Hugging Face's logo

samanvitha7
/
semeval2026-bge_large-bge-large-all-bge_expanded-checkpoints

Sentence Similarity
sentence-transformers
Safetensors
bert
feature-extraction
dense
Generated from Trainer
dataset_size:2733
loss:TripletLoss
text-embeddings-inference
Model card Files
xet
 Community
samanvitha7 commited on
Commit
2b55b9c
·
verified ·
1 Parent(s): 1276b52

Upload model from C:\Users\Lenovo\SemEval2026-task4\bge_large\bge-large-all\bge_expanded\checkpoints

Browse files
Files changed (19) hide show
  1. 1_Pooling/config.json +10 -0
  2. README.md +317 -0
  3. config.json +36 -0
  4. config_sentence_transformers.json +14 -0
  5. final_model/1_Pooling/config.json +10 -0
  6. final_model/README.md +317 -0
  7. final_model/config.json +36 -0
  8. final_model/config_sentence_transformers.json +14 -0
  9. final_model/model.safetensors +3 -0
  10. final_model/modules.json +20 -0
  11. final_model/sentence_bert_config.json +4 -0
  12. final_model/tokenizer.json +0 -0
  13. final_model/tokenizer_config.json +17 -0
  14. model.safetensors +3 -0
  15. modules.json +20 -0
  16. sentence_bert_config.json +4 -0
  17. tokenizer.json +0 -0
  18. tokenizer_config.json +17 -0
  19. training_config.json +15 -0
1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 1024,
3
+ "pooling_mode_cls_token": true,
4
+ "pooling_mode_mean_tokens": false,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
README.md ADDED
@@ -0,0 +1,317 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - sentence-similarity
5
+ - feature-extraction
6
+ - dense
7
+ - generated_from_trainer
8
+ - dataset_size:2733
9
+ - loss:TripletLoss
10
+ base_model: BAAI/bge-large-en-v1.5
11
+ pipeline_tag: sentence-similarity
12
+ library_name: sentence-transformers
13
+ ---
14
+
15
+ # SentenceTransformer based on BAAI/bge-large-en-v1.5
16
+
17
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-large-en-v1.5](https://huggingface.co/BAAI/bge-large-en-v1.5). It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
18
+
19
+ ## Model Details
20
+
21
+ ### Model Description
22
+ - **Model Type:** Sentence Transformer
23
+ - **Base model:** [BAAI/bge-large-en-v1.5](https://huggingface.co/BAAI/bge-large-en-v1.5) <!-- at revision d4aa6901d3a41ba39fb536a557fa166f842b0e09 -->
24
+ - **Maximum Sequence Length:** 512 tokens
25
+ - **Output Dimensionality:** 1024 dimensions
26
+ - **Similarity Function:** Cosine Similarity
27
+ <!-- - **Training Dataset:** Unknown -->
28
+ <!-- - **Language:** Unknown -->
29
+ <!-- - **License:** Unknown -->
30
+
31
+ ### Model Sources
32
+
33
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
34
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
35
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
36
+
37
+ ### Full Model Architecture
38
+
39
+ ```
40
+ SentenceTransformer(
41
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': True, 'architecture': 'BertModel'})
42
+ (1): Pooling({'word_embedding_dimension': 1024, '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})
43
+ (2): Normalize()
44
+ )
45
+ ```
46
+
47
+ ## Usage
48
+
49
+ ### Direct Usage (Sentence Transformers)
50
+
51
+ First install the Sentence Transformers library:
52
+
53
+ ```bash
54
+ pip install -U sentence-transformers
55
+ ```
56
+
57
+ Then you can load this model and run inference.
58
+ ```python
59
+ from sentence_transformers import SentenceTransformer
60
+
61
+ # Download from the 🤗 Hub
62
+ model = SentenceTransformer("sentence_transformers_model_id")
63
+ # Run inference
64
+ sentences = [
65
+ 'The weather is lovely today.',
66
+ "It's so sunny outside!",
67
+ 'He drove to the stadium.',
68
+ ]
69
+ embeddings = model.encode(sentences)
70
+ print(embeddings.shape)
71
+ # [3, 1024]
72
+
73
+ # Get the similarity scores for the embeddings
74
+ similarities = model.similarity(embeddings, embeddings)
75
+ print(similarities)
76
+ # tensor([[1.0000, 0.8407, 0.3897],
77
+ # [0.8407, 1.0000, 0.3653],
78
+ # [0.3897, 0.3653, 1.0000]])
79
+ ```
80
+
81
+ <!--
82
+ ### Direct Usage (Transformers)
83
+
84
+ <details><summary>Click to see the direct usage in Transformers</summary>
85
+
86
+ </details>
87
+ -->
88
+
89
+ <!--
90
+ ### Downstream Usage (Sentence Transformers)
91
+
92
+ You can finetune this model on your own dataset.
