metadata

base_model: sdadas/mmlw-roberta-base
language:
  - en
library_name: sentence-transformers
license: apache-2.0
metrics:
  - cosine_accuracy@1
  - cosine_accuracy@3
  - cosine_accuracy@5
  - cosine_accuracy@10
  - cosine_precision@1
  - cosine_precision@3
  - cosine_precision@5
  - cosine_precision@10
  - cosine_recall@1
  - cosine_recall@3
  - cosine_recall@5
  - cosine_recall@10
  - cosine_ndcg@10
  - cosine_mrr@10
  - cosine_map@100
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dataset_size:1K<n<10K
  - loss:MatryoshkaLoss
  - loss:MultipleNegativesRankingLoss
widget:
  - source_sentence: Żywot św. Stanisława
    sentences:
      - czym różni się Żywot św. Stanisława od Legendy św. Stanisława?
      - w którym kraju w noc sylwestrową je się oliebollen?
      - Pierwsze bloki mieszkalne powstały pod koniec lat 80.
  - source_sentence: Herkules na rozstajach
    sentences:
      - jak zinterpretować wymowę obrazu Herkules na rozstajach?
      - gdzie zginął przedwojenny minister Antoni Olszewski?
      - kiedy konsekrowano katedrę św. Teresy z Avili w Požedze?
  - source_sentence: gdzie rośnie bokkonia?
    sentences:
      - gdzie występuje rogownica szerokolistna?
      - Ochrzcił w sierpniu 1982 ich syna księcia Wilhelma.
      - Pośmiertnie został odznaczony Krzyżem Virtuti Militari.
  - source_sentence: czym jest Kompas Sztuki?
    sentences:
      - ' Projekt Kompas Sztuki: Galeria m2 (m kwadrat).'
      - 'Do rodzaju Caraipa zaliczanych jest ok. 55 gatunków:'
      - kto jest aktualnym rekordzistą Chorwacji w skoku w dal?
  - source_sentence: Dalsze losy relikwii
    sentences:
      - Losy relikwii świętego
      - czemu gra The Saboteur wywołała wiele kontrowersji?
      - >-
        kto jest pierwszym rosyjskim kierowcą wyścigowym startującym w Formule
        1?
model-index:
  - name: mmlw-roberta-base-klej-dyk-v0.1
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 768
          type: dim_768
        metrics:
          - type: cosine_accuracy@1
            value: 0.18990384615384615
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.5865384615384616
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.7692307692307693
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.8533653846153846
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.18990384615384615
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.1955128205128205
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.15384615384615383
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.08533653846153846
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.18990384615384615
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.5865384615384616
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.7692307692307693
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.8533653846153846
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.5204892782178483
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.4127814026251526
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.418150211843158
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 512
          type: dim_512
        metrics:
          - type: cosine_accuracy@1
            value: 0.1875
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.5889423076923077
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.7596153846153846
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.8629807692307693
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.1875
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.19631410256410253
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.15192307692307688
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.08629807692307694
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.1875
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.5889423076923077
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.7596153846153846
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.8629807692307693
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.5204340563935984
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.4100885225885227
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.4147514658961434
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 256
          type: dim_256
        metrics:
          - type: cosine_accuracy@1
            value: 0.19471153846153846
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.5649038461538461
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.7451923076923077
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.8461538461538461
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.19471153846153846
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.18830128205128205
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.1490384615384615
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.08461538461538462
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.19471153846153846
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.5649038461538461
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.7451923076923077
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.8461538461538461
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.5144907264607753
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.4078373015873016
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.413093644747221
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 128
          type: dim_128
        metrics:
          - type: cosine_accuracy@1
            value: 0.18269230769230768
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.5192307692307693
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.7163461538461539
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.8293269230769231
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.18269230769230768
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.17307692307692307
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.14326923076923076
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.08293269230769229
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.18269230769230768
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.5192307692307693
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.7163461538461539
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.8293269230769231
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.4955346842225082
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.38889652014651993
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.39396452853345754
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 64
          type: dim_64
        metrics:
          - type: cosine_accuracy@1
            value: 0.1778846153846154
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.4831730769230769
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.6514423076923077
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.7740384615384616
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.1778846153846154
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.16105769230769232
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.13028846153846152
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.07740384615384614
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.1778846153846154
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.4831730769230769
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.6514423076923077
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.7740384615384616
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.4639263641936578
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.36540083180708166
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.3728380879103276
            name: Cosine Map@100

