This is a sentence-transformers model finetuned from sentence-transformers/all-distilroberta-v1. 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: sentence-transformers/all-distilroberta-v1
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 tokens
Similarity Function: Cosine Similarity

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}) with Transformer model: 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})
  (2): Normalize()
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

python

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("hanwenzhu/all-distilroberta-v1-lr2e-4-bs1024-nneg3-mlbs-mar03")
# Run inference
sentences = [
    'Mathlib.Algebra.Polynomial.HasseDeriv#31',
    'Polynomial.hasseDeriv_coeff',
    'HomologicalComplex.isZero_X_of_isStrictlySupported',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Training Details

Training Dataset

Unnamed Dataset

Size: 5,854,451 training samples
Columns: state_name and premise_name
Approximate statistics based on the first 1000 samples:
state_name premise_name
type string string
details
min: 10 tokens
mean: 17.28 tokens
max: 22 tokens
min: 3 tokens
mean: 11.34 tokens
max: 38 tokens

Samples:

state_name	premise_name
`Mathlib.RingTheory.Ideal.Norm.RelNorm#46`	`RingHomCompTriple.ids`
`Mathlib.RingTheory.Ideal.Norm.RelNorm#46`	`MonoidWithZeroHomClass.toMonoidHomClass`
`Mathlib.RingTheory.Ideal.Norm.RelNorm#46`	`Ideal.subset_span`

Loss: loss.MaskedCachedMultipleNegativesRankingLoss with these parameters:
json
1{ 2 "scale": 20.0, 3 "similarity_fct": "cos_sim" 4}

Evaluation Dataset

Unnamed Dataset

Size: 1,959 evaluation samples
Columns: state_name and premise_name
Approximate statistics based on the first 1000 samples:
state_name premise_name
type string string
details
min: 10 tokens
mean: 17.08 tokens
max: 24 tokens
min: 5 tokens
mean: 11.05 tokens
max: 31 tokens

Samples:

state_name	premise_name
`Mathlib.Algebra.Algebra.Hom#80`	`AlgHom.commutes`
`Mathlib.Algebra.Algebra.NonUnitalSubalgebra#237`	`NonUnitalAlgHom.instNonUnitalAlgSemiHomClass`
`Mathlib.Algebra.Algebra.NonUnitalSubalgebra#237`	`NonUnitalAlgebra.mem_top`

Loss: loss.MaskedCachedMultipleNegativesRankingLoss with these parameters:
json
1{ 2 "scale": 20.0, 3 "similarity_fct": "cos_sim" 4}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 1024
per_device_eval_batch_size: 64
learning_rate: 0.0002
num_train_epochs: 1.0
lr_scheduler_type: cosine
warmup_ratio: 0.03
bf16: True
dataloader_num_workers: 4

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
prediction_loss_only: True
per_device_train_batch_size: 1024
per_device_eval_batch_size: 64
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: 0.0002
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1.0
num_train_epochs: 1.0
max_steps: -1
lr_scheduler_type: cosine
lr_scheduler_kwargs: {}
warmup_ratio: 0.03
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
use_ipex: False
bf16: True
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: 4
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}
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
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: False
hub_always_push: False
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_inputs_for_metrics: False
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
dispatch_batches: None
split_batches: None
include_tokens_per_second: False
include_num_input_tokens_seen: False
neftune_noise_alpha: None
optim_target_modules: None
batch_eval_metrics: False
eval_on_start: False
use_liger_kernel: False
eval_use_gather_object: False
batch_sampler: batch_sampler
multi_dataset_batch_sampler: proportional

