CodeSearch-ModernBERT-Crow-v1.1-Plus

SentenceTransformer based on Shuu12121/CodeModernBERT-Crow-v1.1

This is a sentence-transformers model finetuned from Shuu12121/CodeModernBERT-Crow-v1.1. 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: Shuu12121/CodeModernBERT-Crow-v1.1 <!-- at revision d7baa192c09e1e1da5c39fe9652ea9a4663084f6 -->
• Maximum Sequence Length: 1024 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity

<!-- - **Training Dataset:** Unknown --> <!-- - **Language:** Unknown --> <!-- - **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': 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})
)

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
sentences = [
    'The toggle content, if left empty it will render the default toggle (seen above).',
    "function NavbarToggle() {\n\t    (0, _classCallCheck3['default'])(this, NavbarToggle);\n\t    return (0, _possibleConstructorReturn3['default'])(this, _React$Component.apply(this, arguments));\n\t  }",
    "update = () => {\n\n\t            const timerId = window.requestAnimationFrame( update );\n\t            const elapsed = performance.now() - timestamp;\n\t            const progress = elapsed / duration;\n\t            const opacity = 1.0 - progress > 0 ? 1.0 - progress : 0;\n\t            const radius = progress * canvasWidth * 0.5 / dpr;\n\n\t            context.clearRect( 0, 0, canvasWidth, canvasHeight );\n\t            context.beginPath();\n\t            context.arc( x, y, radius, 0, Math.PI * 2 );\n\t            context.fillStyle = `rgba(${color.r * 255}, ${color.g * 255}, ${color.b * 255}, ${opacity})`;\n\t            context.fill();\n\t            context.closePath();\n\n\t            if ( progress >= 1.0 ) {\n\n\t                window.cancelAnimationFrame( timerId );\n\t                this.updateCanvasArcByProgress( 0 );\n\n\t                /**\n\t                 * Reticle ripple end event\n\t                 * @type {object}\n\t                 * @event Reticle#reticle-ripple-end\n\t                 */\n\t                this.dispatchEvent( { type: 'reticle-ripple-end' } );\n\n\t            }\n\n\t            material.map.needsUpdate = true;\n\n\t        }",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000,  0.6778, -0.0447],
#         [ 0.6778,  1.0000,  0.0303],
#         [-0.0447,  0.0303,  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.* --> <!-- ## 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: 2,392,064 training samples
• Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>label</code>
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1	label
  type	string	string	float
  details	<ul><li>min: 8 tokens</li><li>mean: 74.35 tokens</li><li>max: 1024 tokens</li></ul>	<ul><li>min: 11 tokens</li><li>mean: 182.37 tokens</li><li>max: 1024 tokens</li></ul>	<ul><li>min: 1.0</li><li>mean: 1.0</li><li>max: 1.0</li></ul>

• Samples:

