Add new CrossEncoder model
Browse files- README.md +514 -0
- config.json +36 -0
- model.safetensors +3 -0
- special_tokens_map.json +37 -0
- tokenizer.json +0 -0
- tokenizer_config.json +58 -0
- vocab.txt +0 -0
README.md
ADDED
@@ -0,0 +1,514 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- sentence-transformers
|
4 |
+
- cross-encoder
|
5 |
+
- generated_from_trainer
|
6 |
+
- dataset_size:23770
|
7 |
+
- loss:MultipleNegativesRankingLoss
|
8 |
+
base_model: cross-encoder/ms-marco-MiniLM-L12-v2
|
9 |
+
pipeline_tag: text-ranking
|
10 |
+
library_name: sentence-transformers
|
11 |
+
metrics:
|
12 |
+
- map
|
13 |
+
- mrr@5
|
14 |
+
- ndcg@5
|
15 |
+
model-index:
|
16 |
+
- name: CrossEncoder based on cross-encoder/ms-marco-MiniLM-L12-v2
|
17 |
+
results:
|
18 |
+
- task:
|
19 |
+
type: cross-encoder-reranking
|
20 |
+
name: Cross Encoder Reranking
|
21 |
+
dataset:
|
22 |
+
name: claims evidence dev
|
23 |
+
type: claims-evidence-dev
|
24 |
+
metrics:
|
25 |
+
- type: map
|
26 |
+
value: 0.9904
|
27 |
+
name: Map
|
28 |
+
- type: mrr@5
|
29 |
+
value: 1.0
|
30 |
+
name: Mrr@5
|
31 |
+
- type: ndcg@5
|
32 |
+
value: 0.9882
|
33 |
+
name: Ndcg@5
|
34 |
+
---
|
35 |
+
|
36 |
+
# CrossEncoder based on cross-encoder/ms-marco-MiniLM-L12-v2
|
37 |
+
|
38 |
+
This is a [Cross Encoder](https://www.sbert.net/docs/cross_encoder/usage/usage.html) model finetuned from [cross-encoder/ms-marco-MiniLM-L12-v2](https://huggingface.co/cross-encoder/ms-marco-MiniLM-L12-v2) 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.
|
39 |
+
|
40 |
+
## Model Details
|
41 |
+
|
42 |
+
### Model Description
|
43 |
+
- **Model Type:** Cross Encoder
|
44 |
+
- **Base model:** [cross-encoder/ms-marco-MiniLM-L12-v2](https://huggingface.co/cross-encoder/ms-marco-MiniLM-L12-v2) <!-- at revision a34da8fab3ad458d48778dea3276ce729857efaf -->
|
45 |
+
- **Maximum Sequence Length:** 512 tokens
|
46 |
+
- **Number of Output Labels:** 1 label
|
47 |
+
<!-- - **Training Dataset:** Unknown -->
|
48 |
+
<!-- - **Language:** Unknown -->
|
49 |
+
<!-- - **License:** Unknown -->
|
50 |
+
|
51 |
+
### Model Sources
|
52 |
+
|
53 |
+
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
|
54 |
+
- **Documentation:** [Cross Encoder Documentation](https://www.sbert.net/docs/cross_encoder/usage/usage.html)
|
55 |
+
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
|
56 |
+
- **Hugging Face:** [Cross Encoders on Hugging Face](https://huggingface.co/models?library=sentence-transformers&other=cross-encoder)
|
57 |
+
|
58 |
+
## Usage
|
59 |
+
|
60 |
+
### Direct Usage (Sentence Transformers)
|
61 |
+
|
62 |
+
First install the Sentence Transformers library:
|
63 |
+
|
64 |
+
```bash
|
65 |
+
pip install -U sentence-transformers
|
66 |
+
```
|
67 |
+
|
68 |
+
Then you can load this model and run inference.
