3lvj9jvERvs.json

{
  "id": "wyqxmnjbwmoa7a3f2eksuwljbq",
  "version": "91ee9c0c3df30478510ff8c8a3a545add1ad0259ad3a9f78fba57fbc05ee64f7",
  "input": {
    "audio": "https://upcdn.io/FW25b4F/raw/coding-train/3lvj9jvERvs.m4a"
  },
  "logs": "Transcribe with large-v2 model\nDetected language: English\n  0%|          | 0/78848 [00:00<?, ?frames/s]\n  3%|▎         | 2704/78848 [00:07<03:18, 383.58frames/s]\n  7%|▋         | 5640/78848 [00:13<02:58, 409.94frames/s]\n 10%|█         | 8096/78848 [00:22<03:23, 347.30frames/s]\n 13%|█▎        | 10408/78848 [00:27<03:01, 376.46frames/s]\n 17%|█▋        | 13264/78848 [00:35<02:55, 373.72frames/s]\n 20%|█▉        | 15744/78848 [00:41<02:43, 387.07frames/s]\n 24%|██▎       | 18544/78848 [00:51<02:54, 346.32frames/s]\n 27%|██▋       | 20968/78848 [00:56<02:36, 369.20frames/s]\n 30%|███       | 23904/78848 [01:02<02:15, 405.31frames/s]\n 34%|███▍      | 26848/78848 [01:10<02:13, 389.47frames/s]\n 38%|███▊      | 29624/78848 [01:18<02:09, 380.46frames/s]\n 41%|████      | 32192/78848 [01:26<02:09, 360.18frames/s]\n 44%|████▍     | 35016/78848 [01:34<02:01, 360.42frames/s]\n 48%|████▊     | 38008/78848 [01:43<01:57, 347.70frames/s]\n 52%|█████▏    | 40864/78848 [01:49<01:39, 381.71frames/s]\n 55%|█████▌    | 43760/78848 [02:00<01:43, 338.64frames/s]\n 59%|█████▉    | 46472/78848 [02:08<01:36, 334.62frames/s]\n 63%|██████▎   | 49296/78848 [02:16<01:28, 332.62frames/s]\n 66%|██████▋   | 52296/78848 [02:27<01:24, 315.66frames/s]\n 70%|██████▉   | 55056/78848 [02:37<01:18, 301.48frames/s]\n 73%|███████▎  | 57688/78848 [02:45<01:07, 313.02frames/s]\n 77%|███████▋  | 60560/78848 [02:54<00:57, 317.26frames/s]\n 80%|████████  | 63264/78848 [03:01<00:46, 332.37frames/s]\n 84%|████████▎ | 65984/78848 [03:07<00:35, 362.41frames/s]\n 87%|████████▋ | 68840/78848 [03:14<00:27, 367.10frames/s]\n 91%|█████████ | 71552/78848 [03:21<00:19, 382.63frames/s]\n 95%|█████████▍| 74528/78848 [03:32<00:12, 338.85frames/s]\n 98%|█████████▊| 77520/78848 [03:42<00:04, 320.64frames/s]\n100%|██████████| 78848/78848 [03:44<00:00, 357.34frames/s]\n100%|██████████| 78848/78848 [03:44<00:00, 351.97frames/s]\n",
  "output": {
    "detected_language": "english",
    "segments": [
      {
        "avg_logprob": -0.4077050024721803,
        "compression_ratio": 1.6401515151515151,
        "end": 6.640000000000001,
        "id": 0,
        "no_speech_prob": 0.0025907985400408506,
        "seek": 0,
        "start": 0,
        "temperature": 0,
        "text": " Welcome to part three of my neuro evolution flappy bird coding challenge.",
        "tokens": [
          50364,
          4027,
          281,
          644,
          1045,
          295,
          452,
          16499,
          9303,
          46338,
          7966,
          5255,
          17720,
          3430,
          13,
          50696
        ]
      },
      {
        "avg_logprob": -0.4077050024721803,
        "compression_ratio": 1.6401515151515151,
        "end": 9.6,
        "id": 1,
        "no_speech_prob": 0.0025907985400408506,
        "seek": 0,
        "start": 6.640000000000001,
        "temperature": 0,
        "text": " Alright, so, oh, come on, jump, jump, jump, oh yes!",
        "tokens": [
          50696,
          2798,
          11,
          370,
          11,
          1954,
          11,
          808,
          322,
          11,
          3012,
          11,
          3012,
          11,
          3012,
          11,
          1954,
          2086,
          0,
          50844
        ]
      },
      {
        "avg_logprob": -0.4077050024721803,
        "compression_ratio": 1.6401515151515151,
        "end": 13,
        "id": 2,
        "no_speech_prob": 0.0025907985400408506,
        "seek": 0,
        "start": 9.6,
        "temperature": 0,
        "text": " Oh, that looks like it probably should have hit the pipe there.",
        "tokens": [
          50844,
          876,
          11,
          300,
          1542,
          411,
          309,
          1391,
          820,
          362,
          2045,
          264,
          11240,
          456,
          13,
          51014
        ]
      },
      {
        "avg_logprob": -0.4077050024721803,
        "compression_ratio": 1.6401515151515151,
        "end": 19.52,
        "id": 3,
        "no_speech_prob": 0.0025907985400408506,
        "seek": 0,
        "start": 13,
        "temperature": 0,
        "text": " But anyway, that aside, look at these birds, they're really trying to get through those gaps in the pipes.",
        "tokens": [
          51014,
          583,
          4033,
          11,
          300,
          7359,
          11,
          574,
          412,
          613,
          9009,
          11,
          436,
          434,
          534,
          1382,
          281,
          483,
          807,
          729,
          15031,
          294,
          264,
          21882,
          13,
          51340
        ]
      },
      {
        "avg_logprob": -0.4077050024721803,
        "compression_ratio": 1.6401515151515151,
        "end": 22.8,
        "id": 4,
        "no_speech_prob": 0.0025907985400408506,
        "seek": 0,
        "start": 19.52,
        "temperature": 0,
        "text": " Alright, so, I want to do a few things to improve.",
        "tokens": [
          51340,
          2798,
          11,
          370,
          11,
          286,
          528,
          281,
          360,
          257,
          1326,
          721,
          281,
          3470,
          13,
          51504
        ]
      },
      {
        "avg_logprob": -0.4077050024721803,
        "compression_ratio": 1.6401515151515151,
        "end": 27.04,
        "id": 5,
        "no_speech_prob": 0.0025907985400408506,
        "seek": 0,
        "start": 22.8,
        "temperature": 0,
        "text": " Number one, this is a bit ridiculous that I have to sit here and wait and let it run.",
        "tokens": [
          51504,
          5118,
          472,
          11,
          341,
          307,
          257,
          857,
          11083,
          300,
          286,
          362,
          281,
          1394,
          510,
          293,
          1699,
          293,
          718,
          309,
          1190,
          13,
          51716
        ]
      },
      {
        "avg_logprob": -0.23466581106185913,
        "compression_ratio": 1.7875,
        "end": 32,
        "id": 6,
        "no_speech_prob": 0.0320996455848217,
        "seek": 2704,
        "start": 27.04,
        "temperature": 0,
        "text": " It's kind of enjoyable, I mean, I have to go soon, so I got to speed this up.",
        "tokens": [
          50364,
          467,
          311,
          733,
          295,
          20305,
          11,
          286,
          914,
          11,
          286,
          362,
          281,
          352,
          2321,
          11,
          370,
          286,
          658,
          281,
          3073,
          341,
          493,
          13,
          50612
        ]
      },
      {
        "avg_logprob": -0.23466581106185913,
        "compression_ratio": 1.7875,
        "end": 34.8,
        "id": 7,
        "no_speech_prob": 0.0320996455848217,
        "seek": 2704,
        "start": 32,
        "temperature": 0,
        "text": " So here's one thing I'm going to do to speed this up.",
        "tokens": [
          50612,
          407,
          510,
          311,
          472,
          551,
          286,
          478,
          516,
          281,
          360,
          281,
          3073,
          341,
          493,
          13,
          50752
        ]
      },
      {
        "avg_logprob": -0.23466581106185913,
        "compression_ratio": 1.7875,
        "end": 41.16,
        "id": 8,
        "no_speech_prob": 0.0320996455848217,
        "seek": 2704,
        "start": 34.8,
        "temperature": 0,
        "text": " I am going to, I don't need to, all of the logic that's driving the game,",
        "tokens": [
          50752,
          286,
          669,
          516,
          281,
          11,
          286,
          500,
          380,
          643,
          281,
          11,
          439,
          295,
          264,
          9952,
          300,
          311,
          4840,
          264,
          1216,
          11,
          51070
        ]
      },
      {
        "avg_logprob": -0.23466581106185913,
        "compression_ratio": 1.7875,
        "end": 45.2,
        "id": 9,
        "no_speech_prob": 0.0320996455848217,
        "seek": 2704,
        "start": 41.16,
        "temperature": 0,
        "text": " I could just do that many, many times per frame, right?",
        "tokens": [
          51070,
          286,
          727,
          445,
          360,
          300,
          867,
          11,
          867,
          1413,
          680,
          3920,
          11,
          558,
          30,
          51272
        ]
      },
      {
        "avg_logprob": -0.23466581106185913,
        "compression_ratio": 1.7875,
        "end": 49.72,
        "id": 10,
        "no_speech_prob": 0.0320996455848217,
        "seek": 2704,
        "start": 45.2,
        "temperature": 0,
        "text": " I don't actually need to draw and show the animation.",
        "tokens": [
          51272,
          286,
          500,
          380,
          767,
          643,
          281,
          2642,
          293,
          855,
          264,
          9603,
          13,
          51498
        ]
      },
      {
        "avg_logprob": -0.23466581106185913,
        "compression_ratio": 1.7875,
        "end": 56.4,
        "id": 11,
        "no_speech_prob": 0.0320996455848217,
        "seek": 2704,
        "start": 49.72,
        "temperature": 0,
        "text": " So, for example, if I were, so one thing I need to do is separate out all the logic of the game from the drawing.",
        "tokens": [
          51498,
          407,
          11,
          337,
          1365,
          11,
          498,
          286,
          645,
          11,
          370,
          472,
          551,
          286,
          643,
          281,
          360,
          307,
          4994,
          484,
          439,
          264,
          9952,
          295,
          264,
          1216,
          490,
          264,
          6316,
          13,
          51832
        ]
      },
      {
        "avg_logprob": -0.3387963562011719,
        "compression_ratio": 1.9090909090909092,
        "end": 59.879999999999995,
        "id": 12,
        "no_speech_prob": 0.0001911023719003424,
        "seek": 5640,
        "start": 56.44,
        "temperature": 0,
        "text": " So one thing I'm going to do is I'm going to take out pipes.show here.",
        "tokens": [
          50366,
          407,
          472,
          551,
          286,
          478,
          516,
          281,
          360,
          307,
          286,
          478,
          516,
          281,
          747,
          484,
          21882,
          13,
          34436,
          510,
          13,
          50538
        ]
      },
      {
        "avg_logprob": -0.3387963562011719,
        "compression_ratio": 1.9090909090909092,
        "end": 65.28,
        "id": 13,
        "no_speech_prob": 0.0001911023719003424,
        "seek": 5640,
        "start": 60.879999999999995,
        "temperature": 0,
        "text": " Let me just, this is, like, counter is part of the game logic.",
        "tokens": [
          50588,
          961,
          385,
          445,
          11,
          341,
          307,
          11,
          411,
          11,
          5682,
          307,
          644,
          295,
          264,
          1216,
          9952,
          13,
          50808
        ]
      },
      {
        "avg_logprob": -0.3387963562011719,
        "compression_ratio": 1.9090909090909092,
        "end": 69.75999999999999,
        "id": 14,
        "no_speech_prob": 0.0001911023719003424,
        "seek": 5640,
        "start": 66.4,
        "temperature": 0,
        "text": " Looping through all the pipes and checking them, that's part of the game.",
        "tokens": [
          50864,
          6130,
          26125,
          807,
          439,
          264,
          21882,
          293,
          8568,
          552,
          11,
          300,
          311,
          644,
          295,
          264,
          1216,
          13,
          51032
        ]
      },
      {
        "avg_logprob": -0.3387963562011719,
        "compression_ratio": 1.9090909090909092,
        "end": 71.32,
        "id": 15,
        "no_speech_prob": 0.0001911023719003424,
        "seek": 5640,
        "start": 69.75999999999999,
        "temperature": 0,
        "text": " Looping through all the birds.",
        "tokens": [
          51032,
          6130,
          26125,
          807,
          439,
          264,
          9009,
          13,
          51110
        ]
      },
      {
        "avg_logprob": -0.3387963562011719,
        "compression_ratio": 1.9090909090909092,
        "end": 73.75999999999999,
        "id": 16,
        "no_speech_prob": 0.0001911023719003424,
        "seek": 5640,
        "start": 71.32,
        "temperature": 0,
        "text": " If birds, this is part of the game.",
        "tokens": [
          51110,
          759,
          9009,
          11,
          341,
          307,
          644,
          295,
          264,
          1216,
          13,
          51232
        ]
      },
      {
        "avg_logprob": -0.3387963562011719,
        "compression_ratio": 1.9090909090909092,
        "end": 76.16,
        "id": 17,
        "no_speech_prob": 0.0001911023719003424,
        "seek": 5640,
        "start": 73.75999999999999,
        "temperature": 0,
        "text": " So, what I just, but drawing is not.",
        "tokens": [
          51232,
          407,
          11,
          437,
          286,
          445,
          11,
          457,
          6316,
          307,
          406,
          13,
          51352
        ]
      },
      {
        "avg_logprob": -0.3387963562011719,
        "compression_ratio": 1.9090909090909092,
        "end": 80.96000000000001,
        "id": 18,
        "no_speech_prob": 0.0001911023719003424,
        "seek": 5640,
        "start": 76.16,
        "temperature": 0,
        "text": " So let me take out bird.show, and now I'm just going to, at the end here, set, say let,",
        "tokens": [
          51352,
          407,
          718,
          385,
          747,
          484,
          5255,
          13,
          34436,
          11,
          293,
          586,
          286,
          478,
          445,
          516,
          281,
          11,
          412,
          264,
          917,
          510,
          11,
          992,
          11,
          584,
          718,
          11,
          51592
        ]
      },
      {
        "avg_logprob": -0.27239484497995087,
        "compression_ratio": 1.5032258064516129,
        "end": 93.6,
        "id": 19,
        "no_speech_prob": 0.00008220074960263446,
        "seek": 8096,
        "start": 81.6,
        "temperature": 0,
        "text": " for let bird of birds, bird.show, for let pipe of pipes, pipe.show.",
        "tokens": [
          50396,
          337,
          718,
          5255,
          295,
          9009,
          11,
          5255,
          13,
          34436,
          11,
          337,
          718,
          11240,
          295,
          21882,
          11,
          11240,
          13,
          34436,
          13,
          50996
        ]
      },
      {
        "avg_logprob": -0.27239484497995087,
        "compression_ratio": 1.5032258064516129,
        "end": 95.24,
        "id": 20,
        "no_speech_prob": 0.00008220074960263446,
        "seek": 8096,
        "start": 93.6,
        "temperature": 0,
        "text": " This should be exactly the same.",
        "tokens": [
          50996,
          639,
          820,
          312,
          2293,
          264,
          912,
          13,
          51078
        ]
      },
      {
        "avg_logprob": -0.27239484497995087,
        "compression_ratio": 1.5032258064516129,
        "end": 101.67999999999999,
        "id": 21,
        "no_speech_prob": 0.00008220074960263446,
        "seek": 8096,
        "start": 95.24,
        "temperature": 0,
        "text": " Nothing has changed, but now, and I might as well take the background and put the background here.",
        "tokens": [
          51078,
          6693,
          575,
          3105,
          11,
          457,
          586,
          11,
          293,
          286,
          1062,
          382,
          731,
          747,
          264,
          3678,
          293,
          829,
          264,
          3678,
          510,
          13,
          51400
        ]
      },
      {
        "avg_logprob": -0.27239484497995087,
        "compression_ratio": 1.5032258064516129,
