Hugging Face's logo

turanyigitpazarama
/
qwen-0.6B-turkish-ecommerce-tuned

Sentence Similarity
sentence-transformers
Safetensors
qwen3
feature-extraction
dense
Generated from Trainer
dataset_size:72985
loss:MultipleNegativesRankingLoss
text-embeddings-inference
Model card Files
xet
 Community
qwen-0.6B-turkish-ecommerce-tuned / README.md
turanyigitpazarama's picture
turanyigitpazarama
Fine-tuned Qwen 0.6B on Pazarama triplets for e-commerce general use.
d399230 verified
preview code
|
raw
history blame
13.7 kB
metadata 
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:72985
  - loss:MultipleNegativesRankingLoss
base_model: Qwen/Qwen3-Embedding-0.6B
widget:
  - source_sentence: kablosuz klavye mouse set
    sentences:
      - Dexim KM-317 Slim DKM004 Kablosuz Klavye Mouse Set
      - EXEP  BASİC BEYAZ 200 3 LÜ ANKASTRE SET (ED400-EO320-EF200)
      - Samsung Galaxy A17 128 GB 4 GB Siyah Cep Telefonu
  - source_sentence: masa tenisi filesi
    sentences:
      - Motul 8100 X-Cess Gen2  5W40 5lt
      - >-
        BUFFER® Mekanizmalı Taşınabilir Kaymaz Masa Tenisi Filesi Ağı Portatif
        Tüm Masalara Uyumlu
      - >-
        Robin Home Dragon Oyuncu Masası Gamer Masa Gaming Bilgisayar Masası
        Siyah
  - source_sentence: yapışkanlı banyo rafı
    sentences:
      - Çaykur Tiryaki 5 kg Çay
      - >-
        Yapışkanlı 2 Adet Banyo Düzenleyici Rafı ve Katı Sabunluk Bulaşık
        Süngerlik Seti
      - >-
        HESA Dekoratif 5 Raflı Ahşap Ayaklı Kitaplık Ayakkabılık mutfak rafı -
        Atlantik Çam
  - source_sentence: raflı kitaplık
    sentences:
      - Apple iPhone 16e 128GB Beyaz
      - >-
        Morpanya Sera Metal Kitaplık 5 Raflı Çok Amaçlı Dosya Kitap Rafı Ofis
        Salon Mutfak Raf 150 cm Ceviz
      - >-
        Piev House by House Glow Vitrin 360 Döner Standlı Çekmeceli Raflı
        Hazneli Makyaj Kozmetik Organizer
  - source_sentence: balon standı
    sentences:
      - Mermer Ayaklı Gümüş / Metal Banyo El Havluluk Kağıt Havluluk Standı
      - 2 Adet Balon Süsleme Standı 7li Çubuklu Ikili Set Ayaklı Standı
      - Philips Avent Hızlı Biberon Isıtıcı SCF355/07
pipeline_tag: sentence-similarity
library_name: sentence-transformers

SentenceTransformer based on Qwen/Qwen3-Embedding-0.6B

This is a sentence-transformers model finetuned from Qwen/Qwen3-Embedding-0.6B. 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: Qwen/Qwen3-Embedding-0.6B
  • Maximum Sequence Length: 256 tokens
  • Output Dimensionality: 1024 dimensions
  • Similarity Function: Cosine Similarity

Model Sources

Full Model Architecture 

SentenceTransformer(
  (0): Transformer({'max_seq_length': 256, 'do_lower_case': False, 'architecture': 'Qwen3Model'})
  (1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': False, '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': True, '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("turanyigitpazarama/qwen-0.6B-turkish-ecommerce-tuned")
# Run inference
queries = [
    "balon stand\u0131",
]
documents = [
    '2 Adet Balon Süsleme Standı 7li Çubuklu Ikili Set Ayaklı Standı',
    'Mermer Ayaklı Gümüş / Metal Banyo El Havluluk Kağıt Havluluk Standı',
    'Philips Avent Hızlı Biberon Isıtıcı SCF355/07',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 1024] [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[ 0.6758,  0.3398, -0.0081]], dtype=torch.bfloat16)