|
69 |
+
```python
|
70 |
+
from sentence_transformers import CrossEncoder
|
71 |
+
|
72 |
+
# Download from the 🤗 Hub
|
73 |
+
model = CrossEncoder("Davidsamuel101/ft-ms-marco-MiniLM-L12-v2-claims-reranker-v2")
|
74 |
+
# Get scores for pairs of texts
|
75 |
+
pairs = [
|
76 |
+
['Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.', 'At very high concentrations (100 times atmospheric concentration, or greater), carbon dioxide can be toxic to animal life, so raising the concentration to 10,000 ppm (1%) or higher for several hours will eliminate pests such as whiteflies and spider mites in a greenhouse.'],
|
77 |
+
['Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.', 'Plants can grow as much as 50 percent faster in concentrations of 1,000 ppm CO 2 when compared with ambient conditions, though this assumes no change in climate and no limitation on other nutrients.'],
|
78 |
+
['Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.', 'Higher carbon dioxide concentrations will favourably affect plant growth and demand for water.'],
|
79 |
+
['Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.', "Carbon dioxide in the Earth's atmosphere is essential to life and to most of the planetary biosphere."],
|
80 |
+
['Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.', 'Rennie 2009: "Claim 1: Anthropogenic CO2 can\'t be changing climate, because CO2 is only a trace gas in the atmosphere and the amount produced by humans is dwarfed by the amount from volcanoes and other natural sources.'],
|
81 |
+
]
|
82 |
+
scores = model.predict(pairs)
|
83 |
+
print(scores.shape)
|
84 |
+
# (5,)
|
85 |
+
|
86 |
+
# Or rank different texts based on similarity to a single text
|
87 |
+
ranks = model.rank(
|
88 |
+
'Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.',
|
89 |
+
[
|
90 |
+
'At very high concentrations (100 times atmospheric concentration, or greater), carbon dioxide can be toxic to animal life, so raising the concentration to 10,000 ppm (1%) or higher for several hours will eliminate pests such as whiteflies and spider mites in a greenhouse.',
|
91 |
+
'Plants can grow as much as 50 percent faster in concentrations of 1,000 ppm CO 2 when compared with ambient conditions, though this assumes no change in climate and no limitation on other nutrients.',
|
92 |
+
'Higher carbon dioxide concentrations will favourably affect plant growth and demand for water.',
|
93 |
+
"Carbon dioxide in the Earth's atmosphere is essential to life and to most of the planetary biosphere.",
|
94 |
+
'Rennie 2009: "Claim 1: Anthropogenic CO2 can\'t be changing climate, because CO2 is only a trace gas in the atmosphere and the amount produced by humans is dwarfed by the amount from volcanoes and other natural sources.',
|
95 |
+
]
|
96 |
+
)
|
97 |
+
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]
|
98 |
+
```
|
99 |
+
|
100 |
+
<!--
|
101 |
+
### Direct Usage (Transformers)
|
102 |
+
|
103 |
+
<details><summary>Click to see the direct usage in Transformers</summary>
|
104 |
+
|
105 |
+
</details>
|
106 |
+
-->
|
107 |
+
|
108 |
+
<!--
|
109 |
+
### Downstream Usage (Sentence Transformers)
|
110 |