mmlw-roberta-base-klej-dyk-v0.1

This is a sentence-transformers model finetuned from sdadas/mmlw-roberta-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: sdadas/mmlw-roberta-base
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 tokens
Similarity Function: Cosine Similarity
Language: en
License: apache-2.0

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': True, 'pooling_mode_mean_tokens': False, '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 = [
    'Dalsze losy relikwii',
    'Losy relikwii świętego',
    'czemu gra The Saboteur wywołała wiele kontrowersji?',
]
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]

Evaluation

Metrics

Information Retrieval

Dataset: dim_768
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1899
cosine_accuracy@3	0.5865
cosine_accuracy@5	0.7692
cosine_accuracy@10	0.8534
cosine_precision@1	0.1899
cosine_precision@3	0.1955
cosine_precision@5	0.1538
cosine_precision@10	0.0853
cosine_recall@1	0.1899
cosine_recall@3	0.5865
cosine_recall@5	0.7692
cosine_recall@10	0.8534
cosine_ndcg@10	0.5205
cosine_mrr@10	0.4128
cosine_map@100	0.4182

Information Retrieval

Dataset: dim_512
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1875
cosine_accuracy@3	0.5889
cosine_accuracy@5	0.7596
cosine_accuracy@10	0.863
cosine_precision@1	0.1875
cosine_precision@3	0.1963
cosine_precision@5	0.1519
cosine_precision@10	0.0863
cosine_recall@1	0.1875
cosine_recall@3	0.5889
cosine_recall@5	0.7596
cosine_recall@10	0.863
cosine_ndcg@10	0.5204
cosine_mrr@10	0.4101
cosine_map@100	0.4148

Information Retrieval

Dataset: dim_256
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1947
cosine_accuracy@3	0.5649
cosine_accuracy@5	0.7452
cosine_accuracy@10	0.8462
cosine_precision@1	0.1947
cosine_precision@3	0.1883
cosine_precision@5	0.149
cosine_precision@10	0.0846
cosine_recall@1	0.1947
cosine_recall@3	0.5649
cosine_recall@5	0.7452
cosine_recall@10	0.8462
cosine_ndcg@10	0.5145
cosine_mrr@10	0.4078
cosine_map@100	0.4131

Information Retrieval

Dataset: dim_128
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1827
cosine_accuracy@3	0.5192
cosine_accuracy@5	0.7163
cosine_accuracy@10	0.8293
cosine_precision@1	0.1827
cosine_precision@3	0.1731
cosine_precision@5	0.1433
cosine_precision@10	0.0829
cosine_recall@1	0.1827
cosine_recall@3	0.5192
cosine_recall@5	0.7163
cosine_recall@10	0.8293
cosine_ndcg@10	0.4955
cosine_mrr@10	0.3889
cosine_map@100	0.394

Information Retrieval

Dataset: dim_64
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1779
cosine_accuracy@3	0.4832
cosine_accuracy@5	0.6514
cosine_accuracy@10	0.774
cosine_precision@1	0.1779
cosine_precision@3	0.1611
cosine_precision@5	0.1303
cosine_precision@10	0.0774
cosine_recall@1	0.1779
cosine_recall@3	0.4832
cosine_recall@5	0.6514
cosine_recall@10	0.774
cosine_ndcg@10	0.4639
cosine_mrr@10	0.3654
cosine_map@100	0.3728

