metadata

language: []
library_name: sentence-transformers
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:360
  - loss:CosineSimilarityLoss
base_model: BAAI/bge-large-en
datasets: []
widget:
  - source_sentence: Deadline for submitting project schedule.
    sentences:
      - Variation
      - "The Railway shall have the right to let other contracts in connection with the works. The Contractor shall afford other Contractors reasonable opportunity for the storage of their materials and the execution of their works and shall properly connect and coordinate his work with theirs. If any part of the Contractor\x92s work depends upon proper execution or result upon the work of another Contractor(s), the Contractor shall inspect and promptly report to the Engineer any defects in such works that render it unsuitable for such proper execution and results. The Contractor's failure so-to inspect and report shall constitute an acceptance of the other Contractor's work as fit and proper for the reception of his work, except as to defects which may develop in the other Contractor's work after the execution of his work."
      - >-
        The quantities set out in the accepted Schedule of Rates with items of
        works quantified are the estimated quantities of the works
  - source_sentence: "\_What is the deadline to submit the proposed project schedule?"
    sentences:
      - "having value more than Rs 20 crore and original period of completion 12 months or more, when there is no reduction in original scope of work by more than 10%, and no extension granted on either railway or Contractor\x92s account,"
      - >-
        Can the stones/rocks/bounders obtained during excavation be used for
        construction if found technically satisfactory?
      - >-
        Chart/PERT/CPM. He shall also submit the details of organisation (in
        terms of labour and supervisors), plant and machinery that he intends to
        utilize (from time to time) for execution of the work within stipulated
        date of completion.
  - source_sentence: "Does the contract document contain a \x91third-party liability relationship\x92 provision?"
    sentences:
      - >-
        The Contractor shall indemnify and save harmless the Railway from and
        against all actions, suit, proceedings, losses, costs, damages, charges,
        claims and demands of every nature and description brought or recovered
        against the Railways by reason of any act or omission of the Contractor,
        his agents or employees, in the execution of the works or in his
        guarding of the same. All sums payable by way of compensation under any
        of these conditions shall be considered as reasonable compensation to be
        applied to the actual loss or damage sustained, and whether or not any
        damage shall have been sustained.
      - >-
        the Railway shall not in any way be liable for the supply of materials
        or for the non-supply thereof for any reasons whatsoever nor for any
        loss or damage arising in consequence of such delay or non-supply.
      - >-
        The Railway shall have the right to let other contracts in connection
        with the works.
  - source_sentence: Liquidated Damages
    sentences:
      - >-
        The Contractor shall commence the works within 15 days after the receipt
        by him of an order in writing to this effect from the Engineer and shall
        proceed with the same with due expedition and without delay
      - >-
        Any bribe, commission, gift or advantage given, promised or offered by
        or on behalf of the Contractor or his partner or agent or servant or
        anyone on his behalf
      - purpose of works either free of cost or pay thecost of the same.
  - source_sentence: What is mentioned regarding the patent errors?
    sentences:
      - >-
        the Security Deposit already with railways under the contract shall be
        forfeited.
      - >-
        This clause mentions Special Conditions, which might be additional
        documents relevant to the contract.
      - >-
        shall take upon himself and provide for the risk of any error which may
        subsequently be discovered and shall make no subsequent claim on account
        thereof.
pipeline_tag: sentence-similarity

SentenceTransformer based on BAAI/bge-large-en

This is a sentence-transformers model finetuned from BAAI/bge-large-en. 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: BAAI/bge-large-en
Maximum Sequence Length: 512 tokens
Output Dimensionality: 1024 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': 512, 'do_lower_case': True}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 1024, '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("Ananthu357/Ananthus-BAAI-for-contracts")
# Run inference
sentences = [
    'What is mentioned regarding the patent errors?',
    'shall take upon himself and provide for the risk of any error which may subsequently be discovered and shall make no subsequent claim on account thereof.',
    'This clause mentions Special Conditions, which might be additional documents relevant to the contract.',
]
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]

Training Details

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 16
per_device_eval_batch_size: 16
num_train_epochs: 40
warmup_ratio: 0.1
fp16: 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: 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: 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: 40
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: 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
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss
3.5652	100	0.0564	0.0940
7.1304	200	0.0122	0.0713
10.4348	300	0.0051	0.0655
14.0	400	0.0026	0.0678
17.3043	500	0.001	0.0668
20.8696	600	0.0009	0.0666
24.1739	700	0.0008	0.0671
27.7391	800	0.0007	0.0674
31.0435	900	0.0007	0.0671

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.3.0+cu121
Accelerate: 0.31.0
Datasets: 2.20.0
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",
}