500	0.0503	0.2560	0.7705	-
12.3903	216000	0.044	0.2607	0.7724	-
12.4190	216500	0.045	0.2579	0.7707	-
12.4477	217000	0.0473	0.2605	0.7679	-
12.4763	217500	0.049	0.2557	0.7693	-
12.5050	218000	0.0482	0.2604	0.7725	-
12.5337	218500	0.049	0.2553	0.7751	-
12.5624	219000	0.0448	0.2597	0.7667	-
12.5911	219500	0.0443	0.2550	0.7685	-
12.6197	220000	0.0489	0.2561	0.7706	-
12.6484	220500	0.0448	0.2573	0.7693	-
12.6771	221000	0.0492	0.2565	0.7645	-
12.7058	221500	0.0475	0.2638	0.7674	-
12.7345	222000	0.0467	0.2612	0.7709	-
12.7632	222500	0.0443	0.2589	0.7702	-
12.7918	223000	0.0485	0.2605	0.7720	-
12.8205	223500	0.0437	0.2556	0.7716	-
12.8492	224000	0.0442	0.2558	0.7703	-
12.8779	224500	0.0452	0.2589	0.7711	-
12.9066	225000	0.0472	0.2575	0.7715	-
12.9352	225500	0.0484	0.2595	0.7697	-
12.9639	226000	0.0432	0.2578	0.7663	-
12.9926	226500	0.0479	0.2613	0.7641	-
13.0213	227000	0.04	0.2661	0.7659	-
13.0500	227500	0.0397	0.2573	0.7703	-
13.0786	228000	0.039	0.2697	0.7699	-
13.1073	228500	0.0428	0.2649	0.7680	-
13.1360	229000	0.0376	0.2637	0.7680	-
13.1647	229500	0.0441	0.2656	0.7646	-
13.1934	230000	0.0405	0.2600	0.7668	-
13.2221	230500	0.0438	0.2660	0.7690	-
13.2507	231000	0.0404	0.2635	0.7677	-
13.2794	231500	0.0395	0.2594	0.7684	-
13.3081	232000	0.0404	0.2614	0.7711	-
13.3368	232500	0.0428	0.2617	0.7660	-
13.3655	233000	0.0396	0.2602	0.7675	-
13.3941	233500	0.0433	0.2582	0.7676	-
13.4228	234000	0.0375	0.2602	0.7648	-
13.4515	234500	0.0404	0.2616	0.7666	-
13.4802	235000	0.0395	0.2594	0.7655	-
13.5089	235500	0.0375	0.2635	0.7618	-
13.5375	236000	0.039	0.2619	0.7650	-
13.5662	236500	0.0421	0.2606	0.7628	-
13.5949	237000	0.0416	0.2642	0.7627	-
13.6236	237500	0.0391	0.2636	0.7636	-
13.6523	238000	0.04	0.2632	0.7638	-
13.6809	238500	0.0409	0.2587	0.7632	-
13.7096	239000	0.0386	0.2643	0.7600	-
13.7383	239500	0.0396	0.2616	0.7605	-
13.7670	240000	0.0415	0.2600	0.7652	-
13.7957	240500	0.0403	0.2599	0.7662	-
13.8244	241000	0.0377	0.2621	0.7646	-
13.8530	241500	0.0384	0.2600	0.7623	-
13.8817	242000	0.0392	0.2590	0.7617	-
13.9104	242500	0.0386	0.2588	0.7621	-
13.9391	243000	0.0405	0.2628	0.7616	-
13.9678	243500	0.0379	0.2562	0.7627	-
13.9964	244000	0.0384	0.2611	0.7616	-
14.0251	244500	0.0331	0.2611	0.7596	-
14.0538	245000	0.0372	0.2619	0.7609	-
14.0825	245500	0.0348	0.2646	0.7599	-
14.1112	246000	0.0369	0.2618	0.7610	-
14.1398	246500	0.0351	0.2630	0.7588	-
14.1685	247000	0.0314	0.2639	0.7608	-
14.1972	247500	0.0389	0.2599	0.7606	-
14.2259	248000	0.0397	0.2630	0.7617	-
14.2546	248500	0.0352	0.2613	0.7642	-
14.2833	249000	0.0375	0.2621	0.7654	-
14.3119	249500	0.0408	0.2630	0.7622	-
14.3406	250000	0.0313	0.2633	0.7640	-
14.3693	250500	0.0362	0.2635	0.7629	-
14.3980	251000	0.0368	0.2624	0.7640	-
14.4267	251500	0.0367	0.2647	0.7630	-
14.4553	252000	0.0352	0.2620	0.7646	-
14.4840	252500	0.0312	0.2640	0.7646	-
14.5127	253000	0.0316	0.2636	0.7643	-
14.5414	253500	0.0343	0.2619	0.7638	-
14.5701	254000	0.0363	0.2629	0.7635	-
14.5987	254500	0.0327	0.2619	0.7654	-
14.6274	255000	0.0364	0.2635	0.7659	-
14.6561	255500	0.035	0.2651	0.7640	-
14.6848	256000	0.0369	0.2652	0.7635	-
14.7135	256500	0.037	0.2651	0.7645	-
14.7422	257000	0.0341	0.2643	0.7644	-
14.7708	257500	0.0379	0.2639	0.7648	-
14.7995	258000	0.0331	0.2629	0.7645	-
14.8282	258500	0.0309	0.2628	0.7650	-
14.8569	259000	0.0319	0.2632	0.7651	-
14.8856	259500	0.0342	0.2640	0.7646	-
14.9142	260000	0.0344	0.2637	0.7648	-
14.9429	260500	0.0371	0.2637	0.7648	-
14.9716	261000	0.0364	0.2637	0.7649	-
-1	-1	-	-	-	0.7351

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}
}