Chanisorn/thai-food-mpnet-new-v8

Thai Food Ingredients → Dish Prediction

This is a sentence-transformers model finetuned from sentence-transformers/paraphrase-multilingual-mpnet-base-v2. It maps sentences & paragraphs to a 768-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: sentence-transformers/paraphrase-multilingual-mpnet-base-v2
• Maximum Sequence Length: 128 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity
• Language: th
• License: apache-2.0

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': 128, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (1): Pooling({'word_embedding_dimension': 768, '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})
)

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("thai_food_prediction1")
# Run inference
sentences = [
    'กะปิ,  ข้าวสวย,  หมูสามชั้น,  น้ำตาล,  กระเทียม, สัปปะรด,  กุนเชียง,  ไข่ไก่,  ไข่เค็ม,  ปลาทู,  ถั่วฝักยาว,  หอมแดง,  พริกขี้หนู,  มะม่วง,  กุ้งแห้ง,  ผักชี,  ซีอิ้วดำ',
    'ข้าวคลุกกะปิ',
    'มันบดกระเทียมย่าง',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

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

Evaluation

Metrics

Information Retrieval

• Datasets: thai-food-eval and thai-food-test-eval
• Evaluated with InformationRetrievalEvaluator

Metric	thai-food-eval	thai-food-test-eval
cosine_accuracy@1	0.625	0.6531
cosine_accuracy@3	0.8542	0.8776
cosine_accuracy@5	0.875	0.8776
cosine_accuracy@10	0.9375	0.9388
cosine_precision@1	0.625	0.6531
cosine_precision@3	0.2847	0.2925
cosine_precision@5	0.175	0.1755
cosine_recall@1	0.625	0.6531
cosine_recall@3	0.8542	0.8776
cosine_recall@5	0.875	0.8776
cosine_ndcg@10	0.7846	0.8137
cosine_mrr@10	0.7353	0.7727
cosine_map@100	0.7383	0.7758

Training Details

Training Dataset

Unnamed Dataset

• Size: 1,179 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 4 tokens mean: 31.07 tokens max: 105 tokens	min: 5 tokens mean: 9.97 tokens max: 23 tokens

• Samples:

  anchor	positive
  พริกแห้ง, หอมแดง, กระเทียม, น้ำมัน, หมูสับ, น้ำสะอาด, น้ำตาลปี๊บ, น้ำปลา, มะขามเปียก	น้ำพริกเผาผัดหมูสับ
  มีพริกแห้ง, หอมแดง, กระเทียม, หมูสับ ทำอะไรได้บ้าง	น้ำพริกเผาผัดหมูสับ
  น้ำพริกเผาหมูสับ สูตรผัดราดข้าว	น้ำพริกเผาผัดหมูสับ

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim"
  }
  

Evaluation Dataset

Unnamed Dataset

• Size: 48 evaluation samples
• Columns: anchor and positive
• Approximate statistics based on the first 48 samples:

  	anchor	positive
  type	string	string
  details	min: 12 tokens mean: 41.1 tokens max: 82 tokens	min: 4 tokens mean: 9.83 tokens max: 19 tokens

• Samples:

  anchor	positive
  หมูสับ, หนำเลี๊ยบ, ซีอิ๊วขาว, น้ำมันหอย, พริกไทย, กระเทียม, น้ำมัน, ผงปรุงรส, น้ำตาลทราย	หมูสับผัดหนำเลี๊ยบ
  ใบกะเพรา, เส้นสปาเก็ตตี้, เห็ด, เนื้อหมู, กุ้ง, ปลาหมึก, ผัก, พริก, กระเทียม, ซอสหอยนางรม, ซีอิ๊วขาว, น้ำปลา, ซีอิ๊วดำ, น้ำตาล, น้ำมัน	สปาเก็ตตี้ขี้เมาทะเล
  ไข่ไก่, กุ้ง, ซีอิ้วขาว, น้ำมัน, พริกไทย	ไข่เจียวกุ้ง

• Loss: MultipleNegativesRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• per_device_train_batch_size: 24
• per_device_eval_batch_size: 24
• learning_rate: 5e-06
• num_train_epochs: 7
• warmup_ratio: 0.1
• load_best_model_at_end: True
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: epoch
• prediction_loss_only: True
• per_device_train_batch_size: 24
• per_device_eval_batch_size: 24
• 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-06
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 7
• 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: 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: True
• 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: 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
• 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: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	Validation Loss	thai-food-eval_cosine_ndcg@10	thai-food-test-eval_cosine_ndcg@10
0.2	10	3.0739	-	-	-
0.4	20	2.4736	-	-	-
0.6	30	2.0271	-	-	-
0.8	40	1.6695	-	-	-
1.0	50	1.516	1.6674	0.6194	-
1.2	60	1.3391	-	-	-
1.4	70	1.2024	-	-	-
1.6	80	1.2809	-	-	-
1.8	90	1.1296	-	-	-
2.0	100	0.8932	1.2661	0.6572	-
2.2	110	0.9307	-	-	-
2.4	120	0.7843	-	-	-
2.6	130	0.9402	-	-	-
2.8	140	0.8192	-	-	-
3.0	150	0.9192	1.0902	0.7233	-
3.2	160	0.6749	-	-	-
3.4	170	0.641	-	-	-
3.6	180	0.763	-	-	-
3.8	190	0.8294	-	-	-
4.0	200	0.6798	0.9843	0.7657	-
4.2	210	0.7125	-	-	-
4.4	220	0.5714	-	-	-
4.6	230	0.5759	-	-	-
4.8	240	0.5941	-	-	-
5.0	250	0.5246	0.9562	0.7715	-
5.2	260	0.6761	-	-	-
5.4	270	0.5069	-	-	-
5.6	280	0.5447	-	-	-
5.8	290	0.5582	-	-	-
6.0	300	0.493	0.9472	0.7772	-
6.2	310	0.4803	-	-	-
6.4	320	0.4909	-	-	-
6.6	330	0.4831	-	-	-
6.8	340	0.4818	-	-	-
7.0	350	0.4934	0.9471	0.7846	-
-1	-1	-	-	-	0.8137

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.11.13
• Sentence Transformers: 4.1.0
• Transformers: 4.52.4
• PyTorch: 2.6.0+cu124
• Accelerate: 1.7.0
• Datasets: 2.14.4
• Tokenizers: 0.21.1

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",
}

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}
}