Perpustakaan berat ringan sederhana yang membungkus AI AI terbuka dengan dukungan untuk injeksi ketergantungan.
Fitur yang didukung meliputi:
Untuk penelusuran video aplikasi web Blazor yang dibangun di perpustakaan ini, silakan lihat:
Ini digunakan ke halaman GitHub dan tersedia di: Buka AI UI. Sumber untuk aplikasi web Blazor ada di whetstone.chatgpt.blazor.app.
Contohnya termasuk:
services . Configure < ChatGPTCredentials > ( options =>
{
options . ApiKey = " YOURAPIKEY " ;
options . Organization = " YOURORGANIZATIONID " ;
} ) ;Menggunakan:
services . AddHttpClient ( ) ;atau:
services . AddHttpClient < IChatGPTClient , ChatGPTClient > ( ) ; Konfigurasikan Layanan IChatGPTClient :
services . AddScoped < IChatGPTClient , ChatGPTClient > ( ) ; Penyelesaian obrolan adalah jenis penyelesaian khusus yang dioptimalkan untuk obrolan. Mereka dirancang untuk digunakan dalam konteks percakapan.
Ini menunjukkan penggunaan model turbo GPT-3.5.
using Whetstone . ChatGPT ;
using Whetstone . ChatGPT . Models ;
. . .
var gptRequest = new ChatGPTChatCompletionRequest
{
Model = ChatGPT35Models . Turbo ,
Messages = new List < ChatGPTChatCompletionMessage > ( )
{
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . System , " You are a helpful assistant. " ) ,
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . User , " Who won the world series in 2020? " ) ,
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . Assistant , " The Los Angeles Dodgers won the World Series in 2020. " ) ,
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . User , " Where was it played? " )
} ,
Temperature = 0.9f ,
MaxTokens = 100
} ;
using IChatGPTClient client = new ChatGPTClient ( " YOURAPIKEY " ) ;
var response = await client . CreateChatCompletionAsync ( gptRequest ) ;
Console . WriteLine ( response ? . GetCompletionText ( ) ) ;Model GPT-4 juga dapat digunakan asalkan akun Anda telah diberikan akses ke beta terbatas.
Penyelesaian menggunakan model untuk menjawab berbagai tugas, termasuk tetapi tidak terbatas pada klasifikasi, analisis sentimen, menjawab pertanyaan, dll.
Ini menunjukkan penggunaan langsung model GPT-3.5-turbo-instruct tanpa petunjuk.
Harap dicatat, CreateCompletionAsync sudah usang. Gunakan chatgptChatCompletionRequest , chatgptChtCompletionResponse , dan metode CreateChatCompletionAsync sebagai gantinya.
using Whetstone . ChatGPT ;
using Whetstone . ChatGPT . Models ;
. . .
var gptRequest = new ChatGPTCompletionRequest
{
Model = ChatGPT35Models . Gpt35TurboInstruct ,
Prompt = " How is the weather? "
} ;
using IChatGPTClient client = new ChatGPTClient ( " YOURAPIKEY " ) ;
var response = await client . CreateCompletionAsync ( gptRequest ) ;
Console . WriteLine ( response . GetCompletionText ( ) ) ;GPT-3.5 tidak deterministik. Salah satu uji coba sampel di atas dikembalikan:
Cuacanya sangat bervariasi tergantung pada lokasi. Secara umum, Anda dapat mengharapkan suhu menjadi sedang dan iklim menjadi nyaman, tetapi selalu sebaiknya memeriksa perkiraan untuk area spesifik Anda.
Aplikasi Konsol AC# yang menggunakan penyelesaian tersedia di:
Whetstone.chatgpt.commandlineBot (chatgpt-marv)
Sampel ini termasuk:
Cara membuat unggahan file tuning baru.
List < ChatGPTTurboFineTuneLine > tuningInput = new ( )
{
new ChatGPTTurboFineTuneLine ( )
{
Messages = new List < ChatGPTTurboFineTuneLineMessage > ( )
{
new ( ChatGPTMessageRoles . System , " Marv is a factual chatbot that is also sarcastic. " ) ,
new ( ChatGPTMessageRoles . User , " What's the capital of France? " ) ,
new ( ChatGPTMessageRoles . Assistant , " Paris, as if everyone doesn't know that already. " )
} ,
} ,
new ChatGPTTurboFineTuneLine ( )
{
Messages = new List < ChatGPTTurboFineTuneLineMessage > ( )
{
new ( ChatGPTMessageRoles . System , " Marv is a factual chatbot that is also sarcastic. " ) ,
new ( ChatGPTMessageRoles . User , " Who wrote 'Romeo and Juliet'? " ) ,
new ( ChatGPTMessageRoles . Assistant , " Oh, just some guy named William Shakespeare. Ever heard of him? " )
} ,
} ,
. . .
