Hugging Face's logo

CuongCao
/
oe-bge-reranker-v2-m3-ft-v4

Text Ranking
sentence-transformers
Safetensors
xlm-roberta
cross-encoder
reranker
Generated from Trainer
dataset_size:5400
loss:BinaryCrossEntropyLoss
Eval Results (legacy)
text-embeddings-inference
Model card Files
xet
 Community
CuongCao commited on
Commit
44b09af
·
verified ·
1 Parent(s): 5220360

Upload folder using huggingface_hub

Browse files
Files changed (9) hide show
  1. .gitattributes +1 -0
  2. README.md +538 -0
  3. config.json +36 -0
  4. config_sentence_transformers.json +11 -0
  5. model.safetensors +3 -0
  6. modules.json +8 -0
  7. sentence_bert_config.json +10 -0
  8. tokenizer.json +3 -0
  9. tokenizer_config.json +16 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,538 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - sentence-transformers
4
+ - cross-encoder
5
+ - reranker
6
+ - generated_from_trainer
7
+ - dataset_size:5400
8
+ - loss:BinaryCrossEntropyLoss
9
+ base_model: BAAI/bge-reranker-v2-m3
10
+ pipeline_tag: text-ranking
11
+ library_name: sentence-transformers
12
+ metrics:
13
+ - map
14
+ - mrr@10
15
+ - ndcg@10
16
+ model-index:
17
+ - name: CrossEncoder based on BAAI/bge-reranker-v2-m3
18
+ results:
19
+ - task:
20
+ type: cross-encoder-reranking
21
+ name: Cross Encoder Reranking
22
+ dataset:
23
+ name: support rerank eval
24
+ type: support-rerank-eval
25
+ metrics:
26
+ - type: map
27
+ value: 0.8933
28
+ name: Map
29
+ - type: mrr@10
30
+ value: 0.8933
31
+ name: Mrr@10
32
+ - type: ndcg@10
33
+ value: 0.9203
34
+ name: Ndcg@10
35
+ ---
36
+
37
+ # CrossEncoder based on BAAI/bge-reranker-v2-m3
38
+
39
+ This is a [Cross Encoder](https://www.sbert.net/docs/cross_encoder/usage/usage.html) model finetuned from [BAAI/bge-reranker-v2-m3](https://huggingface.co/BAAI/bge-reranker-v2-m3) using the [sentence-transformers](https://www.SBERT.net) library. It computes scores for pairs of texts, which can be used for text reranking and semantic search.
40
+
41
+ ## Model Details
42
+
43
+ ### Model Description
44
+ - **Model Type:** Cross Encoder
45
+ - **Base model:** [BAAI/bge-reranker-v2-m3](https://huggingface.co/BAAI/bge-reranker-v2-m3) <!-- at revision 953dc6f6f85a1b2dbfca4c34a2796e7dde08d41e -->
46
+ - **Maximum Sequence Length:** 8192 tokens
47
+ - **Number of Output Labels:** 1 label
48
+ - **Supported Modality:** Text
49
+ <!-- - **Training Dataset:** Unknown -->
50
+ <!-- - **Language:** Unknown -->
51
+ <!-- - **License:** Unknown -->
52
+
53
+ ### Model Sources
54
+
55
+ - **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
56
+ - **Documentation:** [Cross Encoder Documentation](https://www.sbert.net/docs/cross_encoder/usage/usage.html)
57
+ - **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
58
+ - **Hugging Face:** [Cross Encoders on Hugging Face](https://huggingface.co/models?library=sentence-transformers&other=cross-encoder)
59
+
60
+ ### Full Model Architecture
61
+
62
+ ```
63
+ CrossEncoder(
64
+ (0): Transformer({'transformer_task': 'sequence-classification', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'logits'}}, 'module_output_name': 'scores', 'architecture': 'XLMRobertaForSequenceClassification'})
65
+ )
66
+ ```
67
+
68
+ ## Usage
69
+
70
+ ### Direct Usage (Sentence Transformers)
71
+
72
+ First install the Sentence Transformers library:
73
+
74
+ ```bash
75
+ pip install -U sentence-transformers
76
+ ```
77
+
78
+ Then you can load this model and run inference.