Training Details

Training Dataset

Unnamed Dataset

Size: 3,738 training samples
Columns: positive and anchor
Approximate statistics based on the first 1000 samples:
positive anchor
type string string
details
min: 5 tokens
mean: 50.1 tokens
max: 466 tokens

min: 6 tokens
mean: 16.62 tokens
max: 49 tokens

	positive	anchor
type	string	string
details	min: 5 tokens mean: 50.1 tokens max: 466 tokens	min: 6 tokens mean: 16.62 tokens max: 49 tokens

Samples:

positive	anchor
`Zespół Blaua (zespół Jabsa, ang. Blau syndrome, BS) – rzadka choroba genetyczna o dziedziczeniu autosomalnym dominującym, charakteryzująca się ziarniniakowym zapaleniem stawów o wczesnym początku, zapaleniem jagodówki (uveitis) i wysypką skórną, a także kamptodaktylią.`	`jakie choroby genetyczne dziedziczą się autosomalnie dominująco?`
`Gorgippia Gorgippia – starożytne miasto bosporańskie nad Morzem Czarnym, którego pozostałości znajdują się obecnie pod współczesną zabudową centralnej części miasta Anapa w Kraju Krasnodarskim w Rosji.`	`gdzie obecnie znajduje się starożytne miasto Gorgippia?`
Ulubionym dystansem Rücker było 400 metrów i to na nim notowała największe indywidualne sukcesy : srebrny medal Mistrzostw Europy juniorów w lekkoatletyce (Saloniki 1991) 6. miejsce w Pucharze Świata w Lekkoatletyce (Hawana 1992) 5. miejsce na Mistrzostwach Europy w Lekkoatletyce (Helsinki 1994) srebro podczas Mistrzostw Świata w Lekkoatletyce (Sewilla 1999) złota medalistka mistrzostw Niemiec Duże sukcesy odnosiła także w sztafecie 4 x 400 metrów : złoto Mistrzostw Europy juniorów w lekkoatletyce (Varaždin 1989) złoty medal Mistrzostw Europy juniorów w lekkoatletyce (Saloniki 1991) brąz na Mistrzostwach Europy w Lekkoatletyce (Helsinki 1994) brązowy medal podczas Igrzysk Olimpijskich (Atlanta 1996) brąz na Halowych Mistrzostwach Świata w Lekkoatletyce (Paryż 1997) złoto Mistrzostw Świata w Lekkoatletyce (Ateny 1997) brązowy medal Mistrzostw Świata w Lekkoatletyce (Sewilla 1999)	`kto zaprojektował medale, które będą wręczane podczas tegorocznych mistrzostw Europy juniorów w lekkoatletyce?`

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: epoch
gradient_accumulation_steps: 8
learning_rate: 2e-05
num_train_epochs: 5
lr_scheduler_type: cosine
warmup_ratio: 0.1
bf16: True
tf32: True
load_best_model_at_end: True
optim: adamw_torch_fused
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: epoch
prediction_loss_only: True
per_device_train_batch_size: 8
per_device_eval_batch_size: 8
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 8
eval_accumulation_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: 5
max_steps: -1
lr_scheduler_type: cosine
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
use_ipex: False
bf16: True
fp16: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: True
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}
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch_fused
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
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Click to expand

