This is a sentence-transformers model finetuned from nomic-ai/modernbert-embed-base. 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: nomic-ai/modernbert-embed-base
Maximum Sequence Length: 8192 tokens
Output Dimensionality: 768 dimensions
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': 8192, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
  (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("sentence_transformers_model_id")
# Run inference
sentences = [
    'ma-a 12 GÍN i-dí-sú-in KI tù-ra-a i-lá-qé 5',
    'Indeed, Iddin-Suen will receive 12 shekels from Turaya.',
    'ina pa-ni-szu-nu i-za#-mu-ru ina _ugu_ szA ni-iq-bu-ni ma-a [x x x x x x] ki',
]
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,  0.6474, -0.1447],
#         [ 0.6474,  1.0000, -0.0218],
#         [-0.1447, -0.0218,  1.0000]])

Evaluation

Metrics

Information Retrieval

Dataset: akkadian-ir
Evaluated with InformationRetrievalEvaluator with these parameters:
json
1{ 2 "truncate_dim": 384 3}

Metric	Value
cosine_accuracy@1	0.5072
cosine_accuracy@3	0.7332
cosine_accuracy@5	0.7707
cosine_accuracy@10	0.8068
cosine_precision@1	0.5072
cosine_precision@3	0.2444
cosine_precision@5	0.1541
cosine_precision@10	0.0807
cosine_recall@1	0.5072
cosine_recall@3	0.7332
cosine_recall@5	0.7707
cosine_recall@10	0.8068
cosine_ndcg@10	0.6698
cosine_mrr@10	0.6246
cosine_map@100	0.6276

Training Details

Training Dataset

Unnamed Dataset

Size: 371,366 training samples
Columns: anchor, positive, and negative
Approximate statistics based on the first 1000 samples:
anchor positive negative
type string string string
details
min: 3 tokens
mean: 38.93 tokens
max: 606 tokens
min: 3 tokens
mean: 52.63 tokens
max: 1105 tokens
min: 2 tokens
mean: 23.5 tokens
max: 335 tokens

Samples:

anchor	positive	negative
`kà-sí-im (kāsu)`	`m. & f.; pl. f. "cup, bowl" [GAL; MB on (DUG.)GÚ.ZI] freq. of metal; for oil, wine; MB, NA as measure of capacity`	`If in Nisannu Month I, for his house ....`
`SIG5 ša-bu-ra-am i-ṣé-er i-dí-a-šur DUMU dan-a-šur PUZUR4.IŠTAR a-hi-šu ù i-ku-pí-a DUMU a-šur-i-mì-tí iš-ma-a-šur i-šu`	`Idī-Aššur s. Dān-Aššur, brother of Puzur-Ištar, and Ikuppiya s. Aššur-imitt owe Išme-Aššur 14 talents broken refined copper.`	`DUMU {1}-ba-da-a.a`
`mì-šu ṣú-ha-ru-ša ša-lim-a-šur ù a-li-ku a-dí šé-ni-šu i-li-ku-ni-ma té-er-ta-ak-nu-ma lá i-li-kà-ni`	`Why is it that Šalim-Aššur's servants and other travellers have come here twice, but no message from you has arrived?`	`servant Ina-šar-Bel-allak`

Loss: MatryoshkaLoss with these parameters:
json
1{ 2 "loss": "MultipleNegativesRankingLoss", 3 "matryoshka_dims": [ 4 768, 5 512, 6 384, 7 256, 8 128 9 ], 10 "matryoshka_weights": [ 11 1, 12 1, 13 1, 14 1, 15 1 16 ], 17 "n_dims_per_step": -1 18}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 32
learning_rate: 2e-05
lr_scheduler_type: cosine
warmup_ratio: 0.1
fp16: True
load_best_model_at_end: True

