e5 cogcache small
This is a sentence-transformers model finetuned from intfloat/multilingual-e5-small. It maps sentences & paragraphs to a 384-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: intfloat/multilingual-e5-small
- Maximum Sequence Length: 512 tokens
- Output Dimensionality: 384 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: BertModel
(1): Pooling({'word_embedding_dimension': 384, '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.
from sentence_transformers import SentenceTransformer
# Download from the 🤗 Hub
model = SentenceTransformer("srikarvar/e5-small-cogcachedata-1")
# Run inference
sentences = [
'How can I improve my English?',
'How can I improve my Spanish?',
'How can I gain weight?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]
# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
Evaluation
Metrics
Binary Classification
- Dataset:
quora-duplicates-dev
- Evaluated with
BinaryClassificationEvaluator
Metric | Value |
---|---|
cosine_accuracy | 0.6846 |
cosine_accuracy_threshold | 0.8909 |
cosine_f1 | 0.8038 |
cosine_f1_threshold | 0.8909 |
cosine_precision | 0.672 |
cosine_recall | 1.0 |
cosine_ap | 0.7428 |
dot_accuracy | 0.6846 |
dot_accuracy_threshold | 0.8909 |
dot_f1 | 0.8038 |
dot_f1_threshold | 0.8909 |
dot_precision | 0.672 |
dot_recall | 1.0 |
dot_ap | 0.7428 |
manhattan_accuracy | 0.6846 |
manhattan_accuracy_threshold | 6.8578 |
manhattan_f1 | 0.8038 |
manhattan_f1_threshold | 7.2272 |
manhattan_precision | 0.672 |
manhattan_recall | 1.0 |
manhattan_ap | 0.743 |
euclidean_accuracy | 0.6846 |
euclidean_accuracy_threshold | 0.4672 |
euclidean_f1 | 0.8038 |
euclidean_f1_threshold | 0.4672 |
euclidean_precision | 0.672 |
euclidean_recall | 1.0 |
euclidean_ap | 0.7428 |
max_accuracy | 0.6846 |
max_accuracy_threshold | 6.8578 |
max_f1 | 0.8038 |
max_f1_threshold | 7.2272 |
max_precision | 0.672 |
max_recall | 1.0 |
max_ap | 0.743 |
Binary Classification
- Dataset:
quora-duplicates-dev
- Evaluated with
BinaryClassificationEvaluator
Metric | Value |
---|---|
cosine_accuracy | 0.8923 |
cosine_accuracy_threshold | 0.7951 |
cosine_f1 | 0.9231 |
cosine_f1_threshold | 0.7486 |
cosine_precision | 0.8571 |
cosine_recall | 1.0 |
cosine_ap | 0.9716 |
dot_accuracy | 0.8923 |
dot_accuracy_threshold | 0.7951 |
dot_f1 | 0.9231 |
dot_f1_threshold | 0.7486 |
dot_precision | 0.8571 |
dot_recall | 1.0 |
dot_ap | 0.9716 |
manhattan_accuracy | 0.8846 |
manhattan_accuracy_threshold | 10.6305 |
manhattan_f1 | 0.9171 |
manhattan_f1_threshold | 10.6305 |
manhattan_precision | 0.8557 |
manhattan_recall | 0.9881 |
manhattan_ap | 0.9702 |
euclidean_accuracy | 0.8923 |
euclidean_accuracy_threshold | 0.6402 |
euclidean_f1 | 0.9231 |
euclidean_f1_threshold | 0.709 |
euclidean_precision | 0.8571 |
euclidean_recall | 1.0 |
euclidean_ap | 0.9716 |
max_accuracy | 0.8923 |
max_accuracy_threshold | 10.6305 |
max_f1 | 0.9231 |
max_f1_threshold | 10.6305 |
max_precision | 0.8571 |
max_recall | 1.0 |
max_ap | 0.9716 |
Binary Classification
- Dataset:
e5-cogcache-dev
- Evaluated with
BinaryClassificationEvaluator
Metric | Value |
---|---|
cosine_accuracy | 0.8923 |
cosine_accuracy_threshold | 0.7951 |
cosine_f1 | 0.9231 |
