SentenceTransformer based on Snowflake/snowflake-arctic-embed-l-v2.0

This is a sentence-transformers model finetuned from Snowflake/snowflake-arctic-embed-l-v2.0 on the csv dataset. 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: Snowflake/snowflake-arctic-embed-l-v2.0
Maximum Sequence Length: 8192 tokens
Output Dimensionality: 1024 dimensions
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': 8192, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (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("omarabb315/makeML-snowflake")
# Run inference
sentences = [
    'بلوزة صيفية نسائية كاجوال بطباعة ثلاثية الأبعاد، تي شيرت بأكمام قصيرة وياقة على شكل V، مقاسات كبيرة، تونيك عصري',
    'OutTop Summer Tops for Women Casual 3D Printing T Shirts Short Sleeve Plus Size Tunic Tops V Neck Tees Fashion Tops Blouses',
    'Omega Psi Phi Pm2.5 Face Bandana Men Women 5-Layer Activated Carbon Filters Breathable Scarf Shield 2 Sizes Medium',
]
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]

Evaluation

Metrics

Knowledge Distillation

Dataset: valid
Evaluated with MSEEvaluator

Metric	Value
negative_mse	-47.9391

Translation

Dataset: valid
Evaluated with TranslationEvaluator

Metric	Value
src2trg_accuracy	0.981
trg2src_accuracy	0.98
mean_accuracy	0.9805

Training Details

Training Dataset

csv

Dataset: csv
Size: 167,423 training samples
Columns: ar, en, and label
Approximate statistics based on the first 1000 samples:
ar en label
type string string list
details
min: 5 tokens
mean: 26.34 tokens
max: 92 tokens

min: 5 tokens
mean: 20.76 tokens
max: 62 tokens

size: 1024 elements

	ar	en	label
type	string	string	list
details	min: 5 tokens mean: 26.34 tokens max: 92 tokens	min: 5 tokens mean: 20.76 tokens max: 62 tokens	size: 1024 elements

Samples:

ar	en	label
`سترة نسائية منفوخة بنقشة مبطنة متعددة من بايونير وومن، مقاس كبير، أخضر`	`The Pioneer Woman Multi-Quilt Print Puffer Vest, Women's, Size: Large, Green`	`[-0.6800060272216797, -0.4665766954421997, 0.0702742338180542, -0.1330789178609848, -0.5821290016174316, ...]`
`جيلدان 64800l بولو بيكيه ناعم للنساء - أبيض، مقاس صغير`	`Gildan 64800l Softstyle Women's Pique Polo - White, S`	`[1.899669885635376, 0.18134558200836182, 1.7645859718322754, -0.37660643458366394, -0.6900650858879089, ...]`
`تايتس Dri-Fit Fast بطول 7/8`	`Dri-Fit Fast 7/8 Tights`	`[0.24706406891345978, 0.21082456409931183, 0.16810369491577148, -0.6220837235450745, -0.019172945991158485, ...]`

Loss: MSELoss

Evaluation Dataset

csv

Dataset: csv
Size: 167,423 evaluation samples
Columns: ar, en, and label
Approximate statistics based on the first 1000 samples:
ar en label
type string string list
details
min: 6 tokens
mean: 26.53 tokens
max: 76 tokens

min: 3 tokens
mean: 21.18 tokens
max: 60 tokens

size: 1024 elements

	ar	en	label
type	string	string	list
details	min: 6 tokens mean: 26.53 tokens max: 76 tokens	min: 3 tokens mean: 21.18 tokens max: 60 tokens	size: 1024 elements

Samples:

ar	en	label
`خاتم خطوبة مميز بقطع وسادة مرصع بالألماس المجروش وحلقة مشقوقة`	`Outstanding Split Shank Crushed Ice Cushion Cut Engagement Ring`	`[0.08735761046409607, 0.38289809226989746, -0.5759294629096985, -0.0157681442797184, -0.05956844612956047, ...]`
`حذاء تومز فالسيا المنصة الإسبادريل باللون الطبيعي من نوردستروم، مقاس 11`	`Toms Valcia Platform Espadrille in Natural at Nordstrom, Size 11`	`[-0.5777891278266907, -1.0087523460388184, -0.5347061157226562, 1.0381789207458496, -1.0726746320724487, ...]`
`جاكيت نسائي مبطن بالفرو الصناعي من أكوي مع غطاء للرأس مقاوم للرياح والماء بسحاب`	`AKOEE Sherpa Fleece Lined Jackets for Women with Hood Windproof Waterproof Zip Up`	`[-1.8026539087295532, -0.15208052098751068, 0.7216541767120361, 0.5569666028022766, -0.7204934358596802, ...]`

Loss: MSELoss

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 2
per_device_eval_batch_size: 64
gradient_accumulation_steps: 32
learning_rate: 2e-05
num_train_epochs: 5
warmup_ratio: 0.1
fp16: True
push_to_hub: 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: 2
per_device_eval_batch_size: 64
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 32
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: 5
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: True
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
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
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