Training Logs

Click to expand

Epoch	Step	Training Loss	loss
0.0017	10	7.5842	-
0.0035	20	6.4567	-
0.0052	30	5.9408	-
0.0070	40	5.7176	-
0.0087	50	5.5353	-
0.0101	58	-	2.2337
0.0105	60	5.4044	-
0.0122	70	5.3384	-
0.0140	80	5.2395	-
0.0157	90	5.1291	-
0.0175	100	5.1093	-
0.0192	110	5.0695	-
0.0203	116	-	1.9949
0.0210	120	4.9664	-
0.0227	130	4.973	-
0.0245	140	4.9065	-
0.0262	150	4.8961	-
0.0280	160	4.839	-
0.0297	170	4.8513	-
0.0304	174	-	1.9119
0.0315	180	4.7662	-
0.0332	190	4.7385	-
0.0350	200	4.7036	-
0.0367	210	4.7013	-
0.0385	220	4.6837	-
0.0402	230	4.6325	-
0.0406	232	-	1.7502
0.0420	240	4.5982	-
0.0437	250	4.5526	-
0.0455	260	4.5793	-
0.0472	270	4.545	-
0.0490	280	4.5173	-
0.0507	290	4.4595	1.6955
0.0525	300	4.4772	-
0.0542	310	4.4038	-
0.0560	320	4.4132	-
0.0577	330	4.4139	-
0.0595	340	4.3585	-
0.0609	348	-	1.6316
0.0612	350	4.3314	-
0.0630	360	4.3805	-
0.0647	370	4.2791	-
0.0665	380	4.2938	-
0.0682	390	4.2591	-
0.0700	400	4.262	-
0.0710	406	-	1.5723
0.0717	410	4.2108	-
0.0735	420	4.1723	-
0.0752	430	4.157	-
0.0769	440	4.1878	-
0.0787	450	4.1644	-
0.0804	460	4.1569	-
0.0811	464	-	1.5368
0.0822	470	4.139	-
0.0839	480	4.0872	-
0.0857	490	4.1169	-
0.0874	500	4.062	-
0.0892	510	4.1138	-
0.0909	520	4.1088	-
0.0913	522	-	1.5232
0.0927	530	4.0526	-
0.0944	540	4.0355	-
0.0962	550	3.9937	-
0.0979	560	3.9647	-
0.0997	570	3.9715	-
0.1014	580	3.9524	1.4901
0.1032	590	3.945	-
0.1049	600	3.9615	-
0.1067	610	3.9713	-
0.1084	620	3.9264	-
0.1102	630	3.9036	-
0.1116	638	-	1.4411
0.1119	640	3.8909	-
0.1137	650	3.901	-
0.1154	660	3.879	-
0.1172	670	3.8696	-
0.1189	680	3.8678	-
0.1207	690	3.8472	-
0.1217	696	-	1.4459
0.1224	700	3.8277	-
0.1242	710	3.8321	-
0.1259	720	3.812	-
0.1277	730	3.8386	-
0.1294	740	3.7583	-
0.1312	750	3.8007	-
0.1319	754	-	1.3644
0.1329	760	3.7337	-
0.1347	770	3.7554	-
0.1364	780	3.7518	-
0.1382	790	3.6993	-
0.1399	800	3.7477	-
0.1417	810	3.6979	-
0.1420	812	-	1.3702
0.1434	820	3.6651	-
0.1452	830	3.7292	-
0.1469	840	3.7005	-
0.1487	850	3.6856	-
0.1504	860	3.631	-
0.1522	870	3.6459	1.3568
0.1539	880	3.6089	-
0.1556	890	3.6134	-
0.1574	900	3.6058	-
0.1591	910	3.6193	-
0.1609	920	3.627	-
0.1623	928	-	1.3072
0.1626	930	3.6202	-
0.1644	940	3.5891	-
0.1661	950	3.6185	-
0.1679	960	3.5984	-
0.1696	970	3.6258	-
0.1714	980	3.5625	-
0.1724	986	-	1.2930
0.1731	990	3.5441	-
0.1749	1000	3.5571	-
0.1766	1010	3.5486	-
0.1784	1020	3.5382	-
