Unsupervised SimCSE on FW make/model strings (distilbert-base-uncased, bs30)

897f46a verified 1 day ago

14.3 kB

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:3810
  - loss:MultipleNegativesRankingLoss
base_model: distilbert/distilbert-base-uncased
widget:
  - source_sentence: bmw ix
    sentences:
      - bmw ix
      - bmw x7
      - bmw i5
  - source_sentence: bmw m2
    sentences:
      - bmw m2
      - tesla s
      - audi s5
  - source_sentence: bmw xm
    sentences:
      - tesla y
      - bmw xm
      - audi rs
  - source_sentence: bmw m8
    sentences:
      - bmw m8
      - audi c7
      - audi a3
  - source_sentence: byd e6
    sentences:
      - audi r8
      - kia ev6
      - byd e6
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - pearson_cosine
  - spearman_cosine
model-index:
  - name: SentenceTransformer based on distilbert/distilbert-base-uncased
    results:
      - task:
          type: semantic-similarity
          name: Semantic Similarity
        dataset:
          name: make model dev
          type: make-model-dev
        metrics:
          - type: pearson_cosine
            value: 0.8766202507026644
            name: Pearson Cosine
          - type: spearman_cosine
            value: 0.8416717372995727
            name: Spearman Cosine

SentenceTransformer based on distilbert/distilbert-base-uncased

This is a sentence-transformers model finetuned from distilbert/distilbert-base-uncased. It maps inputs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Base model: distilbert/distilbert-base-uncased
Maximum Sequence Length: 32 tokens
Output Dimensionality: 768 dimensions
Similarity Function: Cosine Similarity
Supported Modality: Text

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'transformer_task': 'feature-extraction', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'last_hidden_state'}}, 'module_output_name': 'token_embeddings', 'architecture': 'DistilBertModel'})
  (1): Pooling({'embedding_dimension': 768, 'pooling_mode': 'mean', 'include_prompt': True})
)

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("SwarajSolanke-turtle/simcse-make-model-distilbert")
# Run inference
sentences = [
    'byd e6',
    'byd e6',
    'audi r8',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 1.0000, 0.3434],
#         [1.0000, 1.0000, 0.3434],
#         [0.3434, 0.3434, 1.0000]])

Evaluation

Metrics

Semantic Similarity

Dataset: make-model-dev
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.8766
spearman_cosine	0.8417

Training Details

Training Dataset

Unnamed Dataset

Size: 3,810 training samples
Columns: sentence1 and sentence2
Approximate statistics based on the first 100 samples:
sentence1 sentence2
type string string

modality text text
details
min: 4 tokens
mean: 6.43 tokens
max: 12 tokens

min: 4 tokens
mean: 6.43 tokens
max: 12 tokens
Samples:

sentence1 sentence2

mahindra mahindra mm 550 mahindra mahindra mm 550

tata motors aria tata motors aria

mg motor hector tomahawk ev mg motor hector tomahawk ev

	sentence1	sentence2
type	string	string
modality	text	text
details	min: 4 tokens mean: 6.43 tokens max: 12 tokens	min: 4 tokens mean: 6.43 tokens max: 12 tokens

sentence1	sentence2
`mahindra mahindra mm 550`	`mahindra mahindra mm 550`
`tata motors aria`	`tata motors aria`
`mg motor hector tomahawk ev`	`mg motor hector tomahawk ev`

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim",
    "gather_across_devices": false,
    "directions": [
        "query_to_doc"
    ],
    "partition_mode": "joint",
    "hardness_mode": null,
    "hardness_strength": 0.0
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 30
num_train_epochs: 100
learning_rate: 3e-05
warmup_steps: 0.1
fp16: True
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