93
+
94
+ <details><summary>Click to expand</summary>
95
+
96
+ </details>
97
+ -->
98
+
99
+ <!--
100
+ ### Out-of-Scope Use
101
+
102
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
103
+ -->
104
+
105
+ <!--
106
+ ## Bias, Risks and Limitations
107
+
108
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
109
+ -->
110
+
111
+ <!--
112
+ ### Recommendations
113
+
114
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
115
+ -->
116
+
117
+ ## Training Details
118
+
119
+ ### Training Dataset
120
+
121
+ #### Unnamed Dataset
122
+
123
+ * Size: 2,733 training samples
124
+ * Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>sentence_2</code>
125
+ * Approximate statistics based on the first 1000 samples:
126
+ | | sentence_0 | sentence_1 | sentence_2 |
127
+ |:--------|:-------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
128
+ | type | string | string | string |
129
+ | details | <ul><li>min: 15 tokens</li><li>mean: 165.96 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 172.68 tokens</li><li>max: 493 tokens</li></ul> | <ul><li>min: 18 tokens</li><li>mean: 178.63 tokens</li><li>max: 512 tokens</li></ul> |
130
+ * Samples:
131
+ | sentence_0 | sentence_1 | sentence_2 |
132
+ |:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
133
+ | <code>A reclusive tailor in a coastal town discovers that the garments he sews can alter the wearers’ memories. After creating a dress for a grieving widow, he learns that she now recalls a husband who never died but instead vanished into the sea to become a lighthouse keeper. As word spreads, townspeople request clothes to reshape their pasts, leading to conflicting recollections that begin to overwrite shared history. The tailor becomes disturbed when he realizes his own childhood memories are changing without his consent. Seeking answers, he follows a trail of altered recollections to an abandoned lighthouse that appears only at dusk. Inside, he confronts a version of himself who claims to have been guiding the town’s fate for decades. The encounter ends with the tailor sewing one final garment, after which the lighthouse vanishes and the town awakens with no memory of its existence.</code> | <code>In a remote fishing village, a withdrawn cobbler discovers that the shoes he crafts have the power to rewrite the wearer’s memories. After making a pair for a sorrowful widower, he learns the man now remembers a wife who never died but instead sailed away to live on a drifting island. As rumors spread, villagers flock to him, requesting footwear to recast their personal histories, creating a tangle of contradictory recollections that erode the community’s shared past. The cobbler grows uneasy when fragments of his own early life begin to shift without his will. Determined to uncover the cause, he traces a series of warped memories to a weathered pier that appears only under the crimson light of sunset. There, he meets an older version of himself who insists he has been shaping the village’s destiny for generations. Their meeting concludes with the cobbler fashioning one last pair of shoes, after which the pier disappears and the villagers awaken with no awareness it ever existed.</code> | <code>A solitary cartographer in a mountain village discovers that the maps he draws subtly change the terrain for those who follow them. After sketching a route for a lost traveler, the man learns the path now leads to a valley no one remembers existing. As more villagers request maps to forgotten lakes or vanished orchards, the surrounding landscape becomes a patchwork of conflicting geographies. The cartographer begins to notice that certain landmarks from his own youth have disappeared from his charts without his hand ever touching them. Determined to understand, he plots a course to a canyon that appears only under a blood-red moon. There, he meets an older version of himself who claims to have been redrawing the world to keep it from collapsing. The film ends with the cartographer returning home to find the village perched on the edge of an unfamiliar sea, its people unaware they now live on an island.</code> |
134
+ | <code>A young prince becomes embroiled in a scandal after secret letters between him and a commoner are leaked to the press. The letters reveal not only their romantic relationship but also his disdain for the rigid traditions of the monarchy. As public outrage grows, the royal family attempts to suppress the story by arranging the prince’s engagement to a foreign princess. The prince resists, leading to a heated confrontation with his father, the king, who warns him that his actions threaten the stability of the crown. The commoner, meanwhile, is forced into hiding to escape the media frenzy. Ultimately, the prince delivers a televised speech, apologizing for the scandal but declaring his love for the commoner. The story ends ambiguously, with the royal family divided and the prince’s future uncertain.</code> | <code>A young musician finds himself at the center of a controversy after private emails between him and a journalist are published online. The emails expose not only their romantic relationship but also his frustrations with the exploitative practices of the music industry. As public backlash intensifies, his record label attempts to salvage his reputation by announcing a staged relationship between him and a famous actress. The musician rebels, leading to a heated argument with his manager, who warns him that his defiance could destroy his career. The journalist, meanwhile, goes into hiding to avoid relentless media scrutiny. In the end, the musician releases a live-streamed video apologizing for the controversy but professing his love for the journalist. The story concludes ambiguously, with the music industry divided and the musician’s future hanging in the balance.</code> | <code>A reclusive software developer in a futuristic megacity finds themselves at the center of controversy when fragments of their private neural diary are hacked and broadcast across the city’s media networks. The diary entries reveal not only their romantic involvement with a synthetic being—a highly controversial act—but also their disdain for the corporate-controlled government that funds the city’s tech infrastructure. As public debate spirals into chaos, the government pressures the developer to publicly denounce their relationship and assist in the capture of the synthetic, who has since gone into hiding. Torn between their feelings and the looming threat of exile, the developer agrees to a staged reconciliation with the authorities. However, during the live broadcast meant to rehabilitate their image, they use the platform to expose the government's corruption and affirm their loyalty to the synthetic. The broadcast ends abruptly, and the developer disappears into the sprawling unde...</code> |
