Add model card

README.md

@@ -1,431 +1,65 @@
 ---
+language: en
+license: mit
 tags:
-- sentence-transformers
-- sentence-similarity
-- feature-extraction
-- dense
-- generated_from_trainer
-- dataset_size:275838
-- loss:TripletLoss
-base_model: sentence-transformers/all-MiniLM-L12-v2
-widget:
-- source_sentence: Thus saith the LORD of hosts, the God of Israel; As yet they shall
-    use this speech in the land of Judah and in the cities thereof, when I shall bring
-    again their captivity; The LORD bless thee, O habitation of justice, and mountain
-    of holiness.
-  sentences:
-  - 'The LORD shall bless thee out of Zion: and thou shalt see the good of Jerusalem
-    all the days of thy life.'
-  - And he went out to meet Asa, and said unto him, Hear ye me, Asa, and all Judah
-    and Benjamin; The LORD is with you, while ye be with him; and if ye seek him,
-    he will be found of you; but if ye forsake him, he will forsake you.
-  - And say thou unto them, Thus saith the LORD God of Israel; Cursed be the man that
-    obeyeth not the words of this covenant,
-- source_sentence: Because of their wickedness which they have committed to provoke
-    me to anger, in that they went to burn incense, and to serve other gods, whom
-    they knew not, neither they, ye, nor your fathers.
-  sentences:
-  - 'Woe be unto thee, O Moab! the people of Chemosh perisheth: for thy sons are taken
-    captives, and thy daughters captives.'
-  - 'It repenteth me that I have set up Saul to be king: for he is turned back from
-    following me, and hath not performed my commandments. And it grieved Samuel; and
-    he cried unto the LORD all night.'
-  - If thy brother, the son of thy mother, or thy son, or thy daughter, or the wife
-    of thy bosom, or thy friend, which is as thine own soul, entice thee secretly,
-    saying, Let us go and serve other gods, which thou hast not known, thou, nor thy
-    fathers;
-- source_sentence: As the appearance of the bow that is in the cloud in the day of
-    rain, so was the appearance of the brightness round about. This was the appearance
-    of the likeness of the glory of the LORD. And when I saw it, I fell upon my face,
-    and I heard a voice of one that spake.
-  sentences:
-  - 'A new heart also will I give you, and a new spirit will I put within you: and
-    I will take away the stony heart out of your flesh, and I will give you an heart
-    of flesh.'
-  - Thus saith the Lord GOD; If the prince give a gift unto any of his sons, the inheritance
-    thereof shall be his sons’; it shall be their possession by inheritance.
-  - 'And when I saw him, I fell at his feet as dead. And he laid his right hand upon
-    me, saying unto me, Fear not; I am the first and the last:'
-- source_sentence: Masters, give unto your servants that which is just and equal;
-    knowing that ye also have a Master in heaven.
-  sentences:
-  - And not holding the Head, from which all the body by joints and bands having nourishment
-    ministered, and knit together, increaseth with the increase of God.
-  - And he said also unto his disciples, There was a certain rich man, which had a
-    steward; and the same was accused unto him that he had wasted his goods.
-  - And the keeper of the prison awaking out of his sleep, and seeing the prison doors
-    open, he drew out his sword, and would have killed himself, supposing that the
-    prisoners had been fled.
-- source_sentence: But God shall wound the head of his enemies, and the hairy scalp
-    of such an one as goeth on still in his trespasses.
-  sentences:
-  - 'Then again called they the man that was blind, and said unto him, Give God the
-    praise: we know that this man is a sinner.'
-  - Who can utter the mighty acts of the LORD? who can shew forth all his praise?
-  - Again, when the wicked man turneth away from his wickedness that he hath committed,
-    and doeth that which is lawful and right, he shall save his soul alive.
+  - sentence-transformers
+  - bible
+  - cross-reference
+  - semantic-search
+  - intertextuality
 pipeline_tag: sentence-similarity
 library_name: sentence-transformers
-metrics:
-- pearson_cosine
-- spearman_cosine
-model-index:
-- name: SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2
-  results:
-  - task:
-      type: semantic-similarity
-      name: Semantic Similarity
-    dataset:
-      name: intertextual similarity chirho
-      type: intertextual-similarity-chirho
-    metrics:
-    - type: pearson_cosine
-      value: 0.6270927412432303
-      name: Pearson Cosine
-    - type: spearman_cosine
-      value: 0.6326350413656476
-      name: Spearman Cosine
+base_model: sentence-transformers/all-MiniLM-L12-v2
+datasets:
+  - LoveJesus/intertextual-dataset-chirho
 ---
 