        "end": 104.08,
        "id": 22,
        "no_speech_prob": 0.00008220074960263446,
        "seek": 8096,
        "start": 101.67999999999999,
        "temperature": 0,
        "text": " So this is all the drawing stuff.",
        "tokens": [
          51400,
          407,
          341,
          307,
          439,
          264,
          6316,
          1507,
          13,
          51520
        ]
      },
      {
        "avg_logprob": -0.43077986580984934,
        "compression_ratio": 1.7251184834123223,
        "end": 112.44,
        "id": 23,
        "no_speech_prob": 0.0009697424247860909,
        "seek": 10408,
        "start": 104.39999999999999,
        "temperature": 0,
        "text": " And now what I'm going to do is I'm going to say for let n equal zero, n is less than cycles,",
        "tokens": [
          50380,
          400,
          586,
          437,
          286,
          478,
          516,
          281,
          360,
          307,
          286,
          478,
          516,
          281,
          584,
          337,
          718,
          297,
          2681,
          4018,
          11,
          297,
          307,
          1570,
          813,
          17796,
          11,
          50782
        ]
      },
      {
        "avg_logprob": -0.43077986580984934,
        "compression_ratio": 1.7251184834123223,
        "end": 114.52,
        "id": 24,
        "no_speech_prob": 0.0009697424247860909,
        "seek": 10408,
        "start": 112.44,
        "temperature": 0,
        "text": " so I'll call it cycles, n plus plus.",
        "tokens": [
          50782,
          370,
          286,
          603,
          818,
          309,
          17796,
          11,
          297,
          1804,
          1804,
          13,
          50886
        ]
      },
      {
        "avg_logprob": -0.43077986580984934,
        "compression_ratio": 1.7251184834123223,
        "end": 120.75999999999999,
        "id": 25,
        "no_speech_prob": 0.0009697424247860909,
        "seek": 10408,
        "start": 114.52,
        "temperature": 0,
        "text": " I'm going to do all of the game logic stuff many, many times, which goes all the way down to here.",
        "tokens": [
          50886,
          286,
          478,
          516,
          281,
          360,
          439,
          295,
          264,
          1216,
          9952,
          1507,
          867,
          11,
          867,
          1413,
          11,
          597,
          1709,
          439,
          264,
          636,
          760,
          281,
          510,
          13,
          51198
        ]
      },
      {
        "avg_logprob": -0.43077986580984934,
        "compression_ratio": 1.7251184834123223,
        "end": 125.72,
        "id": 26,
        "no_speech_prob": 0.0009697424247860909,
        "seek": 10408,
        "start": 120.75999999999999,
        "temperature": 0,
        "text": " So, and then I'm going to say let cycles equal 100.",
        "tokens": [
          51198,
          407,
          11,
          293,
          550,
          286,
          478,
          516,
          281,
          584,
          718,
          17796,
          2681,
          2319,
          13,
          51446
        ]
      },
      {
        "avg_logprob": -0.43077986580984934,
        "compression_ratio": 1.7251184834123223,
        "end": 126.92,
        "id": 27,
        "no_speech_prob": 0.0009697424247860909,
        "seek": 10408,
        "start": 125.72,
        "temperature": 0,
        "text": " So, watch this.",
        "tokens": [
          51446,
          407,
          11,
          1159,
          341,
          13,
          51506
        ]
      },
      {
        "avg_logprob": -0.43077986580984934,
        "compression_ratio": 1.7251184834123223,
        "end": 132.64,
        "id": 28,
        "no_speech_prob": 0.0009697424247860909,
        "seek": 10408,
        "start": 128.36,
        "temperature": 0,
        "text": " So now this is just, like, doing it really, really fast over time.",
        "tokens": [
          51578,
          407,
          586,
          341,
          307,
          445,
          11,
          411,
          11,
          884,
          309,
          534,
          11,
          534,
          2370,
          670,
          565,
          13,
          51792
        ]
      },
      {
        "avg_logprob": -0.331157568024426,
        "compression_ratio": 1.632183908045977,
        "end": 136.64,
        "id": 29,
        "no_speech_prob": 0.0026729684323072433,
        "seek": 13264,
        "start": 133.2,
        "temperature": 0,
        "text": " But I want to be able to, like, speed it up and slow it down also.",
        "tokens": [
          50392,
          583,
          286,
          528,
          281,
          312,
          1075,
          281,
          11,
          411,
          11,
          3073,
          309,
          493,
          293,
          2964,
          309,
          760,
          611,
          13,
          50564
        ]
      },
      {
        "avg_logprob": -0.331157568024426,
        "compression_ratio": 1.632183908045977,
        "end": 148,
        "id": 30,
        "no_speech_prob": 0.0026729684323072433,
        "seek": 13264,
        "start": 136.64,
        "temperature": 0,
        "text": " So one thing that I can do to do that is I am going to say let slider, slider equals create slider.",
        "tokens": [
          50564,
          407,
          472,
          551,
          300,
          286,
          393,
          360,
          281,
          360,
          300,
          307,
          286,
          669,
          516,
          281,
          584,
          718,
          26046,
          11,
          26046,
          6915,
          1884,
          26046,
          13,
          51132
        ]
      },
      {
        "avg_logprob": -0.331157568024426,
        "compression_ratio": 1.632183908045977,
        "end": 157.44,
        "id": 31,
        "no_speech_prob": 0.0026729684323072433,
        "seek": 13264,
        "start": 148,
        "temperature": 0,
        "text": " I'm going to make a slider that goes from one to 100 and starts at one, and that, I'm just going to say slider.value.",
        "tokens": [
          51132,
          286,
          478,
          516,
          281,
          652,
          257,
          26046,
          300,
          1709,
          490,
          472,
          281,
          2319,
          293,
          3719,
          412,
          472,
          11,
          293,
          300,
          11,
          286,
          478,
          445,
          516,
          281,
          584,
          26046,
          13,
          29155,
          13,
          51604
        ]
      },
      {
        "avg_logprob": -0.2679714417793381,
        "compression_ratio": 1.8297101449275361,
        "end": 163.6,
        "id": 32,
        "no_speech_prob": 0.14413781464099884,
        "seek": 15744,
        "start": 158.4,
        "temperature": 0,
        "text": " So instead of a variable called cycles, I'm just going to use the value of that slider.",
        "tokens": [
          50412,
          407,
          2602,
          295,
          257,
          7006,
          1219,
          17796,
          11,
          286,
          478,
          445,
          516,
          281,
          764,
          264,
          2158,
          295,
          300,
          26046,
          13,
          50672
        ]
      },
      {
        "avg_logprob": -0.2679714417793381,
        "compression_ratio": 1.8297101449275361,
        "end": 167.35999999999999,
        "id": 33,
        "no_speech_prob": 0.14413781464099884,
        "seek": 15744,
        "start": 163.6,
        "temperature": 0,
        "text": " So now I've got a little slider here.",
        "tokens": [
          50672,
          407,
          586,
          286,
          600,
          658,
          257,
          707,
          26046,
          510,
          13,
          50860
        ]
      },
      {
        "avg_logprob": -0.2679714417793381,
        "compression_ratio": 1.8297101449275361,
        "end": 170.4,
        "id": 34,
        "no_speech_prob": 0.14413781464099884,
        "seek": 15744,
        "start": 167.35999999999999,
        "temperature": 0,
        "text": " I can speed things up, and I can slow them back down.",
        "tokens": [
          50860,
          286,
          393,
          3073,
          721,
          493,
          11,
          293,
          286,
          393,
          2964,
          552,
          646,
          760,
          13,
          51012
        ]
      },
      {
        "avg_logprob": -0.2679714417793381,
        "compression_ratio": 1.8297101449275361,
        "end": 175.2,
        "id": 35,
        "no_speech_prob": 0.14413781464099884,
        "seek": 15744,
        "start": 170.4,
        "temperature": 0,
        "text": " So this is going to help me a lot where instead of having to, like, speed up the video that you're watching,",
        "tokens": [
          51012,
          407,
          341,
          307,
          516,
          281,
          854,
          385,
          257,
          688,
          689,
          2602,
          295,
          1419,
          281,
          11,
          411,
          11,
          3073,
          493,
          264,
          960,
          300,
          291,
          434,
          1976,
          11,
          51252
        ]
      },
      {
        "avg_logprob": -0.2679714417793381,
        "compression_ratio": 1.8297101449275361,
        "end": 178.16,
        "id": 36,
        "no_speech_prob": 0.14413781464099884,
        "seek": 15744,
        "start": 175.2,
        "temperature": 0,
        "text": " if I'm editing this after the fact, I can actually just do this.",
        "tokens": [
          51252,
          498,
          286,
          478,
          10000,
          341,
          934,
          264,
          1186,
          11,
          286,
          393,
          767,
          445,
          360,
          341,
          13,
          51400
        ]
      },
      {
        "avg_logprob": -0.2679714417793381,
        "compression_ratio": 1.8297101449275361,
        "end": 183.92,
        "id": 37,
        "no_speech_prob": 0.14413781464099884,
        "seek": 15744,
        "start": 178.16,
        "temperature": 0,
        "text": " So in fact, I could let this go for a little bit, and then I could come back and see, like, how are we doing here?",
        "tokens": [
          51400,
          407,
          294,
          1186,
          11,
          286,
          727,
          718,
          341,
          352,
          337,
          257,
          707,
          857,
          11,
          293,
          550,
          286,
          727,
          808,
          646,
          293,
          536,
          11,
          411,
          11,
          577,
          366,
          321,
          884,
          510,
          30,
          51688
        ]
      },
      {
        "avg_logprob": -0.2679714417793381,
        "compression_ratio": 1.8297101449275361,
        "end": 185.44,
        "id": 38,
        "no_speech_prob": 0.14413781464099884,
        "seek": 15744,
        "start": 183.92,
        "temperature": 0,
        "text": " We got one that's doing pretty well?",
        "tokens": [
          51688,
          492,
          658,
          472,
          300,
          311,
          884,
          1238,
          731,
          30,
          51764
        ]
      },
      {
        "avg_logprob": -0.15565412602526077,
        "compression_ratio": 1.6603773584905661,
        "end": 186.4,
        "id": 39,
        "no_speech_prob": 0.00014653072867076844,
        "seek": 18544,
        "start": 185.44,
        "temperature": 0,
        "text": " So but here's the thing.",
        "tokens": [
          50364,
          407,
          457,
          510,
          311,
          264,
          551,
          13,
          50412
        ]
      },
      {
        "avg_logprob": -0.15565412602526077,
        "compression_ratio": 1.6603773584905661,
        "end": 192.4,
        "id": 40,
        "no_speech_prob": 0.00014653072867076844,
        "seek": 18544,
        "start": 186.4,
        "temperature": 0,
        "text": " I also think there are some things that I really need to do to help the evolutionary process along.",
        "tokens": [
          50412,
          286,
          611,
          519,
          456,
          366,
          512,
          721,
          300,
          286,
          534,
          643,
          281,
          360,
          281,
          854,
          264,
          27567,
          1399,
          2051,
          13,
          50712
        ]
      },
      {
        "avg_logprob": -0.15565412602526077,
        "compression_ratio": 1.6603773584905661,
        "end": 199.76,
        "id": 41,
        "no_speech_prob": 0.00014653072867076844,
        "seek": 18544,
        "start": 193.28,
        "temperature": 0,
        "text": " One thing is I'm going to go into the neural network code and look at how I'm doing mutation.",
        "tokens": [
          50756,
          1485,
          551,
          307,
          286,
          478,
          516,
          281,
          352,
          666,
          264,
          18161,
          3209,
          3089,
          293,
          574,
          412,
          577,
          286,
          478,
          884,
          27960,
          13,
          51080
        ]
      },
      {
        "avg_logprob": -0.15565412602526077,
        "compression_ratio": 1.6603773584905661,
        "end": 201.35999999999999,
        "id": 42,
        "no_speech_prob": 0.00014653072867076844,
        "seek": 18544,
        "start": 199.76,
        "temperature": 0,
        "text": " So this is the mutate function.",
        "tokens": [
          51080,
          407,
          341,
          307,
          264,
          5839,
          473,
          2445,
          13,
          51160
        ]
      },
      {
        "avg_logprob": -0.15565412602526077,
        "compression_ratio": 1.6603773584905661,
        "end": 206.32,
        "id": 43,
        "no_speech_prob": 0.00014653072867076844,
        "seek": 18544,
        "start": 203.12,
        "temperature": 0,
        "text": " So first of all, I'm a little suspicious here.",
        "tokens": [
          51248,
          407,
          700,
          295,
          439,
          11,
          286,
          478,
          257,
          707,
          17931,
          510,
          13,
          51408
        ]
      },
      {
        "avg_logprob": -0.15565412602526077,
        "compression_ratio": 1.6603773584905661,
        "end": 209.68,
        "id": 44,
        "no_speech_prob": 0.00014653072867076844,
        "seek": 18544,
        "start": 207.2,
        "temperature": 0,
        "text": " I want to make sure that mutation is really happening.",
        "tokens": [
          51452,
          286,
          528,
          281,
          652,
          988,
          300,
          27960,
          307,
          534,
          2737,
          13,
          51576
        ]
      },
      {
        "avg_logprob": -0.2377629554134676,
        "compression_ratio": 1.598901098901099,
        "end": 221.20000000000002,
        "id": 45,
        "no_speech_prob": 0.0022871976252645254,
        "seek": 20968,
        "start": 210.64000000000001,
        "temperature": 0,
        "text": " So let me at least console log test.",
        "tokens": [
          50412,
          407,
          718,
          385,
          412,
          1935,
          11076,
          3565,
          1500,
          13,
          50940
        ]
      },
      {
        "avg_logprob": -0.2377629554134676,
        "compression_ratio": 1.598901098901099,
        "end": 223.28,
        "id": 46,
        "no_speech_prob": 0.0022871976252645254,
        "seek": 20968,
        "start": 222.4,
        "temperature": 0,
        "text": " I just want to see.",
        "tokens": [
          51000,
          286,
          445,
          528,
          281,
          536,
          13,
          51044
        ]
      },
      {
        "avg_logprob": -0.2377629554134676,
        "compression_ratio": 1.598901098901099,
        "end": 228.16,
        "id": 47,
        "no_speech_prob": 0.0022871976252645254,
        "seek": 20968,
        "start": 225.12,
        "temperature": 0,
        "text": " I just want to see that this is happening in the code somewhere,",
        "tokens": [
          51136,
          286,
          445,
          528,
          281,
          536,
          300,
          341,
          307,
          2737,
          294,
          264,
          3089,
          4079,
          11,
          51288
        ]
      },
      {