79
+ ```python
80
+ from sentence_transformers import CrossEncoder
81
+
82
+ # Download from the 🤗 Hub
83
+ model = CrossEncoder("cross_encoder_model_id")
84
+ # Get scores for pairs of inputs
85
+ pairs = [
86
+ ['Hi, is there a way to show an image once the personalization toggle is on?', "Hi there, This is Hazel from Tepo Support team. I hope your day is going great! Yes, you can ultilize our conditional settings to show an image based on the toggle's selection, as demonstrated on our demo product [here](https://tepo.app/products/conditional-logic) <3 For step-by-step instructions on how to set up conditional settings on your end, please check out our [guide](https://support.tepo.app/en/article/set-up-conditional-logic-1sp9tac/). If you prefer watching than reading, feel free to take a look at our tutorial video [here](https://www.youtube.com/watch?v=7voAaJ2lzkc)"],
87
+ ['Hi, is there a way to show an image once the personalization toggle is on?', "Hello there, I apologize for getting back to you a bit later, as we've been away during non-working hours. I hope your weekend went great! For your information, we have a feature to change the product image based on the option's selection. I have enabled this feature for your store, so feel free to follow our step-by-step instructions [here](https://support.tepo.app/en/article/change-product-images-based-on-selected-options-1831rew/) to set up this feature for your store. Please let me know if you encounter any difficulties. I am more than happy to help!"],
88
+ ['Hi, is there a way to show an image once the personalization toggle is on?', "Hello, Thank you for contacting Tepo Support team! I hope you're having a wonderful day! To enlarge the size of the image swatches, you can refer to our guide [here](https://support.tepo.app/en/article/customize-style-for-app-options-using-css-26kmin/) for step-by-step instructions. This will allow you to make the necessary adjustments on your end. Feel free to give it a try, and if you encounter any difficulties or if coding isn't your forte, please don't hesitate to share the product URL with me. I'll be more than happy to assist further!"],
89
+ ['Hi, is there a way to show an image once the personalization toggle is on?', 'Good day~ This is Robin from Tepo Support team! I hope your day has been amazing so far! To make sure I’m on the same page, when you say getting the older version of the option set, are you referring to undoing some changes after you’ve already hit save? If I’ve misunderstood anything, please don’t hesitate to correct me. I want to make sure I assist you as best as I can!'],
90
+ ['Hi, is there a way to show an image once the personalization toggle is on?', "> Hello there, > This is Hazel from Tepo Support team. I hope your day is going great! > Thanks for sharing your concern with us. Regarding the packing slip, we do have a step-by-step instruction [here](https://support.tepo.app/en/article/how-to-show-options-in-shopify-packing-slip-1dyay7y/) that you can follow more easily on your end. Feel free to give it a try on your end to show Tepo options in your packing slip In Shopify, a courier slip is also known as a packing slip. I've seen that Hazel has sent you the instruction to add the code into the packing slip template to show the TEPO options already. May I ask if you give it a try on your end yet?"],
91
+ ]
92
+ scores = model.predict(pairs)
93
+ print(scores)
94
+ # [8.1272e-01 4.0087e-02 3.0242e-05 2.8633e-05 2.9541e-05]
95
+
96
+ # Or rank different texts based on similarity to a single text
97
+ ranks = model.rank(
98
