Update README.md

README.md

@@ -3,498 +3,150 @@ tags:
 - sentence-transformers
 - sentence-similarity
 - feature-extraction
-- generated_from_trainer
-- dataset_size:1761750
-- loss:MultipleNegativesRankingLoss
+- code-search
+- modernbert
+- code
+- python
+- java
+- javascript
+- php
+- ruby
+- rust
+- go
+- mteb
 base_model: Shuu12121/CodeModernBERT-Snake
-widget:
-- source_sentence: // MiddlewareHeaders adds headers to a handler
-  sentences:
-  - "function create(options) {\n  console.log('Creating app: ', options.name);\n\n\
-    \  if (!options.status)\n    options.status = (code, msg) => {\n      console.log(msg);\n\
-    \    };\n  if (!options.type) options.type = 'rekit-react';\n\n  const prjDir\
-    \ = path.join(options.location || process.cwd(), options.name);\n  return new\
-    \ Promise(async (resolve, reject) => {\n    try {\n      if (fs.existsSync(prjDir))\
-    \ {\n        reject('FOLDER_EXISTS');\n        return;\n      }\n      fs.mkdirSync(prjDir);\n\
-    \      let gitRepo;\n      if (options.source) {\n        if (/^https?:/.test(options.source))\
-    \ {\n          // It's a git repo\n          gitRepo = options.source;\n     \
-    \   } else {\n          // It's a local folder\n          const srcDir = path.isAbsolute(options.source)\n\
-    \            ? options.source\n            : path.join(process.cwd(), options.source);\n\
-    \          options.status('CREATE_APP_COPY_FILES', `Copy files from ${srcDir}...`);\n\
-    \          await fs.copy(srcDir, prjDir, {\n            filter: src => !/\\/(\\\
-    .git|node_modules\\/|node_modules$)/.test(src) || path.basename(src) === '.gitignore',\n\
-    \          });\n        }\n      } else if (options.type) {\n        // Get gitRepo\n\
-    \        options.status(\n          'QUERY_APP_TYPES_GIT_REPO',\n          `Looking\
-    \ for the git repo for app type ${options.type}...`,\n        );\n        const\
-    \ appTypes = await getAppTypes();\n        const appType = _.find(appTypes, {\
-    \ id: options.type });\n        if (!appType) reject('APP_TYPE_NOT_SUPPORTED');\n\
-    \        gitRepo = appType.repo;\n      } else {\n        await fs.remove(prjDir);\n\
-    \        reject('NO_SOURCE_OR_APP_TYPE');\n      }\n\n      if (gitRepo) {\n \
-    \       options.status('CLONE_PROJECT', `Downloading project from ${gitRepo}...`);\n\
-    \        await cloneRepo(gitRepo, prjDir);\n      }\n\n      postCreate(prjDir,\
-    \ options);\n      options.status('CREATION_SUCCESS', '\U0001F603App creation\
-    \ success.');\n      resolve();\n    } catch (err) {\n      console.log('Failed\
-    \ to create project.');\n      fs.removeSync(prjDir);\n      reject(err);\n  \
-    \  }\n  });\n}"
-  - "@Override\n\tpublic void setFrameworkID(Option<Protos.FrameworkID> frameworkID)\
-    \ throws Exception {\n\t\tsynchronized (startStopLock) {\n\t\t\tverifyIsRunning();\n\
-    \n\t\t\tbyte[] value = frameworkID.isDefined() ? frameworkID.get().getValue().getBytes(ConfigConstants.DEFAULT_CHARSET)\
-    \ :\n\t\t\t\tnew byte[0];\n\t\t\tframeworkIdInZooKeeper.setValue(value);\n\t\t\
-    }\n\t}"
-  - "func MiddlewareHeaders(vs map[string]string) Middleware {\n\treturn func(h http.Handler)\
-    \ http.Handler {\n\t\treturn http.HandlerFunc(func(rw http.ResponseWriter, r *http.Request)\
-    \ {\n\t\t\t// Add headers\n\t\t\thandleHeaders(vs, rw)\n\n\t\t\t// Next handler\n\
-    \t\t\th.ServeHTTP(rw, r)\n\t\t})\n\t}\n}"
-- source_sentence: 'Parses a Plist XML string. Returns an Object.
-
-
-    @param {String} xml - the XML String to decode
-
-    @param {Function} callback - callback function
-
-    @returns {Mixed} the decoded value from the Plist XML
-
-    @api public
-
-    @deprecated use parse() instead'
-  sentences:
-  - "function parseStringSync (xml) {\n  var doc = new DOMParser().parseFromString(xml);\n\
-    \  var plist;\n  if (doc.documentElement.nodeName !== 'plist') {\n    throw new\
-    \ Error('malformed document. First element should be <plist>');\n  }\n  plist\
-    \ = parsePlistXML(doc.documentElement);\n\n  // if the plist is an array with\
-    \ 1 element, pull it out of the array\n  if (plist.length == 1) {\n    plist =\
-    \ plist[0];\n  }\n  return plist;\n}"
-  - "func GetCallStringArgsValues(n ast.Node, ctx *Context) []string {\n\tvalues :=\
-    \ []string{}\n\tswitch node := n.(type) {\n\tcase *ast.CallExpr:\n\t\tfor _, arg\
-    \ := range node.Args {\n\t\t\tswitch param := arg.(type) {\n\t\t\tcase *ast.BasicLit:\n\
