cherifkhalifah/finetuned2-snli-MiniLM-L12-v2

SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L12-v2. 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: sentence-transformers/all-MiniLM-L12-v2
• Maximum Sequence Length: 128 tokens
• Output Dimensionality: 384 tokens
• 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': 128, '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("cherifkhalifah/finetuned2-snli-MiniLM-L12-v2")
# Run inference
sentences = [
    'Many people outside on bicycles.',
    'People are riding bikes in a race.',
    'Protesters are in the back of a photo with a magazine display in the foreground.',
]
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

Semantic Similarity

• Dataset: snli-dev
• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.5041
spearman_cosine	0.4962
pearson_manhattan	0.4848
spearman_manhattan	0.4957
pearson_euclidean	0.4855
spearman_euclidean	0.4962
pearson_dot	0.5041
spearman_dot	0.4962
pearson_max	0.5041
spearman_max	0.4962

Training Details

Training Dataset

Unnamed Dataset

• Size: 100,000 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1	label
  type	string	string	float
  details	min: 7 tokens mean: 16.36 tokens max: 45 tokens	min: 4 tokens mean: 10.62 tokens max: 33 tokens	min: 0.0 mean: 0.5 max: 1.0

• Samples:

  sentence_0	sentence_1	label
  A mother and two children opening gifts on a Christmas morning.	A mother and children cut into a large pizza.	1.0
  Two men in protective gear are in a speed car racing to the finish line.	Two boys are playing tag.	1.0
  A person in a pink jacket is running onto the field.	The woman is running on to the field.	0.5

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• num_train_epochs: 4
• fp16: True
• multi_dataset_batch_sampler: round_robin

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
• torch_empty_cache_steps: None
• learning_rate: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1
• num_train_epochs: 4
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• 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: 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: False
• 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
• eval_use_gather_object: False
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss	snli-dev_spearman_max
0.08	500	0.1782	0.3312
0.16	1000	0.1516	0.3393
0.24	1500	0.1422	0.3798
0.32	2000	0.1405	0.3675
0.4	2500	0.137	0.4029
0.48	3000	0.1398	0.3989
0.56	3500	0.136	0.4216
0.64	4000	0.1351	0.4322
0.72	4500	0.1317	0.4223
0.8	5000	0.1293	0.4331
0.88	5500	0.1318	0.4416
0.96	6000	0.1311	0.4185
1.0	6250	-	0.4522
1.04	6500	0.129	0.4312
1.12	7000	0.1272	0.4544
1.2	7500	0.1271	0.4533
1.28	8000	0.125	0.4456
1.3600	8500	0.1229	0.4570
1.44	9000	0.1241	0.4529
1.52	9500	0.1254	0.4517
1.6	10000	0.1232	0.4563
1.6800	10500	0.1232	0.4565
1.76	11000	0.1198	0.4521
1.8400	11500	0.1201	0.4570
1.92	12000	0.1238	0.4758
2.0	12500	0.1195	0.4671
2.08	13000	0.1155	0.4582
2.16	13500	0.1208	0.4787
2.24	14000	0.1164	0.4733
2.32	14500	0.1164	0.4743
2.4	15000	0.1136	0.4733
2.48	15500	0.1177	0.4704
2.56	16000	0.1152	0.4711
2.64	16500	0.1162	0.4827
2.7200	17000	0.1136	0.4772
2.8	17500	0.1129	0.4853
2.88	18000	0.1161	0.4830
2.96	18500	0.1144	0.4827
3.0	18750	-	0.4850
3.04	19000	0.112	0.4920
3.12	19500	0.1105	0.4901
3.2	20000	0.1122	0.4925
3.2800	20500	0.1114	0.4913
3.36	21000	0.1074	0.4887
3.44	21500	0.1093	0.4819
3.52	22000	0.1107	0.4853
3.6	22500	0.1088	0.4897
3.68	23000	0.1095	0.4922
3.76	23500	0.11	0.4923
3.84	24000	0.1075	0.4950
3.92	24500	0.1107	0.4967
4.0	25000	0.1073	0.4962

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.1.1
• Transformers: 4.44.2
• PyTorch: 2.4.1+cu121
• Accelerate: 0.34.2
• Datasets: 3.0.1
• 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",
}