Hugging Face's logo

Shuu12121
/
CodeSearch-ModernBERT-Owl

Sentence Similarity
sentence-transformers
Safetensors
English
modernbert
feature-extraction
Eval Results
text-embeddings-inference
Model card Files
xet
Community
Shuu12121 commited on
Commit
c727a64
·
verified ·
1 Parent(s): 6a998f1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +76 -783
README.md CHANGED
@@ -3,153 +3,11 @@ tags:
3
  - sentence-transformers
4
  - sentence-similarity
5
  - feature-extraction
6
- - generated_from_trainer
7
- - dataset_size:1431743
8
- - loss:MultipleNegativesRankingLoss
9
  base_model: Shuu12121/CodeModernBERT-Owl
10
- widget:
11
- - source_sentence: return predicted ADEV of noise-type at given tau
12
- sentences:
13
- - "def get_instances(self):\n \n services = []\n for resource\
14
- \ in self._get_instances():\n services.append(resource['entity']['name'])\n\
15
- \n return services"
16
- - "def handle_exception(self, *args):\n \n\n if not self.__enabled:\n\
17
- \ return\n\n cls, instance, trcback = foundations.exceptions.extract_exception(*args)\n\
18
- \n LOGGER.info(\"{0} | Handling '{1}' exception!\".format(\n \
19
- \ self.__class__.__name__, foundations.strings.to_string(cls)))\n\n self.__initialize_context_ui()\n\
20
- \n self.__update_html(self.format_html_exception(cls, instance, trcback))\n\
21
- \n self.show()\n self.__report and self.report_exception_to_crittercism(cls,\
22
- \ instance, trcback)\n foundations.exceptions.base_exception_handler(cls,\
23
- \ instance, trcback)\n self.exec_()"
24
- - "def adev(self, tau0, tau):\n \n prefactor = self.adev_from_qd(tau0=tau0,\
25
- \ tau=tau)\n c = self.c_avar()\n avar = pow(prefactor, 2)*pow(tau,\
26
- \ c)\n return np.sqrt(avar)"
27
- - source_sentence: "Edit a IP4\n\n :param ip4: An IP4 available to save in\
28
- \ format x.x.x.x.\n :param id_ip: IP identifier. Integer value and greater\
29
- \ than zero.\n :param descricao: IP description.\n\n :return: None"
30
- sentences:
31
- - "def _vec_alpha(self, donor_catchments):\n \n return np.dot(linalg.inv(self._matrix_omega(donor_catchments)),\
32
- \ self._vec_b(donor_catchments))"
33
- - "def sync_balancer_files(self):\n \n\n def sync():\n \
34
- \ for balancer in self.configurables[Balancer].values():\n balancer.sync_file(self.configurables[Cluster].values())\n\
35
- \n self.work_pool.submit(sync)"
36
- - "def edit_ipv4(self, ip4, descricao, id_ip):\n \n\n if not is_valid_int_param(id_ip):\n\
37
- \ raise InvalidParameterError(\n u'Ip identifier is\
38
- \ invalid or was not informed.')\n\n if ip4 is None or ip4 == \"\":\n \
39
- \ raise InvalidParameterError(\n u'The IP4 is invalid\
40
- \ or was not informed.')\n\n ip_map = dict()\n ip_map['descricao']\
41
- \ = descricao\n ip_map['ip4'] = ip4\n ip_map['id_ip'] = id_ip\n\n\
42
- \ url = \"ip4/edit/\"\n\n code, xml = self.submit({'ip_map': ip_map},\
43
- \ 'POST', url)\n\n return self.response(code, xml)"
44
- - source_sentence: "Check if health check is disabled.\n\n It logs a message\
45
- \ if health check is disabled and it also adds an item\n to the action\
46
- \ queue based on 'on_disabled' setting.\n\n Returns:\n True\
47
- \ if check is disabled otherwise False."
48
- sentences:
49
- - "def find_guest(name, quiet=False, path=None):\n '''\n Returns the host\
50
- \ for a container.\n\n path\n path to the container parent\n \
51
- \ default: /var/lib/lxc (system default)\n\n .. versionadded:: 2015.8.0\n\
52
- \n\n .. code-block:: bash\n\n salt-run lxc.find_guest name\n '''\n\
53
- \ if quiet:\n log.warning(\"'quiet' argument is being deprecated.\"\n\
54
- \ ' Please migrate to --quiet')\n for data in _list_iter(path=path):\n\
55
- \ host, l = next(six.iteritems(data))\n for x in 'running', 'frozen',\
56
- \ 'stopped':\n if name in l[x]:\n if not quiet:\n \
57
- \ __jid_event__.fire_event(\n {'data':\
58
- \ host,\n 'outputter': 'lxc_find_host'},\n \
59
- \ 'progress')\n return host\n return None"
60
- - "def iter_wave_values(self):\n \n typecode = self.get_typecode(self.samplewidth)\n\
61
- \n if log.level >= 5:\n if self.cfg.AVG_COUNT > 1:\n \
62
- \ # merge samples -> log output in iter_avg_wave_values\n \
63
- \ tlm = None\n else:\n tlm = TextLevelMeter(self.max_value,\
64
- \ 79)\n\n # Use only a read size which is a quare divider of the samplewidth\n\
65
- \ # Otherwise array.array will raise: ValueError: string length not a multiple\
66
- \ of item size\n divider = int(round(float(WAVE_READ_SIZE) / self.samplewidth))\n\
67
- \ read_size = self.samplewidth * divider\n if read_size != WAVE_READ_SIZE:\n\
68
- \ log.info(\"Real use wave read size: %i Bytes\" % read_size)\n\n \
69
- \ get_wave_block_func = functools.partial(self.wavefile.readframes, read_size)\n\
70
- \ skip_count = 0\n\n manually_audioop_bias = self.samplewidth ==\
71
- \ 1 and audioop is None\n\n for frames in iter(get_wave_block_func, \"\"\
72
- ):\n\n if self.samplewidth == 1:\n if audioop is None:\n\
73
- \ log.warning(\"use audioop.bias() work-a-round for missing\
74
- \ audioop.\")\n else:\n # 8 bit samples are\
75
- \ unsigned, see:\n # http://docs.python.org/2/library/audioop.html#audioop.lin2lin\n\
76
- \ frames = audioop.bias(frames, 1, 128)\n\n try:\n\
77
- \ values = array.array(typecode, frames)\n except ValueError,\
78
- \ err:\n # e.g.:\n # ValueError: string length\
79
- \ not a multiple of item size\n # Work-a-round: Skip the last frames\
80
- \ of this block\n frame_count = len(frames)\n divider\
81
- \ = int(math.floor(float(frame_count) / self.samplewidth))\n new_count\
82
- \ = self.samplewidth * divider\n frames = frames[:new_count] #\
83
- \ skip frames\n log.error(\n \"Can't make array\
84
- \ from %s frames: Value error: %s (Skip %i and use %i frames)\" % (\n \
85
- \ frame_count, err, frame_count - new_count, len(frames)\n \
86
- \ ))\n values = array.array(typecode, frames)\n\n \
87
- \ for value in values:\n self.wave_pos += 1 # Absolute\
88
- \ position in the frame stream\n\n if manually_audioop_bias:\n\
89
- \ # audioop.bias can't be used.\n # See:\
90
- \ http://hg.python.org/cpython/file/482590320549/Modules/audioop.c#l957\n \
91
- \ value = value % 0xff - 128\n\n# if abs(value)\
92
- \ < self.min_volume:\n# # log.log(5, \"Ignore to lower amplitude\"\
93
- )\n# skip_count += 1\n# continue\n\n \
94
- \ yield (self.wave_pos, value)\n\n log.info(\"Skip %i samples\
95
- \ that are lower than %i\" % (\n skip_count, self.min_volume\n \
96
- \ ))\n log.info(\"Last readed Frame is: %s\" % self.pformat_pos())"
97
- - "def _check_disabled(self):\n \n if self.config['check_disabled']:\n\
98
- \ if self.config['on_disabled'] == 'withdraw':\n self.log.info(\"\
99
- Check is disabled and ip_prefix will be \"\n \"withdrawn\"\
100
- )\n self.log.info(\"adding %s in the queue\", self.ip_with_prefixlen)\n\
101
- \ self.action.put(self.del_operation)\n self.log.info(\"\
102
- Check is now permanently disabled\")\n elif self.config['on_disabled']\
103
- \ == 'advertise':\n self.log.info(\"check is disabled, ip_prefix\
104
- \ wont be withdrawn\")\n self.log.info(\"adding %s in the queue\"\
105
- , self.ip_with_prefixlen)\n self.action.put(self.add_operation)\n\
106
- \ self.log.info('check is now permanently disabled')\n\n \
107
- \ return True\n\n return False"
108
- - source_sentence: "When serializing an agent distribution, remove the thresholds,\
109
- \ in order\n to avoid cluttering the YAML definition file."
110
- sentences:
111
- - "def serialize_distribution(network_agents, known_modules=[]):\n '''\n When\
112
- \ serializing an agent distribution, remove the thresholds, in order\n to avoid\
113
- \ cluttering the YAML definition file.\n '''\n d = deepcopy(list(network_agents))\n\
114
- \ for v in d:\n if 'threshold' in v:\n del v['threshold']\n\
115
- \ v['agent_type'] = serialize_type(v['agent_type'],\n \
116
- \ known_modules=known_modules)\n return d"
117
- - "def disconnect(self):\n \n if self.root.ref is not None:\n \
118
- \ self.api.disconnect()\n self.root = None"
119
- - "def make_tarball(src_dir):\n \n if type(src_dir) != str:\n raise\
120
- \ TypeError('src_dir must be str')\n output_file = src_dir + \".tar.gz\"\n\
121
- \ log.msg(\"Wrapping tarball '{out}' ...\".format(out=output_file))\n if\
122
- \ not _dry_run:\n with tarfile.open(output_file, \"w:gz\") as tar:\n \
123
- \ tar.add(src_dir, arcname=os.path.basename(src_dir))\n return output_file"
124
- - source_sentence: Encrypts the zip file
125
- sentences:
126
- - "def nsUriMatch(self, value, wanted, strict=0, tt=type(())):\n \n \
127
- \ if value == wanted or (type(wanted) is tt) and value in wanted:\n \
128
- \ return 1\n if not strict and value is not None:\n wanted\
129
- \ = type(wanted) is tt and wanted or (wanted,)\n value = value[-1:]\
130
- \ != '/' and value or value[:-1]\n for item in wanted:\n \
131
- \ if item == value or item[:-1] == value:\n return 1\n\
132
- \ return 0"
133
- - "def transform(self, sents):\n \n\n def convert(tokens):\n \
134
- \ return torch.tensor([self.vocab.stoi[t] for t in tokens], dtype=torch.long)\n\
135
- \n if self.vocab is None:\n raise Exception(\n \
136
- \ \"Must run .fit() for .fit_transform() before \" \"calling .transform().\"\
137
- \n )\n\n seqs = sorted([convert(s) for s in sents], key=lambda\
138
- \ x: -len(x))\n X = torch.LongTensor(pad_sequence(seqs, batch_first=True))\n\
139
- \ return X"
140
- - "def freeze_encrypt(dest_dir, zip_filename, config, opt):\n \n pgp_keys\
141
- \ = grok_keys(config)\n icefile_prefix = \"aomi-%s\" % \\\n \
142
- \ os.path.basename(os.path.dirname(opt.secretfile))\n if opt.icefile_prefix:\n\
143
- \ icefile_prefix = opt.icefile_prefix\n\n timestamp = time.strftime(\"\
144
- %H%M%S-%m-%d-%Y\",\n datetime.datetime.now().timetuple())\n\
145
- \ ice_file = \"%s/%s-%s.ice\" % (dest_dir, icefile_prefix, timestamp)\n \
146
- \ if not encrypt(zip_filename, ice_file, pgp_keys):\n raise aomi.exceptions.GPG(\"\
147
- Unable to encrypt zipfile\")\n\n return ice_file"
148
  pipeline_tag: sentence-similarity
149
  library_name: sentence-transformers
150
  metrics:
151
- - pearson_cosine
152
- - spearman_cosine
153
  model-index:
154
  - name: SentenceTransformer based on Shuu12121/CodeModernBERT-Owl
155
  results:
@@ -161,664 +19,117 @@ model-index:
161
  type: code-docstring-dev
162
  metrics:
163
  - type: pearson_cosine
164
- value: .nan
165
  name: Pearson Cosine
166
  - type: spearman_cosine
167
- value: .nan
168
  name: Spearman Cosine
 
