Classification
- Print
- PDF
Classification
- Print
- PDF
Article Summary
Share feedback
Thanks for sharing your feedback!
Available in Classic and VPC
Introduces referencing examples when classifying entered text according to given settings.
Note
- For details on prompt input and result value, please refer to Playground.
- Recommendations when entering prompts
- Please enter more than one pair of samples into the prompt for better accuracy.
- When entering the contents for creating the result value, make sure no space at the end of the text.
- The generated words and sentences are examples, and the generated sentences may vary each time when the prompt is executed.
Emotion classification
Describe how to set up parameters and configure prompts to analyze which emotions are detected in a given sentence.
Set parameter
We recommend to set the parameters as follows.
Parameter | Value |
---|---|
Maximum token | 30 or lower |
Temperature | 0.5 |
Repetition penalty | 5 |
Stop sequence | ###↵, emotion:, sentence: |
Inject start text | ↵emotion: |
Inject restart text | ↵###↵sentence: |
Configure prompt
Build the prompt as follows.
//You should enter directives to let the generator know the context of what you want.
This is a sentence emotion analyzer.
//Please enter a sample sentence and the emotion of the sample sentence.
Sentence: “I feel really great.”
Emotion: Positive
###
Sentence: “It’s really annoying.”
Emotion: Negative
###
//Please enter a sentence to analyze the emotion.
Sentence: “I’ll send this.”
Result value
The result of executing the prompt is as follows.
This is a sentence emotion analyzer.
Sentence: “I feel really great.”
Emotion: Positive
###
Sentence: “It’s really annoying.”
Emotion: Negative
###
Sentence: “I’ll send this.”
//Output result value
Emotion: Neutral
###
Sentence:
Connect to a medical department
The following describes how to set parameters and enter prompts to analyze a given symptom and point you to the right hospital department.
Set parameter
We recommend to set the parameters as follows.
Parameter | Value |
---|---|
Maximum token | 30 or lower |
Temperature | 0.5 |
Repetition penalty | 5 |
Stop sequence | ###↵, medical department:, symptom: |
Inject start text | ↵medical department: |
Inject restart text | ↵###↵symptom: |
Configure prompt
Build the prompt as follows.
//You should enter directives to let the generator know the context of what you want.
Here we guide you to the appropriate hospital department based on symptoms.
//Please enter sample symptoms and an appropriate department for your symptoms.
Symptom: I have a stuffy nose.
Medical department: Otorhinolaryngology
###
Symptom: I have vocal nodules.
Medical department: Otorhinolaryngology
###
//Please enter the symptoms to check the medical department.
Symptom: I have varicose veins.
Result value
The result of executing the prompt is as follows.
Here we guide you to the appropriate hospital department based on symptoms.
Symptom: I have a stuffy nose.
Medical department: Otorhinolaryngology
###
Symptom: I have vocal nodules.
Medical department: Otorhinolaryngology
###
Symptom: I have varicose veins.
//Result value
Medical department: Surgery
###
Symptom:
Was this article helpful?