+
|
111 |
+
You can finetune this model on your own dataset.
|
112 |
+
|
113 |
+
<details><summary>Click to expand</summary>
|
114 |
+
|
115 |
+
</details>
|
116 |
+
-->
|
117 |
+
|
118 |
+
<!--
|
119 |
+
### Out-of-Scope Use
|
120 |
+
|
121 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
122 |
+
-->
|
123 |
+
|
124 |
+
## Evaluation
|
125 |
+
|
126 |
+
### Metrics
|
127 |
+
|
128 |
+
#### Cross Encoder Reranking
|
129 |
+
|
130 |
+
* Dataset: `claims-evidence-dev`
|
131 |
+
* Evaluated with [<code>CrossEncoderRerankingEvaluator</code>](https://sbert.net/docs/package_reference/cross_encoder/evaluation.html#sentence_transformers.cross_encoder.evaluation.CrossEncoderRerankingEvaluator) with these parameters:
|
132 |
+
```json
|
133 |
+
{
|
134 |
+
"at_k": 5,
|
135 |
+
"always_rerank_positives": true
|
136 |
+
}
|
137 |
+
```
|
138 |
+
|
139 |
+
| Metric | Value |
|
140 |
+
|:-----------|:---------------------|
|
141 |
+
| map | 0.9904 (-0.0096) |
|
142 |
+
| mrr@5 | 1.0000 (+0.0000) |
|
143 |
+
| **ndcg@5** | **0.9882 (-0.0118)** |
|
144 |
+
|
145 |
+
<!--
|
146 |
+
## Bias, Risks and Limitations
|
147 |
+
|
148 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
149 |
+
-->
|
150 |
+
|
151 |
+
<!--
|
152 |
+
### Recommendations
|
153 |
+
|
154 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
155 |
+
-->
|
156 |
+
|
157 |
+
## Training Details
|
158 |
+
|
159 |
+
### Training Dataset
|
160 |
+
|
161 |
+
#### Unnamed Dataset
|
162 |
+
|
163 |
+
* Size: 23,770 training samples
|
164 |
+
* Columns: <code>text1</code>, <code>text2</code>, and <code>label</code>
|
165 |
+
* Approximate statistics based on the first 1000 samples:
|
166 |
+
| | text1 | text2 | label |
|
167 |
+
|:--------|:-------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------------------------------------|:------------------------------------------------|
|
168 |
+
| type | string | string | int |
|
169 |
+
| details | <ul><li>min: 38 characters</li><li>mean: 118.57 characters</li><li>max: 226 characters</li></ul> | <ul><li>min: 14 characters</li><li>mean: 144.96 characters</li><li>max: 1176 characters</li></ul> | <ul><li>0: ~83.70%</li><li>1: ~16.30%</li></ul> |
|
170 |
+
* Samples:
|
171 |
+
| text1 | text2 | label |
|
172 |
+
|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:---------------|
|
173 |
+
| <code>Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.</code> | <code>At very high concentrations (100 times atmospheric concentration, or greater), carbon dioxide can be toxic to animal life, so raising the concentration to 10,000 ppm (1%) or higher for several hours will eliminate pests such as whiteflies and spider mites in a greenhouse.</code> | <code>1</code> |
|
174 |
+
| <code>Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.</code> | <code>Plants can grow as much as 50 percent faster in concentrations of 1,000 ppm CO 2 when compared with ambient conditions, though this assumes no change in climate and no limitation on other nutrients.</code> | <code>1</code> |
|
175 |
+