+ 'Hi, is there a way to show an image once the personalization toggle is on?',
99
+ [
100
+ "Hi there, This is Hazel from Tepo Support team. I hope your day is going great! Yes, you can ultilize our conditional settings to show an image based on the toggle's selection, as demonstrated on our demo product [here](https://tepo.app/products/conditional-logic) <3 For step-by-step instructions on how to set up conditional settings on your end, please check out our [guide](https://support.tepo.app/en/article/set-up-conditional-logic-1sp9tac/). If you prefer watching than reading, feel free to take a look at our tutorial video [here](https://www.youtube.com/watch?v=7voAaJ2lzkc)",
101
+ "Hello there, I apologize for getting back to you a bit later, as we've been away during non-working hours. I hope your weekend went great! For your information, we have a feature to change the product image based on the option's selection. I have enabled this feature for your store, so feel free to follow our step-by-step instructions [here](https://support.tepo.app/en/article/change-product-images-based-on-selected-options-1831rew/) to set up this feature for your store. Please let me know if you encounter any difficulties. I am more than happy to help!",
102
+ "Hello, Thank you for contacting Tepo Support team! I hope you're having a wonderful day! To enlarge the size of the image swatches, you can refer to our guide [here](https://support.tepo.app/en/article/customize-style-for-app-options-using-css-26kmin/) for step-by-step instructions. This will allow you to make the necessary adjustments on your end. Feel free to give it a try, and if you encounter any difficulties or if coding isn't your forte, please don't hesitate to share the product URL with me. I'll be more than happy to assist further!",
103
+ 'Good day~ This is Robin from Tepo Support team! I hope your day has been amazing so far! To make sure I’m on the same page, when you say getting the older version of the option set, are you referring to undoing some changes after you’ve already hit save? If I’ve misunderstood anything, please don’t hesitate to correct me. I want to make sure I assist you as best as I can!',
104
+ "> Hello there, > This is Hazel from Tepo Support team. I hope your day is going great! > Thanks for sharing your concern with us. Regarding the packing slip, we do have a step-by-step instruction [here](https://support.tepo.app/en/article/how-to-show-options-in-shopify-packing-slip-1dyay7y/) that you can follow more easily on your end. Feel free to give it a try on your end to show Tepo options in your packing slip In Shopify, a courier slip is also known as a packing slip. I've seen that Hazel has sent you the instruction to add the code into the packing slip template to show the TEPO options already. May I ask if you give it a try on your end yet?",
105
+ ]
106
+ )
107
+ # [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]
108
+ ```
109
+
110
+ <!--
111
+ ### Direct Usage (Transformers)
112
+
113
+ <details><summary>Click to see the direct usage in Transformers</summary>
114
+
115
+ </details>
116
+ -->
117
+
118
+ <!--
119
+ ### Downstream Usage (Sentence Transformers)
120
+
121
+ You can finetune this model on your own dataset.
122
+
123
+ <details><summary>Click to expand</summary>
124
+
125
+ </details>
126
+ -->
127
+
128
+ <!--
129
+ ### Out-of-Scope Use
130
+
131
+ *List how the model may foreseeably be misused and address what users ought not to do with the model.*
132
+ -->
133
+
134
+ ## Evaluation
135
+
136
+ ### Metrics
137
+
138
+ #### Cross Encoder Reranking
139
+
140
+ * Dataset: `support-rerank-eval`
141