 
 
 
 
 
 
 
169
  ---
170
 
171
- # SentenceTransformer based on Shuu12121/CodeModernBERT-Owl
172
 
173
- This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [Shuu12121/CodeModernBERT-Owl](https://huggingface.co/Shuu12121/CodeModernBERT-Owl). 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.
174
 
175
- ## Model Details
176
 
177
- ### Model Description
178
- - **Model Type:** Sentence Transformer
179
- - **Base model:** [Shuu12121/CodeModernBERT-Owl](https://huggingface.co/Shuu12121/CodeModernBERT-Owl) <!-- at revision d403250d7979eb141409c611c0a39fd7110543a4 -->
180
- - **Maximum Sequence Length:** 2048 tokens
181
- - **Output Dimensionality:** 768 dimensions
182
- - **Similarity Function:** Cosine Similarity
183
- <!-- - **Training Dataset:** Unknown -->
184
- <!-- - **Language:** Unknown -->
185
- <!-- - **License:** Unknown -->
186
 
187
- ### Model Sources
188
 
189
- - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
190
- - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
191
- - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
192
 
193
- ### Full Model Architecture
194
 
195
- ```
196
- SentenceTransformer(
197
- (0): Transformer({'max_seq_length': 2048, 'do_lower_case': False}) with Transformer model: ModernBertModel
198
- (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})
199
- )
200
- ```
 
 
 
 
201
 
202
- ## Usage
203
 
204
- ### Direct Usage (Sentence Transformers)
205
 
206
- First install the Sentence Transformers library:
207
 
208
  ```bash
209
  pip install -U sentence-transformers
210
  ```
211
 
212
- Then you can load this model and run inference.
 
