jimmydzj2006/snowflake-arctic-embed-xs_finetuned_aipolicy

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

This is a sentence-transformers model finetuned from Snowflake/snowflake-arctic-embed-xs. It maps sentences & paragraphs to a 384-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-xs
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 384 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': 384, '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("jimmydzj2006/snowflake-arctic-embed-xs_finetuned_aipolicy")
# Run inference
sentences = [
    'How can human subjects revoke their consent according to the suggested action?',
    'Action ID: MS-2.2-003\nSuggested Action: Provide human subjects with options to withdraw participation or revoke their\nconsent for present or future use of their data in GAI applications.\nGAI Risks: Data Privacy; Human-AI\nConfiguration; Information\nIntegrity',
    'Disinformation and misinformation – both of which may be facilitated by GAI – may erode public trust in true or valid evidence and information, with downstream effects. For example, a synthetic image of a Pentagon blast went viral and briefly caused a drop in the stock market. Generative AI models can also assist malicious actors in creating compelling imagery and propaganda to support disinformation campaigns, which may not be photorealistic, but could enable these campaigns to gain more reach and engagement on social media platforms. Additionally, generative AI models can assist malicious actors in creating fraudulent content intended to impersonate others.\n\n Trustworthy AI Characteristics: Accountable and Transparent, Safe, Valid and Reliable, Interpretable and Explainable\n\n 2.9. Information Security\n\n Information security for computer systems and data is a mature field with widely accepted and standardized practices for offensive and defensive cyber capabilities. GAI-based systems present two primary information security risks: GAI could potentially discover or enable new cybersecurity risks by lowering the barriers for or easing automated exercise of offensive capabilities; simultaneously, it expands the available attack surface, as GAI itself is vulnerable to attacks like prompt injection or data poisoning. \n\n Offensive cyber capabilities advanced by GAI systems may augment cybersecurity attacks such as hacking, malware, and phishing. Reports have indicated that LLMs are already able to discover some vulnerabilities in systems (hardware, software, data) and write code to exploit them. Sophisticated threat actors might further these risks by developing GAI-powered security co-pilots for use in several parts of the attack chain, including informing attackers on how to proactively evade threat detection and escalate privileges after gaining system access.\n\n Information security for GAI models and systems also includes maintaining availability of the GAI system and the integrity and (when applicable) the confidentiality of the GAI code, training data, and model weights. To identify and secure potential attack points in AI systems or specific components of the AI',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# 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	0.6484
cosine_accuracy@3	0.7363
cosine_accuracy@5	0.7802
cosine_accuracy@10	0.8571
cosine_precision@1	0.6484
cosine_precision@3	0.2454
cosine_precision@5	0.156
cosine_precision@10	0.0857
cosine_recall@1	0.6484
cosine_recall@3	0.7363
cosine_recall@5	0.7802
cosine_recall@10	0.8571
cosine_ndcg@10	0.7434
cosine_mrr@10	0.7083
cosine_map@100	0.7158
dot_accuracy@1	0.6484
dot_accuracy@3	0.7363
dot_accuracy@5	0.7802
dot_accuracy@10	0.8571
dot_precision@1	0.6484
dot_precision@3	0.2454
dot_precision@5	0.156
dot_precision@10	0.0857
dot_recall@1	0.6484
dot_recall@3	0.7363
dot_recall@5	0.7802
dot_recall@10	0.8571
dot_ndcg@10	0.7434
dot_mrr@10	0.7083
dot_map@100	0.7158

Training Details

Training Dataset

Unnamed Dataset

• Size: 2,730 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: 8 tokens mean: 15.71 tokens max: 33 tokens	min: 19 tokens mean: 183.25 tokens max: 467 tokens

• Samples:

  sentence_0	sentence_1
  What is the Action ID associated with the suggested action?	Action ID: MS-2.12-004 Suggested Action: Verify effectiveness of carbon capture or offset programs for GAI training and applications, and address green-washing concerns. GAI Risks: Environmental
  What is the suggested action regarding carbon capture or offset programs?	Action ID: MS-2.12-004 Suggested Action: Verify effectiveness of carbon capture or offset programs for GAI training and applications, and address green-washing concerns. GAI Risks: Environmental
  What specific concerns should be addressed in relation to carbon capture programs?	Action ID: MS-2.12-004 Suggested Action: Verify effectiveness of carbon capture or offset programs for GAI training and applications, and address green-washing concerns. GAI Risks: Environmental

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          284,
          256,
          128,
          64,
          32
      ],
      "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: 16
• per_device_eval_batch_size: 16
• 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: 16
• per_device_eval_batch_size: 16
• 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}
• 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
• eval_use_gather_object: False
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss	cosine_map@100
0.2924	50	-	0.5949
0.5848	100	-	0.6455
0.8772	150	-	0.6680
1.0	171	-	0.6721
1.1696	200	-	0.6811
1.4620	250	-	0.6850
1.7544	300	-	0.6959
2.0	342	-	0.7021
2.0468	350	-	0.7008
2.3392	400	-	0.7043
2.6316	450	-	0.7017
2.9240	500	5.9671	0.7018
3.0	513	-	0.7039
3.2164	550	-	0.7014
3.5088	600	-	0.7039
3.8012	650	-	0.7022
4.0	684	-	0.7058
4.0936	700	-	0.7039
4.3860	750	-	0.7061
4.6784	800	-	0.7030
4.9708	850	-	0.7073
5.0	855	-	0.7073
5.2632	900	-	0.7071
5.5556	950	-	0.7095
5.8480	1000	3.5897	0.7103
6.0	1026	-	0.7080
6.1404	1050	-	0.7075
6.4327	1100	-	0.7089
6.7251	1150	-	0.7087
7.0	1197	-	0.7102
7.0175	1200	-	0.7101
7.3099	1250	-	0.7134
7.6023	1300	-	0.7130
7.8947	1350	-	0.7133
8.0	1368	-	0.7142
8.1871	1400	-	0.7125
8.4795	1450	-	0.7163
8.7719	1500	3.0206	0.7124
9.0	1539	-	0.7144
9.0643	1550	-	0.7158
9.3567	1600	-	0.7159
9.6491	1650	-	0.7158
9.9415	1700	-	0.7158
10.0	1710	-	0.7158

Framework Versions

• Python: 3.11.9
• Sentence Transformers: 3.2.0
• Transformers: 4.44.1
• PyTorch: 2.4.0
• Accelerate: 0.34.2
• Datasets: 3.0.0
• Tokenizers: 0.19.1

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}
}