Gurveer05/bge-small-eedi-2024

SentenceTransformer based on BAAI/bge-small-en-v1.5

This is a sentence-transformers model finetuned from BAAI/bge-small-en-v1.5 on the csv dataset. 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: BAAI/bge-small-en-v1.5
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 384 tokens
• Similarity Function: Cosine Similarity
• Training Dataset:
  • csv

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': True}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 384, '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})
  (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("Gurveer05/bge-small-eedi-2024")
# Run inference
sentences = [
    'Express one quantity as a fraction of another A group of 8 friends share £6 equally. What fraction of the money do they each get? 1/8',
    'Thinks the fraction 1/n can express sharing any number of items between n people',
    'Does not recognise the distributive property',
]
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]

Training Details

Training Dataset

csv

• Dataset: csv
• Size: 2,442 training samples
• Columns: sentence1 and sentence2
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	min: 13 tokens mean: 56.55 tokens max: 306 tokens	min: 4 tokens mean: 15.13 tokens max: 40 tokens

• Samples:

  sentence1	sentence2
  Calculate the distance travelled using a speed-time graph Here is a speed-time graph for a car. Which of the following gives the best estimate for the distance travelled between 8 and 10 seconds? 48 m	Believes that when finding area under graph you can use the upper y value rather than average of upper and lower
  Add proper fractions with the same denominator Work out: 4/11+7/11 Write your answer in its simplest form. 11/11	Forgot to simplify the fraction
  Count a number of objects 1,3,5,7, … ? Which pattern matches the sequence above?	When given a linear sequence, cannot match it to a visual pattern

• Loss: MultipleNegativesSymmetricRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• num_train_epochs: 20
• warmup_ratio: 0.1
• fp16: True
• load_best_model_at_end: True
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• 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: 1.0
• num_train_epochs: 20
• 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: 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: 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
• eval_use_gather_object: False
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss
0.3846	15	2.5015
0.7436	29	-
0.7692	30	1.9614
1.1538	45	1.5138
1.4872	58	-
1.5385	60	1.4717
1.9231	75	1.2504
2.2308	87	-
2.3077	90	1.0932
2.6923	105	1.0067
2.9744	116	-
3.0769	120	0.7748
3.4615	135	0.7952
3.7179	145	-
3.8462	150	0.7165
4.2308	165	0.5502
4.4615	174	-
4.6154	180	0.6062
5.0	195	0.4444
5.2051	203	-
5.3846	210	0.4759
5.7692	225	0.4203
5.9487	232	-
6.1538	240	0.3599
6.5385	255	0.3613
6.6923	261	-
6.9231	270	0.3255
7.3077	285	0.2856
7.4359	290	-
7.6923	300	0.2756
8.0769	315	0.2199
8.1795	319	-
8.4615	330	0.2587
8.8462	345	0.2293
8.9231	348	-
9.2308	360	0.1848
9.6154	375	0.2232
9.6667	377	-
10.0	390	0.1755
10.3846	405	0.1898
10.4103	406	-
10.7692	420	0.1673
11.1538	435	0.1546
11.5385	450	0.1692
11.8974	464	-
11.9231	465	0.1471
12.3077	480	0.1299
12.6410	493	-
12.6923	495	0.1269
13.0769	510	0.1201
13.3846	522	-
13.4615	525	0.1306
13.8462	540	0.125
14.1282	551	-
14.2308	555	0.099
14.6154	570	0.1354
14.8718	580	-
15.0	585	0.0967
15.3846	600	0.1159
15.6154	609	-
15.7692	615	0.0996
16.1538	630	0.0933
16.3590	638	-
16.5385	645	0.1007
16.9231	660	0.0972
17.1026	667	-
17.3077	675	0.0978
17.6923	690	0.0874
17.8462	696	-
18.0769	705	0.0843
18.4615	720	0.0986
18.5897	725	-
18.8462	735	0.0862
19.2308	750	0.0751
19.3333	754	-
19.6154	765	0.1038
20.0	780	0.0791

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.10.14
• Sentence Transformers: 3.1.1
• Transformers: 4.44.0
• PyTorch: 2.4.0
• Accelerate: 0.33.0
• Datasets: 2.19.2
• 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",
}