        "avg_logprob": -0.2377629554134676,
        "compression_ratio": 1.598901098901099,
        "end": 235.44,
        "id": 48,
        "no_speech_prob": 0.0022871976252645254,
        "seek": 20968,
        "start": 228.16,
        "temperature": 0,
        "text": " and I'm going to go back to sketch.js and just have, like, three birds because I don't want it to go crazy.",
        "tokens": [
          51288,
          293,
          286,
          478,
          516,
          281,
          352,
          646,
          281,
          12325,
          13,
          25530,
          293,
          445,
          362,
          11,
          411,
          11,
          1045,
          9009,
          570,
          286,
          500,
          380,
          528,
          309,
          281,
          352,
          3219,
          13,
          51652
        ]
      },
      {
        "avg_logprob": -0.2377629554134676,
        "compression_ratio": 1.598901098901099,
        "end": 239.04000000000002,
        "id": 49,
        "no_speech_prob": 0.0022871976252645254,
        "seek": 20968,
        "start": 236.56,
        "temperature": 0,
        "text": " So as soon as the three are done, I should see mutate mutate.",
        "tokens": [
          51708,
          407,
          382,
          2321,
          382,
          264,
          1045,
          366,
          1096,
          11,
          286,
          820,
          536,
          5839,
          473,
          5839,
          473,
          13,
          51832
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 240.88,
        "id": 50,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 239.12,
        "temperature": 0,
        "text": " Yeah, so I got a bunch of mutations.",
        "tokens": [
          50368,
          865,
          11,
          370,
          286,
          658,
          257,
          3840,
          295,
          29243,
          13,
          50456
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 241.44,
        "id": 51,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 240.88,
        "temperature": 0,
        "text": " That's good.",
        "tokens": [
          50456,
          663,
          311,
          665,
          13,
          50484
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 242.72,
        "id": 52,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 241.44,
        "temperature": 0,
        "text": " So that is being called.",
        "tokens": [
          50484,
          407,
          300,
          307,
          885,
          1219,
          13,
          50548
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 245.44,
        "id": 53,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 242.72,
        "temperature": 0,
        "text": " So hopefully, the weights are being changed.",
        "tokens": [
          50548,
          407,
          4696,
          11,
          264,
          17443,
          366,
          885,
          3105,
          13,
          50684
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 247.12,
        "id": 54,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 245.44,
        "temperature": 0,
        "text": " But you know what might work better?",
        "tokens": [
          50684,
          583,
          291,
          458,
          437,
          1062,
          589,
          1101,
          30,
          50768
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 257.92,
        "id": 55,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 247.12,
        "temperature": 0,
        "text": " I think I have a sneaking suspicion is that it might actually help me to, rather than when I mutate, pick an entirely new weight.",
        "tokens": [
          50768,
          286,
          519,
          286,
          362,
          257,
          48525,
          32020,
          307,
          300,
          309,
          1062,
          767,
          854,
          385,
          281,
          11,
          2831,
          813,
          562,
          286,
          5839,
          473,
          11,
          1888,
          364,
          7696,
          777,
          3364,
          13,
          51308
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 259.76,
        "id": 56,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 257.92,
        "temperature": 0,
        "text": " So what's going on here?",
        "tokens": [
          51308,
          407,
          437,
          311,
          516,
          322,
          510,
          30,
          51400
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 264.8,
        "id": 57,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 260.64,
        "temperature": 0,
        "text": " The internals of the neural network are something that I haven't talked about in this series.",
        "tokens": [
          51444,
          440,
          2154,
          1124,
          295,
          264,
          18161,
          3209,
          366,
          746,
          300,
          286,
          2378,
          380,
          2825,
          466,
          294,
          341,
          2638,
          13,
          51652
        ]
      },
      {
        "avg_logprob": -0.16748739227535217,
        "compression_ratio": 1.7375886524822695,
        "end": 268.48,
        "id": 58,
        "no_speech_prob": 0.00006302746623987332,
        "seek": 23904,
        "start": 264.8,
        "temperature": 0,
        "text": " That's really something reserved for my other videos about how neural networks work.",
        "tokens": [
          51652,
          663,
          311,
          534,
          746,
          24819,
          337,
          452,
          661,
          2145,
          466,
          577,
          18161,
          9590,
          589,
          13,
          51836
        ]
      },
      {
        "avg_logprob": -0.18095412747613315,
        "compression_ratio": 1.667883211678832,
        "end": 272.96000000000004,
        "id": 59,
        "no_speech_prob": 0.000005507597961695865,
        "seek": 26848,
        "start": 268.48,
        "temperature": 0,
        "text": " The only thing that's relevant to this series is the inputs of the game come in,",
        "tokens": [
          50364,
          440,
          787,
          551,
          300,
          311,
          7340,
          281,
          341,
          2638,
          307,
          264,
          15743,
          295,
          264,
          1216,
          808,
          294,
          11,
          50588
        ]
      },
      {
        "avg_logprob": -0.18095412747613315,
        "compression_ratio": 1.667883211678832,
        "end": 277.12,
        "id": 60,
        "no_speech_prob": 0.000005507597961695865,
        "seek": 26848,
        "start": 272.96000000000004,
        "temperature": 0,
        "text": " the outputs come out, and the output tells me whether I should jump or not.",
        "tokens": [
          50588,
          264,
          23930,
          808,
          484,
          11,
          293,
          264,
          5598,
          5112,
          385,
          1968,
          286,
          820,
          3012,
          420,
          406,
          13,
          50796
        ]
      },
      {
        "avg_logprob": -0.18095412747613315,
        "compression_ratio": 1.667883211678832,
        "end": 283.68,
        "id": 61,
        "no_speech_prob": 0.000005507597961695865,
        "seek": 26848,
        "start": 279.28000000000003,
        "temperature": 0,
        "text": " But what's actually stored inside the neural network are a bunch of numbers between negative",
        "tokens": [
          50904,
          583,
          437,
          311,
          767,
          12187,
          1854,
          264,
          18161,
          3209,
          366,
          257,
          3840,
          295,
          3547,
          1296,
          3671,
          51124
        ]
      },
      {
        "avg_logprob": -0.18095412747613315,
        "compression_ratio": 1.667883211678832,
        "end": 286.64000000000004,
        "id": 62,
        "no_speech_prob": 0.000005507597961695865,
        "seek": 26848,
        "start": 283.68,
        "temperature": 0,
        "text": " 1 and 1 for all of these connections.",
        "tokens": [
          51124,
          502,
          293,
          502,
          337,
          439,
          295,
          613,
          9271,
          13,
          51272
        ]
      },
      {
        "avg_logprob": -0.18095412747613315,
        "compression_ratio": 1.667883211678832,
        "end": 290.24,
        "id": 63,
        "no_speech_prob": 0.000005507597961695865,
        "seek": 26848,
        "start": 286.64000000000004,
        "temperature": 0,
        "text": " And so mutation means just pluck one of them out and give it a new value.",
        "tokens": [
          51272,
          400,
          370,
          27960,
          1355,
          445,
          41514,
          472,
          295,
          552,
          484,
          293,
          976,
          309,
          257,
          777,
          2158,
          13,
          51452
        ]
      },
      {
        "avg_logprob": -0.18095412747613315,
        "compression_ratio": 1.667883211678832,
        "end": 296.24,
        "id": 64,
        "no_speech_prob": 0.000005507597961695865,
        "seek": 26848,
        "start": 290.88,
        "temperature": 0,
        "text": " But if it's one that's doing pretty well, if I just were to, like, tweak the value a little bit",
        "tokens": [
          51484,
          583,
          498,
          309,
          311,
          472,
          300,
          311,
          884,
          1238,
          731,
          11,
          498,
          286,
          445,
          645,
          281,
          11,
          411,
          11,
          29879,
          264,
          2158,
          257,
          707,
          857,
          51752
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 298.32,
        "id": 65,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 296.24,
        "temperature": 0,
        "text": " instead of just giving it a whole new value.",
        "tokens": [
          50364,
          2602,
          295,
          445,
          2902,
          309,
          257,
          1379,
          777,
          2158,
          13,
          50468
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 301.6,
        "id": 66,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 298.32,
        "temperature": 0,
        "text": " And also, what if I were to just tweak it a tiny bit?",
        "tokens": [
          50468,
          400,
          611,
          11,
          437,
          498,
          286,
          645,
          281,
          445,
          29879,
          309,
          257,
          5870,
          857,
          30,
          50632
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 306.32,
        "id": 67,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 302.16,
        "temperature": 0,
        "text": " So one way I could do that is use a Gaussian distribution of random numbers.",
        "tokens": [
          50660,
          407,
          472,
          636,
          286,
          727,
          360,
          300,
          307,
          764,
          257,
          39148,
          7316,
          295,
          4974,
          3547,
          13,
          50868
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 310.72,
        "id": 68,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 306.32,
        "temperature": 0,
        "text": " Random numbers that cluster around like a mean of 0, but are often like...",
        "tokens": [
          50868,
          37603,
          3547,
          300,
          13630,
          926,
          411,
          257,
          914,
          295,
          1958,
          11,
          457,
          366,
          2049,
          411,
          485,
          51088
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 313.28000000000003,
        "id": 69,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 310.72,
        "temperature": 0,
        "text": " So just like I want to give each a little bit of a nudge.",
        "tokens": [
          51088,
          407,
          445,
          411,
          286,
          528,
          281,
          976,
          1184,
          257,
          707,
          857,
          295,
          257,
          297,
          16032,
          13,
          51216
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 315.92,
        "id": 70,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 314.24,
        "temperature": 0,
        "text": " So let me come back over here.",
        "tokens": [
          51264,
          407,
          718,
          385,
          808,
          646,
          670,
          510,
          13,
          51348
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 320.8,
        "id": 71,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 315.92,
        "temperature": 0,
        "text": " And I'm going to have to build in Gaussian random number generation into the library,",
        "tokens": [
          51348,
          400,
          286,
          478,
          516,
          281,
          362,
          281,
          1322,
          294,
          39148,
          4974,
          1230,
          5125,
          666,
          264,
          6405,
          11,
          51592
        ]
      },
      {
        "avg_logprob": -0.2106757045777376,
        "compression_ratio": 1.6840148698884758,
        "end": 321.92,
        "id": 72,
        "no_speech_prob": 0.00014653087418992072,
        "seek": 29624,
        "start": 320.8,
        "temperature": 0,
        "text": " the neural network library.",
        "tokens": [
          51592,
          264,
          18161,
          3209,
          6405,
          13,
          51648
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 325.44,
        "id": 73,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 322,
        "temperature": 0,
        "text": " Right now, since I have p5 involved, just to fix that,",
        "tokens": [
          50368,
          1779,
          586,
          11,
          1670,
          286,
          362,
          280,
          20,
          3288,
          11,
          445,
          281,
          3191,
          300,
          11,
          50540
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 329.04,
        "id": 74,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 325.44,
        "temperature": 0,
        "text": " I'm going to comment this out and I'm going to say return val.",
        "tokens": [
          50540,
          286,
          478,
          516,
          281,
          2871,
          341,
          484,
          293,
          286,
          478,
          516,
          281,
          584,
          2736,
          1323,
          13,
          50720
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 334.08000000000004,
        "id": 75,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 329.04,
        "temperature": 0,
        "text": " So this would be just like me adjusting the value a little bit, right?",
        "tokens": [
          50720,
          407,
          341,
          576,
          312,
          445,
          411,
          385,
          23559,
          264,
          2158,
          257,
          707,
          857,
          11,
          558,
          30,
          50972
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 338,
        "id": 76,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 334.08000000000004,
        "temperature": 0,
        "text": " Return the actual value changed by a little random amount.",
        "tokens": [
          50972,
          24350,
          264,
          3539,
          2158,
          3105,
          538,
          257,
          707,
          4974,
          2372,
          13,
          51168
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 339.12,
        "id": 77,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 338,
        "temperature": 0,
        "text": " But I think I'm going to...",
        "tokens": [
          51168,
          583,
          286,
          519,
          286,
          478,
          516,
          281,
          485,
          51224
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 345.68,
        "id": 78,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 339.12,
        "temperature": 0,
        "text": " Because when I did this for my class last year, actually, I remember using random Gaussian,",
        "tokens": [
          51224,
          1436,
          562,
          286,
          630,
          341,
          337,
          452,
          1508,
          1036,
          1064,
          11,
