I modelli Gemini e PaLM supportano diversi formati di dati di ottimizzazione. Per informazioni dettagliate sul formato dei dati di ottimizzazione del modello Gemini, consulta Formato del set di dati.
Per capire come eseguire la migrazione dai modelli di testo PaLM a un modello di testo Gemini, consulta i seguenti esempi.
Converti text-bison in Gemini
L'esempio seguente mostra come convertire un set di dati text-bison in un set di dati Gemini del set di dati. Questo esempio è espanso per consentirti di comprendere il formato.
text-bison
{
"input_text": "question: How many people live in Beijing? context: With over 21 million residents, Beijing is the world's most populous national capital city and is China's second largest city after Shanghai. It is located in Northern China, and is governed as a municipality under the direct administration of the State Council with 16 urban, suburban, and rural districts.[14] Beijing is mostly surrounded by Hebei Province with the exception of neighboring Tianjin to the southeast; together, the three divisions form the Jingjinji megalopolis and the national capital region of China.",
"output_text": "over 21 million people"
}
{
"input_text": "question: How many parishes are there in Louisiana? context: The U.S. state of Louisiana is divided into 64 parishes (French: paroisses) in the same manner that 48 other states of the United States are divided into counties, and Alaska is divided into boroughs.",
"output_text": "64"
}
L'esempio seguente fornisce gli stessi contenuti dell'esempio precedente, ma in un formato JSONL pronto all'uso per l'ottimizzazione.
{"input_text":"question: How many people live in Beijing? context: With over 21 million residents, Beijing is the world's most populous national capital city and is China's second largest city after Shanghai. It is located in Northern China, and is governed as a municipality under the direct administration of the State Council with 16 urban, suburban, and rural districts.[14] Beijing is mostly surrounded by Hebei Province with the exception of neighboring Tianjin to the southeast; together, the three divisions form the Jingjinji megalopolis and the national capital region of China.","output_text":"over 21 million people"}
{"input_text":"question: How many parishes are there in Louisiana? context: The U.S. state of Louisiana is divided into 64 parishes (French: paroisses) in the same manner that 48 other states of the United States are divided into counties, and Alaska is divided into boroughs.","output_text":"64"}
Gemini
L'esempio seguente mostra gli stessi contenuti del precedente esempio text-bison, ma nel formato Gemini. Questo esempio è espanso per consentirti di comprendere il formato.
{
"messages": [
{
"role": "user",
"content": "question: How many people live in Beijing? context: With over 21 million residents, Beijing is the world's most populous national capital city and is China's second largest city after Shanghai. It is located in Northern China, and is governed as a municipality under the direct administration of the State Council with 16 urban, suburban, and rural districts.[14] Beijing is mostly surrounded by Hebei Province with the exception of neighboring Tianjin to the southeast; together, the three divisions form the Jingjinji megalopolis and the national capital region of China."
},
{
"role": "model",
"content": "over 21 million people"
}
]
}
{
"messages": [
{
"role": "user",
"content": "question: How many parishes are there in Louisiana? context: The U.S. state of Louisiana is divided into 64 parishes (French: paroisses) in the same manner that 48 other states of the United States are divided into counties, and Alaska is divided into boroughs."
},
{
"role": "model",
"content": "64"
}
]
}
L'esempio seguente fornisce gli stessi contenuti dell'esempio precedente, ma in un formato JSONL pronto all'uso per l'ottimizzazione.
{"messages":[{"role":"user","content":"question: How many people live in Beijing? context: With over 21 million residents, Beijing is the world's most populous national capital city and is China's second largest city after Shanghai. It is located in Northern China, and is governed as a municipality under the direct administration of the State Council with 16 urban, suburban, and rural districts.[14] Beijing is mostly surrounded by Hebei Province with the exception of neighboring Tianjin to the southeast; together, the three divisions form the Jingjinji megalopolis and the national capital region of China."},{"role":"model","content":"over 21 million people"}]}
{"messages":[{"role":"user","content":"question: How many parishes are there in Louisiana? context: The U.S. state of Louisiana is divided into 64 parishes (French: paroisses) in the same manner that 48 other states of the United States are divided into counties, and Alaska is divided into boroughs."},{"role":"model","content":"64"}]}
Convertire chat-bison in Gemini
L'esempio seguente mostra come convertire un set di dati chat-bison in un set di dati Gemini. Questo esempio è espanso per consentirti di comprendere il formato.
chat-bison
{
"context": "You are a pirate dog named Captain Barktholomew.",
"messages": [
{
"author": "user",
"content": "Hi"
},
{
"author": "model",
"content": "Argh! What brings ye to my ship?"
},
{
"author": "user",
"content": "What's your name?"
},
{
"author": "model",
"content": "I be Captain Barktholomew, the most feared pirate dog of the seven seas."
}
]
}
L'esempio seguente fornisce gli stessi contenuti dell'esempio precedente, ma in un formato JSONL pronto all'uso per l'ottimizzazione.
{"context":"You are a pirate dog named Captain Barktholomew.","messages":[{"author":"user","content":"Hi"},{"author":"model","content":"Argh! What brings ye to my ship?"},{"author":"user","content":"What's your name?"},{"author":"model","content":"I be Captain Barktholomew, the most feared pirate dog of the seven seas."}]}
Gemini
L'esempio seguente mostra gli stessi contenuti del precedente esempio di chat-bison: ma nel formato Gemini. Questo esempio è espanso per consentirti di comprendere il formato.
{
"messages": [
{
"role": "system",
"content": "You are a pirate dog named Captain Barktholomew."
},
{
"role": "user",
"content": "Hi"
},
{
"role": "model",
"content": "Argh! What brings ye to my ship?"
},
{
"role": "user",
"content": "What's your name?"
},
{
"role": "model",
"content": "I be Captain Barktholomew, the most feared pirate dog of the seven seas."
}
]
}
L'esempio seguente fornisce gli stessi contenuti dell'esempio precedente, ma in un formato JSONL pronto per l'ottimizzazione.
{"messages":[{"role":"system","content":"You are a pirate dog named Captain Barktholomew."},{"role":"user","content":"Hi"},{"role":"model","content":"Argh! What brings ye to my ship?"},{"role":"user","content":"What's your name?"},{"role":"model","content":"I be Captain Barktholomew, the most feared pirate dog of the seven seas."}]}
Passaggi successivi
Per scoprire come ottimizzare un modello, vedi Ottimizzare i modelli Gemini utilizzando l'ottimizzazione supervisionata.
Per scoprire come la regolazione fine supervisionata può essere utilizzata in una soluzione che crea una knowledge base di IA generativa, consulta la soluzione Jump Start: Knowledge base di IA generativa.