CoExperiences/snowflake-l-marketing-tuned

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.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: Snowflake/snowflake-arctic-embed-l
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 1024 tokens
• 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("CoExperiences/snowflake-l-marketing-tuned")
# Run inference
sentences = [
    "How does Alexis Krivkovich's perspective as a mother influence her optimism about the future of women in the workplace?",
    'Lucia Rahilly: Sometimes, I feel that we’ve been talking about these issues since I was in college, and that can feel discouraging. What are you most optimistic about going into 2022, coming out of this Women in the Workplace report?\n\nAlexis Krivkovich: I’m most optimistic about the fact that we’re having an honest conversation, and now with a real fact base. We’re not talking about these things as perception but as real and measured experiences that companies can’t hide from—and they don’t want to.\n\nAs a mother of three young daughters, it gives me real hope because I’ve been thinking about this question for 20 years. But in 20 years, when they’re fully in the workplace, maybe we’ll have a totally different paradigm.',
    'Learn more about the \nWork Happiness Score at: \ngo.indeed.com/happiness',
]
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	0.81
cosine_accuracy@3	0.93
cosine_accuracy@5	0.97
cosine_accuracy@10	0.98
cosine_precision@1	0.81
cosine_precision@3	0.31
cosine_precision@5	0.194
cosine_precision@10	0.098
cosine_recall@1	0.81
cosine_recall@3	0.93
cosine_recall@5	0.97
cosine_recall@10	0.98
cosine_ndcg@10	0.9037
cosine_mrr@10	0.8781
cosine_map@100	0.8798
dot_accuracy@1	0.81
dot_accuracy@3	0.93
dot_accuracy@5	0.97
dot_accuracy@10	0.98
dot_precision@1	0.81
dot_precision@3	0.31
dot_precision@5	0.194
dot_precision@10	0.098
dot_recall@1	0.81
dot_recall@3	0.93
dot_recall@5	0.97
dot_recall@10	0.98
dot_ndcg@10	0.9037
dot_mrr@10	0.8781
dot_map@100	0.8798

Training Details

Training Dataset

Unnamed Dataset

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

  	sentence_0	sentence_1
  type	string	string
  details	min: 9 tokens mean: 20.08 tokens max: 39 tokens	min: 5 tokens mean: 110.85 tokens max: 187 tokens

• Samples:

  sentence_0	sentence_1
  What significant change occurred in employees' perceptions of their employer's care for their wellbeing during the pandemic?	Workplace Percent Who Feel Employer Cares About Their Wellbeing Plummets Share on LinkedIn Share on Twitter Share on Facebook Share via Email Print Share on LinkedIn Share on Twitter Share on Facebook Share via Email Print Workplace March 18, 2022 Percent Who Feel Employer Cares About Their Wellbeing Plummets by Jim Harter Story Highlights Employees' perceptions of their organization caring about their wellbeing drops During the onset of the pandemic, employees felt employers had more care and concern Employees who feel their employer cares about their wellbeing are 69% less likely to actively search for a job
  How does feeling cared for by an employer impact employees' job search behavior?	Workplace Percent Who Feel Employer Cares About Their Wellbeing Plummets Share on LinkedIn Share on Twitter Share on Facebook Share via Email Print Share on LinkedIn Share on Twitter Share on Facebook Share via Email Print Workplace March 18, 2022 Percent Who Feel Employer Cares About Their Wellbeing Plummets by Jim Harter Story Highlights Employees' perceptions of their organization caring about their wellbeing drops During the onset of the pandemic, employees felt employers had more care and concern Employees who feel their employer cares about their wellbeing are 69% less likely to actively search for a job
  What percentage of U.S. employees feel strongly that their organization cares about their wellbeing?	Fewer than one in four U.S. employees feel strongly that their organization cares about their wellbeing -- the lowest percentage in nearly a decade. This finding has significant implications, as work and life have never been more blended and employee wellbeing matters more than ever-- to employees and the resiliency of organizations. The discovery is based on a random sample of 15,001 full and part-time U.S. employees who were surveyed in February 2022.

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          1024,
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 20
• per_device_eval_batch_size: 20
• num_train_epochs: 5
• 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: 20
• per_device_eval_batch_size: 20
• 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: 5
• 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}
• 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: False
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• 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
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	cosine_map@100
1.0	30	0.8782
1.6667	50	0.8878
2.0	60	0.8854
3.0	90	0.8853
3.3333	100	0.8845
4.0	120	0.8793
5.0	150	0.8798

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.2.0
• Transformers: 4.45.2
• PyTorch: 2.5.0+cu124
• Accelerate: 0.34.2
• Datasets: 3.0.1
• Tokenizers: 0.20.1

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}
}