          767,
          11,
          286,
          1604,
          1228,
          4974,
          39148,
          11,
          51552
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 348.96000000000004,
        "id": 79,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 345.68,
        "temperature": 0,
        "text": " which is this is now a p5 function from the p5 library.",
        "tokens": [
          51552,
          597,
          307,
          341,
          307,
          586,
          257,
          280,
          20,
          2445,
          490,
          264,
          280,
          20,
          6405,
          13,
          51716
        ]
      },
      {
        "avg_logprob": -0.19529640197753906,
        "compression_ratio": 1.6379928315412187,
        "end": 350.16,
        "id": 80,
        "no_speech_prob": 0.1992780864238739,
        "seek": 32192,
        "start": 348.96000000000004,
        "temperature": 0,
        "text": " This doesn't exist in JavaScript.",
        "tokens": [
          51716,
          639,
          1177,
          380,
          2514,
          294,
          15778,
          13,
          51776
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 352,
        "id": 81,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 350.16,
        "temperature": 0,
        "text": " So I've got to implement that into my library.",
        "tokens": [
          50364,
          407,
          286,
          600,
          658,
          281,
          4445,
          300,
          666,
          452,
          6405,
          13,
          50456
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 353.12,
        "id": 82,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 352,
        "temperature": 0,
        "text": " It'll work for now.",
        "tokens": [
          50456,
          467,
          603,
          589,
          337,
          586,
          13,
          50512
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 356.88000000000005,
        "id": 83,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 353.12,
        "temperature": 0,
        "text": " The mean would be 0 and the standard deviation would be like 0.1.",
        "tokens": [
          50512,
          440,
          914,
          576,
          312,
          1958,
          293,
          264,
          3832,
          25163,
          576,
          312,
          411,
          1958,
          13,
          16,
          13,
          50700
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 357.6,
        "id": 84,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 356.88000000000005,
        "temperature": 0,
        "text": " Let's try that.",
        "tokens": [
          50700,
          961,
          311,
          853,
          300,
          13,
          50736
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 360.32000000000005,
        "id": 85,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 358.40000000000003,
        "temperature": 0,
        "text": " So let's see how this goes.",
        "tokens": [
          50776,
          407,
          718,
          311,
          536,
          577,
          341,
          1709,
          13,
          50872
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 361.6,
        "id": 86,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 360.88000000000005,
        "temperature": 0,
        "text": " Oh, and let me...",
        "tokens": [
          50900,
          876,
          11,
          293,
          718,
          385,
          485,
          50936
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 366,
        "id": 87,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 361.6,
        "temperature": 0,
        "text": " I took out the console log and let's give this a whirl now.",
        "tokens": [
          50936,
          286,
          1890,
          484,
          264,
          11076,
          3565,
          293,
          718,
          311,
          976,
          341,
          257,
          35706,
          586,
          13,
          51156
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 368.24,
        "id": 88,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 367.04,
        "temperature": 0,
        "text": " All right, so I'm going to speed this up.",
        "tokens": [
          51208,
          1057,
          558,
          11,
          370,
          286,
          478,
          516,
          281,
          3073,
          341,
          493,
          13,
          51268
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 370.16,
        "id": 89,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 369.36,
        "temperature": 0,
        "text": " And I'm also...",
        "tokens": [
          51324,
          400,
          286,
          478,
          611,
          485,
          51364
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 372.32000000000005,
        "id": 90,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 370.8,
        "temperature": 0,
        "text": " Oh, wait, there's only like two of them.",
        "tokens": [
          51396,
          876,
          11,
          1699,
          11,
          456,
          311,
          787,
          411,
          732,
          295,
          552,
          13,
          51472
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 376.32000000000005,
        "id": 91,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 373.36,
        "temperature": 0,
        "text": " So the other thing I need to do is let's give myself a nice population.",
        "tokens": [
          51524,
          407,
          264,
          661,
          551,
          286,
          643,
          281,
          360,
          307,
          718,
          311,
          976,
          2059,
          257,
          1481,
          4415,
          13,
          51672
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 377.44000000000005,
        "id": 92,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 376.32000000000005,
        "temperature": 0,
        "text": " Let's try 500 of these.",
        "tokens": [
          51672,
          961,
          311,
          853,
          5923,
          295,
          613,
          13,
          51728
        ]
      },
      {
        "avg_logprob": -0.19121386291115147,
        "compression_ratio": 1.7153024911032029,
        "end": 380.08000000000004,
        "id": 93,
        "no_speech_prob": 0.014063319191336632,
        "seek": 35016,
        "start": 378.8,
        "temperature": 0,
        "text": " All right, so there's 500 going.",
        "tokens": [
          51796,
          1057,
          558,
          11,
          370,
          456,
          311,
          5923,
          516,
          13,
          51860
        ]
      },
      {
        "avg_logprob": -0.19185272852579752,
        "compression_ratio": 1.5449735449735449,
        "end": 384.24,
        "id": 94,
        "no_speech_prob": 0.00003269907392677851,
        "seek": 38008,
        "start": 380.56,
        "temperature": 0,
        "text": " I'm going to speed this up and I'm going to let this run for a little while and I will be back.",
        "tokens": [
          50388,
          286,
          478,
          516,
          281,
          3073,
          341,
          493,
          293,
          286,
          478,
          516,
          281,
          718,
          341,
          1190,
          337,
          257,
          707,
          1339,
          293,
          286,
          486,
          312,
          646,
          13,
          50572
        ]
      },
      {
        "avg_logprob": -0.19185272852579752,
        "compression_ratio": 1.5449735449735449,
        "end": 396.24,
        "id": 95,
        "no_speech_prob": 0.00003269907392677851,
        "seek": 38008,
        "start": 394.24,
        "temperature": 0,
        "text": " All right, so it looks to me like we've done it.",
        "tokens": [
          51072,
          1057,
          558,
          11,
          370,
          309,
          1542,
          281,
          385,
          411,
          321,
          600,
          1096,
          309,
          13,
          51172
        ]
      },
      {
        "avg_logprob": -0.19185272852579752,
        "compression_ratio": 1.5449735449735449,
        "end": 403.36,
        "id": 96,
        "no_speech_prob": 0.00003269907392677851,
        "seek": 38008,
        "start": 398.15999999999997,
        "temperature": 0,
        "text": " There's so much variation in here, but I've evolved it for now 76 generations.",
        "tokens": [
          51268,
          821,
          311,
          370,
          709,
          12990,
          294,
          510,
          11,
          457,
          286,
          600,
          14178,
          309,
          337,
          586,
          24733,
          10593,
          13,
          51528
        ]
      },
      {
        "avg_logprob": -0.19185272852579752,
        "compression_ratio": 1.5449735449735449,
        "end": 408.64,
        "id": 97,
        "no_speech_prob": 0.00003269907392677851,
        "seek": 38008,
        "start": 403.36,
        "temperature": 0,
        "text": " And what we see here is that we've got a bunch that are still alive.",
        "tokens": [
          51528,
          400,
          437,
          321,
          536,
          510,
          307,
          300,
          321,
          600,
          658,
          257,
          3840,
          300,
          366,
          920,
          5465,
          13,
          51792
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 411.52,
        "id": 98,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 408.64,
        "temperature": 0,
        "text": " And there's probably one in there that has pretty optimal weights.",
        "tokens": [
          50364,
          400,
          456,
          311,
          1391,
          472,
          294,
          456,
          300,
          575,
          1238,
          16252,
          17443,
          13,
          50508
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 412.64,
        "id": 99,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 411.52,
        "temperature": 0,
        "text": " And it's making it through.",
        "tokens": [
          50508,
          400,
          309,
          311,
          1455,
          309,
          807,
          13,
          50564
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 414.88,
        "id": 100,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 412.64,
        "temperature": 0,
        "text": " Now, one thing that I would really like to do is...",
        "tokens": [
          50564,
          823,
          11,
          472,
          551,
          300,
          286,
          576,
          534,
          411,
          281,
          360,
          307,
          485,
          50676
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 421.68,
        "id": 101,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 414.88,
        "temperature": 0,
        "text": " It's kind of more fun to watch this if I show only the one that's kind of doing best.",
        "tokens": [
          50676,
          467,
          311,
          733,
          295,
          544,
          1019,
          281,
          1159,
          341,
          498,
          286,
          855,
          787,
          264,
          472,
          300,
          311,
          733,
          295,
          884,
          1151,
          13,
          51016
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 423.76,
        "id": 102,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 421.68,
        "temperature": 0,
        "text": " But at some point, there'll only be one left.",
        "tokens": [
          51016,
          583,
          412,
          512,
          935,
          11,
          456,
          603,
          787,
          312,
          472,
          1411,
          13,
          51120
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 425.28,
        "id": 103,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 423.76,
        "temperature": 0,
        "text": " But I think we're pretty much there.",
        "tokens": [
          51120,
          583,
          286,
          519,
          321,
          434,
          1238,
          709,
          456,
          13,
          51196
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 426.88,
        "id": 104,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 425.28,
        "temperature": 0,
        "text": " Here's some things that you might want to...",
        "tokens": [
          51196,
          1692,
          311,
          512,
          721,
          300,
          291,
          1062,
          528,
          281,
          485,
          51276
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 429.12,
        "id": 105,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 426.88,
        "temperature": 0,
        "text": " That I would love to do with this in another part.",
        "tokens": [
          51276,
          663,
          286,
          576,
          959,
          281,
          360,
          365,
          341,
          294,
          1071,
          644,
          13,
          51388
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 431.36,
        "id": 106,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 429.12,
        "temperature": 0,
        "text": " But I'm going to leave them as exercises for the viewer.",
        "tokens": [
          51388,
          583,
          286,
          478,
          516,
          281,
          1856,
          552,
          382,
          11900,
          337,
          264,
          16767,
          13,
          51500
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 432.88,
        "id": 107,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 432.4,
        "temperature": 0,
        "text": " Come on.",
        "tokens": [
          51552,
          2492,
          322,
          13,
          51576
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 434.08,
        "id": 108,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 433.59999999999997,
        "temperature": 0,
        "text": " There you go.",
        "tokens": [
          51612,
          821,
          291,
          352,
          13,
          51636
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 434.56,
        "id": 109,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 434.08,
        "temperature": 0,
        "text": " Oh, you...",
        "tokens": [
          51636,
          876,
          11,
          291,
          485,
          51660
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 435.91999999999996,
        "id": 110,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 434.56,
        "temperature": 0,
        "text": " Oh, it started over.",
        "tokens": [
          51660,
          876,
          11,
          309,
          1409,
          670,
          13,
          51728
        ]
      },
      {
        "avg_logprob": -0.21017332774836842,
        "compression_ratio": 1.7556270096463023,
        "end": 437.59999999999997,
        "id": 111,
        "no_speech_prob": 0.0015487438067793846,
        "seek": 40864,
        "start": 436.64,
        "temperature": 0,
        "text": " So a couple of things.",
        "tokens": [
          51764,
          407,
          257,
          1916,
          295,
          721,
          13,
          51812
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 441.12,
        "id": 112,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 437.6,
        "temperature": 0,
        "text": " One is that I can keep track of the single agent,",
        "tokens": [
          50364,
          1485,
          307,
          300,
          286,
          393,
          1066,
          2837,
          295,
          264,
          2167,
          9461,
          11,
          50540
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 445.52000000000004,
        "id": 113,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 441.12,
        "temperature": 0,
        "text": " the single bird that has done the best over all time, over all generations.",
        "tokens": [
          50540,
          264,
          2167,
          5255,
          300,
          575,
          1096,
          264,
          1151,
          670,
          439,
          565,
          11,
          670,
          439,
          10593,
          13,
          50760
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 449.44,
        "id": 114,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 445.52000000000004,
        "temperature": 0,
        "text": " And I could save that in a variable like the best bird, record bird.",
        "tokens": [