0.1801	1030	3.4519	-
0.1819	1040	3.5072	-
0.1826	1044	-	1.2823
0.1836	1050	3.5042	-
0.1854	1060	3.5005	-
0.1871	1070	3.455	-
0.1889	1080	3.4727	-
0.1906	1090	3.4473	-
0.1924	1100	3.4296	-
0.1927	1102	-	1.2696
0.1941	1110	3.449	-
0.1959	1120	3.4202	-
0.1976	1130	3.4236	-
0.1994	1140	3.414	-
0.2011	1150	3.4264	-
0.2029	1160	3.4005	1.2602
0.2046	1170	3.3801	-
0.2064	1180	3.3543	-
0.2081	1190	3.3866	-
0.2099	1200	3.3831	-
0.2116	1210	3.3691	-
0.2130	1218	-	1.2130
0.2134	1220	3.3607	-
0.2151	1230	3.3659	-
0.2169	1240	3.3538	-
0.2186	1250	3.3336	-
0.2204	1260	3.3403	-
0.2221	1270	3.3062	-
0.2232	1276	-	1.2237
0.2239	1280	3.3251	-
0.2256	1290	3.3475	-
0.2274	1300	3.2729	-
0.2291	1310	3.2872	-
0.2308	1320	3.2778	-
0.2326	1330	3.3147	-
0.2333	1334	-	1.2061
0.2343	1340	3.2477	-
0.2361	1350	3.2871	-
0.2378	1360	3.2458	-
0.2396	1370	3.279	-
0.2413	1380	3.2546	-
0.2431	1390	3.2342	-
0.2434	1392	-	1.1854
0.2448	1400	3.2488	-
0.2466	1410	3.2489	-
0.2483	1420	3.2368	-
0.2501	1430	3.2517	-
0.2518	1440	3.2568	-
0.2536	1450	3.21	1.1616
0.2553	1460	3.1891	-
0.2571	1470	3.1739	-
0.2588	1480	3.2004	-
0.2606	1490	3.1988	-
0.2623	1500	3.1892	-
0.2637	1508	-	1.1306
0.2641	1510	3.1967	-
0.2658	1520	3.1331	-
0.2676	1530	3.155	-
0.2693	1540	3.1564	-
0.2711	1550	3.1912	-
0.2728	1560	3.1005	-
0.2739	1566	-	1.1026
0.2746	1570	3.1166	-
0.2763	1580	3.1453	-
0.2781	1590	3.116	-
0.2798	1600	3.1465	-
0.2816	1610	3.1325	-
0.2833	1620	3.1022	-
0.2840	1624	-	1.1400
0.2851	1630	3.0703	-
0.2868	1640	3.0999	-
0.2886	1650	3.0957	-
0.2903	1660	3.0886	-
0.2921	1670	3.0471	-
0.2938	1680	3.0845	-
0.2942	1682	-	1.1045
0.2956	1690	3.0513	-
0.2973	1700	3.0621	-
0.2991	1710	3.0473	-
0.3008	1720	3.0486	-
0.3026	1730	3.0189	-
0.3043	1740	3.0675	1.1004
0.3061	1750	3.0592	-
0.3078	1760	3.0663	-
0.3095	1770	3.0879	-
0.3113	1780	3.0167	-
0.3130	1790	3.0356	-
0.3144	1798	-	1.0554
0.3148	1800	3.0294	-
0.3165	1810	2.9956	-
0.3183	1820	2.985	-
0.3200	1830	2.9824	-
0.3218	1840	2.9939	-
0.3235	1850	2.9979	-
0.3246	1856	-	1.0561
0.3253	1860	2.9935	-
0.3270	1870	3.0613	-
0.3288	1880	2.9742	-
0.3305	1890	2.9858	-
0.3323	1900	2.9446	-
0.3340	1910	2.9571	-
0.3347	1914	-	1.0333
0.3358	1920	2.9839	-
0.3375	1930	2.9865	-
0.3393	1940	2.9398	-
0.3410	1950	2.9504	-
0.3428	1960	2.9371	-
0.3445	1970	2.9222	-
0.3449	1972	-	1.0322
0.3463	1980	2.8907	-
0.3480	1990	2.9412	-
0.3498	2000	2.944	-
0.3515	2010	2.9168	-
0.3533	2020	2.9076	-
