all-MiniLM-L6-v70-pair_score

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: KhaledReda/all-MiniLM-L6-v69-pair_score
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • pairs_with_scores_v56
• 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': 256, 'do_lower_case': False, 'architecture': '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)

Training Details

Training Dataset

pairs_with_scores_v56

• Dataset: pairs_with_scores_v56 at b377fe7
• Size: 16,807,599 training samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 3 tokens mean: 7.55 tokens max: 25 tokens	min: 3 tokens mean: 57.36 tokens max: 256 tokens	min: 0.0 mean: 0.19 max: 1.0

• Samples:

  sentence1	sentence2	score
  training tees	rich bake multi-grain toast brown toast bread wheat bran toast bread multigrain toast bread rich bake toast bread brown bread contains wheat bran. wheat bran contains more protein than white flour and has a concentration of vitamins and minerals in the kernel. the extra fiber in brown bread greatly aids the digestive processes and studies have shown that it also linked to reducing cholesterol levels in the blood.	0.0
  ornidaz - ornidazole	docavarico 20 mg 5 ampoules	1.0
  nostifix 2tabnew	cellcept 250mg 100/cap new cellcept capsules	0.0

• Loss: CoSENTLoss with these parameters:
  json

  1{ 2 "scale": 20.0, 3 "similarity_fct": "pairwise_cos_sim" 4}

Evaluation Dataset

pairs_with_scores_v56

• Dataset: pairs_with_scores_v56 at b377fe7
• Size: 84,461 evaluation samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 4 tokens mean: 7.56 tokens max: 27 tokens	min: 3 tokens mean: 55.21 tokens max: 256 tokens	min: 0.0 mean: 0.2 max: 1.0

• Samples:

  sentence1	sentence2	score
  elasticated rib edges cardigan	roccobarocco tre edp f/w 100m roccobarocco tre eau de parfum	0.0
  lobster tail -	haloub pure butter ghee - 800 gr haloub ghee	0.0
  molfix junior diaper pants	baby safety corner protector table desk edge baby safety bar furniture corner protector desk corner guard safety edge guard table corner cushion table edge protector	0.0

• Loss: CoSENTLoss with these parameters:
  json

  1{ 2 "scale": 20.0, 3 "similarity_fct": "pairwise_cos_sim" 4}

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• learning_rate: 2e-05
• num_train_epochs: 1
• warmup_ratio: 0.1
• fp16: True

All Hyperparameters

Training Logs