213
  ```python
214
  from sentence_transformers import SentenceTransformer
215
 
216
- # Download from the 🤗 Hub
217
- model = SentenceTransformer("sentence_transformers_model_id")
218
- # Run inference
 
219
  sentences = [
220
  'Encrypts the zip file',
221
  'def freeze_encrypt(dest_dir, zip_filename, config, opt):\n \n pgp_keys = grok_keys(config)\n icefile_prefix = "aomi-%s" % \\\n os.path.basename(os.path.dirname(opt.secretfile))\n if opt.icefile_prefix:\n icefile_prefix = opt.icefile_prefix\n\n timestamp = time.strftime("%H%M%S-%m-%d-%Y",\n datetime.datetime.now().timetuple())\n ice_file = "%s/%s-%s.ice" % (dest_dir, icefile_prefix, timestamp)\n if not encrypt(zip_filename, ice_file, pgp_keys):\n raise aomi.exceptions.GPG("Unable to encrypt zipfile")\n\n return ice_file',
222
  'def transform(self, sents):\n \n\n def convert(tokens):\n return torch.tensor([self.vocab.stoi[t] for t in tokens], dtype=torch.long)\n\n if self.vocab is None:\n raise Exception(\n "Must run .fit() for .fit_transform() before " "calling .transform()."\n )\n\n seqs = sorted([convert(s) for s in sents], key=lambda x: -len(x))\n X = torch.LongTensor(pad_sequence(seqs, batch_first=True))\n return X',
223
  ]
 
 
224
  embeddings = model.encode(sentences)
225
- print(embeddings.shape)
226
- # [3, 768]
227
 
228
- # Get the similarity scores for the embeddings
229
  similarities = model.similarity(embeddings, embeddings)
230
- print(similarities.shape)
231
- # [3, 3]
232
  ```
233
 
