turn pair counting not appropriately reflected in code

[jonmay]

When the seed dialogue is 2x the length of the number of turns to run no turns are offered to the user, even though the logging doesn't indicate this will happen.

setting:

partial log:

Choosing the "blender_90M" model for the bot.
Creating ModelChatWorld for tag conversation_id 53 with 6 turns.
ModelChatWorld:conversation_id 53: is at turn 0, with 6 pairs of turns needed...
Use custom dialogue seeds to start conversations with some context
Context info: {'conversation': [{'speaker_id': 's1', 'text': 't1'}, {'speaker_id': 's2', 'text': 't2'}, {'speaker_id': 's3', 'text': 't3'}, {'speaker_id': 's4', 'text': 't4'}, {'speaker_id': 's5', 'text': 't5'}, {'speaker_id': 's6', 'text': 't6'}, {'speaker_id': 's7', 'text': 't7'}, {'speaker_id': 's8', 'text': 't8'}, {'speaker_id': 's9', 'text': 't9'}, {'speaker_id': 's10', 'text': 't10'}, {'speaker_id': 's11', 'text': 't11'}, {'speaker_id': 's12', 'text': 't12'}], 'target_user': 's12'}

(skipped 12 statements like:)

17:54:19 | TurkLikeAgent: In observe() before semaphore, self.turn_idx is 0 and observation is {'episode_done': False, 'id': 's1', 'text': 't1', 'fake_start': True, 'agent_idx': 1, 'update_id': '71de604b-0666-47f1-b6dc-e781cbbd533e'}
17:54:19 | TurkLikeAgent: In observe() AFTER semaphore, self.turn_idx: 0, observation["text"]: t1

...

ModelChatWorld:conversation_id 53: is at turn 1, with 6 pairs of turns needed...
ModelChatWorld:conversation_id 53: About to act with task turn idx: 1

but the gui skips directly to conversation evaluation.

I was using this test conversation:

    {
        "conversation":
        [
            {
                "speaker_id": "s1",
                "text": "t1"
            },
            {
                "speaker_id": "s2",
                "text": "t2"
            },
            {
                "speaker_id": "s3",
                "text": "t3"
            },
            {
                "speaker_id": "s4",
                "text": "t4"
            },
            {
                "speaker_id": "s5",
                "text": "t5"
            },
            {
                "speaker_id": "s6",
                "text": "t6"
            },
                   {
                "speaker_id": "s7",
                "text": "t7"
                   },
                   {
                "speaker_id": "s8",
                "text": "t8"
                   },
                   {
                "speaker_id": "s9",
                "text": "t9"
                   },
                               {
                "speaker_id": "s10",
                "text": "t10"
                               },
                               {
                "speaker_id": "s11",
                "text": "t11"
                               },
                               {
                "speaker_id": "s12",
                "text": "t12"
            }
        ],
        "target_user": "s12"
    }