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: 8
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: 2e-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: 3
max_steps: -1
lr_scheduler_type: cosine
lr_scheduler_kwargs: None
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: True
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: True
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	akkadian-ir_cosine_ndcg@10
0.0043	50	24.2917	-
0.0086	100	22.3808	-
0.0129	150	19.4952	-
0.0172	200	16.7314	-
0.0215	250	14.4493	-
0.0258	300	12.8579	-
0.0302	350	11.6765	-
0.0345	400	11.056	-
0.0388	450	10.3627	-
0.0431	500	9.5568	-
0.0474	550	9.1752	-
0.0517	600	8.7544	-
0.0560	650	8.7637	-
0.0603	700	8.3496	-
0.0646	750	8.0293	-
0.0689	800	7.5629	-
0.0732	850	7.682	-
0.0775	900	7.2793	-
0.0819	950	7.2354	-
0.0862	1000	7.0245	-
0.0905	1050	6.7	-
0.0948	1100	6.8599	-
0.0991	1150	6.1292	-
0.0999	1160	-	0.1951
0.1034	1200	6.0634	-
0.1077	1250	5.9424	-
0.1120	1300	6.3258	-
0.1163	1350	5.8804	-
0.1206	1400	5.9022	-
0.1249	1450	5.7101	-
0.1292	1500	5.6781	-
0.1336	1550	5.603	-
0.1379	1600	5.4788	-
0.1422	1650	5.5066	-
0.1465	1700	5.6286	-
0.1508	1750	5.2864	-
0.1551	1800	5.399	-
0.1594	1850	5.1604	-
0.1637	1900	5.264	-
0.1680	1950	5.3405	-
0.1723	2000	5.0218	-
0.1766	2050	5.2333	-
0.1809	2100	4.9349	-
0.1852	2150	4.6845	-
0.1896	2200	5.1475	-
0.1939	2250	4.5305	-
0.1982	2300	4.6658	-
0.1999	2320	-	0.2463
0.2025	2350	4.618	-
0.2068	2400	4.6903	-
0.2111	2450	4.4551	-
0.2154	2500	4.7722	-
0.2197	2550	4.3916	-
0.2240	2600	4.1282	-
0.2283	2650	4.3277	-
0.2326	2700	4.6267	-
0.2369	2750	4.3596	-
0.2413	2800	4.3899	-
0.2456	2850	4.2	-
0.2499	2900	4.1903	-
0.2542	2950	4.1434	-
0.2585	3000	4.2724	-
0.2628	3050	4.244	-
0.2671	3100	4.1991	-
0.2714	3150	4.0842	-
0.2757	3200	3.9193	-
0.2800	3250	3.8654	-
0.2843	3300	3.9076	-
0.2886	3350	3.4862	-
0.2930	3400	3.7306	-
0.2973	3450	3.8205	-
0.2998	3480	-	0.2908
0.3016	3500	4.0037	-
0.3059	3550	3.5835	-
0.3102	3600	3.7554	-
0.3145	3650	3.4443	-
0.3188	3700	3.8453	-
0.3231	3750	3.5481	-
0.3274	3800	3.6546	-
0.3317	3850	3.4082	-
0.3360	3900	3.2601	-
0.3403	3950	3.5107	-
0.3446	4000	3.1638	-
0.3490	4050	3.3906	-
0.3533	4100	3.5139	-
0.3576	4150	3.2548	-
0.3619	4200	3.392	-
0.3662	4250	3.292	-
0.3705	4300	3.0331	-
0.3748	4350	2.8747	-
0.3791	4400	3.193	-
0.3834	4450	3.1662	-
0.3877	4500	2.9548	-
0.3920	4550	3.1211	-
0.3963	4600	2.9486	-
0.3998	4640	-	0.3222
0.4007	4650	3.0281	-
0.4050	4700	2.9552	-
0.4093	4750	2.6024	-
0.4136	4800	2.8493	-
0.4179	4850	2.7818	-
0.4222	4900	2.8218	-
0.4265	4950	2.5303	-
0.4308	5000	2.5312	-
0.4351	5050	2.8386	-
0.4394	5100	2.6784	-
0.4437	5150	2.7933	-