  sentence_0	sentence_1	label
  <code>Set the column title<br><br>@param column - column number (first column is: 0)<br>@param title - new column title</code>	<code>setHeader = function(column, newValue) {<br> const obj = this;<br><br> if (obj.headers[column]) {<br> const oldValue = obj.headers[column].textContent;<br> const onchangeheaderOldValue = (obj.options.columns && obj.options.columns[column] && obj.options.columns[column].title) || '';<br><br> if (! newValue) {<br> newValue = getColumnName(column);<br> }<br><br> obj.headers[column].textContent = newValue;<br> // Keep the title property<br> obj.headers[column].setAttribute('title', newValue);<br> // Update title<br> if (!obj.options.columns) {<br> obj.options.columns = [];<br> }<br> if (!obj.options.columns[column]) {<br> obj.options.columns[column] = {};<br> }<br> obj.options.columns[column].title = newValue;<br><br> setHistory.call(obj, {<br> action: 'setHeader',<br> column: column,<br> oldValue: oldValue,<br> newValue: newValue<br> });<br><br> // On onchange header<br> dispatch.c...</code>	<code>1.0</code>
  <code>Elsewhere this is known as a "Weak Value Map". Whereas a std JS WeakMap<br>is weak on its keys, this map is weak on its values. It does not retain these<br>values strongly. If a given value disappears, then the entries for it<br>disappear from every weak-value-map that holds it as a value.<br><br>Just as a WeakMap only allows gc-able values as keys, a weak-value-map<br>only allows gc-able values as values.<br><br>Unlike a WeakMap, a weak-value-map unavoidably exposes the non-determinism of<br>gc to its clients. Thus, both the ability to create one, as well as each<br>created one, must be treated as dangerous capabilities that must be closely<br>held. A program with access to these can read side channels though gc that do<br>not* rely on the ability to measure duration. This is a separate, and bad,<br>timing-independent side channel.<br><br>This non-determinism also enables code to escape deterministic replay. In a<br>blockchain context, this could cause validators to differ from each other,<br>preventing consensus, and thus preventing ...</code>	<code>makeFinalizingMap = (finalizer, opts) => {<br> const { weakValues = false } = opts || {};<br> if (!weakValues || !WeakRef || !FinalizationRegistry) {<br> /** @type Map<K, V> /<br> const keyToVal = new Map();<br> return Far('fakeFinalizingMap', {<br> clearWithoutFinalizing: keyToVal.clear.bind(keyToVal),<br> get: keyToVal.get.bind(keyToVal),<br> has: keyToVal.has.bind(keyToVal),<br> set: (key, val) => {<br> keyToVal.set(key, val);<br> },<br> delete: keyToVal.delete.bind(keyToVal),<br> getSize: () => keyToVal.size,<br> });<br> }<br> /* @type Map<K, WeakRef<any>> */<br> const keyToRef = new Map();<br> const registry = new FinalizationRegistry(key => {<br> // Because this will delete the current binding of key, we need to<br> // be sure that it is not called because a previous binding was collected.<br> // We do this with the unregister in set below, assuming that<br> // unregister immediately suppresses the finalization of the thing<br> // it unregisters. TODO If this is...</code>	<code>1.0</code>
  <code>Creates a function that memoizes the result of func. If resolver is<br>provided, it determines the cache key for storing the result based on the<br>arguments provided to the memoized function. By default, the first argument<br>provided to the memoized function is used as the map cache key. The func<br>is invoked with the this binding of the memoized function.<br><br>Note: The cache is exposed as the cache property on the memoized<br>function. Its creation may be customized by replacing the _.memoize.Cache<br>constructor with one whose instances implement the<br>Map<br>method interface of delete, get, has, and set.<br><br>@static<br>@memberOf _<br>@since 0.1.0<br>@category Function<br>@param {Function} func The function to have its output memoized.<br>@param {Function} [resolver] The function to resolve the cache key.<br>@returns {Function} Returns the new memoized function.<br>@example<br><br>var object = { 'a': 1, 'b': 2 };<br>var othe...</code>	<code>function memoize(func, resolver) {<br> if (typeof func != 'function' || (resolver && typeof resolver != 'function')) {<br> throw new TypeError(FUNC_ERROR_TEXT);<br> }<br> var memoized = function() {<br> var args = arguments,<br> key = resolver ? resolver.apply(this, args) : args[0],<br> cache = memoized.cache;<br><br> if (cache.has(key)) {<br> return cache.get(key);<br> }<br> var result = func.apply(this, args);<br> memoized.cache = cache.set(key, result);<br> return result;<br> };<br> memoized.cache = new (memoize.Cache || MapCache);<br> return memoized;<br> }</code>	<code>1.0</code>

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

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "mini_batch_size": 128,
      "gather_across_devices": false
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 2048
• per_device_eval_batch_size: 2048
• fp16: True
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

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

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: no
• prediction_loss_only: True
• per_device_train_batch_size: 2048
• per_device_eval_batch_size: 2048
• 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: 3
• 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: True
• 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: 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: False
• 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: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

</details>

Training Logs

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 5.1.0
• Transformers: 4.55.3
• PyTorch: 2.7.0+cu128
• Accelerate: 1.7.0
• Datasets: 3.6.0
• Tokenizers: 0.21.4

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",
}

<!-- ## 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.* -->