JsonL Format for Direct Import into Prodigy (generated by LLM from elsewhere) [Newbie Question]

[grahama1970]

Hello. I'm new to Prodigy and SpanCat.
I would like to

• directly import TRAINING_DATA that is generated by an LLM with python
• Directly import that formatted list into Prodigy for touching up labeling errors. B/C prodigy has such a nice interface :)

Question 1: Is the below JSONL the correct format?
Question 2: Are the token attributes necessary for Prodigy to recognize the targeted spans?
Question 3: Is there a complete and simple/intro example somewhere that using SpanCat and EntityRecognition (perhaps off-topic)?

Much appreciation in advance.

{
  "text": "First look at the new MacBook Pro",
  "spans": [
    {"start": 22, "end": 33, "label": "PRODUCT", "token_start": 5, "token_end": 6}
  ],
  "tokens": [
    {"text": "First", "start": 0, "end": 5, "id": 0},
    {"text": "look", "start": 6, "end": 10, "id": 1},
    {"text": "at", "start": 11, "end": 13, "id": 2},
    {"text": "the", "start": 14, "end": 17, "id": 3},
    {"text": "new", "start": 18, "end": 21, "id": 4},
    {"text": "MacBook", "start": 22, "end": 29, "id": 5},
    {"text": "Pro", "start": 30, "end": 33, "id": 6}
  ]
}

[grahama1970]

I got it