Chanisorn/thai-food-mpnet-new-v10

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 = [
    'ปลาทูน่า,  พริกขี้หนู,  ไข่ไก่,  น้ำปลา, เล็กน้อย, น้ำมันพืช',
    'ไข่เจียวทูน่าพริกสับ',
    'ข้าวแต๋น',
]
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

• Dataset: thai-food-eval
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.6053
cosine_accuracy@3	0.8421
cosine_accuracy@5	0.9342
cosine_accuracy@10	0.9737
cosine_precision@1	0.6053
cosine_precision@3	0.2807
cosine_precision@5	0.1868
cosine_recall@1	0.6053
cosine_recall@3	0.8421
cosine_recall@5	0.9342
cosine_ndcg@10	0.789
cosine_mrr@10	0.7292
cosine_map@100	0.7302

Training Details

Training Dataset

Unnamed Dataset

• Size: 2,452 training samples
• Columns: anchor and positive
• Approximate statistics based on the first 1000 samples:

  	anchor	positive
  type	string	string
  details	min: 7 tokens mean: 29.15 tokens max: 125 tokens	min: 4 tokens mean: 9.91 tokens max: 22 tokens

• Samples:

  anchor	positive
  ปลาหมึก, ซีอิ๊วดำ, ผงขมิ้น, น้ำปูนใส, กระเทียมสับ, รากผักชี, พริกแดง, น้ำตาลปี๊บ, เกลือ, น้ำปลา, น้ำมะนาว	ปลาหมึกย่าง
  ไปตกหมึกมา อยากทำอะไรกินง่ายๆ ได้รสชาติของปลาหมึกแท้ๆ	ปลาหมึกย่าง
  อยากกินปลาหมึกๆ ซีฟุ้ด อร่อยๆ	ปลาหมึกย่าง

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Evaluation Dataset

Unnamed Dataset

• Size: 76 evaluation samples
• Columns: anchor and positive
• Approximate statistics based on the first 76 samples:

  	anchor	positive
  type	string	string
  details	min: 4 tokens mean: 47.42 tokens max: 86 tokens	min: 4 tokens mean: 10.24 tokens max: 20 tokens

• Samples:

  anchor	positive
  น้ำมันพืช, กระเทียม, น้ำตาลทราย, น้ำปลา, ซีอิ๊วขาว, ซอสปรุงรส, ซีอิ๊วดำเค็ม, น้ำส้มสายชู, พริกไทย, เส้นหมี่แห้ง, ลูกชิ้น, ถั่วงอก	หมี่คลุก
  น้ำมัน, กระเทียม, หมูหมัก, เส้นใหญ่, ซีอิ้วดำ, คะน้า, กระหล่ำปลี, แครอท, ไข่เป็ด, ไข่ไก่, ผงปรุงรส, น้ำตาลทราย, ซอสหอยนางรม, ซอสปรุงรส, พริกไทย	ผัดซีอิ้วเส้นใหญ่
  สะโพกหมู, น้ำตาลทราย, น้ำตาลปี๊บ, ซีอิ๊วขาว, เกลือ, น้ำเปล่า, ลูกผักชี, ยี่หร่า, กระเทียมไทย, สับละเอียด, น้ำมันพืช	หมูสวรรค์

• 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: 6
• 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: 6
• 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
• hub_revision: None
• 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
• liger_kernel_config: None
• 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
0.0971	10	3.2532	-	-
0.1942	20	2.6975	-	-
0.2913	30	2.3365	-	-
0.3883	40	1.9787	-	-
0.4854	50	1.9125	-	-
0.5825	60	1.7024	-	-
0.6796	70	1.6074	-	-
0.7767	80	1.3358	-	-
0.8738	90	1.4281	-	-
0.9709	100	1.4312	-	-
1.0	103	-	0.9767	0.6681
1.0680	110	1.1873	-	-
1.1650	120	1.1148	-	-
1.2621	130	1.1163	-	-
1.3592	140	1.0429	-	-
1.4563	150	0.9594	-	-
1.5534	160	0.9593	-	-
1.6505	170	1.0314	-	-
1.7476	180	1.0236	-	-
1.8447	190	1.0052	-	-
1.9417	200	1.0062	-	-
2.0	206	-	0.6975	0.7435
2.0388	210	0.8259	-	-
2.1359	220	0.6713	-	-
2.2330	230	0.7833	-	-
2.3301	240	0.8613	-	-
2.4272	250	0.6706	-	-
2.5243	260	0.8971	-	-
2.6214	270	0.7678	-	-
2.7184	280	0.741	-	-
2.8155	290	0.6872	-	-
2.9126	300	0.7854	-	-
3.0	309	-	0.6185	0.7481
3.0097	310	0.7095	-	-
3.1068	320	0.6708	-	-
3.2039	330	0.6311	-	-
3.3010	340	0.6769	-	-
3.3981	350	0.5816	-	-
3.4951	360	0.6604	-	-
3.5922	370	0.6356	-	-
3.6893	380	0.5459	-	-
3.7864	390	0.5856	-	-
3.8835	400	0.6812	-	-
3.9806	410	0.5893	-	-
4.0	412	-	0.5796	0.7742
4.0777	420	0.4721	-	-
4.1748	430	0.4353	-	-
4.2718	440	0.5372	-	-
4.3689	450	0.6343	-	-
4.4660	460	0.6572	-	-
4.5631	470	0.601	-	-
4.6602	480	0.5418	-	-
4.7573	490	0.5312	-	-
4.8544	500	0.5055	-	-
4.9515	510	0.5447	-	-
5.0	515	-	0.5373	0.7877
5.0485	520	0.5501	-	-
5.1456	530	0.5831	-	-
5.2427	540	0.5378	-	-
5.3398	550	0.4975	-	-
5.4369	560	0.5326	-	-
5.5340	570	0.3991	-	-
5.6311	580	0.473	-	-
5.7282	590	0.4915	-	-
5.8252	600	0.4234	-	-
5.9223	610	0.5445	-	-
6.0	618	-	0.5209	0.789

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.11.13
• Sentence Transformers: 4.1.0
• Transformers: 4.53.0
• PyTorch: 2.6.0+cu124
• Accelerate: 1.8.1
• Datasets: 2.14.4
• Tokenizers: 0.21.2

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}
}