vjawa_test_readme_updates (#5)
Browse files- Add Readme.md to model card (10349c9657f59153f0cb3736dd709100b9d861f0)
- Add Readme.md to model card (1dca7875e59127b553ecde21e4c7bcce549d0b2e)
Co-authored-by: Vibhu Jawa <[email protected]>
- README.md +42 -3
- config.json +1 -0
README.md
CHANGED
@@ -3,7 +3,46 @@ tags:
|
|
3 |
- pytorch_model_hub_mixin
|
4 |
- model_hub_mixin
|
5 |
---
|
|
|
6 |
|
7 |
-
This
|
8 |
-
|
9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
- pytorch_model_hub_mixin
|
4 |
- model_hub_mixin
|
5 |
---
|
6 |
+
# nvidia/domain-classifier
|
7 |
|
8 |
+
This repository contains the code for the domain classifier model.
|
9 |
+
|
10 |
+
# How to use in transformers
|
11 |
+
To use the Domain classifier, use the following code:
|
12 |
+
|
13 |
+
```python3
|
14 |
+
|
15 |
+
import torch
|
16 |
+
from torch import nn
|
17 |
+
from transformers import AutoModel, AutoTokenizer, AutoConfig
|
18 |
+
from huggingface_hub import PyTorchModelHubMixin
|
19 |
+
|
20 |
+
class CustomModel(nn.Module, PyTorchModelHubMixin):
|
21 |
+
def __init__(self, config):
|
22 |
+
super(CustomModel, self).__init__()
|
23 |
+
self.model = AutoModel.from_pretrained(config['base_model'])
|
24 |
+
self.dropout = nn.Dropout(config['fc_dropout'])
|
25 |
+
self.fc = nn.Linear(self.model.config.hidden_size, len(config['id2label']))
|
26 |
+
|
27 |
+
def forward(self, input_ids, attention_mask):
|
28 |
+
features = self.model(input_ids=input_ids, attention_mask=attention_mask).last_hidden_state
|
29 |
+
dropped = self.dropout(features)
|
30 |
+
outputs = self.fc(dropped)
|
31 |
+
return torch.softmax(outputs[:, 0, :], dim=1)
|
32 |
+
|
33 |
+
# Setup configuration and model
|
34 |
+
config = AutoConfig.from_pretrained("nvidia/domain-classifier")
|
35 |
+
tokenizer = AutoTokenizer.from_pretrained("nvidia/domain-classifier")
|
36 |
+
model = CustomModel.from_pretrained("nvidia/domain-classifier")
|
37 |
+
|
38 |
+
# Prepare and process inputs
|
39 |
+
text_samples = ["Sports is a popular domain", "Politics is a popular domain"]
|
40 |
+
inputs = tokenizer(text_samples, return_tensors="pt", padding="longest", truncation=True)
|
41 |
+
outputs = model(inputs['input_ids'], inputs['attention_mask'])
|
42 |
+
|
43 |
+
# Predict and display results
|
44 |
+
predicted_classes = torch.argmax(outputs, dim=1)
|
45 |
+
predicted_domains = [config.id2label[class_idx.item()] for class_idx in predicted_classes.cpu().numpy()]
|
46 |
+
print(predicted_domains)
|
47 |
+
# ['Sports', 'News']
|
48 |
+
```
|
config.json
CHANGED
@@ -1,5 +1,6 @@
|
|
1 |
{
|
2 |
"base_model": "microsoft/deberta-v3-base",
|
|
|
3 |
"config_path": null,
|
4 |
"fc_dropout": 0.2,
|
5 |
"id2label": {
|
|
|
1 |
{
|
2 |
"base_model": "microsoft/deberta-v3-base",
|
3 |
+
"model_type": "deberta-v2",
|
4 |
"config_path": null,
|
5 |
"fc_dropout": 0.2,
|
6 |
"id2label": {
|