+ * Evaluated with [<code>CrossEncoderRerankingEvaluator</code>](https://sbert.net/docs/package_reference/cross_encoder/evaluation.html#sentence_transformers.cross_encoder.evaluation.CrossEncoderRerankingEvaluator) with these parameters:
142
+ ```json
143
+ {
144
+ "at_k": 10,
145
+ "always_rerank_positives": false
146
+ }
147
+ ```
148
+
149
+ | Metric | Value |
150
+ |:------------|:---------------------|
151
+ | map | 0.8933 (-0.1067) |
152
+ | mrr@10 | 0.8933 (-0.1067) |
153
+ | **ndcg@10** | **0.9203 (-0.0797)** |
154
+
155
+ <!--
156
+ ## Bias, Risks and Limitations
157
+
158
+ *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
159
+ -->
160
+
161
+ <!--
162
+ ### Recommendations
163
+
164
+ *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
165
+ -->
166
+
167
+ ## Training Details
168
+
169
+ ### Training Dataset
170
+
171
+ #### Unnamed Dataset
172
+
173
+ * Size: 5,400 training samples
174
+ * Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>label</code>
175
+ * Approximate statistics based on the first 1000 samples:
176
+ | | sentence1 | sentence2 | label |
177
+ |:--------|:-----------------------------------------------------------------------------------|:------------------------------------------------------------------------------------|:---------------------------------------------------------------|
178
+ | type | string | string | float |
179
+ | details | <ul><li>min: 9 tokens</li><li>mean: 55.62 tokens</li><li>max: 487 tokens</li></ul> | <ul><li>min: 36 tokens</li><li>mean: 169.9 tokens</li><li>max: 968 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 0.17</li><li>max: 1.0</li></ul> |
180
+ * Samples:
181
+ | sentence1 | sentence2 | label |
182
+ |:----------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------|
183
+ | <code>Hi, is there a way to show an image once the personalization toggle is on?</code> | <code>Hi there, This is Hazel from Tepo Support team. I hope your day is going great! Yes, you can ultilize our conditional settings to show an image based on the toggle's selection, as demonstrated on our demo product [here](https://tepo.app/products/conditional-logic) <3 For step-by-step instructions on how to set up conditional settings on your end, please check out our [guide](https://support.tepo.app/en/article/set-up-conditional-logic-1sp9tac/). If you prefer watching than reading, feel free to take a look at our tutorial video [here](https://www.youtube.com/watch?v=7voAaJ2lzkc)</code> | <code>1.0</code> |
184
+ | <code>Hi, is there a way to show an image once the personalization toggle is on?</code> | <code>Hello there, I apologize for getting back to you a bit later, as we've been away during non-working hours. I hope your weekend went great! For your information, we have a feature to change the product image based on the option's selection. I have enabled this feature for your store, so feel free to follow our step-by-step instructions [here](https://support.tepo.app/en/article/change-product-images-based-on-selected-options-1831rew/) to set up this feature for your store. Please let me know if you encounter any difficulties. I am more than happy to help!</code> | <code>0.0</code> |
185
+ | <code>Hi, is there a way to show an image once the personalization toggle is on?</code> | <code>Hello, Thank you for contacting Tepo Support team! I hope you're having a wonderful day! To enlarge the size of the image swatches, you can refer to our guide [here](https://support.tepo.app/en/article/customize-style-for-app-options-using-css-26kmin/) for step-by-step instructions. This will allow you to make the necessary adjustments on your end. Feel free to give it a try, and if you encounter any difficulties or if coding isn't your forte, please don't hesitate to share the product URL with me. I'll be more than happy to assist further!</code> | <code>0.0</code> |