234
- <!--
235
- ### Direct Usage (Transformers)
236
-
237
- <details><summary>Click to see the direct usage in Transformers</summary>
238
-
239
- </details>
240
- -->
241
-
242
- <!--
243
- ### Downstream Usage (Sentence Transformers)
244
-
245
- You can finetune this model on your own dataset.
246
-
247
- <details><summary>Click to expand</summary>
248
-
249
- </details>
250
- -->
251
-
252
- <!--
253
- ### Out-of-Scope Use
254
-
255
- *List how the model may foreseeably be misused and address what users ought not to do with the model.*
256
- -->
257
-
258
- ## Evaluation
259
-
260
- ### Metrics
261
-
262
- #### Semantic Similarity
263
-
264
- * Dataset: `code-docstring-dev`
265
- * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
266
-
267
- | Metric | Value |
268
- |:--------------------|:--------|
269
- | pearson_cosine | nan |
270
- | **spearman_cosine** | **nan** |
271
-
272
- <!--
273
- ## Bias, Risks and Limitations
274
-
275
- *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
276
- -->
277
-
278
- <!--
279
- ### Recommendations
280
-
281
- *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
282
- -->
283
-
284
- ## Training Details
285
-
286
- ### Training Dataset
287
-
288
- #### Unnamed Dataset
289
-
290
- * Size: 1,431,743 training samples
291
- * Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>label</code>
292
- * Approximate statistics based on the first 1000 samples:
293
- | | sentence_0 | sentence_1 | label |
294
- |:--------|:------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------|:--------------------------------------------------------------|
295
- | type | string | string | float |
296
- | details | <ul><li>min: 5 tokens</li><li>mean: 63.94 tokens</li><li>max: 1310 tokens</li></ul> | <ul><li>min: 29 tokens</li><li>mean: 173.04 tokens</li><li>max: 1801 tokens</li></ul> | <ul><li>min: 1.0</li><li>mean: 1.0</li><li>max: 1.0</li></ul> |
297
- * Samples:
298
- | sentence_0 | sentence_1 | label |
299
- |:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------|
300
- | <code>Serves a cross-domain policy which can allow other policies<br> to exist on the same domain.<br><br> Note that this view, if used, must be the master policy for the<br> domain, and so must be served from the URL ``/crossdomain.xml`` on<br> the domain: setting metapolicy information in other policy files<br> is forbidden by the cross-domain policy specification.<br><br> **Required arguments:**<br><br> ``permitted``<br> A string indicating the extent to which other policies are<br> permitted. A set of constants is available in<br> ``flashpolicies.policies``, defining acceptable values for<br> this argument.<br><br> **Optional arguments:**<br><br> ``domains``<br> A list of domains from which to allow access. Each value may<br> be either a domain name (e.g., ``example.com``) or a wildcard<br> (e.g., ``*.example.com``). Due to serious potential security<br> issues, it is strongly recommended that you not use wildcard<br> domain values.</code> | <code>def metapolicy(request, permitted, domains=None):<br> <br> if domains is None:<br> domains = []<br> policy = policies.Policy(*domains)<br> policy.metapolicy(permitted)<br> return serve(request, policy)</code> | <code>1.0</code> |
301
- | <code>Puts a value from a VEX temporary register into a machine register.<br> This is how the results of operations done to registers get committed to the machine's state.<br><br> :param val: The VexValue to store (Want to store a constant? See Constant() first)<br> :param reg: The integer register number to store into, or register name<br> :return: None</code> | <code>def put(self, val, reg):<br> <br> offset = self.lookup_register(self.irsb_c.irsb.arch, reg)<br> self.irsb_c.put(val.rdt, offset)</code> | <code>1.0</code> |
302
- | <code>Like `get_token`, but using an OAuth 2 authorization code.<br><br> Use this method if you run a webserver that serves as an endpoint for<br> the redirect URI. The webserver can retrieve the authorization code<br> from the URL that is requested by ORCID.<br><br> Parameters<br> ----------<br> :param redirect_uri: string<br> The redirect uri of the institution.<br> :param authorization_code: string<br> The authorization code.<br><br> Returns<br> -------<br> :returns: dict<br> All data of the access token. The access token itself is in the<br> ``"access_token"`` key.</code> | <code>def get_token_from_authorization_code(self,<br> authorization_code, redirect_uri):<br> <br> token_dict = {<br> "client_id": self._key,<br> "client_secret": self._secret,<br> "grant_type": "authorization_code",<br> "code": authorization_code,<br> "redirect_uri": redirect_uri,<br> }<br> response = requests.post(self._token_url, data=token_dict,<br> headers={'Accept': 'application/json'},<br> timeout=self._timeout)<br> response.raise_for_status()<br> if self.do_store_raw_response:<br> self.raw_response = response<br> return json.loads(response.text)</code> | <code>1.0</code> |
303
- * Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
304
- ```json
305
- {
306
- "scale": 20.0,
307
- "similarity_fct": "cos_sim"
308
- }
309
- ```
310
-
311
- ### Training Hyperparameters
312
- #### Non-Default Hyperparameters
313
-
314
- - `eval_strategy`: steps
315
- - `per_device_train_batch_size`: 24
316
- - `per_device_eval_batch_size`: 24
317
- - `fp16`: True
318
- - `multi_dataset_batch_sampler`: round_robin
319
-
320
- #### All Hyperparameters
321
- <details><summary>Click to expand</summary>
322
-
323
- - `overwrite_output_dir`: False
324
- - `do_predict`: False
325
- - `eval_strategy`: steps
326
- - `prediction_loss_only`: True
327
- - `per_device_train_batch_size`: 24
328
- - `per_device_eval_batch_size`: 24
329
- - `per_gpu_train_batch_size`: None
330
- - `per_gpu_eval_batch_size`: None
331
- - `gradient_accumulation_steps`: 1
332
- - `eval_accumulation_steps`: None
333
- - `torch_empty_cache_steps`: None