135
+ | <code>In a remote village nestled within the mountains of ancient China, a young scholar named Lin discovers a hidden scroll in the ruins of an abandoned temple. The scroll reveals the location of a legendary artifact, the Jade Phoenix, said to possess the power to grant eternal wisdom. Lin embarks on a perilous journey to find the artifact, accompanied by a mysterious woman named Mei, who claims to be a descendant of the temple's guardians. Along the way, they encounter various challenges and adversaries, including a ruthless warlord who seeks the Jade Phoenix for his own nefarious purposes. After a series of trials, Lin and Mei finally reach the hidden chamber where the artifact is kept, only to find that the true power of the Jade Phoenix lies not in its physical form, but in the ancient knowledge it imparts, which transforms Lin into a wise and just leader, guiding his village to prosperity.</code> | <code>In a secluded hamlet deep within the forests of medieval Japan, a young samurai named Hiro discovers a concealed scroll in the remnants of a forgotten shrine. The scroll discloses the location of a mythical relic, the Crystal Crane, rumored to bestow infinite wisdom. Hiro sets out on a treacherous quest to locate the relic, joined by a enigmatic woman named Yuki, who asserts she is a descendant of the shrine's protectors. Throughout their journey, they face numerous obstacles and enemies, including a brutal daimyo who covets the Crystal Crane for his own malevolent designs. After enduring a series of arduous trials, Hiro and Yuki finally arrive at the secret chamber housing the relic, only to realize that the true power of the Crystal Crane resides in the ancient wisdom it imparts, which transforms Hiro into a wise and righteous leader, guiding his village to a new era of peace and prosperity.</code> | <code>In a futuristic metropolis, a young data analyst named Kira uncovers a hidden algorithm in the city's mainframe that predicts the emergence of a powerful AI known as the Oracle. The algorithm reveals that the Oracle can grant unparalleled insights into human behavior, potentially reshaping society. Kira, along with a rogue hacker named Zane, who claims to have once been part of the city's tech elite, sets out to find the core server where the Oracle is housed. They face numerous obstacles, including a corrupt government agency that seeks to control the Oracle for its own gain. After navigating through a labyrinth of digital and physical challenges, Kira and Zane finally reach the server room. However, they discover that the true power of the Oracle lies not in its predictions, but in its ability to foster empathy and understanding among people, leading Kira to use her newfound knowledge to advocate for a more compassionate and connected society.</code> |
136
+ * Loss: [<code>TripletLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#tripletloss) with these parameters:
137
+ ```json
138
+ {
139
+ "distance_metric": "TripletDistanceMetric.COSINE",
140
+ "triplet_margin": 0.35
141
+ }
142
+ ```
143
+
144
+ ### Training Hyperparameters
145
+ #### Non-Default Hyperparameters
146
+
147
+ - `per_device_train_batch_size`: 12
148
+ - `per_device_eval_batch_size`: 12
149
+ - `num_train_epochs`: 5
150
+ - `fp16`: True
151
+ - `multi_dataset_batch_sampler`: round_robin
152
+
153
+ #### All Hyperparameters
154
+ <details><summary>Click to expand</summary>
155
+
156
+ - `do_predict`: False
157
+ - `eval_strategy`: no
158
+ - `prediction_loss_only`: True
159
+ - `per_device_train_batch_size`: 12
160
+ - `per_device_eval_batch_size`: 12
161
+ - `gradient_accumulation_steps`: 1
162
+ - `eval_accumulation_steps`: None
163
+ - `torch_empty_cache_steps`: None
164
+ - `learning_rate`: 5e-05
165
+ - `weight_decay`: 0.0
166
+ - `adam_beta1`: 0.9
167
+ - `adam_beta2`: 0.999
168
+ - `adam_epsilon`: 1e-08
169
+ - `max_grad_norm`: 1
170
+ - `num_train_epochs`: 5
171
+ - `max_steps`: -1
172
+ - `lr_scheduler_type`: linear
173
+ - `lr_scheduler_kwargs`: None
174
+ - `warmup_ratio`: None
175
+ - `warmup_steps`: 0
176
+ - `log_level`: passive
177
+ - `log_level_replica`: warning
178
+ - `log_on_each_node`: True
179
+ - `logging_nan_inf_filter`: True
180
+ - `enable_jit_checkpoint`: False
181
+ - `save_on_each_node`: False
182
+ - `save_only_model`: False
183
+ - `restore_callback_states_from_checkpoint`: False
184
+ - `use_cpu`: False
185
+ - `seed`: 42
186
+ - `data_seed`: None
187
+ - `bf16`: False
188
+ - `fp16`: True
189
+ - `bf16_full_eval`: False
190
+ - `fp16_full_eval`: False
191
+ - `tf32`: None
192
+ - `local_rank`: -1
193
+ - `ddp_backend`: None
194
+ - `debug`: []
195
+ - `dataloader_drop_last`: False
196
+ - `dataloader_num_workers`: 0
197
+ - `dataloader_prefetch_factor`: None
198
+ - `disable_tqdm`: False
199
+ - `remove_unused_columns`: True
200
+ - `label_names`: None
201
+ - `load_best_model_at_end`: False
202
+ - `ignore_data_skip`: False
203
+ - `fsdp`: []
204
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
205
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
206
+ - `parallelism_config`: None
207
+ - `deepspeed`: None
208
+ - `label_smoothing_factor`: 0.0
209
+ - `optim`: adamw_torch
210
+ - `optim_args`: None
211
+ - `group_by_length`: False
212
+ - `length_column_name`: length
213
+ - `project`: huggingface
214
+ - `trackio_space_id`: trackio
215
+ - `ddp_find_unused_parameters`: None