-# SentenceTransformer based on sentence-transformers/all-MiniLM-L12-v2
+# Intertextual Embedder (MiniLM-L12) - chirho
 
-This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [sentence-transformers/all-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L12-v2). 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.
+*For God so loved the world that he gave his only begotten Son, that whoever believes in him should not perish but have eternal life. - John 3:16*
 
-## Model Details
+## Description
 
-### Model Description
-- **Model Type:** Sentence Transformer
-- **Base model:** [sentence-transformers/all-MiniLM-L12-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L12-v2) <!-- at revision 936af83a2ecce5fe87a09109ff5cbcefe073173a -->
-- **Maximum Sequence Length:** 128 tokens
-- **Output Dimensionality:** 384 dimensions
-- **Similarity Function:** Cosine Similarity
-<!-- - **Training Dataset:** Unknown -->
-<!-- - **Language:** Unknown -->
-<!-- - **License:** Unknown -->
+A sentence transformer fine-tuned for **biblical verse similarity** and **cross-reference discovery**. Given a verse text, it produces a 384-dimensional embedding that places semantically related verses close together in vector space.
 
-### Model Sources
+## Training
 
-- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
-- **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
-- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
+- **Base model**: sentence-transformers/all-MiniLM-L12-v2
+- **Loss**: Triplet loss (cosine distance, margin=0.5)
+- **Data**: 344,798 triplets from the Treasury of Scripture Knowledge (OpenBible.info)
+  - Anchor: verse A, Positive: cross-referenced verse B, Negative: hard negative (same-book unrelated verse)
+- **Epochs**: 3
+- **Batch size**: 64
+- **Device**: Apple MPS (M4 Pro)
 
-### Full Model Architecture
+## Evaluation
 
-```
-SentenceTransformer(
-  (0): Transformer({'max_seq_length': 128, 'do_lower_case': False, 'architecture': 'BertModel'})
-  (1): Pooling({'word_embedding_dimension': 384, '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})
-  (2): Normalize()
-)
-```
+- **Triplet ranking accuracy**: 86.75% (positive cross-ref ranked higher than negative)
+- **Separation gap**: 0.4213
+- **Pearson cosine**: 0.6271
+- **Spearman cosine**: 0.6326
 
