deman539/food-review-ft-snowflake-l-f18eeff6-7504-48c7-af10-1d2d85ca8caa

SentenceTransformer based on Snowflake/snowflake-arctic-embed-l

This is a sentence-transformers model finetuned from Snowflake/snowflake-arctic-embed-l. 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. This particular checkpoint is finetuned on food and restaurant reviews and is optimized to answer questions from users about this topic.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: Snowflake/snowflake-arctic-embed-l
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 1024 dimensions
• 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': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (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("deman539/food-review-ft-snowflake-l-f18eeff6-7504-48c7-af10-1d2d85ca8caa")
# Run inference
sentences = [
    'What aspects of 10 Downing Street does Ashutosh Tiwari highlight in his review?  ',
    'Restaurant: 10 Downing Street\nReviewer: Ashutosh Tiwari\nReview: 10D is one of the best places to hangout witj friends and families. Great ambience with awesome views. Food and staff behaviour is very kind.\nRating: 4\nMetadata: 4 Reviews , 84 Followers\nTime: 1/5/2019 17:17\nPictures: 0\n7514:',
    'Restaurant: Cafe Eclat\nReviewer: Kamal Prakash\nReview: I really liked the ambience. The blue cushions complimented the tables with wooden finish. The glass doors added to the elegance. The place was very calm. I had the cheesecake here, it literally melted in my mouth, absolutely loved it. One downside is that the place is a bit expensive.\nRating: 4\nMetadata: 14 Reviews , 31 Followers\nTime: 5/10/2018 18:59\nPictures: 2\n7514:',
]
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

Information Retrieval

• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.905
cosine_accuracy@3	0.975
cosine_accuracy@5	0.985
cosine_accuracy@10	0.995
cosine_precision@1	0.905
cosine_precision@3	0.325
cosine_precision@5	0.197
cosine_precision@10	0.0995
cosine_recall@1	0.905
cosine_recall@3	0.975
cosine_recall@5	0.985
cosine_recall@10	0.995
cosine_ndcg@10	0.9548
cosine_mrr@10	0.9413
cosine_map@100	0.9418

Training Details

Training Dataset

Unnamed Dataset

• Size: 1,600 training samples
• Columns: sentence_0 and sentence_1
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1
  type	string	string
  details	min: 11 tokens mean: 20.32 tokens max: 33 tokens	min: 12 tokens mean: 102.78 tokens max: 247 tokens

• Samples:

  sentence_0	sentence_1
  What aspects of Khaan Saab did Dakshay Singh highlight in his review?	Restaurant: Khaan Saab Reviewer: Dakshay Singh Review: Great food. Excellent ambience for a nice quiet dinner for family. Zomato gold benefits can be availed here. Excellent customer service. Great service by Tapan. Very happy Rating: 5 Metadata: 9 Reviews , 9 Followers Time: 4/8/2019 22:23 Pictures: 0 7514:
  Who provided great service according to Dakshay Singh's review of Khaan Saab?	Restaurant: Khaan Saab Reviewer: Dakshay Singh Review: Great food. Excellent ambience for a nice quiet dinner for family. Zomato gold benefits can be availed here. Excellent customer service. Great service by Tapan. Very happy Rating: 5 Metadata: 9 Reviews , 9 Followers Time: 4/8/2019 22:23 Pictures: 0 7514:
  What specific type of parathas did Raj Rohit praise in his review of Triptify?	Restaurant: Triptify Reviewer: Raj Rohit Review: Oh my my. What great parathas. These guys know their game when it comes to parathas. The corn and cheese parathas are brilliant + their sides blend so perfectly with the parathas. Brilliant packaging too. Rating: 5 Metadata: 124 Reviews , 372 Followers Time: 8/25/2018 12:58 Pictures: 1 7514:

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 10
• per_device_eval_batch_size: 10
• num_train_epochs: 10
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 10
• per_device_eval_batch_size: 10
• 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: 10
• 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}
• tp_size: 0
• 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: batch_sampler
• multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss	cosine_ndcg@10
0.3125	50	-	0.9239
0.625	100	-	0.9313
0.9375	150	-	0.9307
1.0	160	-	0.9301
1.25	200	-	0.9382
1.5625	250	-	0.9454
1.875	300	-	0.9501
2.0	320	-	0.9532
2.1875	350	-	0.9501
2.5	400	-	0.9559
2.8125	450	-	0.9505
3.0	480	-	0.9529
3.125	500	0.5558	0.9518
3.4375	550	-	0.9425
3.75	600	-	0.9547
4.0	640	-	0.9551
4.0625	650	-	0.9539
4.375	700	-	0.9637
4.6875	750	-	0.9564
5.0	800	-	0.9624
5.3125	850	-	0.9648
5.625	900	-	0.9577
5.9375	950	-	0.9601
6.0	960	-	0.9632
6.25	1000	0.0655	0.9613
6.5625	1050	-	0.9544
6.875	1100	-	0.9551
7.0	1120	-	0.9558
7.1875	1150	-	0.9562
7.5	1200	-	0.9566
7.8125	1250	-	0.9546
8.0	1280	-	0.9569
8.125	1300	-	0.9584
8.4375	1350	-	0.9573
8.75	1400	-	0.9566
9.0	1440	-	0.9569
9.0625	1450	-	0.9552
9.375	1500	0.0417	0.9549
9.6875	1550	-	0.9548
10.0	1600	-	0.9548

Framework Versions

• Python: 3.13.2
• Sentence Transformers: 4.1.0
• Transformers: 4.51.3
• PyTorch: 2.7.0
• Accelerate: 1.6.0
• Datasets: 3.5.1
• 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",
}

MatryoshkaLoss

@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}

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}
}