-    \t\t\t\tvalue, err := GetString(param)\n\t\t\t\tif err == nil {\n\t\t\t\t\tvalues\
-    \ = append(values, value)\n\t\t\t\t}\n\t\t\tcase *ast.Ident:\n\t\t\t\tvalues =\
-    \ append(values, GetIdentStringValues(param)...)\n\t\t\t}\n\t\t}\n\t}\n\treturn\
-    \ values\n}"
-  - "public static Date beginOfYear(@NotNull final Date date) {\n\t\treturn DateUtils.truncate(date,\
-    \ Calendar.YEAR);\n\t}"
-- source_sentence: '// forbiddenImportsFor determines all of the forbidden
-
-    // imports for a package given the import restrictions
-
-    // and returns a deduplicated list of them'
-  sentences:
-  - "func (i *ImportRestriction) forbiddenImportsFor(pkg Package) []string {\n\tforbiddenImportSet\
-    \ := map[string]struct{}{}\n\timports := pkg.Imports\n\tif !i.ExcludeTests {\n\
-    \t\timports = append(imports, append(pkg.TestImports, pkg.XTestImports...)...)\n\
-    \t}\n\tfor _, imp := range imports {\n\t\tpath := extractVendorPath(imp)\n\t\t\
-    if i.isForbidden(path) {\n\t\t\tforbiddenImportSet[path] = struct{}{}\n\t\t}\n\
-    \t}\n\n\tvar forbiddenImports []string\n\tfor imp := range forbiddenImportSet\
-    \ {\n\t\tforbiddenImports = append(forbiddenImports, imp)\n\t}\n\treturn forbiddenImports\n\
-    }"
-  - "function pick(o, props = []) {\n    return props.reduce((acc, k) => {\n     \
-    \   if (o.hasOwnProperty(k)) {\n            acc[k] = o[k];\n        }\n\n    \
-    \    return acc;\n    }, {});\n}"
-  - "func (s *PutTraceSegmentsOutput) SetUnprocessedTraceSegments(v []*UnprocessedTraceSegment)\
-    \ *PutTraceSegmentsOutput {\n\ts.UnprocessedTraceSegments = v\n\treturn s\n}"
-- source_sentence: 'Validates whether the specified template is syntactically correct
-    and will be accepted by Azure Resource Manager..
-
-
-    @param resourceGroupName The name of the resource group the template will be deployed
-    to. The name is case insensitive.
-
-    @param deploymentName The name of the deployment.
-
-    @param properties The deployment properties.
-
-    @param serviceCallback the async ServiceCallback to handle successful and failed
-    responses.
-
-    @throws IllegalArgumentException thrown if parameters fail the validation
-
-    @return the {@link ServiceFuture} object'
-  sentences:
-  - "func Execute(v string) {\n\tversion = v\n\tif err := rootCmd.Execute(); err !=\
-    \ nil {\n\t\tlog.Fatal(err)\n\t}\n}"
-  - "function( otherPath )\r\n\t{\r\n\t\tvar thisElements = this.elements;\r\n\t\t\
-    var otherElements = otherPath && otherPath.elements;\r\n\r\n\t\tif ( !otherElements\
-    \ || thisElements.length != otherElements.length )\r\n\t\t\treturn false;\r\n\r\
-    \n\t\tfor ( var i = 0 ; i < thisElements.length ; i++ )\r\n\t\t{\r\n\t\t\tif (\
-    \ !thisElements[ i ].equals( otherElements[ i ] ) )\r\n\t\t\t\treturn false;\r\
-    \n\t\t}\r\n\r\n\t\treturn true;\r\n\t}"
-  - "public ServiceFuture<DeploymentValidateResultInner> validateAsync(String resourceGroupName,\
-    \ String deploymentName, DeploymentProperties properties, final ServiceCallback<DeploymentValidateResultInner>\
-    \ serviceCallback) {\n        return ServiceFuture.fromResponse(validateWithServiceResponseAsync(resourceGroupName,\
-    \ deploymentName, properties), serviceCallback);\n    }"
-- source_sentence: This method calculates the turn weight separately.
-  sentences:
-  - "private SingleType parseSingleType() throws TTXPathException {\n\n        final\
-    \ String atomicType = parseAtomicType();\n        final boolean intero = is(TokenType.INTERROGATION,\
-    \ true);\n        return new SingleType(atomicType, intero);\n    }"
-  - "public void putAllWriteable(BeanMap<T> map) {\n        map.types.keySet().stream().filter(key\
-    \ -> getWriteInvoker(key) != null).forEach(key -> this.put(key, map.get(key)));\n\
-    \    }"
-  - "public double calcTurnWeight(int edgeFrom, int nodeVia, int edgeTo) {\n     \
-    \   long turnFlags = turnCostExt.getTurnCostFlags(edgeFrom, nodeVia, edgeTo);\n\
-    \        if (turnCostEncoder.isTurnRestricted(turnFlags))\n            return\
-    \ Double.POSITIVE_INFINITY;\n\n        return turnCostEncoder.getTurnCost(turnFlags);\n\
-    \    }"
 pipeline_tag: sentence-similarity
 library_name: sentence-transformers
+license: apache-2.0
+datasets:
+- code-search-net/code_search_net
+- Shuu12121/python-codesearch-filtered
+- Shuu12121/java-codesearch-filtered
+- Shuu12121/javascript-codesearch-filtered
+- Shuu12121/rust-codesearch-filtered
+- Shuu12121/ruby-codesearch-filtered
+language:
+- en
+
 ---
 
