RexBERT-base-embed-pf-v0.1

SentenceTransformer based on thebajajra/RexBERT-base

This is a sentence-transformers model finetuned from thebajajra/RexBERT-base on the nomic-embed-unsupervised-data dataset. 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: thebajajra/RexBERT-base <!-- at revision 4f66d2977864414371770084b681e00698b98457 -->
• Maximum Sequence Length: 1024 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • nomic-embed-unsupervised-data
• Language: en

<!-- - **License:** Unknown -->

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': 1024, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
  (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("sentence_transformers_model_id")
# Run inference
queries = [
    "Where do you guys go to find used camper shells?",
]
documents = [
    "I've got a newly acquired 1st gen 2005 silvee Toyota tundra trd and am looking for an used camper shell.  Craigslist hasnt been very useful....where do you guys go?\n\nThanks!",
    "I work at a convenience store and the number of Newports I sell a day is insane. Considering buying a couple cartons of em and maybe some parliament menthols if the FDA goes through with this. Should be able to throw em up on craigslist or ebay a week or two later and it'll be like steaks in a piranha pond",
    "Hey guys what is the most optimal tool for pulling long staples out from hardwood flooring? I'm trying to find the most optimal way to do it because I have thousands to pull! Fence pliers did not work too well on account the pointy tip was too thick get in and roll them out and when i tried the gripping/cutting part it broke the staples.\n\nI'm thinking round nose vice grips or a car gasket puller?\n\nThanks",
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 768] [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[0.8108, 0.2481, 0.1200]])

<!-- ### Direct Usage (Transformers) <details><summary>Click to see the direct usage in Transformers</summary> </details> --> <!-- ### Downstream Usage (Sentence Transformers) You can finetune this model on your own dataset. <details><summary>Click to expand</summary> </details> --> <!-- ### Out-of-Scope Use *List how the model may foreseeably be misused and address what users ought not to do with the model.* --> <!-- ## Bias, Risks and Limitations *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* --> <!-- ### Recommendations *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* -->

Training Details

Training Dataset

nomic-embed-unsupervised-data

• Dataset: nomic-embed-unsupervised-data at 917bae6
• Size: 222,490,215 training samples
• Columns: <code>query</code> and <code>document</code>
• Approximate statistics based on the first 1000 samples:

  	query	document
  type	string	string
  details	<ul><li>min: 6 tokens</li><li>mean: 16.83 tokens</li><li>max: 62 tokens</li></ul>	<ul><li>min: 12 tokens</li><li>mean: 162.25 tokens</li><li>max: 1024 tokens</li></ul>

• Samples:

  query	document
  <code>I became a US citizen early this year and this is going to be my first 4th of July as an American!</code>	<code>Because of the current situation, my citizen oath ceremony felt more like a pick up order... Got my certificate, and no guests allowed, so I couldn’t bring anybody to join my ceremony, also no pictures. <br><br>Anyway... I want to celebrate big time this 4th of July, and I’m already planning it! (Any ideas are super welcome!). I say big time but I just really want to do something fun at home with my family. 😊</code>
  <code>"The Kingdom of God for Jesus"; I know you guys know how to answer this overrated question.</code>	<code>Basically what we're talking about is that the "kingdom" of god according to jesus are:<br><br>* "the kingdom as good news (where the kingdom is on earth, whereas by living a beautiful, meaningful life on earth is the meaning of salvation)" <br>* "the kingdom is offered to all"<br>* etc.<br><br>and finally, the question goes like this: "The Kingdom Does Not Ask for Performance; It is a gift, an offer. We can only inherit it. So, what is the point of being good?"</code>
  <code>So I made a "size" chart to go with my weight infograph, all based off that "Relative champ weight/height" thread.</code>	<code>Here's the weight chart I did the other day<br><br><br><br>And here's the size chart I did today. <br><br><br><br>*Anivia, Skarner and Shyvanna (dragon form) are "Dimensions" instead of an actual "height", but I think you can get the jist.<br><br>The original thread this is based off of is located via the link below. I am using these numbers (and my own conversions), so I'm not always sure where they got the numbers!<br><br></code>

• Loss: <code>MultipleNegativesRankingLoss</code> with these parameters:

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

Evaluation Dataset

nomic-embed-unsupervised-data

• Dataset: nomic-embed-unsupervised-data at 917bae6
• Size: 222,727 evaluation samples
• Columns: <code>query</code> and <code>document</code>
• Approximate statistics based on the first 1000 samples:

  	query	document
  type	string	string
  details	<ul><li>min: 6 tokens</li><li>mean: 16.41 tokens</li><li>max: 66 tokens</li></ul>	<ul><li>min: 15 tokens</li><li>mean: 164.47 tokens</li><li>max: 1024 tokens</li></ul>

• Samples:

  query	document
  <code>Do you subscribe to any horror magazines?</code>	<code>I get most of my horror news from blogs and websites and such, but i do subscribe to a bunch of horror mags. With everything being so digital these days, something about flipping through a magazine and reading articles about both classic and upcoming horror movies is refreshing. I get a lot of great recommendations from them, and theres a lot of interesting interviews and behind the scenes stuff that i dont see on the popular websites.</code>
  <code>Missing PDS Laundry Card :(</code>	<code>This is an absolute long shot but I must've accidentally left my laundry card in the dryer card slot because I cant find it anywhere. If someone found a card in there, please DM me. I've already bought a card but I'd like to have my original card back :(</code>
  <code>Talking Bad will be terrible</code>	<code>Talking Dead is horrible and this will be to. Chris Hardwick and the cast of random no name celebrities offer nothing new to the discussion. The only good thing about Breaking Bad ending is that Talking Bad will end soon as well.</code>

• Loss: <code>MultipleNegativesRankingLoss</code> with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 256
• per_device_eval_batch_size: 128
• learning_rate: 2e-06
• num_train_epochs: 4
• warmup_ratio: 0.1
• bf16: True
• batch_sampler: no_duplicates

All Hyperparameters

<details><summary>Click to expand</summary>

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 256
• per_device_eval_batch_size: 128
• 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: 2e-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: 4
• 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
• bf16: True
• 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: True
• 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}
• parallelism_config: None
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• project: huggingface
• trackio_space_id: trackio
• 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: no
• 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: True
• prompts: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

</details>

Training Logs

<details><summary>Click to expand</summary> </details>

Framework Versions

• Python: 3.11.10
• Sentence Transformers: 5.1.2
• Transformers: 4.57.1
• PyTorch: 2.4.1+cu121
• Accelerate: 1.11.0
• Datasets: 4.3.0
• Tokenizers: 0.22.1

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}
}

<!-- ## Glossary *Clearly define terms in order to be accessible across audiences.* --> <!-- ## Model Card Authors *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.* --> <!-- ## Model Card Contact *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.* -->