0.3550	2030	2.8967	1.0103
0.3568	2040	2.8569	-
0.3585	2050	2.8602	-
0.3603	2060	2.8984	-
0.3620	2070	2.8782	-
0.3638	2080	2.8649	-
0.3652	2088	-	1.0136
0.3655	2090	2.8388	-
0.3673	2100	2.8845	-
0.3690	2110	2.8749	-
0.3708	2120	2.8439	-
0.3725	2130	2.8693	-
0.3743	2140	2.8342	-
0.3753	2146	-	0.9949
0.3760	2150	2.8696	-
0.3778	2160	2.872	-
0.3795	2170	2.828	-
0.3813	2180	2.8338	-
0.3830	2190	2.8716	-
0.3847	2200	2.8798	-
0.3854	2204	-	1.0067
0.3865	2210	2.834	-
0.3882	2220	2.7885	-
0.3900	2230	2.8152	-
0.3917	2240	2.8214	-
0.3935	2250	2.8306	-
0.3952	2260	2.8164	-
0.3956	2262	-	0.9845
0.3970	2270	2.8338	-
0.3987	2280	2.8223	-
0.4005	2290	2.8183	-
0.4022	2300	2.7903	-
0.4040	2310	2.7772	-
0.4057	2320	2.7952	0.9900
0.4075	2330	2.7733	-
0.4092	2340	2.8096	-
0.4110	2350	2.771	-
0.4127	2360	2.8178	-
0.4145	2370	2.7539	-
0.4159	2378	-	0.9749
0.4162	2380	2.7488	-
0.4180	2390	2.7592	-
0.4197	2400	2.7385	-
0.4215	2410	2.7564	-
0.4232	2420	2.7573	-
0.4250	2430	2.7686	-
0.4260	2436	-	0.9509
0.4267	2440	2.7147	-
0.4285	2450	2.7375	-
0.4302	2460	2.6995	-
0.4320	2470	2.6888	-
0.4337	2480	2.7171	-
0.4355	2490	2.712	-
0.4362	2494	-	0.9311
0.4372	2500	2.729	-
0.4390	2510	2.6974	-
0.4407	2520	2.7056	-
0.4425	2530	2.7123	-
0.4442	2540	2.701	-
0.4460	2550	2.7211	-
0.4463	2552	-	0.9259
0.4477	2560	2.6974	-
0.4495	2570	2.6823	-
0.4512	2580	2.6968	-
0.4530	2590	2.7126	-
0.4547	2600	2.693	-
0.4565	2610	2.7164	0.9161
0.4582	2620	2.6558	-
0.4600	2630	2.6972	-
0.4617	2640	2.7116	-
0.4634	2650	2.6398	-
0.4652	2660	2.6645	-
0.4666	2668	-	0.8982
0.4669	2670	2.6646	-
0.4687	2680	2.6828	-
0.4704	2690	2.6502	-
0.4722	2700	2.6605	-
0.4739	2710	2.6224	-
0.4757	2720	2.6753	-
0.4767	2726	-	0.8941
0.4774	2730	2.6478	-
0.4792	2740	2.6688	-
0.4809	2750	2.6674	-
0.4827	2760	2.6132	-
0.4844	2770	2.6286	-
0.4862	2780	2.634	-
0.4869	2784	-	0.8756
0.4879	2790	2.6359	-
0.4897	2800	2.6242	-
0.4914	2810	2.6443	-
0.4932	2820	2.59	-
0.4949	2830	2.6166	-
0.4967	2840	2.6249	-
0.4970	2842	-	0.8802
0.4984	2850	2.6257	-
0.5002	2860	2.6286	-
0.5019	2870	2.5671	-
0.5037	2880	2.5959	-
0.5054	2890	2.5962	-
0.5072	2900	2.5521	0.8673
0.5089	2910	2.5833	-
0.5107	2920	2.6015	-
0.5124	2930	2.6446	-
0.5142	2940	2.5655	-
0.5159	2950	2.5802	-
0.5173	2958	-	0.8614
0.5177	2960	2.6124	-
0.5194	2970	2.5372	-
0.5212	2980	2.5108	-
0.5229	2990	2.578	-
0.5247	3000	2.5629	-
0.5264	3010	2.5691	-
0.5275	3016	-	0.8418