0.4480	5200	2.6402	-
0.4524	5250	2.7994	-
0.4567	5300	2.8292	-
0.4610	5350	2.6279	-
0.4653	5400	2.4097	-
0.4696	5450	2.7501	-
0.4739	5500	2.3796	-
0.4782	5550	2.6051	-
0.4825	5600	2.6986	-
0.4868	5650	2.4088	-
0.4911	5700	2.5498	-
0.4954	5750	2.3827	-
0.4997	5800	2.5159	0.3500
0.5040	5850	2.4432	-
0.5084	5900	2.1923	-
0.5127	5950	2.4678	-
0.5170	6000	2.228	-
0.5213	6050	2.2555	-
0.5256	6100	2.4221	-
0.5299	6150	2.3692	-
0.5342	6200	2.5304	-
0.5385	6250	2.2569	-
0.5428	6300	2.0883	-
0.5471	6350	2.2691	-
0.5514	6400	2.2558	-
0.5557	6450	2.2126	-
0.5601	6500	2.1121	-
0.5644	6550	2.12	-
0.5687	6600	2.2115	-
0.5730	6650	1.9303	-
0.5773	6700	1.9711	-
0.5816	6750	2.1382	-
0.5859	6800	1.9612	-
0.5902	6850	1.9234	-
0.5945	6900	2.1105	-
0.5988	6950	1.9214	-
0.5997	6960	-	0.3794
0.6031	7000	1.8454	-
0.6074	7050	2.127	-
0.6118	7100	2.0367	-
0.6161	7150	2.0193	-
0.6204	7200	1.8004	-
0.6247	7250	2.0138	-
0.6290	7300	1.789	-
0.6333	7350	1.9486	-
0.6376	7400	1.9889	-
0.6419	7450	2.0563	-
0.6462	7500	1.9492	-
0.6505	7550	1.8981	-
0.6548	7600	1.8442	-
0.6591	7650	1.852	-
0.6634	7700	1.7902	-
0.6678	7750	1.6871	-
0.6721	7800	1.698	-
0.6764	7850	1.5765	-
0.6807	7900	1.8773	-
0.6850	7950	1.7695	-
0.6893	8000	1.621	-
0.6936	8050	1.492	-
0.6979	8100	1.6412	-
0.6996	8120	-	0.4046
0.7022	8150	1.7606	-
0.7065	8200	1.5547	-
0.7108	8250	1.7866	-
0.7151	8300	1.531	-
0.7195	8350	1.7266	-
0.7238	8400	1.4949	-
0.7281	8450	1.9541	-
0.7324	8500	1.6818	-
0.7367	8550	1.4678	-
0.7410	8600	1.8328	-
0.7453	8650	1.5184	-
0.7496	8700	1.6247	-
0.7539	8750	1.5787	-
0.7582	8800	1.6704	-
0.7625	8850	1.5755	-
0.7668	8900	1.6273	-
0.7712	8950	1.614	-
0.7755	9000	1.5335	-
0.7798	9050	1.461	-
0.7841	9100	1.5011	-
0.7884	9150	1.6853	-
0.7927	9200	1.4713	-
0.7970	9250	1.504	-
0.7996	9280	-	0.4436
0.8013	9300	1.5662	-
0.8056	9350	1.3562	-
0.8099	9400	1.4698	-
0.8142	9450	1.5387	-
0.8185	9500	1.3739	-
0.8229	9550	1.4344	-
0.8272	9600	1.5813	-
0.8315	9650	1.5476	-
0.8358	9700	1.4192	-
0.8401	9750	1.5959	-
0.8444	9800	1.463	-
0.8487	9850	1.5049	-
0.8530	9900	1.5464	-
0.8573	9950	1.5782	-
0.8616	10000	1.4452	-
0.8659	10050	1.3905	-
0.8702	10100	1.5898	-
0.8745	10150	1.3744	-
0.8789	10200	1.2622	-
0.8832	10250	1.1547	-
0.8875	10300	1.3283	-
0.8918	10350	1.4365	-
0.8961	10400	1.5452	-
0.8995	10440	-	0.4659
0.9004	10450	1.3644	-
0.9047	10500	1.4959	-
0.9090	10550	1.4951	-
0.9133	10600	1.3366	-
0.9176	10650	1.5537	-
0.9219	10700	1.2168	-
0.9262	10750	1.2671	-
0.9306	10800	1.2388	-
0.9349	10850	1.4667	-
0.9392	10900	1.2911	-
0.9435	10950	1.2547	-
0.9478	11000	1.4643	-
0.9521	11050	1.4337	-
0.9564	11100	1.2031	-
0.9607	11150	1.3594	-
0.9650	11200	1.3133	-
0.9693	11250	1.2628	-
0.9736	11300	1.116	-