334
- - `learning_rate`: 5e-05
335
- - `weight_decay`: 0.0
336
- - `adam_beta1`: 0.9
337
- - `adam_beta2`: 0.999
338
- - `adam_epsilon`: 1e-08
339
- - `max_grad_norm`: 1
340
- - `num_train_epochs`: 3
341
- - `max_steps`: -1
342
- - `lr_scheduler_type`: linear
343
- - `lr_scheduler_kwargs`: {}
344
- - `warmup_ratio`: 0.0
345
- - `warmup_steps`: 0
346
- - `log_level`: passive
347
- - `log_level_replica`: warning
348
- - `log_on_each_node`: True
349
- - `logging_nan_inf_filter`: True
350
- - `save_safetensors`: True
351
- - `save_on_each_node`: False
352
- - `save_only_model`: False
353
- - `restore_callback_states_from_checkpoint`: False
354
- - `no_cuda`: False
355
- - `use_cpu`: False
356
- - `use_mps_device`: False
357
- - `seed`: 42
358
- - `data_seed`: None
359
- - `jit_mode_eval`: False
360
- - `use_ipex`: False
361
- - `bf16`: False
362
- - `fp16`: True
363
- - `fp16_opt_level`: O1
364
- - `half_precision_backend`: auto
365
- - `bf16_full_eval`: False
366
- - `fp16_full_eval`: False
367
- - `tf32`: None
368
- - `local_rank`: 0
369
- - `ddp_backend`: None
370
- - `tpu_num_cores`: None
371
- - `tpu_metrics_debug`: False
372
- - `debug`: []
373
- - `dataloader_drop_last`: False
374
- - `dataloader_num_workers`: 0
375
- - `dataloader_prefetch_factor`: None
376
- - `past_index`: -1
377
- - `disable_tqdm`: False
378
- - `remove_unused_columns`: True
379
- - `label_names`: None
380
- - `load_best_model_at_end`: False
381
- - `ignore_data_skip`: False
382
- - `fsdp`: []
383
- - `fsdp_min_num_params`: 0
384
- - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
385
- - `tp_size`: 0
386
- - `fsdp_transformer_layer_cls_to_wrap`: None
387
- - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
388
- - `deepspeed`: None
389
- - `label_smoothing_factor`: 0.0
390
- - `optim`: adamw_torch
391
- - `optim_args`: None
392
- - `adafactor`: False
393
- - `group_by_length`: False
394
- - `length_column_name`: length
395
- - `ddp_find_unused_parameters`: None
396
- - `ddp_bucket_cap_mb`: None
397
- - `ddp_broadcast_buffers`: False
398
- - `dataloader_pin_memory`: True
399
- - `dataloader_persistent_workers`: False
400
- - `skip_memory_metrics`: True
401
- - `use_legacy_prediction_loop`: False
402
- - `push_to_hub`: False
403
- - `resume_from_checkpoint`: None
404
- - `hub_model_id`: None
405
- - `hub_strategy`: every_save
406
- - `hub_private_repo`: None
407
- - `hub_always_push`: False
408
- - `gradient_checkpointing`: False
409
- - `gradient_checkpointing_kwargs`: None
410
- - `include_inputs_for_metrics`: False
411
- - `include_for_metrics`: []
412
- - `eval_do_concat_batches`: True
413
- - `fp16_backend`: auto
414
- - `push_to_hub_model_id`: None
415
- - `push_to_hub_organization`: None
416
- - `mp_parameters`:
417
- - `auto_find_batch_size`: False
418
- - `full_determinism`: False
419
- - `torchdynamo`: None
420
- - `ray_scope`: last
421
- - `ddp_timeout`: 1800
422
- - `torch_compile`: False
423
- - `torch_compile_backend`: None
424
- - `torch_compile_mode`: None
425
- - `dispatch_batches`: None
426
- - `split_batches`: None
427
- - `include_tokens_per_second`: False
428
- - `include_num_input_tokens_seen`: False
429
- - `neftune_noise_alpha`: None
430
- - `optim_target_modules`: None
431
- - `batch_eval_metrics`: False
432
- - `eval_on_start`: False
433
- - `use_liger_kernel`: False
434
- - `eval_use_gather_object`: False
435
- - `average_tokens_across_devices`: False
436
- - `prompts`: None
437
- - `batch_sampler`: batch_sampler
438
- - `multi_dataset_batch_sampler`: round_robin
439
-
440
- </details>
441
-
442
- ### Training Logs
443
- <details><summary>Click to expand</summary>
444
-
445
- | Epoch | Step | Training Loss | code-docstring-dev_spearman_cosine |
446
- |:------:|:------:|:-------------:|:----------------------------------:|
447
- | 0.0084 | 500 | 0.9451 | - |
448
- | 0.0168 | 1000 | 0.1154 | - |
449
- | 0.0251 | 1500 | 0.0817 | - |
450
- | 0.0335 | 2000 | 0.0733 | - |
451
- | 0.0419 | 2500 | 0.0751 | - |
452
- | 0.0503 | 3000 | 0.0629 | - |
453
- | 0.0587 | 3500 | 0.0551 | - |
454
- | 0.0671 | 4000 | 0.0604 | - |
455
- | 0.0754 | 4500 | 0.0628 | - |
456
- | 0.0838 | 5000 | 0.0548 | nan |
457
- | 0.0922 | 5500 | 0.054 | - |
458
- | 0.1006 | 6000 | 0.0538 | - |
459
- | 0.1090 | 6500 | 0.0518 | - |
460
- | 0.1173 | 7000 | 0.0543 | - |
461
- | 0.1257 | 7500 | 0.0491 | - |
462
- | 0.1341 | 8000 | 0.0446 | - |
463
- | 0.1425 | 8500 | 0.049 | - |
464
- | 0.1509 | 9000 | 0.0477 | - |
465
- | 0.1592 | 9500 | 0.0458 | - |
466
- | 0.1676 | 10000 | 0.0425 | nan |
467
- | 0.1760 | 10500 | 0.0445 | - |
468
- | 0.1844 | 11000 | 0.0397 | - |
469
- | 0.1928 | 11500 | 0.044 | - |
470
- | 0.2012 | 12000 | 0.0432 | - |
471
- | 0.2095 | 12500 | 0.0402 | - |
472
- | 0.2179 | 13000 | 0.0483 | - |
473
- | 0.2263 | 13500 | 0.0434 | - |
474
- | 0.2347 | 14000 | 0.0425 | - |
475
- | 0.2431 | 14500 | 0.0464 | - |
476
- | 0.2514 | 15000 | 0.038 | nan |
477
- | 0.2598 | 15500 | 0.0391 | - |
478
- | 0.2682 | 16000 | 0.0385 | - |
479
- | 0.2766 | 16500 | 0.0383 | - |
480
- | 0.2850 | 17000 | 0.0396 | - |
481
- | 0.2933 | 17500 | 0.0394 | - |
482
- | 0.3017 | 18000 | 0.0407 | - |
483
- | 0.3101 | 18500 | 0.0437 | - |
484
- | 0.3185 | 19000 | 0.0362 | - |
485
- | 0.3269 | 19500 | 0.0398 | - |
486
- | 0.3353 | 20000 | 0.0379 | nan |
487
- | 0.3436 | 20500 | 0.0418 | - |
488
- | 0.3520 | 21000 | 0.0348 | - |
489
- | 0.3604 | 21500 | 0.0382 | - |
490
- | 0.3688 | 22000 | 0.0374 | - |
491
- | 0.3772 | 22500 | 0.038 | - |
492
- | 0.3855 | 23000 | 0.0365 | - |
493
- | 0.3939 | 23500 | 0.0348 | - |
494
- | 0.4023 | 24000 | 0.0405 | - |
495
- | 0.4107 | 24500 | 0.04 | - |
496
- | 0.4191 | 25000 | 0.0362 | nan |
497
- | 0.4275 | 25500 | 0.0327 | - |
498
- | 0.4358 | 26000 | 0.0331 | - |
499
- | 0.4442 | 26500 | 0.0309 | - |
500
- | 0.4526 | 27000 | 0.0348 | - |
501
- | 0.4610 | 27500 | 0.0295 | - |