-# SentenceTransformer based on Shuu12121/CodeModernBERT-Snake
+# Shuu12121/CodeSearch-ModernBERT-Snake-Plus 🐍
 
-This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Shuu12121/CodeModernBERT-Snake](https://huggingface.co/Shuu12121/CodeModernBERT-Snake). It maps sentences & paragraphs to a 512-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
+このモデルは、`Shuu12121/CodeModernBERT-Snake` をベースにした Sentence Transformer モデルであり、特に**多言語コード検索タスク**において高い性能を発揮するようファインチューニングされています。
 
-## Model Details
+> This is a Sentence Transformer model based on `Shuu12121/CodeModernBERT-Snake`, fine-tuned for high performance on multilingual code search tasks.
 
-### Model Description
-- **Model Type:** Sentence Transformer
-- **Base model:** [Shuu12121/CodeModernBERT-Snake](https://huggingface.co/Shuu12121/CodeModernBERT-Snake) <!-- at revision 73927b7c029b82e13135c02a1d4c5b50564d1e0d -->
-- **Maximum Sequence Length:** 1024 tokens
-- **Output Dimensionality:** 512 dimensions
-- **Similarity Function:** Cosine Similarity
-<!-- - **Training Dataset:** Unknown -->
-<!-- - **Language:** Unknown -->
-<!-- - **License:** Unknown -->
+- **開発者 (Developer)**: [Shuu12121](https://huggingface.co/Shuu12121)  
+- **ベースモデル (Base Model)**: [Shuu12121/CodeModernBERT-Snake](https://huggingface.co/Shuu12121/CodeModernBERT-Snake)  
+- **ライセンス (License)**: Apache-2.0  
 