186
+ * Loss: [<code>BinaryCrossEntropyLoss</code>](https://sbert.net/docs/package_reference/cross_encoder/losses.html#binarycrossentropyloss) with these parameters:
187
+ ```json
188
+ {
189
+ "activation_fn": "torch.nn.modules.linear.Identity",
190
+ "pos_weight": 5
191
+ }
192
+ ```
193
+
194
+ ### Training Hyperparameters
195
+ #### Non-Default Hyperparameters
196
+
197
+ - `eval_strategy`: steps
198
+ - `per_device_train_batch_size`: 1
199
+ - `gradient_accumulation_steps`: 32
200
+ - `learning_rate`: 2e-05
201
+ - `weight_decay`: 0.1
202
+ - `num_train_epochs`: 5
203
+ - `warmup_steps`: 0.25
204
+ - `seed`: 67
205
+ - `fp16`: True
206
+ - `load_best_model_at_end`: True
207
+
208
+ #### All Hyperparameters
209
+ <details><summary>Click to expand</summary>
210
+
211
+ - `do_predict`: False
212
+ - `eval_strategy`: steps
213
+ - `prediction_loss_only`: True
214
+ - `per_device_train_batch_size`: 1
215
+ - `per_device_eval_batch_size`: 8
216
+ - `gradient_accumulation_steps`: 32
217
+ - `eval_accumulation_steps`: None
218
+ - `torch_empty_cache_steps`: None
219
+ - `learning_rate`: 2e-05
220
+ - `weight_decay`: 0.1
221
+ - `adam_beta1`: 0.9
222
+ - `adam_beta2`: 0.999
223
+ - `adam_epsilon`: 1e-08
224
+ - `max_grad_norm`: 1.0
225
+ - `num_train_epochs`: 5
226
+ - `max_steps`: -1
227
+ - `lr_scheduler_type`: linear
228
+ - `lr_scheduler_kwargs`: None
229
+ - `warmup_ratio`: None
230
+ - `warmup_steps`: 0.25
231
+ - `log_level`: passive
232
+ - `log_level_replica`: warning
233
+ - `log_on_each_node`: True
234
+ - `logging_nan_inf_filter`: True
235
+ - `enable_jit_checkpoint`: False
236
+ - `save_on_each_node`: False
237
+ - `save_only_model`: False
238
+ - `restore_callback_states_from_checkpoint`: False
239
+ - `use_cpu`: False
240
+ - `seed`: 67
241
+ - `data_seed`: None
242
+ - `bf16`: False
243
+ - `fp16`: True
244
+ - `bf16_full_eval`: False
245
+ - `fp16_full_eval`: False
246
+ - `tf32`: None
247
+ - `local_rank`: -1
248
+ - `ddp_backend`: None
249
+ - `debug`: []
250
+ - `dataloader_drop_last`: False
251
+ - `dataloader_num_workers`: 0
252
+ - `dataloader_prefetch_factor`: None
253
+ - `disable_tqdm`: True
254
+ - `remove_unused_columns`: True
255
+ - `label_names`: None
256
+ - `load_best_model_at_end`: True
257
+ - `ignore_data_skip`: False
258
+ - `fsdp`: []
259
+ - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
260
+ - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
261
+ - `parallelism_config`: None
262
+ - `deepspeed`: None
263
+ - `label_smoothing_factor`: 0.0
264
+ - `optim`: adamw_torch_fused
265
+ - `optim_args`: None
266
+ - `group_by_length`: False
267
+ - `length_column_name`: length
268
+ - `project`: huggingface
269
+ - `trackio_space_id`: trackio
270
+ - `ddp_find_unused_parameters`: None
271
+ - `ddp_bucket_cap_mb`: None
272
+ - `ddp_broadcast_buffers`: False
273
+ - `dataloader_pin_memory`: True
274
+ - `dataloader_persistent_workers`: False
275
+ - `skip_memory_metrics`: True
276
+ - `push_to_hub`: False
277
+ - `resume_from_checkpoint`: None
278
+ - `hub_model_id`: None
279
+ - `hub_strategy`: every_save
280
+ - `hub_private_repo`: None
281
+ - `hub_always_push`: False
282
+ - `hub_revision`: None
283
+ - `gradient_checkpointing`: False
284
+ - `gradient_checkpointing_kwargs`: None
285
+ - `include_for_metrics`: []
286
+ - `eval_do_concat_batches`: True
287
+ - `auto_find_batch_size`: False
288