Epoch	Step	Training Loss	dim_128_cosine_map@100	dim_256_cosine_map@100	dim_512_cosine_map@100	dim_64_cosine_map@100	dim_768_cosine_map@100
0	0	-	0.3475	0.3675	0.3753	0.2982	0.3798
0.0171	1	2.6683	-	-	-	-	-
0.0342	2	3.2596	-	-	-	-	-
0.0513	3	3.4541	-	-	-	-	-
0.0684	4	2.4201	-	-	-	-	-
0.0855	5	3.5911	-	-	-	-	-
0.1026	6	3.0902	-	-	-	-	-
0.1197	7	2.5999	-	-	-	-	-
0.1368	8	2.892	-	-	-	-	-
0.1538	9	2.8722	-	-	-	-	-
0.1709	10	2.3703	-	-	-	-	-
0.1880	11	2.6833	-	-	-	-	-
0.2051	12	1.9814	-	-	-	-	-
0.2222	13	1.6643	-	-	-	-	-
0.2393	14	1.8493	-	-	-	-	-
0.2564	15	1.5136	-	-	-	-	-
0.2735	16	1.9726	-	-	-	-	-
0.2906	17	1.1505	-	-	-	-	-
0.3077	18	1.3834	-	-	-	-	-
0.3248	19	1.2244	-	-	-	-	-
0.3419	20	1.2107	-	-	-	-	-
0.3590	21	0.8936	-	-	-	-	-
0.3761	22	0.8144	-	-	-	-	-
0.3932	23	0.8353	-	-	-	-	-
0.4103	24	1.572	-	-	-	-	-
0.4274	25	0.9257	-	-	-	-	-
0.4444	26	0.8405	-	-	-	-	-
0.4615	27	0.5621	-	-	-	-	-
0.4786	28	0.4241	-	-	-	-	-
0.4957	29	0.6171	-	-	-	-	-
0.5128	30	0.5989	-	-	-	-	-
0.5299	31	0.2767	-	-	-	-	-
0.5470	32	0.5599	-	-	-	-	-
0.5641	33	0.5964	-	-	-	-	-
0.5812	34	0.9778	-	-	-	-	-
0.5983	35	0.772	-	-	-	-	-
0.6154	36	1.0341	-	-	-	-	-
0.6325	37	0.3503	-	-	-	-	-
0.6496	38	0.8229	-	-	-	-	-
0.6667	39	0.969	-	-	-	-	-
0.6838	40	1.7993	-	-	-	-	-
0.7009	41	0.5542	-	-	-	-	-
0.7179	42	1.332	-	-	-	-	-
0.7350	43	1.1516	-	-	-	-	-
0.7521	44	1.3183	-	-	-	-	-
0.7692	45	1.0865	-	-	-	-	-
0.7863	46	0.6204	-	-	-	-	-
0.8034	47	0.7541	-	-	-	-	-
0.8205	48	0.9362	-	-	-	-	-
0.8376	49	0.3979	-	-	-	-	-
0.8547	50	0.7187	-	-	-	-	-
0.8718	51	0.9217	-	-	-	-	-
0.8889	52	0.4866	-	-	-	-	-
0.9060	53	0.355	-	-	-	-	-
0.9231	54	0.7172	-	-	-	-	-
0.9402	55	0.6007	-	-	-	-	-
0.9573	56	1.1547	-	-	-	-	-
0.9744	57	0.5713	-	-	-	-	-
0.9915	58	0.9089	0.3985	0.4164	0.4264	0.3642	0.4255
1.0085	59	0.594	-	-	-	-	-
1.0256	60	0.6554	-	-	-	-	-
1.0427	61	0.2794	-	-	-	-	-
1.0598	62	0.8654	-	-	-	-	-
1.0769	63	0.9698	-	-	-	-	-
1.0940	64	1.4827	-	-	-	-	-
1.1111	65	0.3159	-	-	-	-	-
1.1282	66	0.255	-	-	-	-	-
1.1453	67	0.9819	-	-	-	-	-
1.1624	68	0.7442	-	-	-	-	-
1.1795	69	0.8199	-	-	-	-	-
1.1966	70	0.2647	-	-	-	-	-
1.2137	71	0.4098	-	-	-	-	-
1.2308	72	0.1608	-	-	-	-	-