-### Model Sources
+---
 
-- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
-- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
-- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
+## 📊 MTEB評価 / MTEB Evaluation
 
-### Full Model Architecture
+このモデルは、Massive Text Embedding Benchmark (MTEB) の**CodeSearchNet Retrieval**タスクにおいて、以下のスコアを記録しています。
 
-```
-SentenceTransformer(
-  (0): Transformer({'max_seq_length': 1024, 'do_lower_case': False}) with Transformer model: ModernBertModel 
-  (1): Pooling({'word_embedding_dimension': 512, '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})
-)
-```
+| 指標 | スコア (標準版) | スコア (COIR版) |
+|:-----|:----------------|:---------------|
+| **main_score (nDCG@10)** | 0.87926 | 0.77199 |
+| ndcg_at_1 | 0.78900 | 0.68372 |
+| ndcg_at_3 | 0.86324 | 0.74734 |
+| ndcg_at_5 | 0.87229 | 0.76061 |
+| ndcg_at_10 | 0.87926 | 0.77199 |
+| recall_at_10 | 0.95667 | 0.85808 |
+| mrr_at_10 | 0.85375 | 0.74433 |
 
-## Usage
+> ※ 標準版：通常のMTEB評価設定、COIR版：より厳しい類似度評価設定  
+> ※ 公式に提出していないためランキング情報は記載していません。(参考順位としては,標準版は10位前後のモデルとほぼ同等ぐらいでCOIR版は８位前後と同等ぐらいです,2025年4月現在)
 
-### Direct Usage (Sentence Transformers)
+---
 
-First install the Sentence Transformers library:
+# 🆚 Crow-Plusとの比較 / Comparison with Crow-Plus
 
-```bash
-pip install -U sentence-transformers
-```
+Crow-Plus（`Shuu12121/CodeSearch-ModernBERT-Crow-Plus`）と比較すると、  
+**Snake-Plusは、パラメータ数を約半分に削減**しながら、非常に高い性能を維持しています。
+
+| 指標 | Crow-Plus | Snake-Plus | コメント |
+|:-----|:----------|:-----------|:---------|
+| **main_score (nDCG@10, 標準版)** | 0.89296 | 0.87926 | SnakeはCrowに迫る水準を維持 |
+| **main_score (nDCG@10, COIR版)** | 0.79884 | 0.77199 | 厳しい設定でもSnakeは健闘 |
+| **パラメータ数 (推定)** | 約150M | 約77M | **SnakeはCrowの約半分** |
+| Recall@10 (標準版) | 96.1% | 95.6% | ほぼ同等のリコール率 |
+
+### 🔥 Snake-Plusの主な特徴
+
+- **パラメータ数はCrow-Plusの約半分（50%）**
+- **推論速度の高速化、メモリ消費量の低減**が可能
+- **性能もほとんど維持**（nDCG@10ではCrow比約98%）
+- 軽量なモデルが必要な環境（例：オンデバイス検索、低リソース推論）に特に適する
+
+---
+
+# 🧠 客観的分析まとめ
+
+Snake-Plusは、**高性能を維持したまま小型化を実現**したモデルです。  
+特に「推論コストを削減しつつ、実用レベルの検索性能を確保したい」ユースケースに非常に適しています。
+
+一方で、**絶対的な最高性能を目指す場合**（特に難易度の高いドメインデータ）では、Crow-Plusの方が優位となる場面もあり得ます。  
+**用途とリソースに応じた使い分け**が推奨されます。
+
+---
+
+## 🔧 モデル詳細 / Model Details
+
+- **ベースモデル**: Shuu12121/CodeModernBERT-Snake
+- **アーキテクチャ**: ModernBERT (hidden\_size: 512, layers: 12, heads: 8)
+- **最大入力長**: 1024トークン
+- **ファインチューニング**: CodeSearchNetなどの自然言語–コードペアを用いた類似性学習
+- **Pooling**: CLS Poolingを使用（SentenceTransformer互換）
+
+---
+
+## 🚀 使用方法 / How to Use
 
