File size: 1,953 Bytes
622b6e1 b5c808f 622b6e1 b5c808f 622b6e1 b5c808f 622b6e1 b5c808f 622b6e1 b5c808f 622b6e1 b5c808f 622b6e1 b5c808f 622b6e1 |
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 |
---
library_name: transformers
base_model: microsoft/infoxlm-large
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: vp-infoxlm-large-dsc
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. -->
# vp-infoxlm-large-dsc
This model is a fine-tuned version of [microsoft/infoxlm-large](https://huggingface.co/microsoft/infoxlm-large) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6113
- Accuracy: 0.8706
- F1: 0.8705
- Precision: 0.8713
- Recall: 0.8706
## 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: 1e-05
- train_batch_size: 8
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.05
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:------:|:---------:|:------:|
| 0.8771 | 1.0 | 3180 | 0.8099 | 0.6890 | 0.6914 | 0.7003 | 0.6890 |
| 0.5911 | 2.0 | 6360 | 0.5717 | 0.8014 | 0.8007 | 0.8107 | 0.8014 |
| 0.4608 | 3.0 | 9540 | 0.5323 | 0.8442 | 0.8442 | 0.8449 | 0.8442 |
| 0.407 | 4.0 | 12720 | 0.5047 | 0.8680 | 0.8679 | 0.8683 | 0.8680 |
| 0.3372 | 5.0 | 15900 | 0.6113 | 0.8706 | 0.8705 | 0.8713 | 0.8706 |
### Framework versions
- Transformers 4.44.2
- Pytorch 2.4.0
- Datasets 3.0.0
- Tokenizers 0.19.1
|