Deehan1866/finetuned-wic-electra-large

SentenceTransformer based on google/electra-large-discriminator

This is a sentence-transformers model finetuned from google/electra-large-discriminator on the Deehan1866/wi_c 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: google/electra-large-discriminator
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 1024 tokens
• Similarity Function: Cosine Similarity
• Training Dataset:
  • Deehan1866/wi_c

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: ElectraModel 
  (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("Deehan1866/finetuned-wic-electra-large")
# Run inference
sentences = [
    "It 's your move ! Roll the dice !",
    'If you roll a six , you can make two moves .',
    'She scrubbed his back .',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# 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	0.5611
cosine_accuracy_threshold	0.9628
cosine_f1	0.6695
cosine_f1_threshold	0.8932
cosine_precision	0.5056
cosine_recall	0.9906
cosine_ap	0.5762
dot_accuracy	0.5219
dot_accuracy_threshold	224.722
dot_f1	0.6688
dot_f1_threshold	128.2531
dot_precision	0.5024
dot_recall	1.0
dot_ap	0.5015
manhattan_accuracy	0.5674
manhattan_accuracy_threshold	62.2153
manhattan_f1	0.6702
manhattan_f1_threshold	123.3105
manhattan_precision	0.5039
manhattan_recall	1.0
manhattan_ap	0.5929
euclidean_accuracy	0.5658
euclidean_accuracy_threshold	2.9031
euclidean_f1	0.6695
euclidean_f1_threshold	8.0959
euclidean_precision	0.5032
euclidean_recall	1.0
euclidean_ap	0.5848
max_accuracy	0.5674
max_accuracy_threshold	224.722
max_f1	0.6702
max_f1_threshold	128.2531
max_precision	0.5056
max_recall	1.0
max_ap	0.5929

Training Details

Training Dataset

Deehan1866/wi_c

• Dataset: Deehan1866/wi_c
• Size: 5,428 training samples
• Columns: sentence1, sentence2, and label
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	label
  type	string	string	int
  details	min: 5 tokens mean: 10.49 tokens max: 28 tokens	min: 5 tokens mean: 10.7 tokens max: 32 tokens	0: ~60.00% 1: ~40.00%

• Samples:

  sentence1	sentence2	label
  You must carry your camping gear .	Sound carries well over water .	0
  Messages must go through diplomatic channels .	Do you think the sofa will go through the door ?	0
  Break an alibi .	The wholesaler broke the container loads into palettes and boxes for local retailers .	0

• Loss: SoftmaxLoss

Evaluation Dataset

Deehan1866/wi_c

• Dataset: Deehan1866/wi_c
• Size: 638 evaluation samples
• Columns: sentence1, sentence2, and label
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	label
  type	string	string	int
  details	min: 5 tokens mean: 11.02 tokens max: 34 tokens	min: 5 tokens mean: 11.45 tokens max: 32 tokens	0: ~50.00% 1: ~50.00%

• Samples:

  sentence1	sentence2	label
  Room and board .	He nailed boards across the windows .	0
  Circulate a rumor .	This letter is being circulated among the faculty .	0
  Hook a fish .	He hooked a snake accidentally , and was so scared he dropped his rod into the water .	1

• Loss: SoftmaxLoss

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• learning_rate: 2e-05
• num_train_epochs: 5
• warmup_ratio: 0.1
• 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: 16
• per_device_eval_batch_size: 16
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• 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: 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
• use_ipex: 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: 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
• 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
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss	quora-duplicates-dev_max_ap
0	0	-	-	0.5844
0.2941	100	-	0.692	0.5929
0.5882	200	-	0.6943	0.5383
0.8824	300	-	0.6937	0.5280
1.1765	400	-	0.6973	0.5232
1.4706	500	0.7014	0.7048	0.5210
1.7647	600	-	0.6932	0.5039
2.0588	700	-	0.6932	0.5021
2.3529	800	-	0.6943	0.5048
2.6471	900	-	0.6940	0.5074
2.9412	1000	0.6939	0.6932	0.5109
3.2353	1100	-	0.6931	0.5143
3.5294	1200	-	0.6934	0.5162
3.8235	1300	-	0.6938	0.5146
4.1176	1400	-	0.6932	0.5167
4.4118	1500	0.6937	0.6933	0.5225
4.7059	1600	-	0.6932	0.5211
5.0	1700	-	0.6932	0.5929

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.10.10
• Sentence Transformers: 3.0.1
• Transformers: 4.42.3
• PyTorch: 2.2.1+cu121
• Accelerate: 0.32.1
• Datasets: 2.20.0
• Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers and SoftmaxLoss

@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",
}