+ - `full_determinism`: False
289
+ - `ddp_timeout`: 1800
290
+ - `torch_compile`: False
291
+ - `torch_compile_backend`: None
292
+ - `torch_compile_mode`: None
293
+ - `include_num_input_tokens_seen`: no
294
+ - `neftune_noise_alpha`: None
295
+ - `optim_target_modules`: None
296
+ - `batch_eval_metrics`: False
297
+ - `eval_on_start`: False
298
+ - `use_liger_kernel`: False
299
+ - `liger_kernel_config`: None
300
+ - `eval_use_gather_object`: False
301
+ - `average_tokens_across_devices`: True
302
+ - `use_cache`: False
303
+ - `prompts`: None
304
+ - `batch_sampler`: batch_sampler
305
+ - `multi_dataset_batch_sampler`: proportional
306
+ - `router_mapping`: {}
307
+ - `learning_rate_mapping`: {}
308
+
309
+ </details>
310
+
311
+ ### Training Logs
312
+ <details><summary>Click to expand</summary>
313
+
314
+ | Epoch | Step | Training Loss | support-rerank-eval_ndcg@10 |
315
+ |:----------:|:-------:|:-------------:|:---------------------------:|
316
+ | -1 | -1 | - | 0.8746 (-0.1254) |
317
+ | 0.0059 | 1 | 1.4427 | - |
318
+ | 0.0296 | 5 | 1.0900 | - |
319
+ | 0.0593 | 10 | 1.8387 | - |
320
+ | 0.0889 | 15 | 1.1265 | - |
321
+ | 0.1185 | 20 | 0.9076 | - |
322
+ | 0.1481 | 25 | 0.8732 | - |
323
+ | 0.1778 | 30 | 0.7739 | - |
324
+ | 0.2074 | 35 | 1.1407 | - |
325
+ | 0.2370 | 40 | 0.7630 | - |
326
+ | 0.2667 | 45 | 0.7938 | - |
327
+ | 0.2963 | 50 | 0.6192 | - |
328
+ | 0.3259 | 55 | 0.7706 | - |
329
+ | 0.3556 | 60 | 0.9821 | - |
330
+ | 0.3852 | 65 | 0.7201 | - |
331
+ | 0.4148 | 70 | 0.6338 | - |
332
+ | 0.4444 | 75 | 0.8196 | - |
333
+ | 0.4741 | 80 | 0.6060 | - |
334
+ | 0.5037 | 85 | 0.8498 | - |
335
+ | 0.5333 | 90 | 0.9620 | - |
336
+ | 0.5630 | 95 | 0.9047 | - |
337
+ | 0.5926 | 100 | 0.8654 | 0.8914 (-0.1086) |
338
+ | 0.6222 | 105 | 0.6017 | - |
339
+ | 0.6519 | 110 | 0.6177 | - |
340
+ | 0.6815 | 115 | 0.5741 | - |
341
+ | 0.7111 | 120 | 0.8225 | - |
342
+ | 0.7407 | 125 | 0.7881 | - |
343
+ | 0.7704 | 130 | 0.8100 | - |
344
+ | 0.8 | 135 | 0.7767 | - |
345
+ | 0.8296 | 140 | 0.9624 | - |
346
+ | 0.8593 | 145 | 0.7971 | - |
347
+ | 0.8889 | 150 | 0.5490 | - |
348
+ | 0.9185 | 155 | 0.5172 | - |
349
+ | 0.9481 | 160 | 0.5474 | - |
350
+ | 0.9778 | 165 | 0.7273 | - |
351
+ | 1.0059 | 170 | 0.6546 | - |
352
+ | 1.0356 | 175 | 0.4965 | - |
353
+ | 1.0652 | 180 | 0.4726 | - |
354
+ | 1.0948 | 185 | 0.7178 | - |
355
+ | 1.1244 | 190 | 0.5441 | - |
356
+ | 1.1541 | 195 | 0.5818 | - |
357
+ | 1.1837 | 200 | 0.4608 | 0.9004 (-0.0996) |
358
+ | 1.2133 | 205 | 0.5216 | - |
359
+ | 1.2430 | 210 | 0.2983 | - |
360
+ | 1.2726 | 215 | 0.4394 | - |
361
+ | 1.3022 | 220 | 0.8886 | - |
362
+ | 1.3319 | 225 | 0.7557 | - |
363
+ | 1.3615 | 230 | 0.5665 | - |
364
+ | 1.3911 | 235 | 0.4785 | - |
365
+ | 1.4207 | 240 | 0.6359 | - |
366
+ | 1.4504 | 245 | 0.4532 | - |
367
+ | 1.48 | 250 | 0.6814 | - |
368
+ | 1.5096 | 255 | 0.5286 | - |
369
+ | 1.5393 | 260 | 0.5322 | - |
370
+ | 1.5689 | 265 | 0.8230 | - |
371
+ | 1.5985 | 270 | 0.4213 | - |
372
+ | 1.6281 | 275 | 0.4357 | - |
373
+ | 1.6578 | 280 | 0.6690 | - |
374
+ | 1.6874 | 285 | 0.6761 | - |
375
+ | 1.7170 | 290 | 0.4560 | - |
376
+ | 1.7467 | 295 | 0.7098 | - |
377
+ | 1.7763 | 300 | 0.6842 | 0.9174 (-0.0826) |
378
+ | 1.8059 | 305 | 0.4919 | - |
379
+ | 1.8356 | 310 | 0.6427 | - |
380
+ | 1.8652 | 315 | 0.3988 | - |
381
+ | 1.8948 | 320 | 0.7871 | - |
382
+ | 1.9244 | 325 | 0.5574 | - |
383
+ | 1.9541 | 330 | 0.5391 | - |
384
+ | 1.9837 | 335 | 0.4867 | - |
385
+ | 2.0119 | 340 | 0.5861 | - |
386
+ | 2.0415 | 345 | 0.2012 | - |
387
+ | 2.0711 | 350 | 0.3108 | - |
388
+ | 2.1007 | 355 | 0.3079 | - |
389
+ | 2.1304 | 360 | 0.1787 | - |
