File size: 2,225 Bytes
bb6157d 21e3cec bb6157d 21e3cec bb6157d 21e3cec |
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 |
---
library_name: transformers
tags:
- generated_from_trainer
model-index:
- name: bert-reg-biencoder-mse
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert-reg-biencoder-mse
This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0817
- Mse: 0.0812
- Mae: 0.2278
- Pearson Corr: 0.2835
- Spearman Corr: 0.2331
- Cosine Sim: 0.9097
## Model description
More information needed
## Intended uses & limitations
More information needed
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 7
### Training results
| Training Loss | Epoch | Step | Validation Loss | Mse | Mae | Pearson Corr | Spearman Corr | Cosine Sim |
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------------:|:-------------:|:----------:|
| 0.1219 | 1.0 | 21 | 0.1124 | 0.1117 | 0.2560 | 0.1406 | 0.0993 | 0.9055 |
| 0.1017 | 2.0 | 42 | 0.0838 | 0.0833 | 0.2248 | 0.1312 | 0.1239 | 0.9045 |
| 0.0872 | 3.0 | 63 | 0.0778 | 0.0775 | 0.2205 | 0.2520 | 0.1374 | 0.9097 |
| 0.0694 | 4.0 | 84 | 0.0860 | 0.0856 | 0.2328 | 0.1923 | 0.1456 | 0.9037 |
| 0.0533 | 5.0 | 105 | 0.0958 | 0.0951 | 0.2418 | 0.3089 | 0.2252 | 0.9132 |
| 0.0478 | 6.0 | 126 | 0.0782 | 0.0778 | 0.2216 | 0.2913 | 0.2325 | 0.9096 |
| 0.0385 | 7.0 | 147 | 0.0817 | 0.0812 | 0.2278 | 0.2835 | 0.2331 | 0.9097 |
### Framework versions
- Transformers 4.45.1
- Pytorch 2.4.0
- Datasets 3.0.1
- Tokenizers 0.20.0
|