 ## Usage
 
-### Direct Usage (Sentence Transformers)
-
-First install the Sentence Transformers library:
-
-```bash
-pip install -U sentence-transformers
-```
-
-Then you can load this model and run inference.
 ```python
 from sentence_transformers import SentenceTransformer
 
-# Download from the 🤗 Hub
-model = SentenceTransformer("sentence_transformers_model_id")
-# Run inference
-sentences = [
-    'But God shall wound the head of his enemies, and the hairy scalp of such an one as goeth on still in his trespasses.',
-    'Again, when the wicked man turneth away from his wickedness that he hath committed, and doeth that which is lawful and right, he shall save his soul alive.',
-    'Who can utter the mighty acts of the LORD? who can shew forth all his praise?',
-]
-embeddings = model.encode(sentences)
-print(embeddings.shape)
-# [3, 384]
-
-# Get the similarity scores for the embeddings
-similarities = model.similarity(embeddings, embeddings)
-print(similarities)
-# tensor([[ 1.0000, -0.0455, -0.3163],
-#         [-0.0455,  1.0000, -0.0269],
-#         [-0.3163, -0.0269,  1.0000]])
-```
-
-<!--
-### 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.*
--->
-
-## Evaluation
-
-### Metrics
-
-#### Semantic Similarity
+model = SentenceTransformer("LoveJesus/intertextual-embedder-chirho")
 
-* Dataset: `intertextual-similarity-chirho`
-* Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
-
-| Metric              | Value      |
-|:--------------------|:-----------|
-| pearson_cosine      | 0.6271     |
-| **spearman_cosine** | **0.6326** |
-
-<!--
-## 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
-
-#### Unnamed Dataset
-
-* Size: 275,838 training samples
-* Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>sentence_2</code>
-* Approximate statistics based on the first 1000 samples:
-  |         | sentence_0                                                                         | sentence_1                                                                         | sentence_2                                                                         |
-  |:--------|:-----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
-  | type    | string                                                                             | string                                                                             | string                                                                             |
-  | details | <ul><li>min: 12 tokens</li><li>mean: 37.48 tokens</li><li>max: 92 tokens</li></ul> | <ul><li>min: 10 tokens</li><li>mean: 36.65 tokens</li><li>max: 83 tokens</li></ul> | <ul><li>min: 9 tokens</li><li>mean: 33.52 tokens</li><li>max: 109 tokens</li></ul> |
-* Samples:
-  | sentence_0                                                                                                                          | sentence_1                                                                                                                                                                                                                       | sentence_2                                                                                                                                                                                                      |
-  |:------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
-  | <code>And the LORD took me as I followed the flock, and the LORD said unto me, Go, prophesy unto my people Israel.</code>           | <code>And as Jesus passed forth from thence, he saw a man, named Matthew, sitting at the receipt of custom: and he saith unto him, Follow me. And he arose, and followed him.</code>                                             | <code>But, behold, I will raise up against you a nation, O house of Israel, saith the LORD the God of hosts; and they shall afflict you from the entering in of Hemath unto the river of the wilderness.</code> |
-  | <code>Let the elders that rule well be counted worthy of double honour, especially they who labour in the word and doctrine.</code> | <code>We then, as workers together with him, beseech you also that ye receive not the grace of God in vain.</code>                                                                                                               | <code>A bishop then must be blameless, the husband of one wife, vigilant, sober, of good behaviour, given to hospitality, apt to teach;</code>                                                                  |
-  | <code>And the chambers and the entries thereof were by the posts of the gates, where they washed the burnt offering.</code>         | <code>He made also ten lavers, and put five on the right hand, and five on the left, to wash in them: such things as they offered for the burnt offering they washed in them; but the sea was for the priests to wash in.</code> | <code>Hath oppressed the poor and needy, hath spoiled by violence, hath not restored the pledge, and hath lifted up his eyes to the idols, hath committed abomination,</code>                                   |
-* Loss: [<code>TripletLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#tripletloss) with these parameters:
-  ```json
-  {
-      "distance_metric": "TripletDistanceMetric.COSINE",
-      "triplet_margin": 0.5
-  }
-  ```
-
-### Training Hyperparameters
-#### Non-Default Hyperparameters
-
-- `eval_strategy`: steps
-- `per_device_train_batch_size`: 64
-- `per_device_eval_batch_size`: 64
-- `multi_dataset_batch_sampler`: round_robin
-
-#### All Hyperparameters
-<details><summary>Click to expand</summary>
-
-- `do_predict`: False
-- `eval_strategy`: steps
-- `prediction_loss_only`: True
-- `per_device_train_batch_size`: 64
-- `per_device_eval_batch_size`: 64
-- `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`: 3
