File size: 26,087 Bytes
de2de5a 4e2ca5f de2de5a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 | ---
tags:
- sentence-transformers
- cross-encoder
- reranker
- generated_from_trainer
- dataset_size:5400
- loss:BinaryCrossEntropyLoss
base_model: BAAI/bge-reranker-large
pipeline_tag: text-ranking
library_name: sentence-transformers
metrics:
- map
- mrr@10
- ndcg@10
model-index:
- name: CrossEncoder based on BAAI/bge-reranker-large
results:
- task:
type: cross-encoder-reranking
name: Cross Encoder Reranking
dataset:
name: support rerank eval
type: support-rerank-eval
metrics:
- type: map
value: 0.9087
name: Map
- type: mrr@10
value: 0.9087
name: Mrr@10
- type: ndcg@10
value: 0.9315
name: Ndcg@10
---
# CrossEncoder based on BAAI/bge-reranker-large
This is a [Cross Encoder](https://www.sbert.net/docs/cross_encoder/usage/usage.html) model finetuned from [BAAI/bge-reranker-large](https://huggingface.co/BAAI/bge-reranker-large) 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.
## Model Details
### Model Description
- **Model Type:** Cross Encoder
- **Base model:** [BAAI/bge-reranker-large](https://huggingface.co/BAAI/bge-reranker-large) <!-- at revision 55611d7bca2a7133960a6d3b71e083071bbfc312 -->
- **Maximum Sequence Length:** 512 tokens
- **Number of Output Labels:** 1 label
<!-- - **Training Dataset:** Unknown -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->
### Model Sources
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Documentation:** [Cross Encoder Documentation](https://www.sbert.net/docs/cross_encoder/usage/usage.html)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/huggingface/sentence-transformers)
- **Hugging Face:** [Cross Encoders on Hugging Face](https://huggingface.co/models?library=sentence-transformers&other=cross-encoder)
## Usage
### Direct Usage (Sentence Transformers)
First install the Sentence Transformers library:
```bash
pip install -U sentence-transformers
```
Then you can load this model and run inference.
```python
from sentence_transformers import CrossEncoder
# Download from the 🤗 Hub
model = CrossEncoder("cross_encoder_model_id")
# Get scores for pairs of texts
pairs = [
['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)"],
['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!"],
['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!"],
['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!'],
['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?"],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)
# Or rank different texts based on similarity to a single text
ranks = model.rank(
'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)",
"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!",
"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!",
'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!',
"> 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?",
]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]
```
<!--
### Direct Usage (Transformers)
<details><summary>Click to see the direct usage in Transformers</summary>
</details>
-->
<!--
### Downstream Usage (Sentence Transformers)
You can finetune this model on your own dataset.
<details><summary>Click to expand</summary>
</details>
-->
<!--
### Out-of-Scope Use
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->
## Evaluation
### Metrics
#### Cross Encoder Reranking
* Dataset: `support-rerank-eval`
* Evaluated with [<code>CrossEncoderRerankingEvaluator</code>](https://sbert.net/docs/package_reference/cross_encoder/evaluation.html#sentence_transformers.cross_encoder.evaluation.CrossEncoderRerankingEvaluator) with these parameters:
```json
{
"at_k": 10,
"always_rerank_positives": false
}
```
| Metric | Value |
|:------------|:---------------------|
| map | 0.9087 (-0.0913) |
| mrr@10 | 0.9087 (-0.0913) |
| **ndcg@10** | **0.9315 (-0.0685)** |
<!--
## Bias, Risks and Limitations
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->
<!--