0.9779	11350	1.2652	-
0.9823	11400	1.2119	-
0.9866	11450	1.1888	-
0.9909	11500	1.2845	-
0.9952	11550	1.399	-
0.9995	11600	1.0896	0.4985
1.0038	11650	1.1697	-
1.0081	11700	1.189	-
1.0124	11750	1.2988	-
1.0167	11800	1.178	-
1.0210	11850	1.4166	-
1.0253	11900	1.1385	-
1.0296	11950	1.1459	-
1.0339	12000	1.2123	-
1.0383	12050	1.0782	-
1.0426	12100	1.2136	-
1.0469	12150	1.2298	-
1.0512	12200	1.2266	-
1.0555	12250	1.1184	-
1.0598	12300	1.1255	-
1.0641	12350	1.2786	-
1.0684	12400	1.2258	-
1.0727	12450	1.2677	-
1.0770	12500	1.1009	-
1.0813	12550	1.3069	-
1.0856	12600	1.1574	-
1.0900	12650	1.232	-
1.0943	12700	1.3349	-
1.0986	12750	1.0868	-
1.0994	12760	-	0.5223
1.1029	12800	1.1968	-
1.1072	12850	1.1317	-
1.1115	12900	1.0791	-
1.1158	12950	1.1399	-
1.1201	13000	1.1907	-
1.1244	13050	1.322	-
1.1287	13100	1.2167	-
1.1330	13150	1.1696	-
1.1373	13200	1.2748	-
1.1417	13250	1.2751	-
1.1460	13300	1.2965	-
1.1503	13350	1.1097	-
1.1546	13400	1.3141	-
1.1589	13450	1.2249	-
1.1632	13500	1.4477	-
1.1675	13550	1.1688	-
1.1718	13600	1.2521	-
1.1761	13650	1.0834	-
1.1804	13700	1.2089	-
1.1847	13750	1.0982	-
1.1890	13800	1.2871	-
1.1933	13850	1.053	-
1.1977	13900	1.1601	-
1.1994	13920	-	0.5383
1.2020	13950	1.2559	-
1.2063	14000	1.076	-
1.2106	14050	1.2375	-
1.2149	14100	1.1363	-
1.2192	14150	1.1253	-
1.2235	14200	1.0961	-
1.2278	14250	1.1226	-
1.2321	14300	1.0251	-
1.2364	14350	1.087	-
1.2407	14400	1.1262	-
1.2450	14450	1.2847	-
1.2494	14500	1.1392	-
1.2537	14550	1.2119	-
1.2580	14600	1.0831	-
1.2623	14650	1.1392	-
1.2666	14700	1.2348	-
1.2709	14750	1.1431	-
1.2752	14800	1.1248	-
1.2795	14850	1.1533	-
1.2838	14900	1.134	-
1.2881	14950	1.1922	-
1.2924	15000	1.2331	-
1.2967	15050	1.1185	-
1.2993	15080	-	0.5594
1.3011	15100	1.3496	-
1.3054	15150	1.0629	-
1.3097	15200	1.2785	-
1.3140	15250	1.2427	-
1.3183	15300	1.2051	-
1.3226	15350	0.9325	-
1.3269	15400	1.0465	-
1.3312	15450	1.1105	-
1.3355	15500	1.1853	-
1.3398	15550	1.1192	-
1.3441	15600	1.0018	-
1.3484	15650	1.1357	-
1.3527	15700	1.2298	-
1.3571	15750	1.0783	-
1.3614	15800	1.271	-
1.3657	15850	1.1724	-
1.3700	15900	1.273	-
1.3743	15950	1.2049	-
1.3786	16000	0.9902	-
1.3829	16050	1.1044	-
1.3872	16100	1.1175	-
1.3915	16150	1.0599	-
1.3958	16200	1.1392	-
1.3993	16240	-	0.5806
1.4001	16250	1.1629	-
1.4044	16300	1.1323	-
1.4088	16350	1.2096	-
1.4131	16400	0.9091	-
1.4174	16450	1.1328	-
1.4217	16500	1.1584	-
1.4260	16550	1.2615	-
1.4303	16600	1.1547	-
1.4346	16650	1.0805	-
1.4389	16700	1.2107	-
1.4432	16750	1.1184	-
1.4475	16800	1.0953	-
1.4518	16850	1.2088	-
1.4561	16900	1.0663	-
1.4605	16950	1.0531	-
1.4648	17000	1.0374	-
1.4691	17050	1.1432	-
1.4734	17100	1.0345	-
1.4777	17150	1.0081	-