0.5282	3020	2.5313	-
0.5299	3030	2.5791	-
0.5317	3040	2.5216	-
0.5334	3050	2.5263	-
0.5352	3060	2.5213	-
0.5369	3070	2.5485	-
0.5376	3074	-	0.8546
0.5386	3080	2.5435	-
0.5404	3090	2.5599	-
0.5421	3100	2.5045	-
0.5439	3110	2.5055	-
0.5456	3120	2.54	-
0.5474	3130	2.5134	-
0.5477	3132	-	0.8515
0.5491	3140	2.5053	-
0.5509	3150	2.4578	-
0.5526	3160	2.517	-
0.5544	3170	2.5061	-
0.5561	3180	2.5262	-
0.5579	3190	2.5787	0.8376
0.5596	3200	2.4855	-
0.5614	3210	2.5058	-
0.5631	3220	2.5279	-
0.5649	3230	2.498	-
0.5666	3240	2.5045	-
0.5680	3248	-	0.8407
0.5684	3250	2.5129	-
0.5701	3260	2.517	-
0.5719	3270	2.4647	-
0.5736	3280	2.4642	-
0.5754	3290	2.4936	-
0.5771	3300	2.4862	-
0.5782	3306	-	0.8310
0.5789	3310	2.4805	-
0.5806	3320	2.4986	-
0.5824	3330	2.481	-
0.5841	3340	2.4747	-
0.5859	3350	2.4939	-
0.5876	3360	2.4691	-
0.5883	3364	-	0.8397
0.5894	3370	2.4798	-
0.5911	3380	2.4439	-
0.5929	3390	2.4849	-
0.5946	3400	2.4653	-
0.5964	3410	2.4795	-
0.5981	3420	2.4681	-
0.5985	3422	-	0.8265
0.5999	3430	2.4671	-
0.6016	3440	2.4579	-
0.6034	3450	2.4319	-
0.6051	3460	2.4235	-
0.6069	3470	2.4447	-
0.6086	3480	2.456	0.8104
0.6104	3490	2.4107	-
0.6121	3500	2.49	-
0.6139	3510	2.4511	-
0.6156	3520	2.4446	-
0.6173	3530	2.4159	-
0.6187	3538	-	0.8086
0.6191	3540	2.4135	-
0.6208	3550	2.4147	-
0.6226	3560	2.4458	-
0.6243	3570	2.4207	-
0.6261	3580	2.4333	-
0.6278	3590	2.3931	-
0.6289	3596	-	0.8036
0.6296	3600	2.4695	-
0.6313	3610	2.4285	-
0.6331	3620	2.4066	-
0.6348	3630	2.414	-
0.6366	3640	2.4229	-
0.6383	3650	2.3916	-
0.6390	3654	-	0.7960
0.6401	3660	2.4376	-
0.6418	3670	2.4196	-
0.6436	3680	2.4132	-
0.6453	3690	2.4016	-
0.6471	3700	2.3749	-
0.6488	3710	2.3963	-
0.6492	3712	-	0.7895
0.6506	3720	2.4223	-
0.6523	3730	2.3787	-
0.6541	3740	2.368	-
0.6558	3750	2.3526	-
0.6576	3760	2.3883	-
0.6593	3770	2.4286	0.7897
0.6611	3780	2.366	-
0.6628	3790	2.3914	-
0.6646	3800	2.416	-
0.6663	3810	2.3731	-
0.6681	3820	2.4097	-
0.6695	3828	-	0.7782
0.6698	3830	2.374	-
0.6716	3840	2.3591	-
0.6733	3850	2.384	-
0.6751	3860	2.398	-
0.6768	3870	2.3712	-
0.6786	3880	2.3936	-
0.6796	3886	-	0.7725
0.6803	3890	2.3895	-
0.6821	3900	2.359	-
0.6838	3910	2.3901	-
0.6856	3920	2.4	-
0.6873	3930	2.3628	-
0.6891	3940	2.3732	-
0.6898	3944	-	0.7658
0.6908	3950	2.3929	-
0.6925	3960	2.3792	-
0.6943	3970	2.3496	-
0.6960	3980	2.3242	-
0.6978	3990	2.3471	-
0.6995	4000	2.3503	-
0.6999	4002	-	0.7617
0.7013	4010	2.3693	-
