File size: 834 Bytes
daac535
 
 
 
 
96dfb72
 
019c7d2
 
 
96dfb72
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
019c7d2
96dfb72
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
---
library_name: transformers
tags: []
---

# Labor Saving Stated Aim Classifier

This is a roberta-base model that is trained to classify whether a set of explicit stated aims extracted from a British [historical patent](https://huggingface.co/datasets/matthewleechen/300YearsOfBritishPatents) includes a labor-saving objective. 

Labels were manually generated and then checked with Gemini 2.0 Flash with the attached [prompt](https://huggingface.co/matthewleechen/labor-saving_stated_aim_classifier/blob/main/labor-saving_prompt.txt).

Hyperparameters:
lr = 6e-5
batch size = 128

Test set results:

```text
{'eval_loss': 0.28036537766456604,
 'eval_accuracy': 0.9,
 'eval_precision': 0.9,
 'eval_recall': 0.9,
 'eval_f1': 0.9,
 'eval_runtime': 0.4135,
 'eval_samples_per_second': 241.832,
 'eval_steps_per_second': 2.418}
```