-- `max_steps`: -1
-- `lr_scheduler_type`: linear
-- `lr_scheduler_kwargs`: None
-- `warmup_ratio`: None
-- `warmup_steps`: 0
-- `log_level`: passive
-- `log_level_replica`: warning
-- `log_on_each_node`: True
-- `logging_nan_inf_filter`: True
-- `enable_jit_checkpoint`: False
-- `save_on_each_node`: False
-- `save_only_model`: False
-- `restore_callback_states_from_checkpoint`: False
-- `use_cpu`: False
-- `seed`: 42
-- `data_seed`: None
-- `bf16`: False
-- `fp16`: False
-- `bf16_full_eval`: False
-- `fp16_full_eval`: False
-- `tf32`: None
-- `local_rank`: -1
-- `ddp_backend`: None
-- `debug`: []
-- `dataloader_drop_last`: False
-- `dataloader_num_workers`: 0
-- `dataloader_prefetch_factor`: None
-- `disable_tqdm`: False
-- `remove_unused_columns`: True
-- `label_names`: None
-- `load_best_model_at_end`: False
-- `ignore_data_skip`: False
-- `fsdp`: []
-- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
-- `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_fused
-- `optim_args`: None
-- `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
-- `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_for_metrics`: []
-- `eval_do_concat_batches`: True
-- `auto_find_batch_size`: False
-- `full_determinism`: False
-- `ddp_timeout`: 1800
-- `torch_compile`: False
-- `torch_compile_backend`: None
-- `torch_compile_mode`: None
-- `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
-- `use_cache`: False
-- `prompts`: None
-- `batch_sampler`: batch_sampler
-- `multi_dataset_batch_sampler`: round_robin
-- `router_mapping`: {}
-- `learning_rate_mapping`: {}
-
-</details>
-
-### Training Logs
-| Epoch  | Step  | Training Loss | intertextual-similarity-chirho_spearman_cosine |
-|:------:|:-----:|:-------------:|:----------------------------------------------:|
-| 0.1160 | 500   | 0.2996        | -                                              |
-| 0.2320 | 1000  | 0.2629        | 0.5336                                         |
-| 0.3480 | 1500  | 0.2529        | -                                              |
-| 0.4640 | 2000  | 0.2434        | 0.5641                                         |
-| 0.5800 | 2500  | 0.2356        | -                                              |
-| 0.6961 | 3000  | 0.2320        | 0.5828                                         |
-| 0.8121 | 3500  | 0.2271        | -                                              |
-| 0.9281 | 4000  | 0.2222        | 0.5963                                         |
-| 1.0    | 4310  | -             | 0.5989                                         |
-| 1.0441 | 4500  | 0.2153        | -                                              |
-| 1.1601 | 5000  | 0.2028        | 0.6041                                         |
-| 1.2761 | 5500  | 0.2025        | -                                              |
-| 1.3921 | 6000  | 0.2006        | 0.6104                                         |
-| 1.5081 | 6500  | 0.1972        | -                                              |
-| 1.6241 | 7000  | 0.1964        | 0.6161                                         |
-| 1.7401 | 7500  | 0.1965        | -                                              |
-| 1.8561 | 8000  | 0.1952        | 0.6213                                         |
-| 1.9722 | 8500  | 0.1935        | -                                              |
-| 2.0    | 8620  | -             | 0.6267                                         |
-| 2.0882 | 9000  | 0.1846        | 0.6282                                         |
-| 2.2042 | 9500  | 0.1783        | -                                              |
-| 2.3202 | 10000 | 0.1797        | 0.6268                                         |
-| 2.4362 | 10500 | 0.1817        | -                                              |
-| 2.5522 | 11000 | 0.1774        | 0.6317                                         |
-| 2.6682 | 11500 | 0.1776        | -                                              |
-| 2.7842 | 12000 | 0.1793        | 0.6325                                         |
-| 2.9002 | 12500 | 0.1758        | -                                              |
-| 3.0    | 12930 | -             | 0.6326                                         |
-
-
-### Framework Versions
-- Python: 3.14.2
-- Sentence Transformers: 5.2.2
-- Transformers: 5.1.0
-- PyTorch: 2.10.0
-- Accelerate: 1.12.0
-- Datasets: 4.5.0
-- Tokenizers: 0.22.2
-
-## Citation
-
-### BibTeX
-
-#### Sentence Transformers
-```bibtex
-@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",
-}
-```
+verses = [
+    "In the beginning God created the heaven and the earth.",
+    "In the beginning was the Word, and the Word was with God, and the Word was God.",
+    "And the children of Israel went into the midst of the sea upon the dry ground.",
+]
 
-#### TripletLoss
-```bibtex
-@misc{hermans2017defense,
-    title={In Defense of the Triplet Loss for Person Re-Identification},
-    author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
-    year={2017},
-    eprint={1703.07737},
-    archivePrefix={arXiv},
-    primaryClass={cs.CV}
-}
+embeddings = model.encode(verses)
+# embeddings[0] will be closest to embeddings[1] (Gen 1:1 <-> John 1:1)
 ```
 
-<!--
-## 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.*
--->
+## Part of models-chirho
 
-<!--
-## Model Card Contact
+This model is part of the [Intertextual Reference Network](https://huggingface.co/LoveJesus) pipeline, paired with:
+- **Classifier**: [LoveJesus/intertextual-classifier-chirho](https://huggingface.co/LoveJesus/intertextual-classifier-chirho)
+- **Dataset**: [LoveJesus/intertextual-dataset-chirho](https://huggingface.co/datasets/LoveJesus/intertextual-dataset-chirho)
 
-*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
--->
+Built with love for Jesus by [loveJesus](https://huggingface.co/LoveJesus).