| <code>Not only is there no scientific evidence that CO2 is a pollutant, higher CO2 concentrations actually help ecosystems support more plant and animal life.</code> | <code>Higher carbon dioxide concentrations will favourably affect plant growth and demand for water.</code> | <code>1</code> |
|
176 |
+
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/cross_encoder/losses.html#multiplenegativesrankingloss) with these parameters:
|
177 |
+
```json
|
178 |
+
{
|
179 |
+
"scale": 10.0,
|
180 |
+
"num_negatives": 4,
|
181 |
+
"activation_fn": "torch.nn.modules.activation.Sigmoid"
|
182 |
+
}
|
183 |
+
```
|
184 |
+
|
185 |
+
### Training Hyperparameters
|
186 |
+
#### Non-Default Hyperparameters
|
187 |
+
|
188 |
+
- `eval_strategy`: steps
|
189 |
+
- `per_device_train_batch_size`: 16
|
190 |
+
- `learning_rate`: 3e-06
|
191 |
+
- `num_train_epochs`: 5
|
192 |
+
- `bf16`: True
|
193 |
+
- `load_best_model_at_end`: True
|
194 |
+
|
195 |
+
#### All Hyperparameters
|
196 |
+
<details><summary>Click to expand</summary>
|
197 |
+
|
198 |
+
- `overwrite_output_dir`: False
|
199 |
+
- `do_predict`: False
|
200 |
+
- `eval_strategy`: steps
|
201 |
+
- `prediction_loss_only`: True
|
202 |
+
- `per_device_train_batch_size`: 16
|
203 |
+
- `per_device_eval_batch_size`: 8
|
204 |
+
- `per_gpu_train_batch_size`: None
|
205 |
+
- `per_gpu_eval_batch_size`: None
|
206 |
+
- `gradient_accumulation_steps`: 1
|
207 |
+
- `eval_accumulation_steps`: None
|
208 |
+
- `torch_empty_cache_steps`: None
|
209 |
+
- `learning_rate`: 3e-06
|
210 |
+
- `weight_decay`: 0.0
|
211 |
+
- `adam_beta1`: 0.9
|
212 |
+
- `adam_beta2`: 0.999
|
213 |
+
- `adam_epsilon`: 1e-08
|
214 |
+
- `max_grad_norm`: 1.0
|
215 |
+
- `num_train_epochs`: 5
|
216 |
+
- `max_steps`: -1
|
217 |
+
- `lr_scheduler_type`: linear
|
218 |
+
- `lr_scheduler_kwargs`: {}
|
219 |
+
- `warmup_ratio`: 0.0
|
220 |
+
- `warmup_steps`: 0
|
221 |
+
- `log_level`: passive
|
222 |
+
- `log_level_replica`: warning
|
223 |
+
- `log_on_each_node`: True
|
224 |
+
- `logging_nan_inf_filter`: True
|
225 |
+
- `save_safetensors`: True
|
226 |
+
- `save_on_each_node`: False
|
227 |
+
- `save_only_model`: False
|
228 |
+
- `restore_callback_states_from_checkpoint`: False
|
229 |
+
- `no_cuda`: False
|
230 |
+
- `use_cpu`: False
|
231 |
+
- `use_mps_device`: False
|
232 |
+
- `seed`: 42
|
233 |
+
- `data_seed`: None
|
234 |
+
- `jit_mode_eval`: False
|
235 |
+
- `use_ipex`: False
|
236 |
+
- `bf16`: True
|
237 |
+
- `fp16`: False
|
238 |
+
- `fp16_opt_level`: O1
|
239 |
+
- `half_precision_backend`: auto
|
240 |
+
- `bf16_full_eval`: False
|
241 |
+
- `fp16_full_eval`: False
|
242 |
+
- `tf32`: None
|
243 |
+
- `local_rank`: 0
|
244 |
+
- `ddp_backend`: None
|
245 |
+
- `tpu_num_cores`: None
|
246 |
+
- `tpu_metrics_debug`: False
|
247 |
+
- `debug`: []
|
248 |
+
- `dataloader_drop_last`: False
|
249 |
+
- `dataloader_num_workers`: 0
|
250 |
+
- `dataloader_prefetch_factor`: None
|
251 |
+
- `past_index`: -1
|
252 |
+
- `disable_tqdm`: False
|
253 |
+
- `remove_unused_columns`: True
|
254 |
+
- `label_names`: None
|
255 |
+
- `load_best_model_at_end`: True
|
256 |
+
- `ignore_data_skip`: False
|
257 |
+
- `fsdp`: []
|
258 |
+
- `fsdp_min_num_params`: 0
|
259 |
+