390
+ | 2.16 | 365 | 0.2861 | - |
391
+ | 2.1896 | 370 | 0.1601 | - |
392
+ | 2.2193 | 375 | 0.5452 | - |
393
+ | 2.2489 | 380 | 0.5525 | - |
394
+ | 2.2785 | 385 | 0.2871 | - |
395
+ | 2.3081 | 390 | 0.4414 | - |
396
+ | 2.3378 | 395 | 0.4438 | - |
397
+ | 2.3674 | 400 | 0.4683 | 0.9131 (-0.0869) |
398
+ | 2.3970 | 405 | 0.2535 | - |
399
+ | 2.4267 | 410 | 0.4270 | - |
400
+ | 2.4563 | 415 | 0.5031 | - |
401
+ | 2.4859 | 420 | 0.4150 | - |
402
+ | 2.5156 | 425 | 0.4957 | - |
403
+ | 2.5452 | 430 | 0.3177 | - |
404
+ | 2.5748 | 435 | 0.4201 | - |
405
+ | 2.6044 | 440 | 0.4555 | - |
406
+ | 2.6341 | 445 | 0.4064 | - |
407
+ | 2.6637 | 450 | 0.3323 | - |
408
+ | 2.6933 | 455 | 0.4686 | - |
409
+ | 2.7230 | 460 | 0.4043 | - |
410
+ | 2.7526 | 465 | 0.3465 | - |
411
+ | 2.7822 | 470 | 0.3980 | - |
412
+ | 2.8119 | 475 | 0.4205 | - |
413
+ | 2.8415 | 480 | 0.5495 | - |
414
+ | 2.8711 | 485 | 0.3356 | - |
415
+ | 2.9007 | 490 | 0.2308 | - |
416
+ | 2.9304 | 495 | 0.2548 | - |
417
+ | 2.96 | 500 | 0.4407 | 0.9122 (-0.0878) |
418
+ | 2.9896 | 505 | 0.3697 | - |
419
+ | 3.0178 | 510 | 0.1593 | - |
420
+ | 3.0474 | 515 | 0.1322 | - |
421
+ | 3.0770 | 520 | 0.1471 | - |
422
+ | 3.1067 | 525 | 0.2635 | - |
423
+ | 3.1363 | 530 | 0.5240 | - |
424
+ | 3.1659 | 535 | 0.3511 | - |
425
+ | 3.1956 | 540 | 0.2239 | - |
426
+ | 3.2252 | 545 | 0.1182 | - |
427
+ | 3.2548 | 550 | 0.3037 | - |
428
+ | 3.2844 | 555 | 0.4826 | - |
429
+ | 3.3141 | 560 | 0.2642 | - |
430
+ | 3.3437 | 565 | 0.2451 | - |
431
+ | 3.3733 | 570 | 0.1562 | - |
432
+ | 3.4030 | 575 | 0.4695 | - |
433
+ | 3.4326 | 580 | 0.2303 | - |
434
+ | 3.4622 | 585 | 0.1626 | - |
435
+ | 3.4919 | 590 | 0.1901 | - |
436
+ | 3.5215 | 595 | 0.1473 | - |
437
+ | **3.5511** | **600** | **0.2546** | **0.9326 (-0.0674)** |
438
+ | 3.5807 | 605 | 0.1199 | - |
439
+ | 3.6104 | 610 | 0.1591 | - |
440
+ | 3.64 | 615 | 0.1855 | - |
441
+ | 3.6696 | 620 | 0.2663 | - |
442
+ | 3.6993 | 625 | 0.3292 | - |
443
+ | 3.7289 | 630 | 0.6523 | - |
444
+ | 3.7585 | 635 | 0.2385 | - |
445
+ | 3.7881 | 640 | 0.3839 | - |
446
+ | 3.8178 | 645 | 0.3784 | - |
447
+ | 3.8474 | 650 | 0.1344 | - |
448
+ | 3.8770 | 655 | 0.2048 | - |
449
+ | 3.9067 | 660 | 0.1156 | - |
450
+ | 3.9363 | 665 | 0.4167 | - |
451
+ | 3.9659 | 670 | 0.4006 | - |
452
+ | 3.9956 | 675 | 0.1582 | - |
453
+ | 4.0237 | 680 | 0.2787 | - |
454
+ | 4.0533 | 685 | 0.0621 | - |
455
+ | 4.0830 | 690 | 0.1142 | - |
456
+ | 4.1126 | 695 | 0.0582 | - |
457
+ | 4.1422 | 700 | 0.0993 | 0.9276 (-0.0724) |
458
+ | 4.1719 | 705 | 0.3279 | - |
459
+ | 4.2015 | 710 | 0.1173 | - |
460
+ | 4.2311 | 715 | 0.2549 | - |
461
+ | 4.2607 | 720 | 0.0817 | - |
462
+ | 4.2904 | 725 | 0.1752 | - |
463
+ | 4.32 | 730 | 0.1363 | - |
464
+ | 4.3496 | 735 | 0.1450 | - |
465
+ | 4.3793 | 740 | 0.3517 | - |
466
+ | 4.4089 | 745 | 0.1702 | - |
467
+ | 4.4385 | 750 | 0.1027 | - |
468
+ | 4.4681 | 755 | 0.0867 | - |
469
+ | 4.4978 | 760 | 0.0804 | - |
470
+ | 4.5274 | 765 | 0.0933 | - |
471
+ | 4.5570 | 770 | 0.0987 | - |
472
+ | 4.5867 | 775 | 0.0811 | - |
473
+ | 4.6163 | 780 | 0.1284 | - |
474
+ | 4.6459 | 785 | 0.1125 | - |
475
+ | 4.6756 | 790 | 0.0418 | - |
476
+ | 4.7052 | 795 | 0.0719 | - |
477
+ | 4.7348 | 800 | 0.0952 | 0.9203 (-0.0797) |
478
+ | 4.7644 | 805 | 0.2026 | - |
479
+ | 4.7941 | 810 | 0.0982 | - |
480
+ | 4.8237 | 815 | 0.3003 | - |
481
+ | 4.8533 | 820 | 0.1127 | - |
482
+ | 4.8830 | 825 | 0.5005 | - |
483
+ | 4.9126 | 830 | 0.0491 | - |
484
+ | 4.9422 | 835 | 0.0535 | - |
485
+ | 4.9719 | 840 | 0.0536 | - |
486
+ | 5.0 | 845 | 0.1931 | - |
487
+
488
+ * The bold row denotes the saved checkpoint.