502
- | 0.4694 | 28000 | 0.0378 | - |
503
- | 0.4777 | 28500 | 0.0318 | - |
504
- | 0.4861 | 29000 | 0.0323 | - |
505
- | 0.4945 | 29500 | 0.0315 | - |
506
- | 0.5029 | 30000 | 0.0336 | nan |
507
- | 0.5113 | 30500 | 0.0334 | - |
508
- | 0.5196 | 31000 | 0.0342 | - |
509
- | 0.5280 | 31500 | 0.0289 | - |
510
- | 0.5364 | 32000 | 0.0332 | - |
511
- | 0.5448 | 32500 | 0.0305 | - |
512
- | 0.5532 | 33000 | 0.0349 | - |
513
- | 0.5616 | 33500 | 0.0309 | - |
514
- | 0.5699 | 34000 | 0.0352 | - |
515
- | 0.5783 | 34500 | 0.035 | - |
516
- | 0.5867 | 35000 | 0.0316 | nan |
517
- | 0.5951 | 35500 | 0.0342 | - |
518
- | 0.6035 | 36000 | 0.0274 | - |
519
- | 0.6118 | 36500 | 0.0333 | - |
520
- | 0.6202 | 37000 | 0.0294 | - |
521
- | 0.6286 | 37500 | 0.029 | - |
522
- | 0.6370 | 38000 | 0.0302 | - |
523
- | 0.6454 | 38500 | 0.0292 | - |
524
- | 0.6537 | 39000 | 0.032 | - |
525
- | 0.6621 | 39500 | 0.03 | - |
526
- | 0.6705 | 40000 | 0.0246 | nan |
527
- | 0.6789 | 40500 | 0.0277 | - |
528
- | 0.6873 | 41000 | 0.0262 | - |
529
- | 0.6957 | 41500 | 0.0293 | - |
530
- | 0.7040 | 42000 | 0.0284 | - |
531
- | 0.7124 | 42500 | 0.028 | - |
532
- | 0.7208 | 43000 | 0.0321 | - |
533
- | 0.7292 | 43500 | 0.0283 | - |
534
- | 0.7376 | 44000 | 0.0295 | - |
535
- | 0.7459 | 44500 | 0.0279 | - |
536
- | 0.7543 | 45000 | 0.0249 | nan |
537
- | 0.7627 | 45500 | 0.0299 | - |
538
- | 0.7711 | 46000 | 0.0258 | - |
539
- | 0.7795 | 46500 | 0.0257 | - |
540
- | 0.7879 | 47000 | 0.0256 | - |
541
- | 0.7962 | 47500 | 0.0281 | - |
542
- | 0.8046 | 48000 | 0.0279 | - |
543
- | 0.8130 | 48500 | 0.0299 | - |
544
- | 0.8214 | 49000 | 0.027 | - |
545
- | 0.8298 | 49500 | 0.0271 | - |
546
- | 0.8381 | 50000 | 0.0281 | nan |
547
- | 0.8465 | 50500 | 0.0274 | - |
548
- | 0.8549 | 51000 | 0.0262 | - |
549
- | 0.8633 | 51500 | 0.0306 | - |
550
- | 0.8717 | 52000 | 0.0262 | - |
551
- | 0.8800 | 52500 | 0.0241 | - |
552
- | 0.8884 | 53000 | 0.0235 | - |
553
- | 0.8968 | 53500 | 0.0268 | - |
554
- | 0.9052 | 54000 | 0.0251 | - |
555
- | 0.9136 | 54500 | 0.0328 | - |
556
- | 0.9220 | 55000 | 0.0235 | nan |
557
- | 0.9303 | 55500 | 0.0261 | - |
558
- | 0.9387 | 56000 | 0.0249 | - |
559
- | 0.9471 | 56500 | 0.0262 | - |
560
- | 0.9555 | 57000 | 0.0231 | - |
561
- | 0.9639 | 57500 | 0.0249 | - |
562
- | 0.9722 | 58000 | 0.0246 | - |
563
- | 0.9806 | 58500 | 0.0299 | - |
564
- | 0.9890 | 59000 | 0.0238 | - |
565
- | 0.9974 | 59500 | 0.0215 | - |
566
- | 1.0 | 59656 | - | nan |
567
- | 1.0058 | 60000 | 0.0157 | nan |
568
- | 1.0141 | 60500 | 0.0095 | - |
569
- | 1.0225 | 61000 | 0.012 | - |
570
- | 1.0309 | 61500 | 0.0105 | - |
571
- | 1.0393 | 62000 | 0.01 | - |
572
- | 1.0477 | 62500 | 0.0101 | - |
573
- | 1.0561 | 63000 | 0.0107 | - |
574
- | 1.0644 | 63500 | 0.0102 | - |
575
- | 1.0728 | 64000 | 0.011 | - |
576
- | 1.0812 | 64500 | 0.0088 | - |
577
- | 1.0896 | 65000 | 0.0106 | nan |
578
- | 1.0980 | 65500 | 0.0108 | - |
579
- | 1.1063 | 66000 | 0.0108 | - |
580
- | 1.1147 | 66500 | 0.011 | - |
581
- | 1.1231 | 67000 | 0.0082 | - |
582
- | 1.1315 | 67500 | 0.0092 | - |
583
- | 1.1399 | 68000 | 0.0106 | - |
584
- | 1.1482 | 68500 | 0.0117 | - |
585
- | 1.1566 | 69000 | 0.0096 | - |
586
- | 1.1650 | 69500 | 0.0094 | - |
587
- | 1.1734 | 70000 | 0.0098 | nan |
588
- | 1.1818 | 70500 | 0.0084 | - |
589
- | 1.1902 | 71000 | 0.0103 | - |
590
- | 1.1985 | 71500 | 0.0112 | - |
591
- | 1.2069 | 72000 | 0.0108 | - |
592
- | 1.2153 | 72500 | 0.0121 | - |
593
- | 1.2237 | 73000 | 0.0103 | - |
594
- | 1.2321 | 73500 | 0.012 | - |
595
- | 1.2404 | 74000 | 0.0134 | - |
596
- | 1.2488 | 74500 | 0.0097 | - |
597
- | 1.2572 | 75000 | 0.0121 | nan |
598
- | 1.2656 | 75500 | 0.0117 | - |
599
- | 1.2740 | 76000 | 0.0108 | - |
600
- | 1.2824 | 76500 | 0.0106 | - |
601
- | 1.2907 | 77000 | 0.0085 | - |
602
- | 1.2991 | 77500 | 0.0119 | - |
603
- | 1.3075 | 78000 | 0.0099 | - |
604
- | 1.3159 | 78500 | 0.0102 | - |
605
- | 1.3243 | 79000 | 0.011 | - |
606
- | 1.3326 | 79500 | 0.0108 | - |
607
- | 1.3410 | 80000 | 0.0097 | nan |
608
- | 1.3494 | 80500 | 0.0101 | - |
609
- | 1.3578 | 81000 | 0.0082 | - |
610
- | 1.3662 | 81500 | 0.0107 | - |
611
- | 1.3745 | 82000 | 0.013 | - |
612
- | 1.3829 | 82500 | 0.0068 | - |
613
- | 1.3913 | 83000 | 0.0102 | - |
614
- | 1.3997 | 83500 | 0.0079 | - |
615
- | 1.4081 | 84000 | 0.0116 | - |
616
- | 1.4165 | 84500 | 0.0095 | - |
617
- | 1.4248 | 85000 | 0.0105 | nan |
618
- | 1.4332 | 85500 | 0.011 | - |
619
- | 1.4416 | 86000 | 0.0131 | - |
620
- | 1.4500 | 86500 | 0.012 | - |
621
- | 1.4584 | 87000 | 0.0105 | - |
622
- | 1.4667 | 87500 | 0.0117 | - |
623
- | 1.4751 | 88000 | 0.0101 | - |
624
- | 1.4835 | 88500 | 0.0108 | - |
625
- | 1.4919 | 89000 | 0.0091 | - |
626
- | 1.5003 | 89500 | 0.0086 | - |
627
- | 1.5086 | 90000 | 0.0093 | nan |
628
- | 1.5170 | 90500 | 0.0102 | - |
629
- | 1.5254 | 91000 | 0.0078 | - |
630
- | 1.5338 | 91500 | 0.0096 | - |
631
- | 1.5422 | 92000 | 0.0103 | - |
632
- | 1.5506 | 92500 | 0.0099 | - |
633
- | 1.5589 | 93000 | 0.011 | - |
634
- | 1.5673 | 93500 | 0.0079 | - |
635
- | 1.5757 | 94000 | 0.0108 | - |
636
- | 1.5841 | 94500 | 0.0089 | - |
637
- | 1.5925 | 95000 | 0.0115 | nan |
638
- | 1.6008 | 95500 | 0.0092 | - |
639
- | 1.6092 | 96000 | 0.0093 | - |
640
- | 1.6176 | 96500 | 0.0083 | - |
641
- | 1.6260 | 97000 | 0.0103 | - |
642
- | 1.6344 | 97500 | 0.01 | - |
643
- | 1.6428 | 98000 | 0.0091 | - |
644
- | 1.6511 | 98500 | 0.0106 | - |
645
- | 1.6595 | 99000 | 0.0105 | - |
646
- | 1.6679 | 99500 | 0.0096 | - |
647
- | 1.6763 | 100000 | 0.0116 | nan |
648
- | 1.6847 | 100500 | 0.0093 | - |
649
- | 1.6930 | 101000 | 0.01 | - |
650
- | 1.7014 | 101500 | 0.0076 | - |
651
- | 1.7098 | 102000 | 0.0078 | - |
652
- | 1.7182 | 102500 | 0.0089 | - |
653
- | 1.7266 | 103000 | 0.0082 | - |
654
- | 1.7349 | 103500 | 0.0081 | - |
655
- | 1.7433 | 104000 | 0.009 | - |
656