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
|
260 |
+
- `tp_size`: 0
|
261 |
+
- `fsdp_transformer_layer_cls_to_wrap`: None
|
262 |
+
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
|
263 |
+
- `deepspeed`: None
|
264 |
+
- `label_smoothing_factor`: 0.0
|
265 |
+
- `optim`: adamw_torch
|
266 |
+
- `optim_args`: None
|
267 |
+
- `adafactor`: False
|
268 |
+
- `group_by_length`: False
|
269 |
+
- `length_column_name`: length
|
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 |
+
- `use_legacy_prediction_loop`: False
|
277 |
+
- `push_to_hub`: False
|
278 |
+
- `resume_from_checkpoint`: None
|
279 |
+
- `hub_model_id`: None
|
280 |
+
- `hub_strategy`: every_save
|
281 |
+
- `hub_private_repo`: None
|
282 |
+
- `hub_always_push`: False
|
283 |
+
- `gradient_checkpointing`: False
|
284 |
+
- `gradient_checkpointing_kwargs`: None
|
285 |
+
- `include_inputs_for_metrics`: False
|
286 |
+
- `include_for_metrics`: []
|
287 |
+
- `eval_do_concat_batches`: True
|
288 |
+
- `fp16_backend`: auto
|
289 |
+
- `push_to_hub_model_id`: None
|
290 |
+
- `push_to_hub_organization`: None
|
291 |
+
- `mp_parameters`:
|
292 |
+
- `auto_find_batch_size`: False
|
293 |
+
- `full_determinism`: False
|
294 |
+
- `torchdynamo`: None
|
295 |
+
- `ray_scope`: last
|
296 |
+
- `ddp_timeout`: 1800
|
297 |
+
- `torch_compile`: False
|
298 |
+
- `torch_compile_backend`: None
|
299 |
+
- `torch_compile_mode`: None
|
300 |
+
- `include_tokens_per_second`: False
|
301 |
+
- `include_num_input_tokens_seen`: False
|
302 |
+
- `neftune_noise_alpha`: None
|
303 |
+
- `optim_target_modules`: None
|
304 |
+
- `batch_eval_metrics`: False
|
305 |
+
- `eval_on_start`: False
|
306 |
+
- `use_liger_kernel`: False
|
307 |
+
- `eval_use_gather_object`: False
|
308 |
+
- `average_tokens_across_devices`: False
|
309 |
+
- `prompts`: None
|
310 |
+
- `batch_sampler`: batch_sampler
|
311 |
+
- `multi_dataset_batch_sampler`: proportional
|
312 |
+
|
313 |
+
</details>
|
314 |
+
|
315 |
+
### Training Logs
|
316 |
+
<details><summary>Click to expand</summary>
|
317 |
+
|
318 |
+
| Epoch | Step | Training Loss | claims-evidence-dev_ndcg@5 |
|
319 |
+
|:----------:|:-------:|:-------------:|:--------------------------:|
|
320 |
+
| 0.0336 | 50 | 1.2496 | - |
|
321 |
+
| **0.0673** | **100** | **1.2605** | **0.9523 (-0.0477)** |
|
322 |
+
| 0.1009 | 150 | 1.1969 | - |
|
323 |
+
| 0.1346 | 200 | 1.2353 | 0.9529 (-0.0471) |
|
324 |
+
| 0.1682 | 250 | 1.2114 | - |
|
325 |
+
| 0.2019 | 300 | 1.1438 | 0.9551 (-0.0449) |
|
326 |
+
| 0.2355 | 350 | 1.2062 | - |
|
327 |
+
| 0.2692 | 400 | 1.1631 | 0.9568 (-0.0432) |
|
328 |
+
| 0.3028 | 450 | 1.115 | - |
|
329 |
+
| 0.3365 | 500 | 1.2029 | 0.9582 (-0.0418) |
|
330 |
+
| 0.3701 | 550 | 1.0615 | - |
|
331 |
+
| 0.4038 | 600 | 1.185 | 0.9649 (-0.0351) |
|
332 |
+
| 0.4374 | 650 | 1.0651 | - |
|
333 |
+
| 0.4711 | 700 | 1.0951 | 0.9682 (-0.0318) |
|
334 |
+
| 0.5047 | 750 | 1.1267 | - |
|
335 |
+
| 0.5384 | 800 | 1.0822 | 0.9727 (-0.0273) |
|
336 |
+
| 0.5720 | 850 | 1.0658 | - |
|
337 |
+
| 0.6057 | 900 | 1.0113 | 0.9785 (-0.0215) |
|
338 |
+
| 0.6393 | 950 | 1.0578 | - |
|
339 |
+