-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 = [
-    'This method calculates the turn weight separately.',
-    'public double calcTurnWeight(int edgeFrom, int nodeVia, int edgeTo) {\n        long turnFlags = turnCostExt.getTurnCostFlags(edgeFrom, nodeVia, edgeTo);\n        if (turnCostEncoder.isTurnRestricted(turnFlags))\n            return Double.POSITIVE_INFINITY;\n\n        return turnCostEncoder.getTurnCost(turnFlags);\n    }',
-    'public void putAllWriteable(BeanMap<T> map) {\n        map.types.keySet().stream().filter(key -> getWriteInvoker(key) != null).forEach(key -> this.put(key, map.get(key)));\n    }',
+model = SentenceTransformer("Shuu12121/CodeSearch-ModernBERT-Snake-Plus")
+
+code_snippets = [
+    "def factorial(n): if n == 0: return 1 else: return n * factorial(n-1)",
+    "function binarySearch(arr, target) { let left = 0, right = arr.length - 1; while (left <= right) { const mid = Math.floor((left + right) / 2); if (arr[mid] === target) return mid; if (arr[mid] < target) left = mid + 1; else right = mid - 1; } return -1; }"
 ]
-embeddings = model.encode(sentences)
-print(embeddings.shape)
-# [3, 512]
-
-# Get the similarity scores for the embeddings
-similarities = model.similarity(embeddings, embeddings)
-print(similarities.shape)
-# [3, 3]
-```
 
-<!--
-### 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: 1,761,750 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: 3 tokens</li><li>mean: 47.87 tokens</li><li>max: 633 tokens</li></ul> | <ul><li>min: 28 tokens</li><li>mean: 164.44 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>// Read reads from serial port.<br>// It is blocked until data received or timeout after p.timeout.</code> | <code>func (p *port) Read(b []byte) (n int, err error) {<br>	var done uint32<br>	if err = syscall.ReadFile(p.handle, b, &done, nil); err != nil {<br>		return<br>	}<br>	if done == 0 {<br>		err = ErrTimeout<br>		return<br>	}<br>	n = int(done)<br>	return<br>}</code>                                                                                                                                                                                                                                                                                                                                                                                                                         | <code>1.0</code> |
-  | <code>// _NET_WM_STRUT_PARTIAL set</code>                                                                        | <code>func WmStrutPartialSet(xu *xgbutil.XUtil, win xproto.Window,<br>	struts *WmStrutPartial) error {<br><br>	rawStruts := make([]uint, 12)<br>	rawStruts[0] = struts.Left<br>	rawStruts[1] = struts.Right<br>	rawStruts[2] = struts.Top<br>	rawStruts[3] = struts.Bottom<br>	rawStruts[4] = struts.LeftStartY<br>	rawStruts[5] = struts.LeftEndY<br>	rawStruts[6] = struts.RightStartY<br>	rawStruts[7] = struts.RightEndY<br>	rawStruts[8] = struts.TopStartX<br>	rawStruts[9] = struts.TopEndX<br>	rawStruts[10] = struts.BottomStartX<br>	rawStruts[11] = struts.BottomEndX<br><br>	return xprop.ChangeProp32(xu, win, "_NET_WM_STRUT_PARTIAL", "CARDINAL",<br>		rawStruts...)<br>}</code> | <code>1.0</code> |
-  | <code>// Union returns a new geometry representing all points in this geometry and the<br>// other.</code>       | <code>func (g *Geometry) Union(other *Geometry) (*Geometry, error) {<br>	return g.binaryTopo("Union", cGEOSUnion, other)<br>}</code>                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            | <code>1.0</code> |
-* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
-  ```json
-  {
-      "scale": 20.0,
-      "similarity_fct": "cos_sim"
-  }
-  ```
-
-### Training Hyperparameters
-#### Non-Default Hyperparameters
-
-- `per_device_train_batch_size`: 400
-- `per_device_eval_batch_size`: 400
-- `num_train_epochs`: 5
-- `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`: 400
-- `per_device_eval_batch_size`: 400
-- `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`: 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}
-- `tp_size`: 0
-- `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
-- `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
-- `eval_use_gather_object`: False
-- `average_tokens_across_devices`: False
-- `prompts`: None
-- `batch_sampler`: batch_sampler
-- `multi_dataset_batch_sampler`: round_robin
-
-</details>
-
-### Training Logs
-| Epoch  | Step  | Training Loss |
-|:------:|:-----:|:-------------:|
-| 0.1135 | 500   | 1.0064        |
-| 0.2270 | 1000  | 0.1985        |
-| 0.3405 | 1500  | 0.1802        |
-| 0.4540 | 2000  | 0.1659        |
-| 0.5675 | 2500  | 0.1583        |
-| 0.6810 | 3000  | 0.153         |
-| 0.7946 | 3500  | 0.1478        |
-| 0.9081 | 4000  | 0.1425        |
-| 1.0216 | 4500  | 0.132         |
-| 1.1351 | 5000  | 0.097         |
-| 1.2486 | 5500  | 0.1           |
-| 1.3621 | 6000  | 0.0972        |
-| 1.4756 | 6500  | 0.0958        |
-| 1.5891 | 7000  | 0.0968        |
-| 1.7026 | 7500  | 0.0945        |
-| 1.8161 | 8000  | 0.0943        |
-| 1.9296 | 8500  | 0.0938        |
-| 2.0431 | 9000  | 0.0831        |
-| 2.1566 | 9500  | 0.0634        |
-| 2.2701 | 10000 | 0.0642        |
-| 2.3837 | 10500 | 0.0639        |
-| 2.4972 | 11000 | 0.0646        |
-| 2.6107 | 11500 | 0.065         |
-| 2.7242 | 12000 | 0.0637        |
-| 2.8377 | 12500 | 0.062         |
-| 2.9512 | 13000 | 0.0626        |
-| 3.0647 | 13500 | 0.0522        |
-| 3.1782 | 14000 | 0.0443        |
-| 3.2917 | 14500 | 0.0435        |
-| 3.4052 | 15000 | 0.0447        |
-| 3.5187 | 15500 | 0.0441        |
-| 3.6322 | 16000 | 0.045         |
-| 3.7457 | 16500 | 0.0443        |
-| 3.8593 | 17000 | 0.0441        |
-| 3.9728 | 17500 | 0.0433        |
-| 4.0863 | 18000 | 0.0368        |
-| 4.1998 | 18500 | 0.0333        |
-| 4.3133 | 19000 | 0.0332        |
-| 4.4268 | 19500 | 0.0335        |
-| 4.5403 | 20000 | 0.033         |
-| 4.6538 | 20500 | 0.0334        |
-| 4.7673 | 21000 | 0.0325        |
-| 4.8808 | 21500 | 0.0342        |
-| 4.9943 | 22000 | 0.0341        |
-
-
-### Framework Versions
-- Python: 3.11.12
-- Sentence Transformers: 3.4.1
-- Transformers: 4.51.3
-- PyTorch: 2.6.0+cu124
-- Accelerate: 1.5.2
-- Datasets: 3.5.0
-- Tokenizers: 0.21.1
-
-## 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",
-}
-```
+queries = [
+    "calculate the factorial of a number recursively",
+    "find an element in a sorted array using binary search"
+]
 