          50760,
          400,
          286,
          727,
          3155,
          300,
          294,
          257,
          7006,
          411,
          264,
          1151,
          5255,
          11,
          2136,
          5255,
          13,
          50956
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 453.12,
        "id": 115,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 449.44,
        "temperature": 0,
        "text": " And what I could do is actually I could download those weights.",
        "tokens": [
          50956,
          400,
          437,
          286,
          727,
          360,
          307,
          767,
          286,
          727,
          5484,
          729,
          17443,
          13,
          51140
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 455.6,
        "id": 116,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 453.12,
        "temperature": 0,
        "text": " I could save those weights to like a JSON file.",
        "tokens": [
          51140,
          286,
          727,
          3155,
          729,
          17443,
          281,
          411,
          257,
          31828,
          3991,
          13,
          51264
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 458.08000000000004,
        "id": 117,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 455.6,
        "temperature": 0,
        "text": " It's neural network configuration to a JSON file.",
        "tokens": [
          51264,
          467,
          311,
          18161,
          3209,
          11694,
          281,
          257,
          31828,
          3991,
          13,
          51388
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 460.64000000000004,
        "id": 118,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 459.04,
        "temperature": 0,
        "text": " Or even just save it in memory.",
        "tokens": [
          51436,
          1610,
          754,
          445,
          3155,
          309,
          294,
          4675,
          13,
          51516
        ]
      },
      {
        "avg_logprob": -0.18608417818623205,
        "compression_ratio": 1.833976833976834,
        "end": 464.72,
        "id": 119,
        "no_speech_prob": 0.00006814843800384551,
        "seek": 43760,
        "start": 460.64000000000004,
        "temperature": 0,
        "text": " And I could always just click a button or something and say, let me just see this one.",
        "tokens": [
          51516,
          400,
          286,
          727,
          1009,
          445,
          2052,
          257,
          2960,
          420,
          746,
          293,
          584,
          11,
          718,
          385,
          445,
          536,
          341,
          472,
          13,
          51720
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 467.44000000000005,
        "id": 120,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 465.6,
        "temperature": 0,
        "text": " So that's something I would really love to add to this.",
        "tokens": [
          50408,
          407,
          300,
          311,
          746,
          286,
          576,
          534,
          959,
          281,
          909,
          281,
          341,
          13,
          50500
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 468,
        "id": 121,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 467.44000000000005,
        "temperature": 0,
        "text": " It's actually...",
        "tokens": [
          50500,
          467,
          311,
          767,
          485,
          50528
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 470.48,
        "id": 122,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 468.88000000000005,
        "temperature": 0,
        "text": " I might as well just show it to you.",
        "tokens": [
          50572,
          286,
          1062,
          382,
          731,
          445,
          855,
          309,
          281,
          291,
          13,
          50652
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 478.8,
        "id": 123,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 470.48,
        "temperature": 0,
        "text": " Because the example that I made for my course, which is here,",
        "tokens": [
          50652,
          1436,
          264,
          1365,
          300,
          286,
          1027,
          337,
          452,
          1164,
          11,
          597,
          307,
          510,
          11,
          51068
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 480.64000000000004,
        "id": 124,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 478.8,
        "temperature": 0,
        "text": " is essentially exactly the same.",
        "tokens": [
          51068,
          307,
          4476,
          2293,
          264,
          912,
          13,
          51160
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 482.40000000000003,
        "id": 125,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 480.64000000000004,
        "temperature": 0,
        "text": " But there's a couple different things.",
        "tokens": [
          51160,
          583,
          456,
          311,
          257,
          1916,
          819,
          721,
          13,
          51248
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 483.36,
        "id": 126,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 482.40000000000003,
        "temperature": 0,
        "text": " So one is I'm going to...",
        "tokens": [
          51248,
          407,
          472,
          307,
          286,
          478,
          516,
          281,
          485,
          51296
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 485.04,
        "id": 127,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 483.36,
        "temperature": 0,
        "text": " So let me just ramp up the speed here.",
        "tokens": [
          51296,
          407,
          718,
          385,
          445,
          12428,
          493,
          264,
          3073,
          510,
          13,
          51380
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 487.84000000000003,
        "id": 128,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 486.16,
        "temperature": 0,
        "text": " So let me let this run for a little while.",
        "tokens": [
          51436,
          407,
          718,
          385,
          718,
          341,
          1190,
          337,
          257,
          707,
          1339,
          13,
          51520
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 490.24,
        "id": 129,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 488.48,
        "temperature": 0,
        "text": " The game's maybe a little bit different in this one.",
        "tokens": [
          51552,
          440,
          1216,
          311,
          1310,
          257,
          707,
          857,
          819,
          294,
          341,
          472,
          13,
          51640
        ]
      },
      {
        "avg_logprob": -0.1879072333827163,
        "compression_ratio": 1.7366412213740459,
        "end": 492.96000000000004,
        "id": 130,
        "no_speech_prob": 0.000025466311853961088,
        "seek": 46472,
        "start": 490.88000000000005,
        "temperature": 0,
        "text": " But so one of the things that I like about this is",
        "tokens": [
          51672,
          583,
          370,
          472,
          295,
          264,
          721,
          300,
          286,
          411,
          466,
          341,
          307,
          51776
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 498.47999999999996,
        "id": 131,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 492.96,
        "temperature": 0,
        "text": " once I've let it run for a little while, I can just click run best so far.",
        "tokens": [
          50364,
          1564,
          286,
          600,
          718,
          309,
          1190,
          337,
          257,
          707,
          1339,
          11,
          286,
          393,
          445,
          2052,
          1190,
          1151,
          370,
          1400,
          13,
          50640
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 499.2,
        "id": 132,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 498.47999999999996,
        "temperature": 0,
        "text": " So now...",
        "tokens": [
          50640,
          407,
          586,
          485,
          50676
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 500.4,
        "id": 133,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 499.2,
        "temperature": 0,
        "text": " And I could speed this up a little bit.",
        "tokens": [
          50676,
          400,
          286,
          727,
          3073,
          341,
          493,
          257,
          707,
          857,
          13,
          50736
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 504.88,
        "id": 134,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 501.59999999999997,
        "temperature": 0,
        "text": " But I think I better train it more because I didn't get a really good one.",
        "tokens": [
          50796,
          583,
          286,
          519,
          286,
          1101,
          3847,
          309,
          544,
          570,
          286,
          994,
          380,
          483,
          257,
          534,
          665,
          472,
          13,
          50960
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 508.96,
        "id": 135,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 505.52,
        "temperature": 0,
        "text": " So this is the idea of being able to just run the best one that's done,",
        "tokens": [
          50992,
          407,
          341,
          307,
          264,
          1558,
          295,
          885,
          1075,
          281,
          445,
          1190,
          264,
          1151,
          472,
          300,
          311,
          1096,
          11,
          51164
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 510.56,
        "id": 136,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 508.96,
        "temperature": 0,
        "text": " you know, that has the best high score.",
        "tokens": [
          51164,
          291,
          458,
          11,
          300,
          575,
          264,
          1151,
          1090,
          6175,
          13,
          51244
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 511.12,
        "id": 137,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 510.56,
        "temperature": 0,
        "text": " Oh, look at this.",
        "tokens": [
          51244,
          876,
          11,
          574,
          412,
          341,
          13,
          51272
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 511.44,
        "id": 138,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 511.12,
        "temperature": 0,
        "text": " Okay.",
        "tokens": [
          51272,
          1033,
          13,
          51288
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 512,
        "id": 139,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 511.44,
        "temperature": 0,
        "text": " So I'm going to...",
        "tokens": [
          51288,
          407,
          286,
          478,
          516,
          281,
          485,
          51316
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 512.88,
        "id": 140,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 512,
        "temperature": 0,
        "text": " Now I'm going to do this.",
        "tokens": [
          51316,
          823,
          286,
          478,
          516,
          281,
          360,
          341,
          13,
          51360
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 519.28,
        "id": 141,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 513.92,
        "temperature": 0,
        "text": " So we can see this one that kind of has a pretty good sense of how to play the game.",
        "tokens": [
          51412,
          407,
          321,
          393,
          536,
          341,
          472,
          300,
          733,
          295,
          575,
          257,
          1238,
          665,
          2020,
          295,
          577,
          281,
          862,
          264,
          1216,
          13,
          51680
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 520.0799999999999,
        "id": 142,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 519.28,
        "temperature": 0,
        "text": " Kind of running...",
        "tokens": [
          51680,
          9242,
          295,
          2614,
          485,
          51720
        ]
      },
      {
        "avg_logprob": -0.1979679202441103,
        "compression_ratio": 1.7086092715231789,
        "end": 522.48,
        "id": 143,
        "no_speech_prob": 0.00004469377745408565,
        "seek": 49296,
        "start": 520.88,
        "temperature": 0,
        "text": " Just watch that one on its own.",
        "tokens": [
          51760,
          1449,
          1159,
          300,
          472,
          322,
          1080,
          1065,
          13,
          51840
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 525.6800000000001,
        "id": 144,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 523.9200000000001,
        "temperature": 0,
        "text": " It's very satisfying to watch.",
        "tokens": [
          50412,
          467,
          311,
          588,
          18348,
          281,
          1159,
          13,
          50500
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 527.0400000000001,
        "id": 145,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 525.6800000000001,
        "temperature": 0,
        "text": " I'm like amazed that it's doing it.",
        "tokens": [
          50500,
          286,
          478,
          411,
          20507,
          300,
          309,
          311,
          884,
          309,
          13,
          50568
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 530.32,
        "id": 146,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 527.6800000000001,
        "temperature": 0,
        "text": " So that's one thing that would be worth as an exercise",
        "tokens": [
          50600,
          407,
          300,
          311,
          472,
          551,
          300,
          576,
          312,
          3163,
          382,
          364,
          5380,
          50732
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 532.48,
        "id": 147,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 530.32,
        "temperature": 0,
        "text": " trying to add to this implementation that I made.",
        "tokens": [
          50732,
          1382,
          281,
          909,
          281,
          341,
          11420,
          300,
          286,
          1027,
          13,
          50840
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 536.24,
        "id": 148,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 533.12,
        "temperature": 0,
        "text": " Now, once I even do that, I could save it to a file.",
        "tokens": [
          50872,
          823,
          11,
          1564,
          286,
          754,
          360,
          300,
          11,
          286,
          727,
          3155,
          309,
          281,
          257,
          3991,
          13,
          51028
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 538.64,
        "id": 149,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 536.24,
        "temperature": 0,
        "text": " So there's actually in the toy neural network library,",
        "tokens": [
          51028,
          407,
          456,
          311,
          767,
          294,
          264,
          12058,
          18161,
          3209,
          6405,
          11,
          51148
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 540.96,
        "id": 150,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 538.64,
        "temperature": 0,
        "text": " there's like a save JSON, I think, function.",
        "tokens": [
          51148,
          456,
          311,
          411,
          257,
          3155,
          31828,
          11,
          286,
          519,
          11,
          2445,
          13,
          51264
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 543.2,
        "id": 151,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 540.96,
        "temperature": 0,
        "text": " So I could save it to a file to reload later.",
        "tokens": [
          51264,
          407,
          286,