| 0.6729 | 1000 | 1.074 | 0.9829 (-0.0171) |
|
340 |
+
| 0.7066 | 1050 | 1.0287 | - |
|
341 |
+
| 0.7402 | 1100 | 0.9337 | 0.9873 (-0.0127) |
|
342 |
+
| 0.7739 | 1150 | 0.9798 | - |
|
343 |
+
| 0.8075 | 1200 | 0.9697 | 0.9899 (-0.0101) |
|
344 |
+
| 0.8412 | 1250 | 0.984 | - |
|
345 |
+
| 0.8748 | 1300 | 0.9913 | 0.9898 (-0.0102) |
|
346 |
+
| 0.9085 | 1350 | 1.0126 | - |
|
347 |
+
| 0.9421 | 1400 | 0.9458 | 0.9897 (-0.0103) |
|
348 |
+
| 0.9758 | 1450 | 0.9594 | - |
|
349 |
+
| 1.0094 | 1500 | 0.9798 | 0.9896 (-0.0104) |
|
350 |
+
| 1.0431 | 1550 | 0.9599 | - |
|
351 |
+
| 1.0767 | 1600 | 0.9485 | 0.9887 (-0.0113) |
|
352 |
+
| 1.1104 | 1650 | 0.9021 | - |
|
353 |
+
| 1.1440 | 1700 | 0.9778 | 0.9887 (-0.0113) |
|
354 |
+
| 1.1777 | 1750 | 0.9836 | - |
|
355 |
+
| 1.2113 | 1800 | 0.939 | 0.9912 (-0.0088) |
|
356 |
+
| 1.2450 | 1850 | 0.9476 | - |
|
357 |
+
| 1.2786 | 1900 | 0.964 | 0.9914 (-0.0086) |
|
358 |
+
| 1.3122 | 1950 | 0.9238 | - |
|
359 |
+
| 1.3459 | 2000 | 0.9811 | 0.9895 (-0.0105) |
|
360 |
+
| 1.3795 | 2050 | 0.905 | - |
|
361 |
+
| 1.4132 | 2100 | 0.8979 | 0.9896 (-0.0104) |
|
362 |
+
| 1.4468 | 2150 | 0.8998 | - |
|
363 |
+
| 1.4805 | 2200 | 0.9016 | 0.9896 (-0.0104) |
|
364 |
+
| 1.5141 | 2250 | 0.9183 | - |
|
365 |
+
| 1.5478 | 2300 | 0.8805 | 0.9896 (-0.0104) |
|
366 |
+
| 1.5814 | 2350 | 0.8672 | - |
|
367 |
+
| 1.6151 | 2400 | 0.8822 | 0.9896 (-0.0104) |
|
368 |
+
| 1.6487 | 2450 | 0.8724 | - |
|
369 |
+
| 1.6824 | 2500 | 0.9397 | 0.9883 (-0.0117) |
|
370 |
+
| 1.7160 | 2550 | 0.8903 | - |
|
371 |
+
| 1.7497 | 2600 | 0.9305 | 0.9882 (-0.0118) |
|
372 |
+
| 1.7833 | 2650 | 0.8741 | - |
|
373 |
+
| 1.8170 | 2700 | 0.8951 | 0.9874 (-0.0126) |
|
374 |
+
| 1.8506 | 2750 | 0.8958 | - |
|
375 |
+
| 1.8843 | 2800 | 0.8529 | 0.9873 (-0.0127) |
|
376 |
+
| 1.9179 | 2850 | 0.9468 | - |
|
377 |
+
| 1.9515 | 2900 | 0.8683 | 0.9882 (-0.0118) |
|
378 |
+
| 1.9852 | 2950 | 0.9145 | - |
|
379 |
+
| 2.0188 | 3000 | 0.9137 | 0.9883 (-0.0117) |
|
380 |
+
| 2.0525 | 3050 | 0.8175 | - |
|
381 |
+
| 2.0861 | 3100 | 0.911 | 0.9883 (-0.0117) |
|
382 |
+
| 2.1198 | 3150 | 0.8749 | - |
|
383 |
+
| 2.1534 | 3200 | 0.8491 | 0.9883 (-0.0117) |
|
384 |
+
| 2.1871 | 3250 | 0.9057 | - |
|
385 |
+
| 2.2207 | 3300 | 0.9034 | 0.9882 (-0.0118) |
|
386 |
+
| 2.2544 | 3350 | 0.8505 | - |
|
387 |
+
| 2.2880 | 3400 | 0.8762 | 0.9883 (-0.0117) |
|
388 |
+
| 2.3217 | 3450 | 0.8974 | - |
|
389 |
+
| 2.3553 | 3500 | 0.8832 | 0.9884 (-0.0116) |
|
390 |
+
| 2.3890 | 3550 | 0.851 | - |
|
391 |
+
| 2.4226 | 3600 | 0.8584 | 0.9890 (-0.0110) |
|
392 |
+
| 2.4563 | 3650 | 0.9032 | - |
|
393 |
+
| 2.4899 | 3700 | 0.8963 | 0.9893 (-0.0107) |
|
394 |
+
| 2.5236 | 3750 | 0.8756 | - |
|
395 |
+
| 2.5572 | 3800 | 0.843 | 0.9882 (-0.0118) |
|
396 |
+
| 2.5908 | 3850 | 0.8778 | - |
|
397 |
+
| 2.6245 | 3900 | 0.8434 | 0.9882 (-0.0118) |
|
398 |
+
| 2.6581 | 3950 | 0.9193 | - |
|
399 |
+
| 2.6918 | 4000 | 0.8724 | 0.9875 (-0.0125) |
|
400 |
+
| 2.7254 | 4050 | 0.9062 | - |
|
401 |
+
| 2.7591 | 4100 | 0.8807 | 0.9875 (-0.0125) |
|
402 |
+
| 2.7927 | 4150 | 0.8252 | - |
|
403 |
+
| 2.8264 | 4200 | 0.8725 | 0.9875 (-0.0125) |
|
404 |
+
