EditChatLogRequestSchema

idintegerrequired

Numeric ID of the training log entry to update.

Example: 1001

kwargsobjectrequired

Additional metadata key-value pairs to attach to the log entry.

messagesstringrequired

Replacement messages content as a JSON-encoded string. Each message must have role and content fields. The total token count must not exceed 8,192 tokens.

Example: [{"role": "user", "content": "What is 2+2?"}, {"role": "assistant", "content": "4"}]

modelstringrequired

Model name to associate with this log entry.

tagsstring[]required

Tags to apply to this log entry.

has_been_synthesizedboolean

Marks whether this log entry has already been used for synthetic data generation.

Default: null

synthetic_messagesobject[]

Synthetic version of the conversation messages for this entry.

Default: null

usageobject

Token usage statistics for this log entry (prompt tokens, completion tokens, total tokens).

Default: null

EditChatLogRequestSchema

{
  "id": 1001,
  "messages": "[{\"role\": \"user\", \"content\": \"What is 2+2?\"}, {\"role\": \"assistant\", \"content\": \"4\"}]",
  "model": "Llama-3.3-70B-Instruct",
  "kwargs": {},
  "tags": [
    "apitest"
  ]
}