cosine_f1_threshold | 0.7486 |
cosine_precision | 0.8571 |
cosine_recall | 1.0 |
cosine_ap | 0.9716 |
dot_accuracy | 0.8923 |
dot_accuracy_threshold | 0.7951 |
dot_f1 | 0.9231 |
dot_f1_threshold | 0.7486 |
dot_precision | 0.8571 |
dot_recall | 1.0 |
dot_ap | 0.9716 |
manhattan_accuracy | 0.8846 |
manhattan_accuracy_threshold | 10.6305 |
manhattan_f1 | 0.9171 |
manhattan_f1_threshold | 10.6305 |
manhattan_precision | 0.8557 |
manhattan_recall | 0.9881 |
manhattan_ap | 0.9702 |
euclidean_accuracy | 0.8923 |
euclidean_accuracy_threshold | 0.6402 |
euclidean_f1 | 0.9231 |
euclidean_f1_threshold | 0.709 |
euclidean_precision | 0.8571 |
euclidean_recall | 1.0 |
euclidean_ap | 0.9716 |
max_accuracy | 0.8923 |
max_accuracy_threshold | 10.6305 |
max_f1 | 0.9231 |
max_f1_threshold | 10.6305 |
max_precision | 0.8571 |
max_recall | 1.0 |
max_ap | 0.9716 |
Training Details
Training Dataset
Unnamed Dataset
- Size: 2,000 training samples
- Columns:
label
,sentence1
, andsentence2
- Approximate statistics based on the first 1000 samples:
label sentence1 sentence2 type int string string details - 0: ~55.10%
- 1: ~44.90%
- min: 6 tokens
- mean: 13.24 tokens
- max: 66 tokens
- min: 4 tokens
- mean: 13.29 tokens
- max: 55 tokens
- Samples:
label sentence1 sentence2 1
What are the ingredients of a pizza?
What are the ingredients of a pizza
1
What are the ingredients of a pizza?
What are the ingredients of pizza
1
What are the ingredients of a pizza?
What are ingredients of pizza
- Loss:
OnlineContrastiveLoss
Evaluation Dataset
Unnamed Dataset
- Size: 130 evaluation samples
- Columns:
label
,sentence1
, andsentence2
- Approximate statistics based on the first 1000 samples:
label sentence1 sentence2 type int string string details - 0: ~35.38%
- 1: ~64.62%
- min: 6 tokens
- mean: 10.85 tokens
- max: 20 tokens
- min: 5 tokens
- mean: 11.48 tokens
- max: 22 tokens
- Samples:
label sentence1 sentence2 1
What are the ingredients of a pizza?
What are the ingredients of a pizza
1
What are the ingredients of a pizza?
What are the ingredients of pizza
1
What are the ingredients of a pizza?
What are ingredients of pizza
- Loss:
OnlineContrastiveLoss
Training Hyperparameters
Non-Default Hyperparameters
eval_strategy
: epochper_device_train_batch_size
: 16per_device_eval_batch_size
: 16num_train_epochs
: 6warmup_ratio
: 0.1batch_sampler
: no_duplicates
All Hyperparameters
Click to expand
overwrite_output_dir
: Falsedo_predict
: Falseeval_strategy
: epochprediction_loss_only
: Trueper_device_train_batch_size
: 16per_device_eval_batch_size
: 16per_gpu_train_batch_size
: Noneper_gpu_eval_batch_size
: Nonegradient_accumulation_steps
: 1eval_accumulation_steps
: Nonelearning_rate
: 5e-05weight_decay
: 0.0adam_beta1
: 0.9adam_beta2
: 0.999adam_epsilon
: 1e-08max_grad_norm
: 1.0num_train_epochs
: 6max_steps
: -1lr_scheduler_type
: linearlr_scheduler_kwargs
: {}warmup_ratio
: 0.1warmup_steps
: 0log_level
: passivelog_level_replica
: warninglog_on_each_node
: Truelogging_nan_inf_filter
: Truesave_safetensors
: Truesave_on_each_node
: Falsesave_only_model
: Falserestore_callback_states_from_checkpoint
: Falseno_cuda
: Falseuse_cpu
: Falseuse_mps_device
: Falseseed
: 42data_seed
: Nonejit_mode_eval