1.4820	17200	1.0979	-
1.4863	17250	1.0554	-
1.4906	17300	1.1095	-
1.4949	17350	1.1157	-
1.4992	17400	1.0901	0.5940
1.5035	17450	1.2183	-
1.5078	17500	1.1127	-
1.5121	17550	0.9928	-
1.5165	17600	1.0612	-
1.5208	17650	1.2894	-
1.5251	17700	1.0407	-
1.5294	17750	1.0467	-
1.5337	17800	1.1305	-
1.5380	17850	1.2103	-
1.5423	17900	1.0317	-
1.5466	17950	0.8727	-
1.5509	18000	1.0039	-
1.5552	18050	1.1078	-
1.5595	18100	0.8985	-
1.5638	18150	1.073	-
1.5682	18200	1.1185	-
1.5725	18250	1.1867	-
1.5768	18300	1.0053	-
1.5811	18350	1.0772	-
1.5854	18400	1.1199	-
1.5897	18450	1.1933	-
1.5940	18500	1.1376	-
1.5983	18550	1.0323	-
1.5992	18560	-	0.6092
1.6026	18600	1.1533	-
1.6069	18650	1.1542	-
1.6112	18700	0.8537	-
1.6155	18750	1.2019	-
1.6199	18800	0.9037	-
1.6242	18850	1.1072	-
1.6285	18900	0.9368	-
1.6328	18950	0.8755	-
1.6371	19000	1.0589	-
1.6414	19050	1.2077	-
1.6457	19100	1.0273	-
1.6500	19150	0.9574	-
1.6543	19200	0.9654	-
1.6586	19250	0.9936	-
1.6629	19300	0.936	-
1.6672	19350	1.1334	-
1.6715	19400	1.1132	-
1.6759	19450	0.9652	-
1.6802	19500	0.9999	-
1.6845	19550	1.0588	-
1.6888	19600	0.8735	-
1.6931	19650	1.0931	-
1.6974	19700	0.9329	-
1.6991	19720	-	0.6159
1.7017	19750	1.0249	-
1.7060	19800	0.9529	-
1.7103	19850	1.0974	-
1.7146	19900	1.156	-
1.7189	19950	1.2541	-
1.7232	20000	1.1157	-
1.7276	20050	0.9739	-
1.7319	20100	0.8053	-
1.7362	20150	0.9672	-
1.7405	20200	0.9638	-
1.7448	20250	1.0336	-
1.7491	20300	1.0707	-
1.7534	20350	1.1464	-
1.7577	20400	0.9545	-
1.7620	20450	1.0381	-
1.7663	20500	1.217	-
1.7706	20550	1.1779	-
1.7749	20600	0.8474	-
1.7793	20650	1.062	-
1.7836	20700	0.8884	-
1.7879	20750	1.1615	-
1.7922	20800	1.0987	-
1.7965	20850	1.126	-
1.7991	20880	-	0.6320
1.8008	20900	1.0833	-
1.8051	20950	1.049	-
1.8094	21000	1.0177	-
1.8137	21050	1.1588	-
1.8180	21100	0.9397	-
1.8223	21150	0.9947	-
1.8266	21200	1.0446	-
1.8309	21250	1.1826	-
1.8353	21300	0.9498	-
1.8396	21350	1.3614	-
1.8439	21400	1.1025	-
1.8482	21450	1.028	-
1.8525	21500	1.0175	-
1.8568	21550	0.8465	-
1.8611	21600	0.9803	-
1.8654	21650	0.8592	-
1.8697	21700	0.9792	-
1.8740	21750	1.0933	-
1.8783	21800	0.8312	-
1.8826	21850	1.0615	-
1.8870	21900	1.0027	-
1.8913	21950	1.1034	-
1.8956	22000	1.0831	-
1.8990	22040	-	0.6385
1.8999	22050	0.9895	-
1.9042	22100	1.1019	-
1.9085	22150	1.1036	-
1.9128	22200	0.9039	-
1.9171	22250	1.0744	-
1.9214	22300	1.1484	-
1.9257	22350	1.0977	-
1.9300	22400	1.091	-
1.9343	22450	0.8213	-
1.9387	22500	1.0402	-
1.9430	22550	1.1233	-
1.9473	22600	1.0408	-
1.9516	22650	1.1515	-
1.9559	22700	1.1289	-
1.9602	22750	0.8997	-
1.9645	22800	0.9587	-
1.9688	22850	0.9716	-
1.9731	22900	0.9622	-
1.9774	22950	1.0119	-
1.9817	23000	1.0433	-
1.9860	23050	1.1165	-
