File size: 2,268 Bytes
8758d51
522b4ff
 
8758d51
5a4f0d1
8758d51
a4570fe
87f2b37
8758d51
 
829bc8b
220304b
87f2b37
220304b
87f2b37
 
 
 
77f9e05
87f2b37
77f9e05
 
 
 
 
 
 
 
 
 
 
 
 
 
87f2b37
 
d9c50cc
b578004
 
2a3cc95
 
b578004
092bbad
 
 
 
 
b578004
092bbad
b578004
 
 
 
 
d9c50cc
 
87f2b37
 
 
 
 
 
 
 
 
2c197b8
 
 
8758d51
 
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
---
tags: 
  - salesken
license: apache-2.0
inference: true
datasets: google_wellformed_query
widget:
  text: "She present paper today"
---

**model name**: Query Wellformedness Scoring

**Description** : Evaluate the well-formedness of sentences by checking grammatical correctness and completeness. Sensitive to case and penalizes sentences for incorrect grammar and case.

**Features**:
  - *Wellformedness Score*: Provides a score indicating grammatical correctness and completeness.
  - *Case Sensitivity*: Recognizes and penalizes incorrect casing in sentences.
  - *Broad Applicability*: Can be used on a wide range of sentences.

**Example**:
1. Dogs are mammals.
2. she loves to read books on history.
3. When the rain in Spain.
4. Eating apples are healthy for you.
5. The Eiffel Tower is in Paris.

Among these sentences:

Sentences 1 and 5 are well-formed and have correct grammar and case.
Sentence 2 starts with a lowercase letter.
Sentence 3 is a fragment and is not well-formed.
Sentence 4 has a subject-verb agreement error.


**example_usage:**
*library: HuggingFace transformers*
```python
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("Ashishkr/query_wellformedness_score")
model = AutoModelForSequenceClassification.from_pretrained("Ashishkr/query_wellformedness_score")
sentences = [
    "The quarterly financial report are showing an increase.",  # Incorrect
    "Him has completed the audit for last fiscal year.",  # Incorrect
    "Please to inform the board about the recent developments.",  # Incorrect
    "The team successfully achieved all its targets for the last quarter.",  # Correct
    "Our company is exploring new ventures in the European market."  # Correct
]

features = tokenizer(sentences, padding=True, truncation=True, return_tensors="pt")
model.eval()
with torch.no_grad():
    scores = model(**features).logits
print(scores)
```

**Intended Use Cases**

*Content Creation*: Validate the well-formedness of written content.

*Educational Platforms*: Helps students check the grammaticality of their sentences.

*Chatbots & Virtual Assistants*: To validate user queries or generate well-formed responses.

  contact: ashish@salesken.ai