| 2.8600 | 4250 | 0.9094 | - |
|
405 |
+
| 2.8937 | 4300 | 0.8589 | 0.9874 (-0.0126) |
|
406 |
+
| 2.9273 | 4350 | 0.8625 | - |
|
407 |
+
| 2.9610 | 4400 | 0.8138 | 0.9874 (-0.0126) |
|
408 |
+
| 2.9946 | 4450 | 0.9217 | - |
|
409 |
+
| 3.0283 | 4500 | 0.8871 | 0.9872 (-0.0128) |
|
410 |
+
| 3.0619 | 4550 | 0.8504 | - |
|
411 |
+
| 3.0956 | 4600 | 0.944 | 0.9873 (-0.0127) |
|
412 |
+
| 3.1292 | 4650 | 0.8258 | - |
|
413 |
+
| 3.1629 | 4700 | 0.9054 | 0.9874 (-0.0126) |
|
414 |
+
| 3.1965 | 4750 | 0.8297 | - |
|
415 |
+
| 3.2301 | 4800 | 0.8483 | 0.9875 (-0.0125) |
|
416 |
+
| 3.2638 | 4850 | 0.909 | - |
|
417 |
+
| 3.2974 | 4900 | 0.8486 | 0.9892 (-0.0108) |
|
418 |
+
| 3.3311 | 4950 | 0.8937 | - |
|
419 |
+
| 3.3647 | 5000 | 0.8821 | 0.9874 (-0.0126) |
|
420 |
+
| 3.3984 | 5050 | 0.873 | - |
|
421 |
+
| 3.4320 | 5100 | 0.8773 | 0.9874 (-0.0126) |
|
422 |
+
| 3.4657 | 5150 | 0.8592 | - |
|
423 |
+
| 3.4993 | 5200 | 0.8449 | 0.9882 (-0.0118) |
|
424 |
+
| 3.5330 | 5250 | 0.8651 | - |
|
425 |
+
| 3.5666 | 5300 | 0.8943 | 0.9882 (-0.0118) |
|
426 |
+
| 3.6003 | 5350 | 0.8535 | - |
|
427 |
+
| 3.6339 | 5400 | 0.8687 | 0.9882 (-0.0118) |
|
428 |
+
| 3.6676 | 5450 | 0.9213 | - |
|
429 |
+
| 3.7012 | 5500 | 0.887 | 0.9882 (-0.0118) |
|
430 |
+
| 3.7349 | 5550 | 0.8787 | - |
|
431 |
+
| 3.7685 | 5600 | 0.8466 | 0.9882 (-0.0118) |
|
432 |
+
| 3.8022 | 5650 | 0.8517 | - |
|
433 |
+
| 3.8358 | 5700 | 0.8349 | 0.9883 (-0.0117) |
|
434 |
+
| 3.8694 | 5750 | 0.8647 | - |
|
435 |
+
| 3.9031 | 5800 | 0.8406 | 0.9882 (-0.0118) |
|
436 |
+
| 3.9367 | 5850 | 0.8385 | - |
|
437 |
+
| 3.9704 | 5900 | 0.8631 | 0.9882 (-0.0118) |
|
438 |
+
| 4.0040 | 5950 | 0.823 | - |
|
439 |
+
| 4.0377 | 6000 | 0.9163 | 0.9881 (-0.0119) |
|
440 |
+
| 4.0713 | 6050 | 0.8373 | - |
|
441 |
+
| 4.1050 | 6100 | 0.892 | 0.9882 (-0.0118) |
|
442 |
+
| 4.1386 | 6150 | 0.8666 | - |
|
443 |
+
| 4.1723 | 6200 | 0.8536 | 0.9882 (-0.0118) |
|
444 |
+
| 4.2059 | 6250 | 0.8784 | - |
|
445 |
+
| 4.2396 | 6300 | 0.9616 | 0.9882 (-0.0118) |
|
446 |
+
| 4.2732 | 6350 | 0.8464 | - |
|
447 |
+
| 4.3069 | 6400 | 0.865 | 0.9882 (-0.0118) |
|
448 |
+
| 4.3405 | 6450 | 0.8411 | - |
|
449 |
+
| 4.3742 | 6500 | 0.8943 | 0.9882 (-0.0118) |
|
450 |
+
| 4.4078 | 6550 | 0.8577 | - |
|
451 |
+
| 4.4415 | 6600 | 0.8683 | 0.9882 (-0.0118) |
|
452 |
+
| 4.4751 | 6650 | 0.8706 | - |
|
453 |
+
| 4.5087 | 6700 | 0.8645 | 0.9882 (-0.0118) |
|
454 |
+
| 4.5424 | 6750 | 0.8899 | - |
|
455 |
+
| 4.5760 | 6800 | 0.8593 | 0.9882 (-0.0118) |
|
456 |
+
| 4.6097 | 6850 | 0.8838 | - |
|
457 |
+
| 4.6433 | 6900 | 0.8379 | 0.9882 (-0.0118) |
|
458 |
+
| 4.6770 | 6950 | 0.8759 | - |
|
459 |
+
| 4.7106 | 7000 | 0.8608 | 0.9882 (-0.0118) |
|
460 |
+
| 4.7443 | 7050 | 0.8858 | - |
|
461 |
+
| 4.7779 | 7100 | 0.8594 | 0.9882 (-0.0118) |
|
462 |
+
| 4.8116 | 7150 | 0.8403 | - |
|
463 |
+
| 4.8452 | 7200 | 0.8898 | 0.9882 (-0.0118) |
|
464 |
+
| 4.8789 | 7250 | 0.8382 | - |
|
465 |
+
| 4.9125 | 7300 | 0.8307 | 0.9882 (-0.0118) |
|
466 |
+
| 4.9462 | 7350 | 0.8601 | - |
|
467 |
+
| 4.9798 | 7400 | 0.8076 | 0.9882 (-0.0118) |
|
468 |
+
|
469 |
+
* The bold row denotes the saved checkpoint.