1.9903	23100	0.9443	-
1.9947	23150	1.0661	-
1.9990	23200	1.0166	0.6490
2.0033	23250	0.982	-
2.0076	23300	1.1731	-
2.0119	23350	1.0112	-
2.0162	23400	1.0373	-
2.0205	23450	0.8866	-
2.0248	23500	0.9581	-
2.0291	23550	1.2335	-
2.0334	23600	0.9536	-
2.0377	23650	0.9767	-
2.0420	23700	1.0382	-
2.0464	23750	1.1288	-
2.0507	23800	0.8292	-
2.0550	23850	1.1083	-
2.0593	23900	0.9252	-
2.0636	23950	1.1108	-
2.0679	24000	1.1602	-
2.0722	24050	0.9616	-
2.0765	24100	1.0108	-
2.0808	24150	1.0974	-
2.0851	24200	0.9542	-
2.0894	24250	0.9269	-
2.0937	24300	1.0494	-
2.0981	24350	1.074	-
2.0989	24360	-	0.6508
2.1024	24400	0.8881	-
2.1067	24450	1.0372	-
2.1110	24500	1.0833	-
2.1153	24550	1.1226	-
2.1196	24600	1.1199	-
2.1239	24650	0.9263	-
2.1282	24700	0.9799	-
2.1325	24750	0.9388	-
2.1368	24800	1.1606	-
2.1411	24850	0.998	-
2.1454	24900	1.1349	-
2.1498	24950	1.1257	-
2.1541	25000	0.9132	-
2.1584	25050	1.068	-
2.1627	25100	0.9177	-
2.1670	25150	1.0174	-
2.1713	25200	1.1028	-
2.1756	25250	0.9742	-
2.1799	25300	0.9095	-
2.1842	25350	0.9706	-
2.1885	25400	1.1514	-
2.1928	25450	1.1459	-
2.1971	25500	1.2085	-
2.1989	25520	-	0.6582
2.2014	25550	0.9208	-
2.2058	25600	1.1146	-
2.2101	25650	1.003	-
2.2144	25700	0.9197	-
2.2187	25750	1.029	-
2.2230	25800	1.1035	-
2.2273	25850	1.0754	-
2.2316	25900	1.1675	-
2.2359	25950	0.9714	-
2.2402	26000	1.116	-
2.2445	26050	1.0347	-
2.2488	26100	1.027	-
2.2531	26150	0.9373	-
2.2575	26200	1.1202	-
2.2618	26250	0.8809	-
2.2661	26300	1.0182	-
2.2704	26350	1.0033	-
2.2747	26400	0.9937	-
2.2790	26450	1.0004	-
2.2833	26500	1.0014	-
2.2876	26550	1.2473	-
2.2919	26600	1.0909	-
2.2962	26650	1.1588	-
2.2988	26680	-	0.6620
2.3005	26700	0.9797	-
2.3048	26750	1.1163	-
2.3092	26800	1.1619	-
2.3135	26850	0.9237	-
2.3178	26900	0.9336	-
2.3221	26950	1.0761	-
2.3264	27000	1.0624	-
2.3307	27050	0.9467	-
2.3350	27100	1.2416	-
2.3393	27150	0.8832	-
2.3436	27200	1.0419	-
2.3479	27250	0.8805	-
2.3522	27300	1.063	-
2.3565	27350	1.0	-
2.3608	27400	0.9411	-
2.3652	27450	1.2561	-
2.3695	27500	1.0111	-
2.3738	27550	0.9595	-
2.3781	27600	0.8381	-
2.3824	27650	1.0234	-
2.3867	27700	0.8935	-
2.3910	27750	0.8965	-
2.3953	27800	1.0653	-
2.3988	27840	-	0.6641
2.3996	27850	1.0907	-
2.4039	27900	1.0517	-
2.4082	27950	0.9392	-
2.4125	28000	0.9978	-
2.4169	28050	1.0318	-
2.4212	28100	0.9021	-
2.4255	28150	0.9216	-
2.4298	28200	1.0857	-
2.4341	28250	0.9689	-
2.4384	28300	1.0085	-
2.4427	28350	1.0434	-
2.4470	28400	1.1309	-
2.4513	28450	0.9319	-
2.4556	28500	0.9562	-
2.4599	28550	0.9197	-
2.4642	28600	1.2111	-
2.4686	28650	1.0983	-
2.4729	28700	0.9562	-
2.4772	28750	0.9327	-
2.4815	28800	0.9716	-
2.4858	28850	1.0202	-
2.4901	28900	1.1367	-