Training Details

Training Dataset

Unnamed Dataset

  • Size: 72,985 training samples
  • Columns: sentence_0, sentence_1, and sentence_2
  • Approximate statistics based on the first 1000 samples:
    sentence_0 sentence_1 sentence_2
    type string string string
    details
    • min: 2 tokens
    • mean: 7.15 tokens
    • max: 18 tokens
    • min: 6 tokens
    • mean: 26.01 tokens
    • max: 87 tokens
    • min: 6 tokens
    • mean: 27.99 tokens
    • max: 102 tokens
  • Samples:
    sentence_0 sentence_1 sentence_2
    ayçiçek yağı Komili Ayçiçek Yağı 4 lt Shiffa Home Katı Hindistan Cevizi Yağı 330 ml.
    the purest solutions The Purest Solutions Bha %2 Oil Control Toner & Siyah Nokta Hedefli, Yağlanma, Gözenek Dengeleyici T Stanley The AeroLight Transit Lacivert 0.35 lt Termos Bardak
    banyo paspası Eko Trend Djt 3 Lü Yıkanabilir Kaymaz Taban Banyo Paspas Seti 747 Klasik Sepetli Bursa Banyo Dolabı
  • Loss: MultipleNegativesRankingLoss with these parameters:
    {
    "scale": 20.0,
    "similarity_fct": "cos_sim",
    "gather_across_devices": false
}

Training Hyperparameters

Non-Default Hyperparameters

  • per_device_train_batch_size: 16
  • per_device_eval_batch_size: 16
  • multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand
  • per_device_train_batch_size: 16
  • num_train_epochs: 3
  • max_steps: -1
  • learning_rate: 5e-05
  • lr_scheduler_type: linear
  • lr_scheduler_kwargs: None
  • warmup_steps: 0
  • 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
  • label_smoothing_factor: 0.0
  • bf16: False
  • fp16: False
  • 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: trackio
  • eval_strategy: no
  • per_device_eval_batch_size: 16
  • 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
  • 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_backend: None
  • ddp_timeout: 1800
  • fsdp: []
  • fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
  • deepspeed: None
  • debug: []
  • skip_memory_metrics: True
  • do_predict: False
  • resume_from_checkpoint: None
  • warmup_ratio: None
  • local_rank: -1
  • prompts: None
  • batch_sampler: batch_sampler
  • multi_dataset_batch_sampler: round_robin
  • router_mapping: {}
  • learning_rate_mapping: {}

Training Logs

Epoch Step Training Loss
0.1096 500 0.3013
0.2192 1000 0.1969
0.3288 1500 0.1656
0.4384 2000 0.1587
0.5480 2500 0.1445
0.6576 3000 0.1511
0.7672 3500 0.1298
0.8768 4000 0.1384
0.9864 4500 0.1397
1.0960 5000 0.0981
1.2056 5500 0.0914
1.3152 6000 0.0862
1.4248 6500 0.0882
1.5344 7000 0.0902
1.6440 7500 0.0855
1.7536 8000 0.0863
1.8632 8500 0.0929
1.9728 9000 0.0868
2.0824 9500 0.0795
2.1920 10000 0.0801
2.3016 10500 0.0799
2.4112 11000 0.0778
2.5208 11500 0.0777
2.6304 12000 0.0821
2.7400 12500 0.0730
2.8496 13000 0.0742
2.9592 13500 0.0763

Framework Versions

  • Python: 3.13.2
  • Sentence Transformers: 5.1.1
  • Transformers: 5.2.0
  • PyTorch: 2.10.0+cu128
  • Accelerate: 1.10.1
  • Datasets: 3.6.0
  • Tokenizers: 0.22.2

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{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}