|
470 |
+
</details>
|
471 |
+
|
472 |
+
### Framework Versions
|
473 |
+
- Python: 3.13.2
|
474 |
+
- Sentence Transformers: 4.1.0
|
475 |
+
- Transformers: 4.51.3
|
476 |
+
- PyTorch: 2.7.0+cu128
|
477 |
+
- Accelerate: 1.6.0
|
478 |
+
- Datasets: 3.6.0
|
479 |
+
- Tokenizers: 0.21.1
|
480 |
+
|
481 |
+
## Citation
|
482 |
+
|
483 |
+
### BibTeX
|
484 |
+
|
485 |
+
#### Sentence Transformers
|
486 |
+
```bibtex
|
487 |
+
@inproceedings{reimers-2019-sentence-bert,
|
488 |
+
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
|
489 |
+
author = "Reimers, Nils and Gurevych, Iryna",
|
490 |
+
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
|
491 |
+
month = "11",
|
492 |
+
year = "2019",
|
493 |
+
publisher = "Association for Computational Linguistics",
|
494 |
+
url = "https://arxiv.org/abs/1908.10084",
|
495 |
+
}
|
496 |
+
```
|
497 |
+
|
498 |
+
<!--
|
499 |
+
## Glossary
|
500 |
+
|
501 |
+
*Clearly define terms in order to be accessible across audiences.*
|
502 |
+
-->
|
503 |
+
|
504 |
+
<!--
|
505 |
+
## Model Card Authors
|
506 |
+
|
507 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
508 |
+
-->
|
509 |
+
|
510 |
+
<!--
|
511 |
+
## Model Card Contact
|
512 |
+
|
513 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
514 |
+
-->
|
config.json
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"BertForSequenceClassification"
|
4 |
+
],
|
5 |
+
"attention_probs_dropout_prob": 0.1,
|
6 |
+
"classifier_dropout": null,
|
7 |
+
"gradient_checkpointing": false,
|
8 |
+
"hidden_act": "gelu",
|
9 |
+
"hidden_dropout_prob": 0.1,
|
10 |
+
"hidden_size": 384,
|
11 |
+
"id2label": {
|
12 |
+
"0": "LABEL_0"
|
13 |
+
},
|
14 |
+
"initializer_range": 0.02,
|
15 |
+
"intermediate_size": 1536,
|
16 |
+
"label2id": {
|
17 |
+
"LABEL_0": 0
|
18 |
+
},
|
19 |
+
"layer_norm_eps": 1e-12,
|
20 |
+
"max_position_embeddings": 512,
|
21 |
+
"model_type": "bert",
|
22 |
+
"num_attention_heads": 12,
|
23 |
+
"num_hidden_layers": 12,
|
24 |
+
"pad_token_id": 0,
|
25 |
+
"position_embedding_type": "absolute",
|
26 |
+
"sbert_ce_default_activation_function": "torch.nn.modules.linear.Identity",
|
27 |
+
"sentence_transformers": {
|
28 |
+
"activation_fn": "torch.nn.modules.activation.Sigmoid",
|
29 |
+
"version": "4.1.0"
|
30 |
+
},
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"transformers_version": "4.51.3",
|
33 |
+
"type_vocab_size": 2,
|
34 |
+
"use_cache": true,
|
35 |
+
"vocab_size": 30522
|
36 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cef312f794c75951055c03231a2b1e07598f7e5e332f8bacfa93f3500a610ced
|
3 |
+
size 133464836
|
special_tokens_map.json
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": {
|
3 |
+
"content": "[CLS]",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": false,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"mask_token": {
|
10 |
+
"content": "[MASK]",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": {
|
17 |
+
"content": "[PAD]",
|
18 |
+
"lstrip": false,
|
19 |
+
"normalized": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"single_word": false
|
22 |
+
},
|
23 |
+
"sep_token": {
|
24 |
+
"content": "[SEP]",
|
25 |
+
"lstrip": false,
|
26 |
+
"normalized": false,
|
27 |
+
"rstrip": false,
|
28 |
+
"single_word": false
|
29 |
+
},
|
30 |
+
"unk_token": {
|
31 |
+
"content": "[UNK]",
|
32 |
+
"lstrip": false,
|
33 |
+
"normalized": false,
|
34 |
+
"rstrip": false,
|
35 |
+
"single_word": false
|
36 |
+
}
|
37 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"100": {
|
12 |
+
"content": "[UNK]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"101": {
|
20 |
+
"content": "[CLS]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"102": {
|
28 |
+
"content": "[SEP]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"103": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"clean_up_tokenization_spaces": true,
|
45 |
+
"cls_token": "[CLS]",
|
46 |
+
"do_basic_tokenize": true,
|
47 |
+
"do_lower_case": true,
|
48 |
+
"extra_special_tokens": {},
|
49 |
+
"mask_token": "[MASK]",
|
50 |
+
"model_max_length": 512,
|
51 |
+
"never_split": null,
|
52 |
+
"pad_token": "[PAD]",
|
53 |
+
"sep_token": "[SEP]",
|
54 |
+
"strip_accents": null,
|
55 |
+
"tokenize_chinese_chars": true,
|
56 |
+
"tokenizer_class": "BertTokenizer",
|
57 |
+
"unk_token": "[UNK]"
|
58 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|