} ;
byte [ ] tuningText = tuningInput . ToJsonLBinary ( ) ;
string fileName = " marvin.jsonl " ;
ChatGPTUploadFileRequest ? uploadRequest = new ChatGPTUploadFileRequest
{
File = new ChatGPTFileContent
{
FileName = fileName ,
Content = tuningText
}
} ;
ChatGPTFileInfo ? newTurboTestFile ;
using ( IChatGPTClient client = new ChatGPTClient ( " YOURAPIKEY " ) )
{
newTurboTestFile = await client . UploadFileAsync ( uploadRequest ) ;
} Setelah file dibuat, dapatkan FileID, dan referensi saat membuat fine tuning baru.
IChatGPTClient client = new ChatGPTClient ( " YOURAPIKEY " ) ;
uploadedFileInfo = await client . UploadFileAsync ( uploadRequest ) ;
var fileList = await client . ListFilesAsync ( ) ;
var foundFile = fileList ? . Data ? . First ( x => x . Filename . Equals ( " marvin.jsonl " ) ) ;
ChatGPTCreateFineTuneRequest tuningRequest = new ChatGPTCreateFineTuneRequest
{
TrainingFileId = foundFile ? . Id ,
Model = " gpt-3.5-turbo-1106 "
} ;
ChatGPTFineTuneJob ? tuneResponse = await client . CreateFineTuneAsync ( tuningRequest ) ;
string ? fineTuneId = tuneResponse ? . Id ;Memproses permintaan tuning fine akan memakan waktu. Setelah selesai, status akan melaporkan "berhasil" dan siap digunakan dalam permintaan penyelesaian.
using IChatGPTClient client = new ChatGPTClient ( " YOURAPIKEY " ) ;
ChatGPTFineTuneJob ? tuneResponse = await client . RetrieveFineTuneAsync ( " FINETUNEID " ) ;
if ( tuneResponse . Status . Equals ( " succeeded " ) )
{
var gptRequest = new ChatGPTChatCompletionRequest
{
Model = " FINETUNEID " ,
Messages = new List < ChatGPTChatCompletionMessage > ( )
{
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . System , " You are a helpful assistant. " ) ,
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . User , " Who won the world series in 2020? " ) ,
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . Assistant , " The Los Angeles Dodgers won the World Series in 2020. " ) ,
new ChatGPTChatCompletionMessage ( ChatGPTMessageRoles . User , " Where was it played? " )
} ,
Temperature = 0.9f ,
MaxTokens = 100
} ;
var response = await client . CreateChatCompletionAsync ( gptRequest ) ;
Console . WriteLine ( response ? . GetCompletionText ( ) ) ; Berikut adalah contoh yang menghasilkan gambar 1024x1024.
ChatGPTCreateImageRequest imageRequest = new ( )
{
Prompt = " A sail boat " ,
Size = CreatedImageSize . Size1024 ,
ResponseFormat = CreatedImageFormat . Base64
} ;
using IChatGPTClient client = new ChatGPTClient ( " YOURAPIKEY " ) ;
ChatGPTImageResponse ? imageResponse = await client . CreateImageAsync ( imageRequest ) ;
var imageData = imageResponse ? . Data ? [ 0 ] ;
if ( imageData != null )
{
byte [ ] ? imageBytes = await client . DownloadImageAsync ( imageData ) ;
} Dia adalah contoh yang menyalin file audio menggunakan bisikan.
string audioFile = @"audiofilestranscriptiontest.mp3" ;
byte [ ] fileContents = File . ReadAllBytes ( audioFile ) ;
ChatGPTFileContent gptFile = new ChatGPTFileContent
{
FileName = audioFile ,
Content = fileContents
} ;
ChatGPTAudioTranscriptionRequest ? transcriptionRequest = new ChatGPTAudioTranscriptionRequest
{
File = gptFile
} ;
using IChatGPTClient client = new ChatGPTClient ( " YOURAPIKEY " ) ;
ChatGPTAudioResponse ? audioResponse = await client . CreateTranscriptionAsync ( transcriptionRequest , true ) ;
Console . WriteLine ( audioResponse ? . Text ) ;