1.2479	73	0.2092	-	-	-	-	-
1.2650	74	0.1231	-	-	-	-	-
1.2821	75	0.3203	-	-	-	-	-
1.2991	76	0.1435	-	-	-	-	-
1.3162	77	0.2293	-	-	-	-	-
1.3333	78	0.131	-	-	-	-	-
1.3504	79	0.1662	-	-	-	-	-
1.3675	80	0.094	-	-	-	-	-
1.3846	81	0.1454	-	-	-	-	-
1.4017	82	0.3096	-	-	-	-	-
1.4188	83	0.3188	-	-	-	-	-
1.4359	84	0.1156	-	-	-	-	-
1.4530	85	0.0581	-	-	-	-	-
1.4701	86	0.0543	-	-	-	-	-
1.4872	87	0.0427	-	-	-	-	-
1.5043	88	0.07	-	-	-	-	-
1.5214	89	0.0451	-	-	-	-	-
1.5385	90	0.0646	-	-	-	-	-
1.5556	91	0.1152	-	-	-	-	-
1.5726	92	0.1292	-	-	-	-	-
1.5897	93	0.1591	-	-	-	-	-
1.6068	94	0.1194	-	-	-	-	-
1.6239	95	0.0876	-	-	-	-	-
1.6410	96	0.1018	-	-	-	-	-
1.6581	97	0.3309	-	-	-	-	-
1.6752	98	0.2214	-	-	-	-	-
1.6923	99	0.1536	-	-	-	-	-
1.7094	100	0.1543	-	-	-	-	-
1.7265	101	0.3663	-	-	-	-	-
1.7436	102	0.2719	-	-	-	-	-
1.7607	103	0.1379	-	-	-	-	-
1.7778	104	0.0479	-	-	-	-	-
1.7949	105	0.0757	-	-	-	-	-
1.8120	106	0.059	-	-	-	-	-
1.8291	107	0.119	-	-	-	-	-
1.8462	108	0.1295	-	-	-	-	-
1.8632	109	0.115	-	-	-	-	-
1.8803	110	0.142	-	-	-	-	-
1.8974	111	0.1064	-	-	-	-	-
1.9145	112	0.0959	-	-	-	-	-
1.9316	113	0.0839	-	-	-	-	-
1.9487	114	0.1762	-	-	-	-	-
1.9658	115	0.1986	-	-	-	-	-
1.9829	116	0.0599	-	-	-	-	-
2.0	117	0.1145	0.3869	0.4095	0.4135	0.3664	0.4195
2.0171	118	0.0815	-	-	-	-	-
2.0342	119	0.1052	-	-	-	-	-
2.0513	120	0.1348	-	-	-	-	-
2.0684	121	0.255	-	-	-	-	-
2.0855	122	0.251	-	-	-	-	-
2.1026	123	0.3033	-	-	-	-	-
2.1197	124	0.0385	-	-	-	-	-
2.1368	125	0.0687	-	-	-	-	-
2.1538	126	0.1682	-	-	-	-	-
2.1709	127	0.0774	-	-	-	-	-
2.1880	128	0.0944	-	-	-	-	-
2.2051	129	0.036	-	-	-	-	-
2.2222	130	0.0393	-	-	-	-	-
2.2393	131	0.0387	-	-	-	-	-
2.2564	132	0.0273	-	-	-	-	-
2.2735	133	0.056	-	-	-	-	-
2.2906	134	0.0279	-	-	-	-	-
2.3077	135	0.0557	-	-	-	-	-
2.3248	136	0.0197	-	-	-	-	-
2.3419	137	0.0216	-	-	-	-	-
2.3590	138	0.0212	-	-	-	-	-
2.3761	139	0.0239	-	-	-	-	-
2.3932	140	0.0526	-	-	-	-	-
2.4103	141	0.1072	-	-	-	-	-
2.4274	142	0.0347	-	-	-	-	-
2.4444	143	0.024	-	-	-	-	-
2.4615	144	0.0128	-	-	-	-	-
2.4786	145	0.0089	-	-	-	-	-
2.4957	146	0.0101	-	-	-	-	-
2.5128	147	0.0124	-	-	-	-	-
2.5299	148	0.011	-	-	-	-	-