216
+ - `ddp_bucket_cap_mb`: None
217
+ - `ddp_broadcast_buffers`: False
218
+ - `dataloader_pin_memory`: True
219
+ - `dataloader_persistent_workers`: False
220
+ - `skip_memory_metrics`: True
221
+ - `push_to_hub`: False
222
+ - `resume_from_checkpoint`: None
223
+ - `hub_model_id`: None
224
+ - `hub_strategy`: every_save
225
+ - `hub_private_repo`: None
226
+ - `hub_always_push`: False
227
+ - `hub_revision`: None
228
+ - `gradient_checkpointing`: False
229
+ - `gradient_checkpointing_kwargs`: None
230
+ - `include_for_metrics`: []
231
+ - `eval_do_concat_batches`: True
232
+ - `auto_find_batch_size`: False
233
+ - `full_determinism`: False
234
+ - `ddp_timeout`: 1800
235
+ - `torch_compile`: False
236
+ - `torch_compile_backend`: None
237
+ - `torch_compile_mode`: None
238
+ - `include_num_input_tokens_seen`: no
239
+ - `neftune_noise_alpha`: None
240
+ - `optim_target_modules`: None
241
+ - `batch_eval_metrics`: False
242
+ - `eval_on_start`: False
243
+ - `use_liger_kernel`: False
244
+ - `liger_kernel_config`: None
245
+ - `eval_use_gather_object`: False
246
+ - `average_tokens_across_devices`: True
247
+ - `use_cache`: False
248
+ - `prompts`: None
249
+ - `batch_sampler`: batch_sampler
250
+ - `multi_dataset_batch_sampler`: round_robin
251
+ - `router_mapping`: {}
252
+ - `learning_rate_mapping`: {}
253
+
254
+ </details>
255
+
256
+ ### Training Logs
257
+ | Epoch | Step | Training Loss |
258
+ |:------:|:----:|:-------------:|
259
+ | 2.1930 | 500 | 0.0153 |
260
+ | 4.3860 | 1000 | 0.0003 |
261
+
262
+
263
+ ### Framework Versions
264
+ - Python: 3.11.7
265
+ - Sentence Transformers: 5.2.2
266
+ - Transformers: 5.0.0
267
+ - PyTorch: 2.5.1+cu121
268
+ - Accelerate: 1.10.1
269
+ - Datasets: 4.2.0
270
+ - Tokenizers: 0.22.2
271
+
272
+ ## Citation
273
+
274
+ ### BibTeX
275
+
276
+ #### Sentence Transformers
277
+ ```bibtex
278
+ @inproceedings{reimers-2019-sentence-bert,
279
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
280
+ author = "Reimers, Nils and Gurevych, Iryna",
281
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
282
+ month = "11",
283
+ year = "2019",
284
+ publisher = "Association for Computational Linguistics",
285
+ url = "https://arxiv.org/abs/1908.10084",
286
+ }
287
+ ```
288
+
289
+ #### TripletLoss
290
+ ```bibtex
291
+ @misc{hermans2017defense,
292
+ title={In Defense of the Triplet Loss for Person Re-Identification},
293
+ author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
294
+ year={2017},
295
+ eprint={1703.07737},
296
+ archivePrefix={arXiv},
297
+ primaryClass={cs.CV}
298
+ }
299
+ ```
300
+
301
+ <!--
302
+ ## Glossary
303
+
304
+ *Clearly define terms in order to be accessible across audiences.*
305
+ -->
306
+
307
+ <!--
308
+ ## Model Card Authors
309
+
310
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
311
+ -->
312
+
313
+ <!--
314
+ ## Model Card Contact
315
+
316
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
317
+ -->
config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "dtype": "float32",
10
+ "eos_token_id": null,
11
+ "gradient_checkpointing": false,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 1024,
15
+ "id2label": {
16
+ "0": "LABEL_0"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "is_decoder": false,
21
+ "label2id": {
22
+ "LABEL_0": 0
23
+ },
24
+ "layer_norm_eps": 1e-12,
25
+ "max_position_embeddings": 512,
26
+ "model_type": "bert",
27
+ "num_attention_heads": 16,
28
+ "num_hidden_layers": 24,
29
+ "pad_token_id": 0,
30
+ "position_embedding_type": "absolute",
31
+ "tie_word_embeddings": true,
32
+ "transformers_version": "5.0.0",
33
+ "type_vocab_size": 2,
34
+ "use_cache": true,
35
+ "vocab_size": 30522
36
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "5.2.2",
4
+ "transformers": "5.0.0",
5
+ "pytorch": "2.5.1+cu121"
6
+ },
7
+ "model_type": "SentenceTransformer",
8
+ "prompts": {
9
+ "query": "",
10
+ "document": ""
11
+ },
12
+ "default_prompt_name": null,
13
+ "similarity_fn_name": "cosine"
14
+ }
final_model/1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 1024,
3
+ "pooling_mode_cls_token": true,
4
+ "pooling_mode_mean_tokens": false,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
final_model/README.md ADDED
@@ -0,0 +1,317 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - sentence-similarity
5
+ - feature-extraction
6
+ - dense
7
+ - generated_from_trainer
8
+ - dataset_size:2733
9
+ - loss:TripletLoss
10
+ base_model: BAAI/bge-large-en-v1.5
11
+ pipeline_tag: sentence-similarity
12
+ library_name: sentence-transformers
13
+ ---
14
+
15
+ # SentenceTransformer based on BAAI/bge-large-en-v1.5
16
+
17
+ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [BAAI/bge-large-en-v1.5](https://huggingface.co/BAAI/bge-large-en-v1.5). It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
18
+
19
+ ## Model Details
20
+
21
+ ### Model Description
22
+ - **Model Type:** Sentence Transformer
23
+ - **Base model:** [BAAI/bge-large-en-v1.5](https://huggingface.co/BAAI/bge-large-en-v1.5) <!-- at revision d4aa6901d3a41ba39fb536a557fa166f842b0e09 -->
24
+ - **Maximum Sequence Length:** 512 tokens
25
+ - **Output Dimensionality:** 1024 dimensions
26
+ - **Similarity Function:** Cosine Similarity
27
+ <!-- - **Training Dataset:** Unknown -->
28
+ <!-- - **Language:** Unknown -->
29
+ <!-- - **License:** Unknown -->
30
+
31
+ ### Model Sources
32
+
33
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
34
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
35
+ - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
36
+
37
+ ### Full Model Architecture
38
+
39
+ ```
40
+ SentenceTransformer(
41
+ (0): Transformer({'max_seq_length': 512, 'do_lower_case': True, 'architecture': 'BertModel'})
42
+ (1): Pooling({'word_embedding_dimension': 1024, '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})
43
+ (2): Normalize()
44
+ )
45
+ ```
46
+
47
+ ## Usage
48
+
49
+ ### Direct Usage (Sentence Transformers)
50
+
51
+ First install the Sentence Transformers library:
52
+
53
+ ```bash
54
+ pip install -U sentence-transformers
55
+ ```
56
+
57
+ Then you can load this model and run inference.