### Recommendations
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->
## Training Details
### Training Dataset
#### Unnamed Dataset
* Size: 5,400 training samples
* Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>label</code>
* Approximate statistics based on the first 1000 samples:
| | sentence1 | sentence2 | label |
|:--------|:--------------------------------------------------------------------------------------------------|:---------------------------------------------------------------------------------------------------|:---------------------------------------------------------------|
| type | string | string | float |
| details | <ul><li>min: 28 characters</li><li>mean: 218.29 characters</li><li>max: 1961 characters</li></ul> | <ul><li>min: 141 characters</li><li>mean: 643.39 characters</li><li>max: 3722 characters</li></ul> | <ul><li>min: 0.0</li><li>mean: 0.17</li><li>max: 1.0</li></ul> |
* Samples:
| sentence1 | sentence2 | label |
|:----------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------|
| <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> |
| <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> |
| <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> |
* Loss: [<code>BinaryCrossEntropyLoss</code>](https://sbert.net/docs/package_reference/cross_encoder/losses.html#binarycrossentropyloss) with these parameters:
```json
{
"activation_fn": "torch.nn.modules.linear.Identity",
"pos_weight": 5
}
```
### Training Hyperparameters
#### Non-Default Hyperparameters
- `eval_strategy`: steps
- `per_device_train_batch_size`: 1
- `gradient_accumulation_steps`: 32
- `learning_rate`: 1e-05
- `num_train_epochs`: 5
- `warmup_ratio`: 0.2
- `warmup_steps`: 0.2
- `seed`: 67
- `fp16`: True
- `load_best_model_at_end`: True
#### All Hyperparameters
<details><summary>Click to expand</summary>
- `do_predict`: False
- `eval_strategy`: steps
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 1
- `per_device_eval_batch_size`: 8
- `gradient_accumulation_steps`: 32
- `eval_accumulation_steps`: None
- `torch_empty_cache_steps`: None
- `learning_rate`: 1e-05
- `weight_decay`: 0.0
- `adam_beta1`: 0.9
- `adam_beta2`: 0.999
- `adam_epsilon`: 1e-08
- `max_grad_norm`: 1.0
- `num_train_epochs`: 5
- `max_steps`: -1
- `lr_scheduler_type`: linear
- `lr_scheduler_kwargs`: None
- `warmup_ratio`: 0.2
- `warmup_steps`: 0.2
- `log_level`: passive
- `log_level_replica`: warning
- `log_on_each_node`: True
- `logging_nan_inf_filter`: True
- `enable_jit_checkpoint`: False
- `save_on_each_node`: False
- `save_only_model`: False
- `restore_callback_states_from_checkpoint`: False
- `use_cpu`: False
- `seed`: 67
- `data_seed`: None
- `bf16`: False
- `fp16`: True
- `bf16_full_eval`: False
- `fp16_full_eval`: False
- `tf32`: None
- `local_rank`: -1
- `ddp_backend`: None
- `debug`: []
- `dataloader_drop_last`: False
- `dataloader_num_workers`: 0
- `dataloader_prefetch_factor`: None
- `disable_tqdm`: True
- `remove_unused_columns`: True
- `label_names`: None
- `load_best_model_at_end`: True
- `ignore_data_skip`: False
- `fsdp`: []
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
- `parallelism_config`: None
- `deepspeed`: None
- `label_smoothing_factor`: 0.0
- `optim`: adamw_torch_fused
- `optim_args`: None
- `group_by_length`: False
- `length_column_name`: length
- `project`: huggingface
- `trackio_space_id`: trackio
- `ddp_find_unused_parameters`: None
- `ddp_bucket_cap_mb`: None
- `ddp_broadcast_buffers`: False
- `dataloader_pin_memory`: True
- `dataloader_persistent_workers`: False
- `skip_memory_metrics`: True
- `push_to_hub`: False
- `resume_from_checkpoint`: None
- `hub_model_id`: None
- `hub_strategy`: every_save
- `hub_private_repo`: None
- `hub_always_push`: False
- `hub_revision`: None
- `gradient_checkpointing`: False
- `gradient_checkpointing_kwargs`: None
- `include_for_metrics`: []
- `eval_do_concat_batches`: True
- `auto_find_batch_size`: False
- `full_determinism`: False
- `ddp_timeout`: 1800