2.4944	28950	0.9014	-
2.4987	29000	1.1313	0.6672
2.5030	29050	1.148	-
2.5073	29100	0.799	-
2.5116	29150	1.0012	-
2.5159	29200	0.7844	-
2.5202	29250	1.1639	-
2.5246	29300	0.9905	-
2.5289	29350	1.0579	-
2.5332	29400	0.9329	-
2.5375	29450	0.9496	-
2.5418	29500	0.9521	-
2.5461	29550	0.8535	-
2.5504	29600	1.019	-
2.5547	29650	1.1031	-
2.5590	29700	1.0894	-
2.5633	29750	1.0078	-
2.5676	29800	0.8403	-
2.5719	29850	0.916	-
2.5763	29900	1.2096	-
2.5806	29950	0.9969	-
2.5849	30000	1.1598	-
2.5892	30050	0.8849	-
2.5935	30100	1.0619	-
2.5978	30150	0.9554	-
2.5987	30160	-	0.6693
2.6021	30200	1.1025	-
2.6064	30250	1.1252	-
2.6107	30300	0.8108	-
2.6150	30350	0.927	-
2.6193	30400	1.1574	-
2.6236	30450	1.0098	-
2.6280	30500	0.8702	-
2.6323	30550	0.9672	-
2.6366	30600	0.9361	-
2.6409	30650	0.9801	-
2.6452	30700	1.114	-
2.6495	30750	0.8666	-
2.6538	30800	0.9648	-
2.6581	30850	0.9423	-
2.6624	30900	1.059	-
2.6667	30950	0.9149	-
2.6710	31000	0.8954	-
2.6753	31050	0.8769	-
2.6796	31100	0.8124	-
2.6840	31150	1.151	-
2.6883	31200	1.0145	-
2.6926	31250	0.9653	-
2.6969	31300	1.136	-
2.6986	31320	-	0.6693
2.7012	31350	0.9122	-
2.7055	31400	1.0161	-
2.7098	31450	1.0152	-
2.7141	31500	1.1181	-
2.7184	31550	0.8969	-
2.7227	31600	1.2101	-
2.7270	31650	1.0958	-
2.7313	31700	0.9548	-
2.7357	31750	0.9755	-
2.7400	31800	0.9796	-
2.7443	31850	1.0564	-
2.7486	31900	0.9581	-
2.7529	31950	0.8607	-
2.7572	32000	0.8933	-
2.7615	32050	0.9828	-
2.7658	32100	1.1992	-
2.7701	32150	1.0162	-
2.7744	32200	0.8406	-
2.7787	32250	0.7896	-
2.7830	32300	1.0311	-
2.7874	32350	1.0507	-
2.7917	32400	1.136	-
2.7960	32450	1.0504	-
2.7986	32480	-	0.6697
2.8003	32500	0.9271	-
2.8046	32550	1.0412	-
2.8089	32600	0.8542	-
2.8132	32650	1.1015	-
2.8175	32700	0.9957	-
2.8218	32750	1.0845	-
2.8261	32800	1.1226	-
2.8304	32850	1.0235	-
2.8347	32900	0.996	-
2.8390	32950	1.0855	-
2.8434	33000	1.2322	-
2.8477	33050	0.999	-
2.8520	33100	1.04	-
2.8563	33150	1.1466	-
2.8606	33200	0.9061	-
2.8649	33250	1.0011	-
2.8692	33300	1.0205	-
2.8735	33350	1.0136	-
2.8778	33400	0.8956	-
2.8821	33450	0.9722	-
2.8864	33500	0.8962	-
2.8907	33550	0.9545	-
2.8951	33600	0.8474	-
2.8985	33640	-	0.6700
2.8994	33650	0.782	-
2.9037	33700	0.9551	-
2.9080	33750	1.0217	-
2.9123	33800	0.8188	-
2.9166	33850	1.0652	-
2.9209	33900	1.1314	-
2.9252	33950	0.9487	-
2.9295	34000	0.9906	-
2.9338	34050	1.1317	-
2.9381	34100	0.9139	-
2.9424	34150	0.9394	-
2.9468	34200	0.9904	-
2.9511	34250	1.0758	-
2.9554	34300	0.9388	-
2.9597	34350	0.9417	-
2.9640	34400	0.9871	-
2.9683	34450	1.0431	-
2.9726	34500	1.0538	-
2.9769	34550	1.078	-
2.9812	34600	1.0972	-
2.9855	34650	1.0294	-
2.9898	34700	1.0387	-
2.9941	34750	0.8923	-
2.9984	34800	1.0937	0.6698