58
+ ```python
59
+ from sentence_transformers import SentenceTransformer
60
+
61
+ # Download from the 🤗 Hub
62
+ model = SentenceTransformer("sentence_transformers_model_id")
63
+ # Run inference
64
+ sentences = [
65
+ 'The weather is lovely today.',
66
+ "It's so sunny outside!",
67
+ 'He drove to the stadium.',
68
+ ]
69
+ embeddings = model.encode(sentences)
70
+ print(embeddings.shape)
71
+ # [3, 1024]
72
+
73
+ # Get the similarity scores for the embeddings
74
+ similarities = model.similarity(embeddings, embeddings)
75
+ print(similarities)
76
+ # tensor([[1.0000, 0.8407, 0.3897],
77
+ # [0.8407, 1.0000, 0.3653],
78
+ # [0.3897, 0.3653, 1.0000]])
79
+ ```
80
+
81
+ <!--
82
+ ### Direct Usage (Transformers)
83
+
84
+ <details><summary>Click to see the direct usage in Transformers</summary>
85
+
86
+ </details>
87
+ -->
88
+
89
+ <!--
90
+ ### Downstream Usage (Sentence Transformers)
91
+
92
+ You can finetune this model on your own dataset.
93
+
94
+ <details><summary>Click to expand</summary>
95
+
96
+ </details>
97
+ -->
98
+
99
+ <!--
100
+ ### Out-of-Scope Use
101
+
102
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
103
+ -->
104
+
105
+ <!--
106
+ ## Bias, Risks and Limitations
107
+
108
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
109
+ -->
110
+
111
+ <!--
112
+ ### Recommendations
113
+
114
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
115
+ -->
116
+
117
+ ## Training Details
118
+
119
+ ### Training Dataset
120
+
121
+ #### Unnamed Dataset
122
+
123
+ * Size: 2,733 training samples
124
+ * Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>sentence_2</code>
125
+ * Approximate statistics based on the first 1000 samples:
126
+ | | sentence_0 | sentence_1 | sentence_2 |
127
+ |:--------|:-------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------|
128
+ | type | string | string | string |
129
+ | details | <ul><li>min: 15 tokens</li><li>mean: 165.96 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 14 tokens</li><li>mean: 172.68 tokens</li><li>max: 493 tokens</li></ul> | <ul><li>min: 18 tokens</li><li>mean: 178.63 tokens</li><li>max: 512 tokens</li></ul> |
130
+ * Samples:
131
+ | sentence_0 | sentence_1 | sentence_2 |
132
+ |:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
133
+ | <code>A reclusive tailor in a coastal town discovers that the garments he sews can alter the wearers’ memories. After creating a dress for a grieving widow, he learns that she now recalls a husband who never died but instead vanished into the sea to become a lighthouse keeper. As word spreads, townspeople request clothes to reshape their pasts, leading to conflicting recollections that begin to overwrite shared history. The tailor becomes disturbed when he realizes his own childhood memories are changing without his consent. Seeking answers, he follows a trail of altered recollections to an abandoned lighthouse that appears only at dusk. Inside, he confronts a version of himself who claims to have been guiding the town’s fate for decades. The encounter ends with the tailor sewing one final garment, after which the lighthouse vanishes and the town awakens with no memory of its existence.</code> | <code>In a remote fishing village, a withdrawn cobbler discovers that the shoes he crafts have the power to rewrite the wearer’s memories. After making a pair for a sorrowful widower, he learns the man now remembers a wife who never died but instead sailed away to live on a drifting island. As rumors spread, villagers flock to him, requesting footwear to recast their personal histories, creating a tangle of contradictory recollections that erode the community’s shared past. The cobbler grows uneasy when fragments of his own early life begin to shift without his will. Determined to uncover the cause, he traces a series of warped memories to a weathered pier that appears only under the crimson light of sunset. There, he meets an older version of himself who insists he has been shaping the village’s destiny for generations. Their meeting concludes with the cobbler fashioning one last pair of shoes, after which the pier disappears and the villagers awaken with no awareness it ever existed.</code> | <code>A solitary cartographer in a mountain village discovers that the maps he draws subtly change the terrain for those who follow them. After sketching a route for a lost traveler, the man learns the path now leads to a valley no one remembers existing. As more villagers request maps to forgotten lakes or vanished orchards, the surrounding landscape becomes a patchwork of conflicting geographies. The cartographer begins to notice that certain landmarks from his own youth have disappeared from his charts without his hand ever touching them. Determined to understand, he plots a course to a canyon that appears only under a blood-red moon. There, he meets an older version of himself who claims to have been redrawing the world to keep it from collapsing. The film ends with the cartographer returning home to find the village perched on the edge of an unfamiliar sea, its people unaware they now live on an island.</code> |