- | 1.7517 | 104500 | 0.0082 | - |
657
- | 1.7601 | 105000 | 0.008 | nan |
658
- | 1.7685 | 105500 | 0.0082 | - |
659
- | 1.7769 | 106000 | 0.0077 | - |
660
- | 1.7852 | 106500 | 0.0103 | - |
661
- | 1.7936 | 107000 | 0.0103 | - |
662
- | 1.8020 | 107500 | 0.0103 | - |
663
- | 1.8104 | 108000 | 0.0079 | - |
664
- | 1.8188 | 108500 | 0.0082 | - |
665
- | 1.8271 | 109000 | 0.0088 | - |
666
- | 1.8355 | 109500 | 0.0096 | - |
667
- | 1.8439 | 110000 | 0.0097 | nan |
668
- | 1.8523 | 110500 | 0.0085 | - |
669
- | 1.8607 | 111000 | 0.01 | - |
670
- | 1.8690 | 111500 | 0.0114 | - |
671
- | 1.8774 | 112000 | 0.0075 | - |
672
- | 1.8858 | 112500 | 0.0083 | - |
673
- | 1.8942 | 113000 | 0.0113 | - |
674
- | 1.9026 | 113500 | 0.0077 | - |
675
- | 1.9110 | 114000 | 0.0077 | - |
676
- | 1.9193 | 114500 | 0.0107 | - |
677
- | 1.9277 | 115000 | 0.0077 | nan |
678
- | 1.9361 | 115500 | 0.0094 | - |
679
- | 1.9445 | 116000 | 0.0082 | - |
680
- | 1.9529 | 116500 | 0.0089 | - |
681
- | 1.9612 | 117000 | 0.0066 | - |
682
- | 1.9696 | 117500 | 0.0102 | - |
683
- | 1.9780 | 118000 | 0.0097 | - |
684
- | 1.9864 | 118500 | 0.0081 | - |
685
- | 1.9948 | 119000 | 0.0086 | - |
686
- | 2.0 | 119312 | - | nan |
687
- | 2.0032 | 119500 | 0.0063 | - |
688
- | 2.0115 | 120000 | 0.0051 | nan |
689
- | 2.0199 | 120500 | 0.0037 | - |
690
- | 2.0283 | 121000 | 0.0062 | - |
691
- | 2.0367 | 121500 | 0.0045 | - |
692
- | 2.0451 | 122000 | 0.0046 | - |
693
- | 2.0534 | 122500 | 0.0038 | - |
694
- | 2.0618 | 123000 | 0.0044 | - |
695
- | 2.0702 | 123500 | 0.0042 | - |
696
- | 2.0786 | 124000 | 0.0029 | - |
697
- | 2.0870 | 124500 | 0.0029 | - |
698
- | 2.0953 | 125000 | 0.0067 | nan |
699
- | 2.1037 | 125500 | 0.0067 | - |
700
- | 2.1121 | 126000 | 0.005 | - |
701
- | 2.1205 | 126500 | 0.005 | - |
702
- | 2.1289 | 127000 | 0.0037 | - |
703
- | 2.1373 | 127500 | 0.0043 | - |
704
- | 2.1456 | 128000 | 0.0036 | - |
705
- | 2.1540 | 128500 | 0.0042 | - |
706
- | 2.1624 | 129000 | 0.0039 | - |
707
- | 2.1708 | 129500 | 0.0032 | - |
708
- | 2.1792 | 130000 | 0.0046 | nan |
709
- | 2.1875 | 130500 | 0.0037 | - |
710
- | 2.1959 | 131000 | 0.0036 | - |
711
- | 2.2043 | 131500 | 0.0042 | - |
712
- | 2.2127 | 132000 | 0.0044 | - |
713
- | 2.2211 | 132500 | 0.0028 | - |
714
- | 2.2294 | 133000 | 0.0043 | - |
715
- | 2.2378 | 133500 | 0.0052 | - |
716
- | 2.2462 | 134000 | 0.0031 | - |
717
- | 2.2546 | 134500 | 0.0048 | - |
718
- | 2.2630 | 135000 | 0.0031 | nan |
719
- | 2.2714 | 135500 | 0.0054 | - |
720
- | 2.2797 | 136000 | 0.0033 | - |
721
- | 2.2881 | 136500 | 0.0036 | - |
722
- | 2.2965 | 137000 | 0.0033 | - |
723
- | 2.3049 | 137500 | 0.0039 | - |
724
- | 2.3133 | 138000 | 0.0044 | - |
725
- | 2.3216 | 138500 | 0.0034 | - |
726
- | 2.3300 | 139000 | 0.0058 | - |
727
- | 2.3384 | 139500 | 0.0036 | - |
728
- | 2.3468 | 140000 | 0.0033 | nan |
729
- | 2.3552 | 140500 | 0.0034 | - |
730
- | 2.3636 | 141000 | 0.0032 | - |
731
- | 2.3719 | 141500 | 0.0036 | - |
732
- | 2.3803 | 142000 | 0.0038 | - |
733
- | 2.3887 | 142500 | 0.0036 | - |
734
- | 2.3971 | 143000 | 0.0045 | - |
735
- | 2.4055 | 143500 | 0.0035 | - |
736
- | 2.4138 | 144000 | 0.0042 | - |
737
- | 2.4222 | 144500 | 0.0029 | - |
738
- | 2.4306 | 145000 | 0.005 | nan |
739
- | 2.4390 | 145500 | 0.0045 | - |
740
- | 2.4474 | 146000 | 0.0035 | - |
741
- | 2.4557 | 146500 | 0.004 | - |
742
- | 2.4641 | 147000 | 0.0044 | - |
743
- | 2.4725 | 147500 | 0.0036 | - |
744
- | 2.4809 | 148000 | 0.0047 | - |
745
- | 2.4893 | 148500 | 0.0035 | - |
746
- | 2.4977 | 149000 | 0.0048 | - |
747
- | 2.5060 | 149500 | 0.0041 | - |
748
- | 2.5144 | 150000 | 0.0029 | nan |
749
- | 2.5228 | 150500 | 0.0038 | - |
750
- | 2.5312 | 151000 | 0.0032 | - |
751
- | 2.5396 | 151500 | 0.0043 | - |
752
- | 2.5479 | 152000 | 0.0038 | - |
753
- | 2.5563 | 152500 | 0.0037 | - |
754
- | 2.5647 | 153000 | 0.0023 | - |
755
- | 2.5731 | 153500 | 0.0041 | - |
756
- | 2.5815 | 154000 | 0.0049 | - |
757
- | 2.5898 | 154500 | 0.0048 | - |
758
- | 2.5982 | 155000 | 0.0034 | nan |
759
- | 2.6066 | 155500 | 0.0031 | - |
760
- | 2.6150 | 156000 | 0.0036 | - |
761
- | 2.6234 | 156500 | 0.0034 | - |
762
- | 2.6318 | 157000 | 0.0037 | - |
763
- | 2.6401 | 157500 | 0.0035 | - |
764
- | 2.6485 | 158000 | 0.0037 | - |
765
- | 2.6569 | 158500 | 0.0043 | - |
766
- | 2.6653 | 159000 | 0.0042 | - |
767
- | 2.6737 | 159500 | 0.0049 | - |
768
- | 2.6820 | 160000 | 0.0035 | nan |
769
- | 2.6904 | 160500 | 0.0026 | - |
770
- | 2.6988 | 161000 | 0.0049 | - |
771
- | 2.7072 | 161500 | 0.0034 | - |
772
- | 2.7156 | 162000 | 0.0039 | - |
773
- | 2.7240 | 162500 | 0.0042 | - |
774
- | 2.7323 | 163000 | 0.0052 | - |
775
- | 2.7407 | 163500 | 0.0045 | - |
776
- | 2.7491 | 164000 | 0.0043 | - |
777
- | 2.7575 | 164500 | 0.0034 | - |
778
- | 2.7659 | 165000 | 0.0038 | nan |
779
- | 2.7742 | 165500 | 0.0029 | - |
780
- | 2.7826 | 166000 | 0.0041 | - |
781
- | 2.7910 | 166500 | 0.0041 | - |
782
- | 2.7994 | 167000 | 0.0048 | - |
783
- | 2.8078 | 167500 | 0.0044 | - |
784
- | 2.8161 | 168000 | 0.0041 | - |
785
- | 2.8245 | 168500 | 0.0035 | - |
786
- | 2.8329 | 169000 | 0.0026 | - |
787
- | 2.8413 | 169500 | 0.0033 | - |
788
- | 2.8497 | 170000 | 0.0048 | nan |
789
- | 2.8581 | 170500 | 0.0046 | - |
790
- | 2.8664 | 171000 | 0.0027 | - |
791
- | 2.8748 | 171500 | 0.0037 | - |
792
- | 2.8832 | 172000 | 0.0028 | - |
793
- | 2.8916 | 172500 | 0.0032 | - |
794
- | 2.9000 | 173000 | 0.0029 | - |
795
- | 2.9083 | 173500 | 0.0043 | - |
796
- | 2.9167 | 174000 | 0.0048 | - |
797
- | 2.9251 | 174500 | 0.0037 | - |
798
- | 2.9335 | 175000 | 0.003 | nan |
799
- | 2.9419 | 175500 | 0.0034 | - |
800
- | 2.9502 | 176000 | 0.0035 | - |
801
- | 2.9586 | 176500 | 0.0042 | - |
802
- | 2.9670 | 177000 | 0.005 | - |
803
- | 2.9754 | 177500 | 0.0038 | - |
804
- | 2.9838 | 178000 | 0.0032 | - |
805
- | 2.9922 | 178500 | 0.0028 | - |
806
- | 3.0 | 178968 | - | nan |
807
 
