GphaHoa/academic_distilroberta_base

CrossEncoder based on cross-encoder/stsb-distilroberta-base

This is a Cross Encoder model finetuned from cross-encoder/stsb-distilroberta-base using the sentence-transformers library. It computes scores for pairs of texts, which can be used for text reranking and semantic search.

Model Details

Model Description

• Model Type: Cross Encoder
• Base model: cross-encoder/stsb-distilroberta-base
• Maximum Sequence Length: 128 tokens
• Number of Output Labels: 1 label

Model Sources

• Documentation: Sentence Transformers Documentation
• Documentation: Cross Encoder Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Cross Encoders on Hugging Face

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 CrossEncoder

# Download from the 🤗 Hub
model = CrossEncoder("cross_encoder_model_id")
# Get scores for pairs of texts
pairs = [
    ['Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and', 'Discussing different types of quadrilaterals. Quadrilateral is a closed figure with four line segments. Each point where the two line segments meet is called a vertex. The closed figure also form four angles.. Discussing different types of quadrilaterals Albert Mhango, Mzimba Introduction: Quad'],
    ['Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and', 'Discussing properties of quadrilaterals. The common properties that you will see in every quadrilateral include; all quadrilaterals have four sides, they all consist of four vertices and the sum of interior angles is equal to 360 degrees.. Discussing properties of quadrilaterals Albert Mhango,'],
    ['Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and', 'Calculating interior and exterior angles of a triangle. The exterior angle of a triangle is equal to the sum of two opposite interior angles. This property will help you to find angles in a triangle and exterior angles.. Calculating interior and exterior angles of a triangle Albert Mhango, Mzimba Introduction: The'],
    ['Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and', 'Using properties of quadrilaterals to solve problems. A quadrilateral is a geometric figure with four sides. The general properties of quadrilaterals include; they all have four sides, have two diagonals, have four interior angles and the sum of their interior angles is equal to 360 degrees.. Using properties'],
    ['Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and', 'Solving Problems Involving Polygons. In this lesson, you will learn how we can use the exterior angles of polygon formula to solve problems.. Solving Problems Involving Polygons Mary Chagwa, Blantyre Introduction: In the previous lesson, you were deriving the formula for finding the sum'],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)

# Or rank different texts based on similarity to a single text
ranks = model.rank(
    'Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and',
    [
        'Discussing different types of quadrilaterals. Quadrilateral is a closed figure with four line segments. Each point where the two line segments meet is called a vertex. The closed figure also form four angles.. Discussing different types of quadrilaterals Albert Mhango, Mzimba Introduction: Quad',
        'Discussing properties of quadrilaterals. The common properties that you will see in every quadrilateral include; all quadrilaterals have four sides, they all consist of four vertices and the sum of interior angles is equal to 360 degrees.. Discussing properties of quadrilaterals Albert Mhango,',
        'Calculating interior and exterior angles of a triangle. The exterior angle of a triangle is equal to the sum of two opposite interior angles. This property will help you to find angles in a triangle and exterior angles.. Calculating interior and exterior angles of a triangle Albert Mhango, Mzimba Introduction: The',
        'Using properties of quadrilaterals to solve problems. A quadrilateral is a geometric figure with four sides. The general properties of quadrilaterals include; they all have four sides, have two diagonals, have four interior angles and the sum of their interior angles is equal to 360 degrees.. Using properties',
        'Solving Problems Involving Polygons. In this lesson, you will learn how we can use the exterior angles of polygon formula to solve problems.. Solving Problems Involving Polygons Mary Chagwa, Blantyre Introduction: In the previous lesson, you were deriving the formula for finding the sum',
    ]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]

Evaluation

Metrics

Cross Encoder Reranking

• Dataset: reranking-dev
• Evaluated with CrossEncoderRerankingEvaluator with these parameters:

  {
      "at_k": 50,
      "always_rerank_positives": false
  }
  

Metric	Value
map	0.7207 (+0.0992)
mrr@50	0.7903 (+0.0528)
ndcg@50	0.8072 (+0.0817)

Training Details

Training Dataset

Unnamed Dataset

• Size: 1,314,940 training samples
• Columns: topic, content, and label
• Approximate statistics based on the first 1000 samples:

  	topic	content	label
  type	string	string	int
  details	min: 42 characters mean: 158.98 characters max: 336 characters	min: 7 characters mean: 150.81 characters max: 353 characters	0: ~76.30% 1: ~23.70%

• Samples:

  topic	content	label
  Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and	Discussing different types of quadrilaterals. Quadrilateral is a closed figure with four line segments. Each point where the two line segments meet is called a vertex. The closed figure also form four angles.. Discussing different types of quadrilaterals Albert Mhango, Mzimba Introduction: Quad	1
  Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and	Discussing properties of quadrilaterals. The common properties that you will see in every quadrilateral include; all quadrilaterals have four sides, they all consist of four vertices and the sum of interior angles is equal to 360 degrees.. Discussing properties of quadrilaterals Albert Mhango,	1
  Triangles and polygons. Space, shape and measurement. Form 1. Malawi Mathematics Syllabus. Learning outcomes: students must be able to solve problems involving angles, triangles and polygons including: types of triangles, calculate the interior and exterior angles of a triangle, different types of polygons, interior angles and	Calculating interior and exterior angles of a triangle. The exterior angle of a triangle is equal to the sum of two opposite interior angles. This property will help you to find angles in a triangle and exterior angles.. Calculating interior and exterior angles of a triangle Albert Mhango, Mzimba Introduction: The	1

• Loss: BinaryCrossEntropyLoss with these parameters:

  {
      "activation_fn": "torch.nn.modules.linear.Identity",
      "pos_weight": 3.7016043663024902
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 128
• per_device_eval_batch_size: 128
• learning_rate: 2e-05
• warmup_ratio: 0.1
• seed: 12
• bf16: True
• dataloader_num_workers: 4
• 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: 128
• per_device_eval_batch_size: 128
• 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: 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: 3
• 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: 12
• data_seed: None
• jit_mode_eval: False
• use_ipex: False
• bf16: True
• 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: 4
• 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: None
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• 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
• 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
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	reranking-dev_ndcg@50
0.0001	1	1.0174	-
0.9999	10272	0.6234	0.7913 (+0.0658)
1.9998	20544	0.3901	0.8041 (+0.0786)
2.9997	30816	0.2978	0.8072 (+0.0817)

Framework Versions

• Python: 3.11.13
• Sentence Transformers: 4.1.0
• Transformers: 4.52.4
• PyTorch: 2.6.0+cu124
• Accelerate: 1.7.0
• Datasets: 2.14.4
• Tokenizers: 0.21.1

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",
}