-#### MultipleNegativesRankingLoss
-```bibtex
-@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}
-}
+code_embeddings = model.encode(code_snippets)
+query_embeddings = model.encode(queries)
 ```
 
-<!--
-## Glossary
+---
+
+## 🧩 想定用途と制限 / Intended Use & Limitations
+
+**用途例 / Intended Use:**
+- 多言語コード検索 (Natural Language to Code, Code to Code)
+- コードの類似性判定
+- コード分類・クラスタリング
+- コード推薦システム構築
 
-*Clearly define terms in order to be accessible across audiences.*
--->
+**対象言語 / Target Languages:**
+- Python, Java, JavaScript, PHP, Ruby, Go, Rust
+
+**制限 / Limitations:**
+- 主に**関数レベルのコード**に最適化。非常に長いファイルや構文エラーを含むコードには弱い傾向あり。
+- ドメイン特化タスクには追加ファインチューニングが有効な場合がある。
+- **生成タスク**（例：コード補完・生成）には向いていない（エンコーダ専用）。
+
+---
 
-<!--
-## Model Card Authors
+## 📩 連絡先 / Contact
 
-*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
--->
+ご質問・ご提案はこちらへどうぞ。  
+For questions or suggestions, please contact:
 
-<!--
-## Model Card Contact
+**📧 [email protected]**
 
-*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
--->