per_device_train_batch_size: 30
num_train_epochs: 100
max_steps: -1
learning_rate: 3e-05
lr_scheduler_type: linear
lr_scheduler_kwargs: None
warmup_steps: 0.1
optim: adamw_torch_fused
optim_args: None
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
optim_target_modules: None
gradient_accumulation_steps: 1
average_tokens_across_devices: True
max_grad_norm: 1.0
label_smoothing_factor: 0.0
bf16: False
fp16: True
bf16_full_eval: False
fp16_full_eval: False
tf32: None
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
use_liger_kernel: False
liger_kernel_config: None
use_cache: False
neftune_noise_alpha: None
torch_empty_cache_steps: None
auto_find_batch_size: False
log_on_each_node: True
logging_nan_inf_filter: True
include_num_input_tokens_seen: no
log_level: passive
log_level_replica: warning
disable_tqdm: False
project: huggingface
trackio_space_id: None
trackio_bucket_id: None
trackio_static_space_id: None
per_device_eval_batch_size: 8
prediction_loss_only: True
eval_on_start: False
eval_do_concat_batches: True
eval_use_gather_object: False
eval_accumulation_steps: None
include_for_metrics: []
batch_eval_metrics: False
save_only_model: False
save_on_each_node: False
enable_jit_checkpoint: False
push_to_hub: False
hub_private_repo: None
hub_model_id: None
hub_strategy: every_save
hub_always_push: False
hub_revision: None
load_best_model_at_end: False
ignore_data_skip: False
restore_callback_states_from_checkpoint: False
full_determinism: False
seed: 42
data_seed: None
use_cpu: False
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
parallelism_config: None
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_pin_memory: True
dataloader_persistent_workers: False
dataloader_prefetch_factor: None
dataloader_multiprocessing_context: None
dataloader_in_order: True
remove_unused_columns: True
label_names: None
train_sampling_strategy: random
length_column_name: length
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
ddp_static_graph: None
ddp_backend: None
ddp_timeout: 1800
fsdp: None
fsdp_config: None
deepspeed: None
debug: []
skip_memory_metrics: True
do_predict: False
resume_from_checkpoint: None
local_rank: -1
prompts: None
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional
router_mapping: {}
learning_rate_mapping: {}
warmup_ratio: None

Training Logs

Epoch	Step	Training Loss	make-model-dev_spearman_cosine
-1	-1	-	0.8198
5.0	635	0.0413	-
10.0	1270	0.0012	0.8535
15.0	1905	0.0006	-
20.0	2540	0.0006	0.8349
25.0	3175	0.0005	-
30.0	3810	0.0003	0.8346
35.0	4445	0.0002	-
40.0	5080	0.0003	0.8460
45.0	5715	0.0003	-
50.0	6350	0.0003	0.8307
55.0	6985	0.0002	-
60.0	7620	0.0002	0.8426
65.0	8255	0.0002	-
70.0	8890	0.0002	0.8389
75.0	9525	0.0001	-
80.0	10160	0.0001	0.8402
85.0	10795	0.0001	-
90.0	11430	0.0001	0.8393
95.0	12065	0.0001	-
100.0	12700	0.0001	0.8417
-1	-1	-	0.8417

Training Time

Training: 9.5 minutes
Evaluation: 29.9 seconds
Total: 10.0 minutes

Framework Versions

Python: 3.12.0
Sentence Transformers: 6.1.0
Transformers: 5.20.1
PyTorch: 2.14.1+cu130
Accelerate: 1.15.0
Datasets: 5.1.0
Tokenizers: 0.23.3

Additional Resources

Training and Finetuning Embedding Models with Sentence Transformers: the end-to-end guide for training or finetuning Sentence Transformer models.
Introduction to Matryoshka Embedding Models: variable-size embeddings that can be truncated with minimal quality loss.
Binary and Scalar Embedding Quantization for Significantly Faster & Cheaper Retrieval: post-training compression of embedding vectors.
Multimodal Embedding & Reranker Models with Sentence Transformers: use text, image, audio, and video models through the same API.
Training and Finetuning Multimodal Embedding & Reranker Models with Sentence Transformers: train multimodal embedding models, with a Visual Document Retrieval walkthrough.

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",
}

MultipleNegativesRankingLoss

@misc{oord2019representationlearningcontrastivepredictive,
      title={Representation Learning with Contrastive Predictive Coding},
      author={Aaron van den Oord and Yazhe Li and Oriol Vinyals},
      year={2019},
      eprint={1807.03748},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/1807.03748},
}