134
+ | <code>A young prince becomes embroiled in a scandal after secret letters between him and a commoner are leaked to the press. The letters reveal not only their romantic relationship but also his disdain for the rigid traditions of the monarchy. As public outrage grows, the royal family attempts to suppress the story by arranging the prince’s engagement to a foreign princess. The prince resists, leading to a heated confrontation with his father, the king, who warns him that his actions threaten the stability of the crown. The commoner, meanwhile, is forced into hiding to escape the media frenzy. Ultimately, the prince delivers a televised speech, apologizing for the scandal but declaring his love for the commoner. The story ends ambiguously, with the royal family divided and the prince’s future uncertain.</code> | <code>A young musician finds himself at the center of a controversy after private emails between him and a journalist are published online. The emails expose not only their romantic relationship but also his frustrations with the exploitative practices of the music industry. As public backlash intensifies, his record label attempts to salvage his reputation by announcing a staged relationship between him and a famous actress. The musician rebels, leading to a heated argument with his manager, who warns him that his defiance could destroy his career. The journalist, meanwhile, goes into hiding to avoid relentless media scrutiny. In the end, the musician releases a live-streamed video apologizing for the controversy but professing his love for the journalist. The story concludes ambiguously, with the music industry divided and the musician’s future hanging in the balance.</code> | <code>A reclusive software developer in a futuristic megacity finds themselves at the center of controversy when fragments of their private neural diary are hacked and broadcast across the city’s media networks. The diary entries reveal not only their romantic involvement with a synthetic being—a highly controversial act—but also their disdain for the corporate-controlled government that funds the city’s tech infrastructure. As public debate spirals into chaos, the government pressures the developer to publicly denounce their relationship and assist in the capture of the synthetic, who has since gone into hiding. Torn between their feelings and the looming threat of exile, the developer agrees to a staged reconciliation with the authorities. However, during the live broadcast meant to rehabilitate their image, they use the platform to expose the government's corruption and affirm their loyalty to the synthetic. The broadcast ends abruptly, and the developer disappears into the sprawling unde...</code> |
135
+ | <code>In a remote village nestled within the mountains of ancient China, a young scholar named Lin discovers a hidden scroll in the ruins of an abandoned temple. The scroll reveals the location of a legendary artifact, the Jade Phoenix, said to possess the power to grant eternal wisdom. Lin embarks on a perilous journey to find the artifact, accompanied by a mysterious woman named Mei, who claims to be a descendant of the temple's guardians. Along the way, they encounter various challenges and adversaries, including a ruthless warlord who seeks the Jade Phoenix for his own nefarious purposes. After a series of trials, Lin and Mei finally reach the hidden chamber where the artifact is kept, only to find that the true power of the Jade Phoenix lies not in its physical form, but in the ancient knowledge it imparts, which transforms Lin into a wise and just leader, guiding his village to prosperity.</code> | <code>In a secluded hamlet deep within the forests of medieval Japan, a young samurai named Hiro discovers a concealed scroll in the remnants of a forgotten shrine. The scroll discloses the location of a mythical relic, the Crystal Crane, rumored to bestow infinite wisdom. Hiro sets out on a treacherous quest to locate the relic, joined by a enigmatic woman named Yuki, who asserts she is a descendant of the shrine's protectors. Throughout their journey, they face numerous obstacles and enemies, including a brutal daimyo who covets the Crystal Crane for his own malevolent designs. After enduring a series of arduous trials, Hiro and Yuki finally arrive at the secret chamber housing the relic, only to realize that the true power of the Crystal Crane resides in the ancient wisdom it imparts, which transforms Hiro into a wise and righteous leader, guiding his village to a new era of peace and prosperity.