          727,
          3155,
          309,
          281,
          257,
          3991,
          281,
          25628,
          1780,
          13,
          51376
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 545.6800000000001,
        "id": 152,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 543.2,
        "temperature": 0,
        "text": " So if I'm training it all day or for an hour,",
        "tokens": [
          51376,
          407,
          498,
          286,
          478,
          3097,
          309,
          439,
          786,
          420,
          337,
          364,
          1773,
          11,
          51500
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 547.52,
        "id": 153,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 545.6800000000001,
        "temperature": 0,
        "text": " I could save it and bring it into another environment.",
        "tokens": [
          51500,
          286,
          727,
          3155,
          309,
          293,
          1565,
          309,
          666,
          1071,
          2823,
          13,
          51592
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 548.08,
        "id": 154,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 547.52,
        "temperature": 0,
        "text": " That's one thing.",
        "tokens": [
          51592,
          663,
          311,
          472,
          551,
          13,
          51620
        ]
      },
      {
        "avg_logprob": -0.1583314375443892,
        "compression_ratio": 1.8150684931506849,
        "end": 550.5600000000001,
        "id": 155,
        "no_speech_prob": 0.00013765368203166872,
        "seek": 52296,
        "start": 549.0400000000001,
        "temperature": 0,
        "text": " Another thing that I might want to do is",
        "tokens": [
          51668,
          3996,
          551,
          300,
          286,
          1062,
          528,
          281,
          360,
          307,
          51744
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 553.52,
        "id": 156,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 550.56,
        "temperature": 0,
        "text": " I could probably optimize the fitness function quite a bit.",
        "tokens": [
          50364,
          286,
          727,
          1391,
          19719,
          264,
          15303,
          2445,
          1596,
          257,
          857,
          13,
          50512
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 555.28,
        "id": 157,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 553.52,
        "temperature": 0,
        "text": " For example, what if...",
        "tokens": [
          50512,
          1171,
          1365,
          11,
          437,
          498,
          485,
          50600
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 556.16,
        "id": 158,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 555.28,
        "temperature": 0,
        "text": " And this is a...",
        "tokens": [
          50600,
          400,
          341,
          307,
          257,
          485,
          50644
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 560.16,
        "id": 159,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 556.16,
        "temperature": 0,
        "text": " Thank you to Versatilis for this suggestion.",
        "tokens": [
          50644,
          1044,
          291,
          281,
          12226,
          267,
          388,
          271,
          337,
          341,
          16541,
          13,
          50844
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 563.1999999999999,
        "id": 160,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 560.16,
        "temperature": 0,
        "text": " What if I actually scored not just how long it lasts,",
        "tokens": [
          50844,
          708,
          498,
          286,
          767,
          18139,
          406,
          445,
          577,
          938,
          309,
          20669,
          11,
          50996
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 567.04,
        "id": 161,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 563.1999999999999,
        "temperature": 0,
        "text": " but how well it maintains its Y location",
        "tokens": [
          50996,
          457,
          577,
          731,
          309,
          33385,
          1080,
          398,
          4914,
          51188
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 569.5999999999999,
        "id": 162,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 567.04,
        "temperature": 0,
        "text": " in between the top and bottom of those two pipes?",
        "tokens": [
          51188,
          294,
          1296,
          264,
          1192,
          293,
          2767,
          295,
          729,
          732,
          21882,
          30,
          51316
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 570.16,
        "id": 163,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 569.5999999999999,
        "temperature": 0,
        "text": " So that would be...",
        "tokens": [
          51316,
          407,
          300,
          576,
          312,
          485,
          51344
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 574.4,
        "id": 164,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 571.04,
        "temperature": 0,
        "text": " You could probably think of many other ways that you could optimize this.",
        "tokens": [
          51388,
          509,
          727,
          1391,
          519,
          295,
          867,
          661,
          2098,
          300,
          291,
          727,
          19719,
          341,
          13,
          51556
        ]
      },
      {
        "avg_logprob": -0.2369232515318204,
        "compression_ratio": 1.596958174904943,
        "end": 576.88,
        "id": 165,
        "no_speech_prob": 0.00042388279689475894,
        "seek": 55056,
        "start": 574.4,
        "temperature": 0,
        "text": " So saving it, making the interface.",
        "tokens": [
          51556,
          407,
          6816,
          309,
          11,
          1455,
          264,
          9226,
          13,
          51680
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 580.08,
        "id": 166,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 577.68,
        "temperature": 0,
        "text": " Now we could also take this and we could go and add",
        "tokens": [
          50404,
          823,
          321,
          727,
          611,
          747,
          341,
          293,
          321,
          727,
          352,
          293,
          909,
          50524
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 582.64,
        "id": 167,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 580.08,
        "temperature": 0,
        "text": " all the fun coding train design elements",
        "tokens": [
          50524,
          439,
          264,
          1019,
          17720,
          3847,
          1715,
          4959,
          50652
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 585.68,
        "id": 168,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 582.64,
        "temperature": 0,
        "text": " from the original, from that Flappy Bird clone repository.",
        "tokens": [
          50652,
          490,
          264,
          3380,
          11,
          490,
          300,
          479,
          875,
          7966,
          15931,
          26506,
          25841,
          13,
          50804
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 587.04,
        "id": 169,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 585.68,
        "temperature": 0,
        "text": " Pull requests now start.",
        "tokens": [
          50804,
          15074,
          12475,
          586,
          722,
          13,
          50872
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 588.48,
        "id": 170,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 587.04,
        "temperature": 0,
        "text": " The floodgates are open.",
        "tokens": [
          50872,
          440,
          10481,
          70,
          1024,
          366,
          1269,
          13,
          50944
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 592.4,
        "id": 171,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 588.48,
        "temperature": 0,
        "text": " So we can make a version of this that maybe has even",
        "tokens": [
          50944,
          407,
          321,
          393,
          652,
          257,
          3037,
          295,
          341,
          300,
          1310,
          575,
          754,
          51140
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 596,
        "id": 172,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 592.96,
        "temperature": 0,
        "text": " a saved pre-trained Flappy Bird agent.",
        "tokens": [
          51168,
          257,
          6624,
          659,
          12,
          17227,
          2001,
          479,
          875,
          7966,
          15931,
          9461,
          13,
          51320
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 598.32,
        "id": 173,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 596,
        "temperature": 0,
        "text": " Another suggestion that just came in from the chat",
        "tokens": [
          51320,
          3996,
          16541,
          300,
          445,
          1361,
          294,
          490,
          264,
          5081,
          51436
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 601.68,
        "id": 174,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 598.32,
        "temperature": 0,
        "text": " is the actual Y velocity as an input as well.",
        "tokens": [
          51436,
          307,
          264,
          3539,
          398,
          9269,
          382,
          364,
          4846,
          382,
          731,
          13,
          51604
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 603.2,
        "id": 175,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 601.68,
        "temperature": 0,
        "text": " Because if I'm already moving up,",
        "tokens": [
          51604,
          1436,
          498,
          286,
          478,
          1217,
          2684,
          493,
          11,
          51680
        ]
      },
      {
        "avg_logprob": -0.22027741056500058,
        "compression_ratio": 1.6401384083044983,
        "end": 605.6,
        "id": 176,
        "no_speech_prob": 0.012624586001038551,
        "seek": 57688,
        "start": 603.76,
        "temperature": 0,
        "text": " maybe I shouldn't jump versus if I'm moving down.",
        "tokens": [
          51708,
          1310,
          286,
          4659,
          380,
          3012,
          5717,
          498,
          286,
          478,
          2684,
          760,
          13,
          51800
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 606.64,
        "id": 177,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 605.6,
        "temperature": 0,
        "text": " So that's something you can add.",
        "tokens": [
          50364,
          407,
          300,
          311,
          746,
          291,
          393,
          909,
          13,
          50416
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 607.76,
        "id": 178,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 606.64,
        "temperature": 0,
        "text": " Oh, and here's another one.",
        "tokens": [
          50416,
          876,
          11,
          293,
          510,
          311,
          1071,
          472,
          13,
          50472
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 615.2,
        "id": 179,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 608.32,
        "temperature": 0,
        "text": " What if actually, instead of my own manual feature extraction,",
        "tokens": [
          50500,
          708,
          498,
          767,
          11,
          2602,
          295,
          452,
          1065,
          9688,
          4111,
          30197,
          11,
          50844
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 619.6,
        "id": 180,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 615.2,
        "temperature": 0,
        "text": " where I'm kind of evaluating the Flappy Bird scene to make inputs,",
        "tokens": [
          50844,
          689,
          286,
          478,
          733,
          295,
          27479,
          264,
          479,
          875,
          7966,
          15931,
          4145,
          281,
          652,
          15743,
          11,
          51064
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 621.76,
        "id": 181,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 619.6,
        "temperature": 0,
        "text": " what if just the pixels,",
        "tokens": [
          51064,
          437,
          498,
          445,
          264,
          18668,
          11,
          51172
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 625.84,
        "id": 182,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 621.76,
        "temperature": 0,
        "text": " what if it was just the pixels of the game itself that became the input?",
        "tokens": [
          51172,
          437,
          498,
          309,
          390,
          445,
          264,
          18668,
          295,
          264,
          1216,
          2564,
          300,
          3062,
          264,
          4846,
          30,
          51376
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 627.84,
        "id": 183,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 625.84,
        "temperature": 0,
        "text": " But now we have a problem, right?",
        "tokens": [
          51376,
          583,
          586,
          321,
          362,
          257,
          1154,
          11,
          558,
          30,
          51476
        ]
      },
      {
        "avg_logprob": -0.18532896041870117,
        "compression_ratio": 1.6309012875536482,
        "end": 632.64,
        "id": 184,
        "no_speech_prob": 0.00003647842459031381,
        "seek": 60560,
        "start": 627.84,
        "temperature": 0,
        "text": " We have a problem, which is that I am looking at a scene.",
        "tokens": [
          51476,
          492,
          362,
          257,
          1154,
          11,
          597,
          307,
          300,
          286,
          669,
          1237,
          412,
          257,
          4145,
          13,
          51716
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 634,
        "id": 185,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 632.64,
        "temperature": 0,
        "text": " Let me just speed this up.",
        "tokens": [
          50364,
          961,
          385,
          445,
          3073,
          341,
          493,
          13,
          50432
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 636.88,
        "id": 186,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 635.04,
        "temperature": 0,
        "text": " With, ah, come on.",
        "tokens": [
          50484,
          2022,
          11,
          3716,
          11,
          808,
          322,
          13,
          50576
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 639.28,
        "id": 187,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 637.76,
        "temperature": 0,
        "text": " Let me just refresh, right?",
        "tokens": [
          50620,
          961,
          385,
          445,
          15134,
          11,
          558,
          30,
          50696
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 643.4399999999999,
        "id": 188,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 639.28,
        "temperature": 0,
        "text": " I am looking at a scene with many, many, many, many birds playing all at once.",
        "tokens": [
          50696,
          286,
          669,
          1237,
          412,
          257,
          4145,
          365,
          867,
          11,
          867,
          11,
          867,
          11,
          867,
          9009,
          2433,
          439,
          412,
          1564,
          13,
          50904
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 648.48,
        "id": 189,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 644.48,
        "temperature": 0,
        "text": " If I want to use the pixels of the game as input,",
        "tokens": [
          50956,
          759,
          286,
          528,
          281,
          764,
          264,
          18668,
          295,