: Falseuse_ipex
: Falsebf16
: Falsefp16
: Falsefp16_opt_level
: O1half_precision_backend
: autobf16_full_eval
: Falsefp16_full_eval
: Falsetf32
: Nonelocal_rank
: 0ddp_backend
: Nonetpu_num_cores
: Nonetpu_metrics_debug
: Falsedebug
: []dataloader_drop_last
: Falsedataloader_num_workers
: 0dataloader_prefetch_factor
: Nonepast_index
: -1disable_tqdm
: Falseremove_unused_columns
: Truelabel_names
: Noneload_best_model_at_end
: Falseignore_data_skip
: Falsefsdp
: []fsdp_min_num_params
: 0fsdp_config
: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}fsdp_transformer_layer_cls_to_wrap
: Noneaccelerator_config
: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}deepspeed
: Nonelabel_smoothing_factor
: 0.0optim
: adamw_torchoptim_args
: Noneadafactor
: Falsegroup_by_length
: Falselength_column_name
: lengthddp_find_unused_parameters
: Noneddp_bucket_cap_mb
: Noneddp_broadcast_buffers
: Falsedataloader_pin_memory
: Truedataloader_persistent_workers
: Falseskip_memory_metrics
: Trueuse_legacy_prediction_loop
: Falsepush_to_hub
: Falseresume_from_checkpoint
: Nonehub_model_id
: Nonehub_strategy
: every_savehub_private_repo
: Falsehub_always_push
: Falsegradient_checkpointing
: Falsegradient_checkpointing_kwargs
: Noneinclude_inputs_for_metrics
: Falseeval_do_concat_batches
: Truefp16_backend
: autopush_to_hub_model_id
: Nonepush_to_hub_organization
: Nonemp_parameters
:auto_find_batch_size
: Falsefull_determinism
: Falsetorchdynamo
: Noneray_scope
: lastddp_timeout
: 1800torch_compile
: Falsetorch_compile_backend
: Nonetorch_compile_mode
: Nonedispatch_batches
: Nonesplit_batches
: Noneinclude_tokens_per_second
: Falseinclude_num_input_tokens_seen
: Falseneftune_noise_alpha
: Noneoptim_target_modules
: Nonebatch_eval_metrics
: Falsebatch_sampler
: no_duplicatesmulti_dataset_batch_sampler
: proportional
Training Logs
Epoch | Step | Training Loss | loss | e5-cogcache-dev_max_ap | quora-duplicates-dev_max_ap |
---|---|---|---|---|---|
0 | 0 | - | - | - | 0.7430 |
1.0 | 125 | - | 0.4486 | - | 0.8547 |
2.0 | 250 | - | 0.2319 | - | 0.9373 |
3.0 | 375 | - | 0.1411 | - | 0.9634 |
4.0 | 500 | 0.2324 | 0.1785 | - | 0.9687 |
5.0 | 625 | - | 0.1681 | - | 0.9713 |
6.0 | 750 | - | 0.1477 | 0.9716 | 0.9716 |
Framework Versions
- Python: 3.10.12
- Sentence Transformers: 3.0.1
- Transformers: 4.41.2
- PyTorch: 2.1.2+cu121
- Accelerate: 0.32.1
- 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",
}
- Downloads last month
- 9
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for srikarvar/e5-small-cogcachedata-1
Base model
intfloat/multilingual-e5-smallEvaluation results
- Cosine Accuracy on quora duplicates devself-reported0.685
- Cosine Accuracy Threshold on quora duplicates devself-reported0.891
- Cosine F1 on quora duplicates devself-reported0.804
- Cosine F1 Threshold on quora duplicates devself-reported0.891
- Cosine Precision on quora duplicates devself-reported0.672
- Cosine Recall on quora duplicates devself-reported1.000
- Cosine Ap on quora duplicates devself-reported0.743
- Dot Accuracy on quora duplicates devself-reported0.685
- Dot Accuracy Threshold on quora duplicates devself-reported0.891
- Dot F1 on quora duplicates devself-reported0.804