808
- </details>
809
 
810
- ### Framework Versions
811
- - Python: 3.11.11
812
- - Sentence Transformers: 3.4.1
813
- - Transformers: 4.50.0
814
- - PyTorch: 2.6.0+cu124
815
- - Accelerate: 1.5.2
816
- - Datasets: 3.4.1
817
- - Tokenizers: 0.21.1
818
 
819
- ## Citation
820
 
821
- ### BibTeX
822
 
823
  #### Sentence Transformers
824
  ```bibtex
@@ -843,22 +154,4 @@ You can finetune this model on your own dataset.
843
  archivePrefix={arXiv},
844
  primaryClass={cs.CL}
845
  }
846
- ```
847
-
848
- <!--
849
- ## Glossary
850
-
851
- *Clearly define terms in order to be accessible across audiences.*
852
- -->
853
-
854
- <!--
855
- ## Model Card Authors
856
-
857
- *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
858
- -->
859
-
860
- <!--
861
- ## Model Card Contact
862
-
863
- *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
864
- -->
 
3
  - sentence-transformers
4
  - sentence-similarity
5
  - feature-extraction
 
 
 
6
  base_model: Shuu12121/CodeModernBERT-Owl
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7
  pipeline_tag: sentence-similarity
8
  library_name: sentence-transformers
9
  metrics:
10
+ - code_eval
 
11
  model-index:
12
  - name: SentenceTransformer based on Shuu12121/CodeModernBERT-Owl
13
  results:
 
19
  type: code-docstring-dev
20
  metrics:
21
  - type: pearson_cosine
22
+ value: null
23
  name: Pearson Cosine
24
  - type: spearman_cosine
25
+ value: null
26
  name: Spearman Cosine
27
+ license: apache-2.0
28
+ datasets:
29
+ - code-search-net/code_search_net
30
+ - Shuu12121/java-codesearch-dataset-open
31
+ - Shuu12121/rust-codesearch-dataset-open
32
+ - google/code_x_glue_ct_code_to_text
33
+ language:
34
+ - en
35
  ---
36
 
 
37
 
 
38
 
39
+ ## SentenceTransformer based on Shuu12121/CodeModernBERT-Owl
40
 
41
+ このモデルは、[Shuu12121/CodeModernBERT-Owl](https://huggingface.co/Shuu12121/CodeModernBERT-Owl) をベースにファインチューニングされた [sentence-transformers](https://www.SBERT.net) モデルです。
42
+ **特にコードサーチに特化しており、コード片やドキュメントから効果的に意味的類似性を計算できる** ように設計されています。
 
 
 
 
 
 
 
43
 
44
+ ---
45
 
46
+ ### モデル評価
47
+
48
+ #### CoIRにおける評価結果
49
+
50
+ 本モデルは、150M程度と比較的小さいモデルながら**コードサーチタスクにおける評価指標である CodeSearchNet で 76.89** を達成しました。
51
+ 他のタスクには対応していないため、評価値は提供されていません。
52
+ CodeSearchNetタスクにおける評価値としては、他の有名なモデルと比較しても高いパフォーマンスを示しています。
53
+
54
+ | モデル名 | CodeSearchNet 評価値 |
55
+ |----------------------------------------------|-----------------------|
56
+ | **Shuu12121/CodeModernBERT-Owl** | **76.89** |
57
+ | Salesforce/SFR-Embedding-Code-2B_R | 73.5 |
58
+ | CodeSage-large-v2 | 94.26 |
59
+ | Salesforce/SFR-Embedding-Code-400M_R | 72.53 |
60
+ | CodeSage-large | 90.58 |
61
+ | Voyage-Code-002 | 81.79 |
62
+ | E5-Mistral | 54.25 |
63
+ | E5-Base-v2 | 67.99 |
64
+ | OpenAI-Ada-002 | 74.21 |
65
+ | BGE-Base-en-v1.5 | 69.6 |
66
+ | BGE-M3 | 43.23 |
67
+ | UniXcoder | 60.2 |
68
+ | GTE-Base-en-v1.5 | 43.35 |
69
+ | Contriever | 34.72 |
70
 
71
+ ---
72
 
73
+ ### モデル詳細
74
+
75
+ - **モデルタイプ:** Sentence Transformer
76
+ - **ベースモデル:** [Shuu12121/CodeModernBERT-Owl](https://huggingface.co/Shuu12121/CodeModernBERT-Owl)
77
+ - **最大シーケンス長:** 2048トークン
78
+ - **出力次元:** 768次元
79
+ - **類似度関数:** コサイン類似度
80
+ - **ライセンス:** Apache-2.0
81
+
82
+ ---
83
 
84
+ ### 使用方法
85
 
86
+ #### Sentence Transformers ライブラリのインストール
87
 
88
+ 以下のコマンドで Sentence Transformers をインストールします。
89
 
90
  ```bash
91
  pip install -U sentence-transformers
92
  ```
93
 
94
+ #### モデルのロードと推論
95
+
96
  ```python
97
  from sentence_transformers import SentenceTransformer
98
 
99
+ # モデルをダウンロードしてロード
100
+ model = SentenceTransformer("Shuu12121/CodeSearch-ModernBERT-Owl")
101
+
102
+ # 推論用の文リスト
103
  sentences = [
104
  'Encrypts the zip file',
105
  'def freeze_encrypt(dest_dir, zip_filename, config, opt):\n \n pgp_keys = grok_keys(config)\n icefile_prefix = "aomi-%s" % \\\n os.path.basename(os.path.dirname(opt.secretfile))\n if opt.icefile_prefix:\n icefile_prefix = opt.icefile_prefix\n\n timestamp = time.strftime("%H%M%S-%m-%d-%Y",\n datetime.datetime.now().timetuple())\n ice_file = "%s/%s-%s.ice" % (dest_dir, icefile_prefix, timestamp)\n if not encrypt(zip_filename, ice_file, pgp_keys):\n raise aomi.exceptions.GPG("Unable to encrypt zipfile")\n\n return ice_file',
106
  'def transform(self, sents):\n \n\n def convert(tokens):\n return torch.tensor([self.vocab.stoi[t] for t in tokens], dtype=torch.long)\n\n if self.vocab is None:\n raise Exception(\n "Must run .fit() for .fit_transform() before " "calling .transform()."\n )\n\n seqs = sorted([convert(s) for s in sents], key=lambda x: -len(x))\n X = torch.LongTensor(pad_sequence(seqs, batch_first=True))\n return X',
107
  ]
108
+
109
+ # 埋め込みベクトルの生成
110
  embeddings = model.encode(sentences)
111
+ print(embeddings.shape) # [3, 768]
 
112
 
113
+ # 類似度スコアの計算
114
  similarities = model.similarity(embeddings, embeddings)
115
+ print(similarities.shape) # [3, 3]
 
116
  ```
117
 
118
+ ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
119
 
120
+ ### ライブラリバージョン
121
 
122
+ - Python: 3.11.11
123
+ - Sentence Transformers: 3.4.1
124
+ - Transformers: 4.50.0
125
+ - PyTorch: 2.6.0+cu124
126
+ - Accelerate: 1.5.2
127
+ - Datasets: 3.4.1
128
+ - Tokenizers: 0.21.1
 
129
 
130
+ ---
131
 
132
+ ### 引用情報
133
 
134
  #### Sentence Transformers
135
  ```bibtex
 
154
  archivePrefix={arXiv},
155
  primaryClass={cs.CL}
156
  }
157
+ ```