          264,
          1216,
          382,
          4846,
          11,
          51156
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 650.8,
        "id": 190,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 648.48,
        "temperature": 0,
        "text": " I cannot have the whole population there.",
        "tokens": [
          51156,
          286,
          2644,
          362,
          264,
          1379,
          4415,
          456,
          13,
          51272
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 655.04,
        "id": 191,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 651.6,
        "temperature": 0,
        "text": " I need to have just the one current bird that is playing.",
        "tokens": [
          51312,
          286,
          643,
          281,
          362,
          445,
          264,
          472,
          2190,
          5255,
          300,
          307,
          2433,
          13,
          51484
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 657.52,
        "id": 192,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 655.04,
        "temperature": 0,
        "text": " So I have to create a graphics context",
        "tokens": [
          51484,
          407,
          286,
          362,
          281,
          1884,
          257,
          11837,
          4319,
          51608
        ]
      },
      {
        "avg_logprob": -0.34264555922499645,
        "compression_ratio": 1.648068669527897,
        "end": 659.84,
        "id": 193,
        "no_speech_prob": 0.000054759679187554866,
        "seek": 63264,
        "start": 657.52,
        "temperature": 0,
        "text": " for every single member of the population.",
        "tokens": [
          51608,
          337,
          633,
          2167,
          4006,
          295,
          264,
          4415,
          13,
          51724
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 663.12,
        "id": 194,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 659.84,
        "temperature": 0,
        "text": " And in fact, this Flappy Bird game is so simple",
        "tokens": [
          50364,
          400,
          294,
          1186,
          11,
          341,
          479,
          875,
          7966,
          15931,
          1216,
          307,
          370,
          2199,
          50528
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 666.1600000000001,
        "id": 195,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 663.12,
        "temperature": 0,
        "text": " that actually, as the player plays the game,",
        "tokens": [
          50528,
          300,
          767,
          11,
          382,
          264,
          4256,
          5749,
          264,
          1216,
          11,
          50680
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 668,
        "id": 196,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 666.1600000000001,
        "temperature": 0,
        "text": " it doesn't affect how the game plays out.",
        "tokens": [
          50680,
          309,
          1177,
          380,
          3345,
          577,
          264,
          1216,
          5749,
          484,
          13,
          50772
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 670.72,
        "id": 197,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 668,
        "temperature": 0,
        "text": " Imagine if I were doing my asteroids coding challenge.",
        "tokens": [
          50772,
          11739,
          498,
          286,
          645,
          884,
          452,
          50230,
          17720,
          3430,
          13,
          50908
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 674.08,
        "id": 198,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 671.36,
        "temperature": 0,
        "text": " I can't have them all play in the same world",
        "tokens": [
          50940,
          286,
          393,
          380,
          362,
          552,
          439,
          862,
          294,
          264,
          912,
          1002,
          51076
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 677.44,
        "id": 199,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 674.08,
        "temperature": 0,
        "text": " because if one shoots an asteroid, the other one hasn't shot that asteroid.",
        "tokens": [
          51076,
          570,
          498,
          472,
          20704,
          364,
          33833,
          11,
          264,
          661,
          472,
          6132,
          380,
          3347,
          300,
          33833,
          13,
          51244
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 679.52,
        "id": 200,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 677.44,
        "temperature": 0,
        "text": " The game is changing for all of them.",
        "tokens": [
          51244,
          440,
          1216,
          307,
          4473,
          337,
          439,
          295,
          552,
          13,
          51348
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 682.24,
        "id": 201,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 679.52,
        "temperature": 0,
        "text": " So another thought experiment could be,",
        "tokens": [
          51348,
          407,
          1071,
          1194,
          5120,
          727,
          312,
          11,
          51484
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 684.5600000000001,
        "id": 202,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 682.24,
        "temperature": 0,
        "text": " how could I actually make this work",
        "tokens": [
          51484,
          577,
          727,
          286,
          767,
          652,
          341,
          589,
          51600
        ]
      },
      {
        "avg_logprob": -0.2659315656321917,
        "compression_ratio": 1.7581227436823104,
        "end": 688.4000000000001,
        "id": 203,
        "no_speech_prob": 0.050328899174928665,
        "seek": 65984,
        "start": 684.5600000000001,
        "temperature": 0,
        "text": " where I see, instead of all of them playing within one canvas,",
        "tokens": [
          51600,
          689,
          286,
          536,
          11,
          2602,
          295,
          439,
          295,
          552,
          2433,
          1951,
          472,
          16267,
          11,
          51792
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 691.12,
        "id": 204,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 688.88,
        "temperature": 0,
        "text": " a grid of the whole population,",
        "tokens": [
          50388,
          257,
          10748,
          295,
          264,
          1379,
          4415,
          11,
          50500
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 694,
        "id": 205,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 691.12,
        "temperature": 0,
        "text": " lots of little, little, little versions of the game all playing together?",
        "tokens": [
          50500,
          3195,
          295,
          707,
          11,
          707,
          11,
          707,
          9606,
          295,
          264,
          1216,
          439,
          2433,
          1214,
          30,
          50644
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 697.12,
        "id": 206,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 694,
        "temperature": 0,
        "text": " You know, to be honest, to make this work, I don't have to see anything at all.",
        "tokens": [
          50644,
          509,
          458,
          11,
          281,
          312,
          3245,
          11,
          281,
          652,
          341,
          589,
          11,
          286,
          500,
          380,
          362,
          281,
          536,
          1340,
          412,
          439,
          13,
          50800
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 699.12,
        "id": 207,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 697.12,
        "temperature": 0,
        "text": " But try doing this with another game.",
        "tokens": [
          50800,
          583,
          853,
          884,
          341,
          365,
          1071,
          1216,
          13,
          50900
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 704.4,
        "id": 208,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 699.12,
        "temperature": 0,
        "text": " Can you find my snake challenge, my asteroids challenge, my frogger challenge?",
        "tokens": [
          50900,
          1664,
          291,
          915,
          452,
          12650,
          3430,
          11,
          452,
          50230,
          3430,
          11,
          452,
          17259,
          1321,
          3430,
          30,
          51164
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 707.92,
        "id": 209,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 704.4,
        "temperature": 0,
        "text": " Any of those now could be trained,",
        "tokens": [
          51164,
          2639,
          295,
          729,
          586,
          727,
          312,
          8895,
          11,
          51340
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 712.72,
        "id": 210,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 707.92,
        "temperature": 0,
        "text": " a bot could be trained to play that game using the exact same technique.",
        "tokens": [
          51340,
          257,
          10592,
          727,
          312,
          8895,
          281,
          862,
          300,
          1216,
          1228,
          264,
          1900,
          912,
          6532,
          13,
          51580
        ]
      },
      {
        "avg_logprob": -0.28247178395589195,
        "compression_ratio": 1.7207547169811321,
        "end": 715.52,
        "id": 211,
        "no_speech_prob": 0.0000748459278838709,
        "seek": 68840,
        "start": 713.6,
        "temperature": 0,
        "text": " All right, from the chat, I also just learned",
        "tokens": [
          51624,
          1057,
          558,
          11,
          490,
          264,
          5081,
          11,
          286,
          611,
          445,
          3264,
          51720
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 717.6,
        "id": 212,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 715.6,
        "temperature": 0,
        "text": " that I have a couple mistakes.",
        "tokens": [
          50368,
          300,
          286,
          362,
          257,
          1916,
          8038,
          13,
          50468
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 720.64,
        "id": 213,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 717.6,
        "temperature": 0,
        "text": " One is that, supposedly, I guess in the game Flappy Bird,",
        "tokens": [
          50468,
          1485,
          307,
          300,
          11,
          20581,
          11,
          286,
          2041,
          294,
          264,
          1216,
          479,
          875,
          7966,
          15931,
          11,
          50620
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 722.64,
        "id": 214,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 720.64,
        "temperature": 0,
        "text": " if you go to the bottom, you die.",
        "tokens": [
          50620,
          498,
          291,
          352,
          281,
          264,
          2767,
          11,
          291,
          978,
          13,
          50720
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 724.48,
        "id": 215,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 723.36,
        "temperature": 0,
        "text": " So I don't have that implemented.",
        "tokens": [
          50756,
          407,
          286,
          500,
          380,
          362,
          300,
          12270,
          13,
          50812
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 727.04,
        "id": 216,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 724.48,
        "temperature": 0,
        "text": " It'd be interesting to see if it could solve itself with that.",
        "tokens": [
          50812,
          467,
          1116,
          312,
          1880,
          281,
          536,
          498,
          309,
          727,
          5039,
          2564,
          365,
          300,
          13,
          50940
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 729.76,
        "id": 217,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 727.04,
        "temperature": 0,
        "text": " Also, the way that I'm finding the closest pipe,",
        "tokens": [
          50940,
          2743,
          11,
          264,
          636,
          300,
          286,
          478,
          5006,
          264,
          13699,
          11240,
          11,
          51076
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 734,
        "id": 218,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 730.3199999999999,
        "temperature": 0,
        "text": " it's no longer the closest pipe right as it gets right up to it.",
        "tokens": [
          51104,
          309,
          311,
          572,
          2854,
          264,
          13699,
          11240,
          558,
          382,
          309,
          2170,
          558,
          493,
          281,
          309,
          13,
          51288
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 736.88,
        "id": 219,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 734,
        "temperature": 0,
        "text": " And so I really should not consider it to be the closest pipe",
        "tokens": [
          51288,
          400,
          370,
          286,
          534,
          820,
          406,
          1949,
          309,
          281,
          312,
          264,
          13699,
          11240,
          51432
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 738.24,
        "id": 220,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 736.88,
        "temperature": 0,
        "text": " until it's gone all the way past.",
        "tokens": [
          51432,
          1826,
          309,
          311,
          2780,
          439,
          264,
          636,
          1791,
          13,
          51500
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 740.96,
        "id": 221,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 738.24,
        "temperature": 0,
        "text": " But anyway, there's lots of little tweaks and minor things.",
        "tokens": [
          51500,
          583,
          4033,
          11,
          456,
          311,
          3195,
          295,
          707,
          46664,
          293,
          6696,
          721,
          13,
          51636
        ]
      },
      {
        "avg_logprob": -0.323727656633426,
        "compression_ratio": 1.7306501547987616,
        "end": 745.28,
        "id": 222,
        "no_speech_prob": 0.0009697253117337823,
        "seek": 71552,
        "start": 740.96,
        "temperature": 0,
        "text": " So I would gladly encourage you to create a little bit of a challenge",
        "tokens": [
          51636,
          407,
          286,
          576,
          47307,
          5373,
          291,
          281,
          1884,
          257,
          707,
          857,
          295,
          257,
          3430,
          51852
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 747.6,
        "id": 223,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 745.6,
        "temperature": 0,
        "text": " and maybe do a little bit of a code challenge.",
        "tokens": [
          50380,
          293,
          1310,
          360,
          257,
          707,
          857,
          295,
          257,
          3089,
          3430,
          13,
          50480
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 748.64,
        "id": 224,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 747.6,
        "temperature": 0,
        "text": " I'm not sure if I'm going to do that.",
        "tokens": [
          50480,
          286,
          478,
          406,
          988,
          498,
          286,
          478,
          516,
          281,
          360,
          300,
          13,
          50532
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 751.52,
        "id": 225,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 748.64,
        "temperature": 0,
        "text": " But I would love to see you create your own versions of this",
        "tokens": [
          50532,
          583,
          286,