</code> | <code>In a futuristic metropolis, a young data analyst named Kira uncovers a hidden algorithm in the city's mainframe that predicts the emergence of a powerful AI known as the Oracle. The algorithm reveals that the Oracle can grant unparalleled insights into human behavior, potentially reshaping society. Kira, along with a rogue hacker named Zane, who claims to have once been part of the city's tech elite, sets out to find the core server where the Oracle is housed. They face numerous obstacles, including a corrupt government agency that seeks to control the Oracle for its own gain. After navigating through a labyrinth of digital and physical challenges, Kira and Zane finally reach the server room. However, they discover that the true power of the Oracle lies not in its predictions, but in its ability to foster empathy and understanding among people, leading Kira to use her newfound knowledge to advocate for a more compassionate and connected society.</code> |
136
+ * Loss: [<code>TripletLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#tripletloss) with these parameters:
137
+ ```json
138
+ {
139
+ "distance_metric": "TripletDistanceMetric.COSINE",
140
+ "triplet_margin": 0.35
141
+ }
142
+ ```
143
+
144
+ ### Training Hyperparameters
145
+ #### Non-Default Hyperparameters
146
+
147
+ - `per_device_train_batch_size`: 12
148
+ - `per_device_eval_batch_size`: 12
149
+ - `num_train_epochs`: 5
150
+ - `fp16`: True
151
+ - `multi_dataset_batch_sampler`: round_robin
152
+
153
+ #### All Hyperparameters
154
+ <details><summary>Click to expand</summary>
155
+
156
+ - `do_predict`: False
157
+ - `eval_strategy`: no
158
+ - `prediction_loss_only`: True
159
+ - `per_device_train_batch_size`: 12
160
+ - `per_device_eval_batch_size`: 12
161
+ - `gradient_accumulation_steps`: 1
162
+ - `eval_accumulation_steps`: None
163
+ - `torch_empty_cache_steps`: None
164
+ - `learning_rate`: 5e-05
165
+ - `weight_decay`: 0.0
166
+ - `adam_beta1`: 0.9
167
+ - `adam_beta2`: 0.999
168
+ - `adam_epsilon`: 1e-08
169
+ - `max_grad_norm`: 1
170
+ - `num_train_epochs`: 5
171
+ - `max_steps`: -1
172
+ - `lr_scheduler_type`: linear
173
+ - `lr_scheduler_kwargs`: None
174
+ - `warmup_ratio`: None
175
+ - `warmup_steps`: 0
176
+ - `log_level`: passive
177
+ - `log_level_replica`: warning
178
+ - `log_on_each_node`: True
179
+ - `logging_nan_inf_filter`: True
180
+ - `enable_jit_checkpoint`: False
181
+ - `save_on_each_node`: False
182
+ - `save_only_model`: False
183
+ - `restore_callback_states_from_checkpoint`: False
184
+ - `use_cpu`: False
185
+ - `seed`: 42
186
+ - `data_seed`: None
187
+ - `bf16`: False
188
+ - `fp16`: True
189
+ - `bf16_full_eval`: False
190
+ - `fp16_full_eval`: False
191
+ - `tf32`: None
192
+ - `local_rank`: -1
193
+ - `ddp_backend`: None
194
+ - `debug`: []
195
+ - `dataloader_drop_last`: False
196
+ - `dataloader_num_workers`: 0
197
+ - `dataloader_prefetch_factor`: None
198
+ - `disable_tqdm`: False
199
+ - `remove_unused_columns`: True
200
+ - `label_names`: None
201
+ - `load_best_model_at_end`: False
202
+ - `ignore_data_skip`: False
203
+ - `fsdp`: []
204
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
205
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
206
+ - `parallelism_config`: None
207
+ - `deepspeed`: None
208
+ - `label_smoothing_factor`: 0.0
209
+ - `optim`: adamw_torch
210
+ - `optim_args`: None
211
+ - `group_by_length`: False
212
+ - `length_column_name`: length
213
+ - `project`: huggingface
214
+ - `trackio_space_id`: trackio
215
+ - `ddp_find_unused_parameters`: None
216
+ - `ddp_bucket_cap_mb`: None
217
+ - `ddp_broadcast_buffers`: False
218
+ - `dataloader_pin_memory`: True
219
+ - `dataloader_persistent_workers`: False
220
+ - `skip_memory_metrics`: True
221
+ - `push_to_hub`: False
222
+ - `resume_from_checkpoint`: None
223
+ - `hub_model_id`: None
224
+ - `hub_strategy`: every_save
225
+ - `hub_private_repo`: None
226
+ - `hub_always_push`: False
227
+ - `hub_revision`: None
228
+ - `gradient_checkpointing`: False
229
+ - `gradient_checkpointing_kwargs`: None
230
+ - `include_for_metrics`: []
231
+ - `eval_do_concat_batches`: True
232
+ - `auto_find_batch_size`: False
233
+ - `full_determinism`: False
234
+ - `ddp_timeout`: 1800
235
+ - `torch_compile`: False
236
+ - `torch_compile_backend`: None
237
+ - `torch_compile_mode`: None
238
+ - `include_num_input_tokens_seen`: no
239
+ - `neftune_noise_alpha`: None
240
+ - `optim_target_modules`: None
241
+ - `batch_eval_metrics`: False
242
+ - `eval_on_start`: False
243
+ - `use_liger_kernel`: False
244
+ - `liger_kernel_config`: None
245
+ - `eval_use_gather_object`: False
246
+ - `average_tokens_across_devices`: True
247
+ - `use_cache`: False
248
+ - `prompts`: None
249
+ - `batch_sampler`: batch_sampler
250
+ - `multi_dataset_batch_sampler`: round_robin
251
+ - `router_mapping`: {}
252
+ - `learning_rate_mapping`: {}
253
+
254
+ </details>
255
+
256
+ ### Training Logs
257
+ | Epoch | Step | Training Loss |
258
+ |:------:|:----:|:-------------:|
259
+ | 2.1930 | 500 | 0.0153 |
260
+ | 4.3860 | 1000 | 0.0003 |
261
+
262
+
263
+ ### Framework Versions
264
+ - Python: 3.11.7
265
+ - Sentence Transformers: 5.2.2
266