2.5470	149	0.0182	-	-	-	-	-
2.5641	150	0.0379	-	-	-	-	-
2.5812	151	0.0395	-	-	-	-	-
2.5983	152	0.0372	-	-	-	-	-
2.6154	153	0.031	-	-	-	-	-
2.6325	154	0.0136	-	-	-	-	-
2.6496	155	0.0355	-	-	-	-	-
2.6667	156	0.0296	-	-	-	-	-
2.6838	157	0.0473	-	-	-	-	-
2.7009	158	0.0295	-	-	-	-	-
2.7179	159	0.0576	-	-	-	-	-
2.7350	160	0.0592	-	-	-	-	-
2.7521	161	0.0571	-	-	-	-	-
2.7692	162	0.0221	-	-	-	-	-
2.7863	163	0.0179	-	-	-	-	-
2.8034	164	0.0195	-	-	-	-	-
2.8205	165	0.0291	-	-	-	-	-
2.8376	166	0.024	-	-	-	-	-
2.8547	167	0.0396	-	-	-	-	-
2.8718	168	0.0352	-	-	-	-	-
2.8889	169	0.0431	-	-	-	-	-
2.9060	170	0.0222	-	-	-	-	-
2.9231	171	0.016	-	-	-	-	-
2.9402	172	0.0307	-	-	-	-	-
2.9573	173	0.0439	-	-	-	-	-
2.9744	174	0.0197	-	-	-	-	-
2.9915	175	0.0181	0.3928	0.4120	0.4152	0.3717	0.4180
3.0085	176	0.03	-	-	-	-	-
3.0256	177	0.0325	-	-	-	-	-
3.0427	178	0.0286	-	-	-	-	-
3.0598	179	0.0746	-	-	-	-	-
3.0769	180	0.0677	-	-	-	-	-
3.0940	181	0.0574	-	-	-	-	-
3.1111	182	0.0158	-	-	-	-	-
3.1282	183	0.0092	-	-	-	-	-
3.1453	184	0.0412	-	-	-	-	-
3.1624	185	0.0308	-	-	-	-	-
3.1795	186	0.022	-	-	-	-	-
3.1966	187	0.0157	-	-	-	-	-
3.2137	188	0.0109	-	-	-	-	-
3.2308	189	0.0059	-	-	-	-	-
3.2479	190	0.0206	-	-	-	-	-
3.2650	191	0.0135	-	-	-	-	-
3.2821	192	0.0199	-	-	-	-	-
3.2991	193	0.0124	-	-	-	-	-
3.3162	194	0.0081	-	-	-	-	-
3.3333	195	0.0052	-	-	-	-	-
3.3504	196	0.006	-	-	-	-	-
3.3675	197	0.0074	-	-	-	-	-
3.3846	198	0.0085	-	-	-	-	-
3.4017	199	0.0273	-	-	-	-	-
3.4188	200	0.0363	-	-	-	-	-
3.4359	201	0.0077	-	-	-	-	-
3.4530	202	0.0046	-	-	-	-	-
3.4701	203	0.0067	-	-	-	-	-
3.4872	204	0.0054	-	-	-	-	-
3.5043	205	0.0055	-	-	-	-	-
3.5214	206	0.0052	-	-	-	-	-
3.5385	207	0.004	-	-	-	-	-
3.5556	208	0.0102	-	-	-	-	-
3.5726	209	0.0228	-	-	-	-	-
3.5897	210	0.0315	-	-	-	-	-
3.6068	211	0.0095	-	-	-	-	-
3.6239	212	0.0069	-	-	-	-	-
3.6410	213	0.0066	-	-	-	-	-
3.6581	214	0.0395	-	-	-	-	-
3.6752	215	0.0176	-	-	-	-	-
3.6923	216	0.0156	-	-	-	-	-
3.7094	217	0.0168	-	-	-	-	-
3.7265	218	0.0376	-	-	-	-	-
3.7436	219	0.0149	-	-	-	-	-
3.7607	220	0.0179	-	-	-	-	-
3.7778	221	0.0059	-	-	-	-	-
3.7949	222	0.013	-	-	-	-	-
3.8120	223	0.0081	-	-	-	-	-
3.8291	224	0.0136	-	-	-	-	-