Framework Versions

Python: 3.12.3
Sentence Transformers: 5.3.0
Transformers: 4.57.6
PyTorch: 2.9.1+cu128
Accelerate: 1.13.0
Datasets: 4.6.1
Tokenizers: 0.22.2

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",
}

MatryoshkaLoss

bibtex

@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

bibtex

@misc{oord2019representationlearningcontrastivepredictive,
      title={Representation Learning with Contrastive Predictive Coding},
      author={Aaron van den Oord and Yazhe Li and Oriol Vinyals},
      year={2019},
      eprint={1807.03748},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/1807.03748},
}

	anchor	positive	negative
type	string	string	string
details	min: 3 tokens mean: 38.93 tokens max: 606 tokens	min: 3 tokens mean: 52.63 tokens max: 1105 tokens	min: 2 tokens mean: 23.5 tokens max: 335 tokens

Noddybear/akkadian-embed

SentenceTransformer based on nomic-ai/modernbert-embed-base

Model Details

Model Description

Model Sources

Full Model Architecture

Usage

Direct Usage (Sentence Transformers)

Evaluation

Metrics

Information Retrieval

Training Details

Training Dataset

Unnamed Dataset

Training Hyperparameters

Non-Default Hyperparameters

All Hyperparameters

Training Logs

Framework Versions

Citation

BibTeX

Sentence Transformers

MatryoshkaLoss

MultipleNegativesRankingLoss

Noddybear/
akkadian-embed