+ - Transformers: 5.0.0
267
+ - PyTorch: 2.5.1+cu121
268
+ - Accelerate: 1.10.1
269
+ - Datasets: 4.2.0
270
+ - Tokenizers: 0.22.2
271
+
272
+ ## Citation
273
+
274
+ ### BibTeX
275
+
276
+ #### Sentence Transformers
277
+ ```bibtex
278
+ @inproceedings{reimers-2019-sentence-bert,
279
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
280
+ author = "Reimers, Nils and Gurevych, Iryna",
281
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
282
+ month = "11",
283
+ year = "2019",
284
+ publisher = "Association for Computational Linguistics",
285
+ url = "https://arxiv.org/abs/1908.10084",
286
+ }
287
+ ```
288
+
289
+ #### TripletLoss
290
+ ```bibtex
291
+ @misc{hermans2017defense,
292
+ title={In Defense of the Triplet Loss for Person Re-Identification},
293
+ author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
294
+ year={2017},
295
+ eprint={1703.07737},
296
+ archivePrefix={arXiv},
297
+ primaryClass={cs.CV}
298
+ }
299
+ ```
300
+
301
+ <!--
302
+ ## Glossary
303
+
304
+ *Clearly define terms in order to be accessible across audiences.*
305
+ -->
306
+
307
+ <!--
308
+ ## Model Card Authors
309
+
310
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
311
+ -->
312
+
313
+ <!--
314
+ ## Model Card Contact
315
+
316
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
317
+ -->
final_model/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "BertModel"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": null,
8
+ "classifier_dropout": null,
9
+ "dtype": "float32",
10
+ "eos_token_id": null,
11
+ "gradient_checkpointing": false,
12
+ "hidden_act": "gelu",
13
+ "hidden_dropout_prob": 0.1,
14
+ "hidden_size": 1024,
15
+ "id2label": {
16
+ "0": "LABEL_0"
17
+ },
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 4096,
20
+ "is_decoder": false,
21
+ "label2id": {
22
+ "LABEL_0": 0
23
+ },
24
+ "layer_norm_eps": 1e-12,
25
+ "max_position_embeddings": 512,
26
+ "model_type": "bert",
27
+ "num_attention_heads": 16,
28
+ "num_hidden_layers": 24,
29
+ "pad_token_id": 0,
30
+ "position_embedding_type": "absolute",
31
+ "tie_word_embeddings": true,
32
+ "transformers_version": "5.0.0",
33
+ "type_vocab_size": 2,
34
+ "use_cache": true,
35
+ "vocab_size": 30522
36
+ }
final_model/config_sentence_transformers.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "5.2.2",
4
+ "transformers": "5.0.0",
5
+ "pytorch": "2.5.1+cu121"
6
+ },
7
+ "model_type": "SentenceTransformer",
8
+ "prompts": {
9
+ "query": "",
10
+ "document": ""
11
+ },
12
+ "default_prompt_name": null,
13
+ "similarity_fn_name": "cosine"
14
+ }
final_model/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27f46a08374a161b92eb699c69e83072907866f72358372b5af7389324957135
3
+ size 358088704
final_model/modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ },
14
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Normalize",
18
+ "type": "sentence_transformers.models.Normalize"
19
+ }
20
+ ]
final_model/sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": true
4
+ }
final_model/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
final_model/tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "clean_up_tokenization_spaces": true,
4
+ "cls_token": "[CLS]",
5
+ "do_basic_tokenize": true,
6
+ "do_lower_case": true,
7
+ "is_local": false,
8
+ "mask_token": "[MASK]",
9
+ "model_max_length": 512,
10
+ "never_split": null,
11
+ "pad_token": "[PAD]",
12
+ "sep_token": "[SEP]",
13
+ "strip_accents": null,
14
+ "tokenize_chinese_chars": true,
15
+ "tokenizer_class": "BertTokenizer",
16
+ "unk_token": "[UNK]"
17
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0510f8ed7fce73b4c2f522795a2bb90f7875095b6572129161ad3f50cef0cef
3
+ size 358875136
modules.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ },
14
+ {
15
+ "idx": 2,
16
+ "name": "2",
17
+ "path": "2_Normalize",
18
+ "type": "sentence_transformers.models.Normalize"
19
+ }
20
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": true
4
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "clean_up_tokenization_spaces": true,
4
+ "cls_token": "[CLS]",
5
+ "do_basic_tokenize": true,
6
+ "do_lower_case": true,
7
+ "is_local": false,
8
+ "mask_token": "[MASK]",
9
+ "model_max_length": 512,
10
+ "never_split": null,
11
+ "pad_token": "[PAD]",
12
+ "sep_token": "[SEP]",
13
+ "strip_accents": null,
14
+ "tokenize_chinese_chars": true,
15
+ "tokenizer_class": "BertTokenizer",
16
+ "unk_token": "[UNK]"
17
+ }
training_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "BAAI/bge-large-en-v1.5",
3
+ "train_file": "/home/pruthwikmishra/bge_large/data/final_train_all.jsonl",
4
+ "dev_file": "/home/pruthwikmishra/bge_large/data/dev.jsonl",
5
+ "output_dir": "/home/pruthwikmishra/bge_large/checkpoints_expanded",
6
+ "log_dir": "/home/pruthwikmishra/bge_large/logs",
7
+ "num_epochs": 5,
8
+ "batch_size": 12,
9
+ "learning_rate": 1.5e-05,
10
+ "warmup_ratio": 0.1,
11
+ "margin": 0.35,
12
+ "save_steps": 100,
13
+ "eval_steps": 100,
14
+ "logging_steps": 20
15
+ }