3.8462	225	0.0129	-	-	-	-	-
3.8632	226	0.0132	-	-	-	-	-
3.8803	227	0.0228	-	-	-	-	-
3.8974	228	0.0091	-	-	-	-	-
3.9145	229	0.0112	-	-	-	-	-
3.9316	230	0.0124	-	-	-	-	-
3.9487	231	0.0224	-	-	-	-	-
3.9658	232	0.0191	-	-	-	-	-
3.9829	233	0.0078	-	-	-	-	-
4.0	234	0.0145	0.3959	0.411	0.4154	0.3741	0.4179
4.0171	235	0.0089	-	-	-	-	-
4.0342	236	0.0157	-	-	-	-	-
4.0513	237	0.019	-	-	-	-	-
4.0684	238	0.0315	-	-	-	-	-
4.0855	239	0.0311	-	-	-	-	-
4.1026	240	0.0155	-	-	-	-	-
4.1197	241	0.0078	-	-	-	-	-
4.1368	242	0.0069	-	-	-	-	-
4.1538	243	0.0246	-	-	-	-	-
4.1709	244	0.011	-	-	-	-	-
4.1880	245	0.0169	-	-	-	-	-
4.2051	246	0.0065	-	-	-	-	-
4.2222	247	0.0093	-	-	-	-	-
4.2393	248	0.0059	-	-	-	-	-
4.2564	249	0.0072	-	-	-	-	-
4.2735	250	0.0114	-	-	-	-	-
4.2906	251	0.0048	-	-	-	-	-
4.3077	252	0.0099	-	-	-	-	-
4.3248	253	0.0061	-	-	-	-	-
4.3419	254	0.005	-	-	-	-	-
4.3590	255	0.0077	-	-	-	-	-
4.3761	256	0.0057	-	-	-	-	-
4.3932	257	0.0106	-	-	-	-	-
4.4103	258	0.0176	-	-	-	-	-
4.4274	259	0.0085	-	-	-	-	-
4.4444	260	0.0059	-	-	-	-	-
4.4615	261	0.0063	-	-	-	-	-
4.4786	262	0.003	-	-	-	-	-
4.4957	263	0.0041	-	-	-	-	-
4.5128	264	0.0048	-	-	-	-	-
4.5299	265	0.0037	-	-	-	-	-
4.5470	266	0.0052	-	-	-	-	-
4.5641	267	0.0084	-	-	-	-	-
4.5812	268	0.0183	-	-	-	-	-
4.5983	269	0.0065	-	-	-	-	-
4.6154	270	0.0074	-	-	-	-	-
4.6325	271	0.0046	-	-	-	-	-
4.6496	272	0.009	-	-	-	-	-
4.6667	273	0.01	-	-	-	-	-
4.6838	274	0.0158	-	-	-	-	-
4.7009	275	0.0077	-	-	-	-	-
4.7179	276	0.0259	-	-	-	-	-
4.7350	277	0.0204	-	-	-	-	-
4.7521	278	0.0155	-	-	-	-	-
4.7692	279	0.0101	-	-	-	-	-
4.7863	280	0.0062	-	-	-	-	-
4.8034	281	0.0065	-	-	-	-	-
4.8205	282	0.0115	-	-	-	-	-
4.8376	283	0.0088	-	-	-	-	-
4.8547	284	0.0157	-	-	-	-	-
4.8718	285	0.0145	-	-	-	-	-
4.8889	286	0.0122	-	-	-	-	-
4.9060	287	0.007	-	-	-	-	-
4.9231	288	0.0126	-	-	-	-	-
4.9402	289	0.0094	-	-	-	-	-
4.9573	290	0.016	0.3940	0.4131	0.4148	0.3728	0.4182

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.12.2
Sentence Transformers: 3.0.0
Transformers: 4.41.2
PyTorch: 2.3.1
Accelerate: 0.27.2
Datasets: 2.19.1
Tokenizers: 0.19.1

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}
}