0.7030	4020	2.3608	-
0.7048	4030	2.3419	-
0.7065	4040	2.3577	-
0.7083	4050	2.3403	-
0.7100	4060	2.3491	0.7549
0.7118	4070	2.3175	-
0.7135	4080	2.3513	-
0.7153	4090	2.3767	-
0.7170	4100	2.371	-
0.7188	4110	2.3103	-
0.7202	4118	-	0.7585
0.7205	4120	2.3048	-
0.7223	4130	2.3406	-
0.7240	4140	2.3551	-
0.7258	4150	2.3309	-
0.7275	4160	2.3565	-
0.7293	4170	2.3111	-
0.7303	4176	-	0.7527
0.7310	4180	2.2925	-
0.7328	4190	2.281	-
0.7345	4200	2.3131	-
0.7363	4210	2.3568	-
0.7380	4220	2.3645	-
0.7398	4230	2.3283	-
0.7405	4234	-	0.7497
0.7415	4240	2.3098	-
0.7433	4250	2.3136	-
0.7450	4260	2.3141	-
0.7468	4270	2.2717	-
0.7485	4280	2.325	-
0.7503	4290	2.3358	-
0.7506	4292	-	0.7449
0.7520	4300	2.296	-
0.7538	4310	2.3211	-
0.7555	4320	2.3035	-
0.7573	4330	2.3114	-
0.7590	4340	2.3076	-
0.7608	4350	2.334	0.7416
0.7625	4360	2.2805	-
0.7643	4370	2.3302	-
0.7660	4380	2.2753	-
0.7678	4390	2.3084	-
0.7695	4400	2.308	-
0.7709	4408	-	0.7463
0.7712	4410	2.2909	-
0.7730	4420	2.2796	-
0.7747	4430	2.2868	-
0.7765	4440	2.3021	-
0.7782	4450	2.2977	-
0.7800	4460	2.2885	-
0.7810	4466	-	0.7391
0.7817	4470	2.2967	-
0.7835	4480	2.2774	-
0.7852	4490	2.3178	-
0.7870	4500	2.2785	-
0.7887	4510	2.2493	-
0.7905	4520	2.2866	-
0.7912	4524	-	0.7325
0.7922	4530	2.2632	-
0.7940	4540	2.289	-
0.7957	4550	2.2782	-
0.7975	4560	2.2607	-
0.7992	4570	2.2914	-
0.8010	4580	2.2593	-
0.8013	4582	-	0.7318
0.8027	4590	2.3077	-
0.8045	4600	2.2793	-
0.8062	4610	2.3051	-
0.8080	4620	2.2914	-
0.8097	4630	2.2646	-
0.8115	4640	2.2574	0.7308
0.8132	4650	2.2654	-
0.8150	4660	2.235	-
0.8167	4670	2.258	-
0.8185	4680	2.2935	-
0.8202	4690	2.281	-
0.8216	4698	-	0.7281
0.8220	4700	2.295	-
0.8237	4710	2.3095	-
0.8255	4720	2.2516	-
0.8272	4730	2.2292	-
0.8290	4740	2.2635	-
0.8307	4750	2.2522	-
0.8318	4756	-	0.7330
0.8325	4760	2.248	-
0.8342	4770	2.3082	-
0.8360	4780	2.2447	-
0.8377	4790	2.2596	-
0.8395	4800	2.2747	-
0.8412	4810	2.2343	-
0.8419	4814	-	0.7319
0.8430	4820	2.2521	-
0.8447	4830	2.2642	-
0.8464	4840	2.2492	-
0.8482	4850	2.2788	-
0.8499	4860	2.2925	-
0.8517	4870	2.2491	-
0.8520	4872	-	0.7304
0.8534	4880	2.2666	-
0.8552	4890	2.2261	-
0.8569	4900	2.2504	-
0.8587	4910	2.2567	-
0.8604	4920	2.2813	-
0.8622	4930	2.244	0.7277
0.8639	4940	2.2645	-
0.8657	4950	2.228	-
0.8674	4960	2.2322	-
0.8692	4970	2.2547	-
0.8709	4980	2.2722	-
0.8723	4988	-	0.7272
0.8727	4990	2.227	-
0.8744	5000	2.2407	-
0.8762	5010	2.2269	-
0.8779	5020	2.2428	-