- `torch_compile`: False
- `torch_compile_backend`: None
- `torch_compile_mode`: None
- `include_num_input_tokens_seen`: no
- `neftune_noise_alpha`: None
- `optim_target_modules`: None
- `batch_eval_metrics`: False
- `eval_on_start`: False
- `use_liger_kernel`: False
- `liger_kernel_config`: None
- `eval_use_gather_object`: False
- `average_tokens_across_devices`: True
- `use_cache`: False
- `prompts`: None
- `batch_sampler`: batch_sampler
- `multi_dataset_batch_sampler`: proportional
- `router_mapping`: {}
- `learning_rate_mapping`: {}
</details>
### Training Logs
#### Image
| Epoch | Step | Training Loss | support-rerank-eval_ndcg@10 |
|:----------:|:-------:|:-------------:|:---------------------------:|
| -1 | -1 | - | 0.8413 (-0.1587) |
| 0.0119 | 1 | 1.3864 | - |
| 0.0593 | 5 | 2.2658 | - |
| 0.1185 | 10 | 1.5262 | - |
| 0.1778 | 15 | 1.5737 | - |
| 0.2370 | 20 | 1.3171 | - |
| 0.2963 | 25 | 1.0806 | - |
| 0.3556 | 30 | 1.0587 | - |
| 0.4148 | 35 | 0.8149 | - |
| 0.4741 | 40 | 0.8706 | - |
| 0.5333 | 45 | 1.1330 | - |
| 0.5926 | 50 | 1.1210 | - |
| 0.6519 | 55 | 0.7105 | - |
| 0.7111 | 60 | 0.8596 | - |
| 0.7704 | 65 | 0.8416 | - |
| 0.8296 | 70 | 0.6543 | - |
| 0.8889 | 75 | 0.7784 | - |
| 0.9481 | 80 | 0.5754 | - |
| 1.0 | 85 | 0.7647 | - |
| 1.0593 | 90 | 0.5856 | - |
| 1.1185 | 95 | 0.5812 | - |
| 1.1778 | 100 | 0.5459 | 0.8906 (-0.1094) |
| 1.2370 | 105 | 0.5259 | - |
| 1.2963 | 110 | 0.6609 | - |
| 1.3556 | 115 | 0.5932 | - |
| 1.4148 | 120 | 0.5482 | - |
| 1.4741 | 125 | 0.7539 | - |
| 1.5333 | 130 | 0.6486 | - |
| 1.5926 | 135 | 0.6475 | - |
| 1.6519 | 140 | 0.5627 | - |
| 1.7111 | 145 | 0.6848 | - |
| 1.7704 | 150 | 0.6890 | - |
| 1.8296 | 155 | 0.5712 | - |
| 1.8889 | 160 | 0.5949 | - |
| 1.9481 | 165 | 0.7066 | - |
| 2.0 | 170 | 0.7013 | - |
| 2.0593 | 175 | 0.3729 | - |
| 2.1185 | 180 | 0.4327 | - |
| 2.1778 | 185 | 0.3703 | - |
| 2.2370 | 190 | 0.4843 | - |
| 2.2963 | 195 | 0.3460 | - |
| 2.3556 | 200 | 0.3679 | 0.9241 (-0.0759) |
| 2.4148 | 205 | 0.4412 | - |
| 2.4741 | 210 | 0.5300 | - |
| 2.5333 | 215 | 0.4922 | - |
| 2.5926 | 220 | 0.4407 | - |
| 2.6519 | 225 | 0.3835 | - |
| 2.7111 | 230 | 0.4581 | - |
| 2.7704 | 235 | 0.4687 | - |
| 2.8296 | 240 | 0.3968 | - |
| 2.8889 | 245 | 0.3373 | - |
| 2.9481 | 250 | 0.3882 | - |
| 3.0 | 255 | 0.5394 | - |
| 3.0593 | 260 | 0.3229 | - |
| 3.1185 | 265 | 0.2746 | - |
| 3.1778 | 270 | 0.4415 | - |
| 3.2370 | 275 | 0.2434 | - |
| 3.2963 | 280 | 0.3033 | - |
| 3.3556 | 285 | 0.3377 | - |
| 3.4148 | 290 | 0.3316 | - |
| 3.4741 | 295 | 0.2425 | - |
| 3.5333 | 300 | 0.3449 | 0.9141 (-0.0859) |
| 3.5926 | 305 | 0.2569 | - |
| 3.6519 | 310 | 0.3693 | - |
| 3.7111 | 315 | 0.2660 | - |
| 3.7704 | 320 | 0.3553 | - |
| 3.8296 | 325 | 0.3099 | - |
| 3.8889 | 330 | 0.2923 | - |
| 3.9481 | 335 | 0.3726 | - |
| 4.0 | 340 | 0.5681 | - |
| 4.0593 | 345 | 0.1964 | - |
| 4.1185 | 350 | 0.2471 | - |
| 4.1778 | 355 | 0.3227 | - |
| 4.2370 | 360 | 0.2911 | - |
| 4.2963 | 365 | 0.2578 | - |
| 4.3556 | 370 | 0.2402 | - |
| 4.4148 | 375 | 0.2367 | - |
| 4.4741 | 380 | 0.1998 | - |
| 4.5333 | 385 | 0.2359 | - |
| 4.5926 | 390 | 0.2177 | - |
| 4.6519 | 395 | 0.1999 | - |
| **4.7111** | **400** | **0.2293** | **0.9315 (-0.0685)** |
| 4.7704 | 405 | 0.2748 | - |
| 4.8296 | 410 | 0.2312 | - |
| 4.8889 | 415 | 0.3013 | - |
| 4.9481 | 420 | 0.1602 | - |
| 5.0 | 425 | 0.3283 | - |
* The bold row denotes the saved checkpoint.
### Framework Versions
- Python: 3.12.12
- Sentence Transformers: 5.2.3
- Transformers: 5.0.0
- PyTorch: 2.10.0+cu128
- Accelerate: 1.12.0
- Datasets: 4.8.3
- Tokenizers: 0.22.2
## Citation
### BibTeX
#### Sentence Transformers
```bibtex
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}
```
<!--
## Glossary
*Clearly define terms in order to be accessible across audiences.*
-->
<!--
## Model Card Authors
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->
<!--
## Model Card Contact
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
--> |