          576,
          959,
          281,
          536,
          291,
          1884,
          428,
          1065,
          9606,
          295,
          341,
          50676
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 753.8399999999999,
        "id": 226,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 751.52,
        "temperature": 0,
        "text": " and submit them as community contributions.",
        "tokens": [
          50676,
          293,
          10315,
          552,
          382,
          1768,
          15725,
          13,
          50792
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 756.8,
        "id": 227,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 753.8399999999999,
        "temperature": 0,
        "text": " And I'll look at them and maybe do another live stream next week",
        "tokens": [
          50792,
          400,
          286,
          603,
          574,
          412,
          552,
          293,
          1310,
          360,
          1071,
          1621,
          4309,
          958,
          1243,
          50940
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 758.72,
        "id": 228,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 756.8,
        "temperature": 0,
        "text": " or whenever I do the next live stream,",
        "tokens": [
          50940,
          420,
          5699,
          286,
          360,
          264,
          958,
          1621,
          4309,
          11,
          51036
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 761.76,
        "id": 229,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 758.72,
        "temperature": 0,
        "text": " going through and showing some of these community contributions.",
        "tokens": [
          51036,
          516,
          807,
          293,
          4099,
          512,
          295,
          613,
          1768,
          15725,
          13,
          51188
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 768.3199999999999,
        "id": 230,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 761.76,
        "temperature": 0,
        "text": " All right, so thank you very much for watching this three-part coding challenge.",
        "tokens": [
          51188,
          1057,
          558,
          11,
          370,
          1309,
          291,
          588,
          709,
          337,
          1976,
          341,
          1045,
          12,
          6971,
          17720,
          3430,
          13,
          51516
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 770.72,
        "id": 231,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 768.3199999999999,
        "temperature": 0,
        "text": " If it was really in three parts, I think it was two, three.",
        "tokens": [
          51516,
          759,
          309,
          390,
          534,
          294,
          1045,
          3166,
          11,
          286,
          519,
          309,
          390,
          732,
          11,
          1045,
          13,
          51636
        ]
      },
      {
        "avg_logprob": -0.5677702043439958,
        "compression_ratio": 1.86,
        "end": 775.1999999999999,
        "id": 232,
        "no_speech_prob": 0.004982127342373133,
        "seek": 74528,
        "start": 771.28,
        "temperature": 0,
        "text": " I look forward to doing more, many more coding challenges.",
        "tokens": [
          51664,
          286,
          574,
          2128,
          281,
          884,
          544,
          11,
          867,
          544,
          17720,
          4759,
          13,
          51860
        ]
      },
      {
        "avg_logprob": -0.4637036604039809,
        "compression_ratio": 0.875,
        "end": 778.24,
        "id": 233,
        "no_speech_prob": 0.004133284091949463,
        "seek": 77520,
        "start": 775.2,
        "temperature": 0,
        "text": " Thanks for watching, and I'll see you in a future video.",
        "tokens": [
          50404,
          2561,
          337,
          1976,
          11,
          293,
          286,
          603,
          536,
          291,
          294,
          257,
          2027,
          960,
          13,
          50516
        ]
      }
    ],
    "transcription": " Welcome to part three of my neuro evolution flappy bird coding challenge. Alright, so, oh, come on, jump, jump, jump, oh yes! Oh, that looks like it probably should have hit the pipe there. But anyway, that aside, look at these birds, they're really trying to get through those gaps in the pipes. Alright, so, I want to do a few things to improve. Number one, this is a bit ridiculous that I have to sit here and wait and let it run. It's kind of enjoyable, I mean, I have to go soon, so I got to speed this up. So here's one thing I'm going to do to speed this up. I am going to, I don't need to, all of the logic that's driving the game, I could just do that many, many times per frame, right? I don't actually need to draw and show the animation. So, for example, if I were, so one thing I need to do is separate out all the logic of the game from the drawing. So one thing I'm going to do is I'm going to take out pipes.show here. Let me just, this is, like, counter is part of the game logic. Looping through all the pipes and checking them, that's part of the game. Looping through all the birds. If birds, this is part of the game. So, what I just, but drawing is not. So let me take out bird.show, and now I'm just going to, at the end here, set, say let, for let bird of birds, bird.show, for let pipe of pipes, pipe.show. This should be exactly the same. Nothing has changed, but now, and I might as well take the background and put the background here. So this is all the drawing stuff. And now what I'm going to do is I'm going to say for let n equal zero, n is less than cycles, so I'll call it cycles, n plus plus. I'm going to do all of the game logic stuff many, many times, which goes all the way down to here. So, and then I'm going to say let cycles equal 100. So, watch this. So now this is just, like, doing it really, really fast over time. But I want to be able to, like, speed it up and slow it down also. So one thing that I can do to do that is I am going to say let slider, slider equals create slider. I'm going to make a slider that goes from one to 100 and starts at one, and that, I'm just going to say slider.value. So instead of a variable called cycles, I'm just going to use the value of that slider. So now I've got a little slider here. I can speed things up, and I can slow them back down. So this is going to help me a lot where instead of having to, like, speed up the video that you're watching, if I'm editing this after the fact, I can actually just do this. So in fact, I could let this go for a little bit, and then I could come back and see, like, how are we doing here? We got one that's doing pretty well? So but here's the thing. I also think there are some things that I really need to do to help the evolutionary process along. One thing is I'm going to go into the neural network code and look at how I'm doing mutation. So this is the mutate function. So first of all, I'm a little suspicious here. I want to make sure that mutation is really happening. So let me at least console log test. I just want to see. I just want to see that this is happening in the code somewhere, and I'm going to go back to sketch.js and just have, like, three birds because I don't want it to go crazy. So as soon as the three are done, I should see mutate mutate. Yeah, so I got a bunch of mutations. That's good. So that is being called. So hopefully, the weights are being changed. But you know what might work better? I think I have a sneaking suspicion is that it might actually help me to, rather than when I mutate, pick an entirely new weight. So what's going on here? The internals of the neural network are something that I haven't talked about in this series. That's really something reserved for my other videos about how neural networks work. The only thing that's relevant to this series is the inputs of the game come in, the outputs come out, and the output tells me whether I should jump or not. But what's actually stored inside the neural network are a bunch of numbers between negative 1 and 1 for all of these connections. And so mutation means just pluck one of them out and give it a new value. But if it's one that's doing pretty well, if I just were to, like, tweak the value a little bit instead of just giving it a whole new value. And also, what if I were to just tweak it a tiny bit? So one way I could do that is use a Gaussian distribution of random numbers. Random numbers that cluster around like a mean of 0, but are often like... So just like I want to give each a little bit of a nudge. So let me come back over here. And I'm going to have to build in Gaussian random number generation into the library, the neural network library. Right now, since I have p5 involved, just to fix that, I'm going to comment this out and I'm going to say return val. So this would be just like me adjusting the value a little bit, right? Return the actual value changed by a little random amount. But I think I'm going to... Because when I did this for my class last year, actually, I remember using random Gaussian, which is this is now a p5 function from the p5 library. This doesn't exist in JavaScript. So I've got to implement that into my library. It'll work for now. The mean would be 0 and the standard deviation would be like 0.1. Let's try that. So let's see how this goes. Oh, and let me... I took out the console log and let's give this a whirl now. All right, so I'm going to speed this up. And I'm also... Oh, wait, there's only like two of them. So the other thing I need to do is let's give myself a nice population. Let's try 500 of these. All right, so there's 500 going. I'm going to speed this up and I'm going to let this run for a little while and I will be back. All right, so it looks to me like we've done it. There's so much variation in here, but I've evolved it for now 76 generations. And what we see here is that we've got a bunch that are still alive. And there's probably one in there that has pretty optimal weights. And it's making it through. Now, one thing that I would really like to do is... It's kind of more fun to watch this if I show only the one that's kind of doing best. But at some point, there'll only be one left. But I think we're pretty much there. Here's some things that you might want to... That I would love to do with this in another part. But I'm going to leave them as exercises for the viewer. Come on. There you go. Oh, you... Oh, it started over. So a couple of things. One is that I can keep track of the single agent, the single bird that has done the best over all time, over all generations. And I could save that in a variable like the best bird, record bird. And what I could do is actually I could download those weights. I could save those weights to like a JSON file. It's neural network configuration to a JSON file. Or even just save it in memory. And I could always just click a button or something and say, let me just see this one. So that's something I would really love to add to this. It's actually... I might as well just show it to you. Because the example that I made for my course, which is here, is essentially exactly the same. But there's a couple different things. So one is I'm going to... So let me just ramp up the speed here. So let me let this run for a little while. The game's maybe a little bit different in this one. But so one of the things that I like about this is once I've let it run for a little while, I can just click run best so far. So now... And I could speed this up a little bit. But I think I better train it more because I didn't get a really good one. So this is the idea of being able to just run the best one that's done, you know, that has the best high score. Oh, look at this. Okay. So I'm going to... Now I'm going to do this. So we can see this one that kind of has a pretty good sense of how to play the game. Kind of running... Just watch that one on its own. It's very satisfying to watch. I'm like amazed that it's doing it. So that's one thing that would be worth as an exercise trying to add to this implementation that I made. Now, once I even do that, I could save it to a file. So there's actually in the toy neural network library, there's like a save JSON, I think, function. So I could save it to a file to reload later. So if I'm training it all day or for an hour, I could save it and bring it into another environment. That's one thing. Another thing that I might want to do is I could probably optimize the fitness function quite a bit. For example, what if... And this is a... Thank you to Versatilis for this suggestion. What if I actually scored not just how long it lasts, but how well it maintains its Y location in between the top and bottom of those two pipes? So that would be... You could probably think of many other ways that you could optimize this. So saving it, making the interface. Now we could also take this and we could go and add all the fun coding train design elements from the original, from that Flappy Bird clone repository. Pull requests now start. The floodgates are open. So we can make a version of this that maybe has even a saved pre-trained Flappy Bird agent. Another suggestion that just came in from the chat is the actual Y velocity as an input as well. Because if I'm already moving up, maybe I shouldn't jump versus if I'm moving down. So that's something you can add. Oh, and here's another one. What if actually, instead of my own manual feature extraction, where I'm kind of evaluating the Flappy Bird scene to make inputs, what if just the pixels, what if it was just the pixels of the game itself that became the input? But now we have a problem, right? We have a problem, which is that I am looking at a scene. Let me just speed this up. With, ah, come on. Let me just refresh, right? I am looking at a scene with many, many, many, many birds playing all at once. If I want to use the pixels of the game as input, I cannot have the whole population there. I need to have just the one current bird that is playing. So I have to create a graphics context for every single member of the population. And in fact, this Flappy Bird game is so simple that actually, as the player plays the game, it doesn't affect how the game plays out. Imagine if I were doing my asteroids coding challenge. I can't have them all play in the same world because if one shoots an asteroid, the other one hasn't shot that asteroid. The game is changing for all of them. So another thought experiment could be, how could I actually make this work where I see, instead of all of them playing within one canvas, a grid of the whole population, lots of little, little, little versions of the game all playing together? You know, to be honest, to make this work, I don't have to see anything at all. But try doing this with another game. Can you find my snake challenge, my asteroids challenge, my frogger challenge? Any of those now could be trained, a bot could be trained to play that game using the exact same technique. All right, from the chat, I also just learned that I have a couple mistakes. One is that, supposedly, I guess in the game Flappy Bird, if you go to the bottom, you die. So I don't have that implemented. It'd be interesting to see if it could solve itself with that. Also, the way that I'm finding the closest pipe, it's no longer the closest pipe right as it gets right up to it. And so I really should not consider it to be the closest pipe until it's gone all the way past. But anyway, there's lots of little tweaks and minor things. So I would gladly encourage you to create a little bit of a challenge and maybe do a little bit of a code challenge. I'm not sure if I'm going to do that. But I would love to see you create your own versions of this and submit them as community contributions. And I'll look at them and maybe do another live stream next week or whenever I do the next live stream, going through and showing some of these community contributions. All right, so thank you very much for watching this three-part coding challenge. If it was really in three parts, I think it was two, three. I look forward to doing more, many more coding challenges. Thanks for watching, and I'll see you in a future video.",
    "translation": null
  },
  "error": null,
  "status": "succeeded",
  "created_at": "2023-09-26T21:03:58.466805Z",
  "started_at": "2023-09-26T21:25:39.501662Z",
  "completed_at": "2023-09-26T21:29:27.683504Z",
  "webhook": "https://83ceaa0b612c.ngrok.app/?video_id=3lvj9jvERvs",
  "webhook_events_filter": [
    "completed"
  ],
  "metrics": {
    "predict_time": 228.181842
  },
  "urls": {
    "cancel": "https://api.replicate.com/v1/predictions/wyqxmnjbwmoa7a3f2eksuwljbq/cancel",
    "get": "https://api.replicate.com/v1/predictions/wyqxmnjbwmoa7a3f2eksuwljbq"
  }
}