0.8797	5030	2.2448	-
0.8814	5040	2.2562	-
0.8825	5046	-	0.7256
0.8832	5050	2.2364	-
0.8849	5060	2.2445	-
0.8867	5070	2.2409	-
0.8884	5080	2.2261	-
0.8902	5090	2.2613	-
0.8919	5100	2.2718	-
0.8926	5104	-	0.7233
0.8937	5110	2.2544	-
0.8954	5120	2.2276	-
0.8972	5130	2.2385	-
0.8989	5140	2.2401	-
0.9007	5150	2.2769	-
0.9024	5160	2.2399	-
0.9028	5162	-	0.7231
0.9042	5170	2.2205	-
0.9059	5180	2.2303	-
0.9077	5190	2.231	-
0.9094	5200	2.2356	-
0.9112	5210	2.2386	-
0.9129	5220	2.2233	0.7233
0.9147	5230	2.2509	-
0.9164	5240	2.2201	-
0.9182	5250	2.2189	-
0.9199	5260	2.1992	-
0.9217	5270	2.2362	-
0.9231	5278	-	0.7221
0.9234	5280	2.2293	-
0.9251	5290	2.2302	-
0.9269	5300	2.2216	-
0.9286	5310	2.2191	-
0.9304	5320	2.2504	-
0.9321	5330	2.2447	-
0.9332	5336	-	0.7221
0.9339	5340	2.2326	-
0.9356	5350	2.2315	-
0.9374	5360	2.244	-
0.9391	5370	2.2369	-
0.9409	5380	2.2312	-
0.9426	5390	2.2739	-
0.9433	5394	-	0.7206
0.9444	5400	2.2598	-
0.9461	5410	2.2319	-
0.9479	5420	2.2312	-
0.9496	5430	2.2592	-
0.9514	5440	2.2503	-
0.9531	5450	2.232	-
0.9535	5452	-	0.7208
0.9549	5460	2.2341	-
0.9566	5470	2.2564	-
0.9584	5480	2.2087	-
0.9601	5490	2.257	-
0.9619	5500	2.2524	-
0.9636	5510	2.253	0.7204
0.9654	5520	2.2424	-
0.9671	5530	2.2459	-
0.9689	5540	2.2387	-
0.9706	5550	2.2482	-
0.9724	5560	2.2156	-
0.9738	5568	-	0.7200
0.9741	5570	2.2343	-
0.9759	5580	2.2426	-
0.9776	5590	2.2154	-
0.9794	5600	2.2365	-
0.9811	5610	2.275	-
0.9829	5620	2.2689	-
0.9839	5626	-	0.7200
0.9846	5630	2.2356	-
0.9864	5640	2.2517	-
0.9881	5650	2.2436	-
0.9899	5660	2.2229	-
0.9916	5670	2.2617	-
0.9934	5680	2.2359	-
0.9941	5684	-	0.7201
0.9951	5690	2.2444	-
0.9969	5700	2.2505	-
0.9986	5710	2.2713	-

Framework Versions

Python: 3.11.8
Sentence Transformers: 3.1.1
Transformers: 4.45.1
PyTorch: 2.4.0+cu121
Accelerate: 0.34.2
Datasets: 3.0.0
Tokenizers: 0.20.0

Citation

BibTeX

Sentence Transformers

bibtex

@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",
}

MaskedCachedMultipleNegativesRankingLoss

bibtex

@misc{gao2021scaling,
    title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
    author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
    year={2021},
    eprint={2101.06983},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

	state_name	premise_name
type	string	string
details	min: 10 tokens mean: 17.28 tokens max: 22 tokens	min: 3 tokens mean: 11.34 tokens max: 38 tokens