SentenceTransformer based on uitnlp/CafeBERT

This is a sentence-transformers model finetuned from uitnlp/CafeBERT. It maps sentences & paragraphs to a 256-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: uitnlp/CafeBERT
Maximum Sequence Length: 256 tokens
Output Dimensionality: 256 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': 256, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (1): Pooling({'word_embedding_dimension': 1024, '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): Dense({'in_features': 1024, 'out_features': 256, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
)

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("ThuanPhong/sentence_CafeBERT")
# Run inference
sentences = [
    'Chúng tôi đang tiến vào sa mạc.',
    'Chúng tôi chuyển đến sa mạc.',
    'Người phụ nữ này đang chạy vì cô ta đến muộn.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 256]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Binary Classification

Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.5404
cosine_accuracy_threshold	1.0
cosine_f1	0.6299
cosine_f1_threshold	1.0
cosine_precision	0.4597
cosine_recall	1.0
cosine_ap	0.4597
dot_accuracy	0.5403
dot_accuracy_threshold	46.2905
dot_f1	0.6299
dot_f1_threshold	46.2905
dot_precision	0.4597
dot_recall	0.9999
dot_ap	0.4578
manhattan_accuracy	0.5411
manhattan_accuracy_threshold	0.0
manhattan_f1	0.6299
manhattan_f1_threshold	0.0002
manhattan_precision	0.4597
manhattan_recall	1.0
manhattan_ap	0.4604
euclidean_accuracy	0.5412
euclidean_accuracy_threshold	0.0
euclidean_f1	0.6299
euclidean_f1_threshold	0.0
euclidean_precision	0.4597
euclidean_recall	1.0
euclidean_ap	0.4602
max_accuracy	0.5412
max_accuracy_threshold	46.2905
max_f1	0.6299
max_f1_threshold	46.2905
max_precision	0.4597
max_recall	1.0
max_ap	0.4604

Training Details

Training Dataset

Unnamed Dataset

Size: 461,625 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 int
details
min: 4 tokens
mean: 21.87 tokens
max: 121 tokens

min: 4 tokens
mean: 32.19 tokens
max: 162 tokens

0: ~55.90%
1: ~44.10%

	sentence_0	sentence_1	label
type	string	string	int
details	min: 4 tokens mean: 21.87 tokens max: 121 tokens	min: 4 tokens mean: 32.19 tokens max: 162 tokens	0: ~55.90% 1: ~44.10%

Samples:

sentence_0	sentence_1	label
`Khi nào William Caxton giới thiệu máy in ép vào nước Anh?`	`Những đặc điểm mà độc giả của Shakespeare ngày nay có thể thấy kỳ quặc hay lỗi thời thường đại diện cho những nét đặc trưng của tiếng Anh trung Đại.`	`0`
`Nhưng tôi không biết rằng tôi phải, " Dorcas do dự.`	`Dorcas sợ phản ứng của họ.`	`0`
`Đông Đức là tên gọi thường được sử dụng để chỉ quốc gia nào?`	`Cộng hòa Dân chủ Đức (tiếng Đức: Deutsche Demokratische Republik, DDR; thường được gọi là Đông Đức) là một quốc gia nay không còn nữa, tồn tại từ 1949 đến 1990 theo định hướng xã hội chủ nghĩa tại phần phía đông nước Đức ngày nay.`	`1`

Loss: OnlineContrastiveLoss

Training Hyperparameters

Non-Default Hyperparameters

num_train_epochs: 2
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 8
per_device_eval_batch_size: 8
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: 2
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: 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: 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