489
+ </details>
490
+
491
+ ### Training Time
492
+ - **Training**: 1.1 hours
493
+ - **Evaluation**: 1.3 minutes
494
+ - **Total**: 1.2 hours
495
+
496
+ ### Framework Versions
497
+ - Python: 3.12.13
498
+ - Sentence Transformers: 5.4.0
499
+ - Transformers: 5.0.0
500
+ - PyTorch: 2.10.0+cu128
501
+ - Accelerate: 1.13.0
502
+ - Datasets: 4.0.0
503
+ - Tokenizers: 0.22.2
504
+
505
+ ## Citation
506
+
507
+ ### BibTeX
508
+
509
+ #### Sentence Transformers
510
+ ```bibtex
511
+ @inproceedings{reimers-2019-sentence-bert,
512
+ title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
513
+ author = "Reimers, Nils and Gurevych, Iryna",
514
+ booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
515
+ month = "11",
516
+ year = "2019",
517
+ publisher = "Association for Computational Linguistics",
518
+ url = "https://arxiv.org/abs/1908.10084",
519
+ }
520
+ ```
521
+
522
+ <!--
523
+ ## Glossary
524
+
525
+ *Clearly define terms in order to be accessible across audiences.*
526
+ -->
527
+
528
+ <!--
529
+ ## Model Card Authors
530
+
531
+ *Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
532
+ -->
533
+
534
+ <!--
535
+ ## Model Card Contact
536
+
537
+ *Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
538
+ -->
config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_cross_attention": false,
3
+ "architectures": [
4
+ "XLMRobertaForSequenceClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "bos_token_id": 0,
8
+ "classifier_dropout": null,
9
+ "dtype": "float32",
10
+ "eos_token_id": 2,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.1,
13
+ "hidden_size": 1024,
14
+ "id2label": {
15
+ "0": "LABEL_0"
16
+ },
17
+ "initializer_range": 0.02,
18
+ "intermediate_size": 4096,
19
+ "is_decoder": false,
20
+ "label2id": {
21
+ "LABEL_0": 0
22
+ },
23
+ "layer_norm_eps": 1e-05,
24
+ "max_position_embeddings": 8194,
25
+ "model_type": "xlm-roberta",
26
+ "num_attention_heads": 16,
27
+ "num_hidden_layers": 24,
28
+ "output_past": true,
29
+ "pad_token_id": 1,
30
+ "position_embedding_type": "absolute",
31
+ "tie_word_embeddings": true,
32
+ "transformers_version": "5.0.0",
33
+ "type_vocab_size": 1,
34
+ "use_cache": false,
35
+ "vocab_size": 250002
36
+ }
config_sentence_transformers.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "pytorch": "2.10.0+cu128",
4
+ "sentence_transformers": "5.4.0",
5
+ "transformers": "5.0.0"
6
+ },
7
+ "activation_fn": "torch.nn.modules.activation.Sigmoid",
8
+ "default_prompt_name": null,
9
+ "model_type": "CrossEncoder",
10
+ "prompts": {}
11
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9d15684d16d2dd1f2f705ebddf58474df9343e7fd72091d1ccdc1871258fed1
3
+ size 2271071804
modules.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.base.modules.transformer.Transformer"
7
+ }
8
+ ]
sentence_bert_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "transformer_task": "sequence-classification",
3
+ "modality_config": {
4
+ "text": {
5
+ "method": "forward",
6
+ "method_output_name": "logits"
7
+ }
8
+ },
9
+ "module_output_name": "scores"
10
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:925c5f322a1f4ed5f8518497d5ed2cad884bb6dd97fcdcbe312d6e99ad2eaa0a
3
+ size 16781739
tokenizer_config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": true,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<s>",
5
+ "clean_up_tokenization_spaces": true,
6
+ "cls_token": "<s>",
7
+ "eos_token": "</s>",
8
+ "is_local": false,
9
+ "mask_token": "<mask>",
10
+ "model_max_length": 8192,
11
+ "pad_token": "<pad>",
12
+ "sep_token": "</s>",
13
+ "sp_model_kwargs": {},
14
+ "tokenizer_class": "XLMRobertaTokenizer",
15
+ "unk_token": "<unk>"
16
+ }