id stringlengths 66 88 | text stringlengths 46 1.7k | duration float64 2.32 30 | dnsmos float64 2.8 3.65 | sampling_rate int64 24k 24k | path stringlengths 79 101 | audio_content_token_indices listlengths 58 749 | audio_global_embedding listlengths 128 128 | audio_token_len int64 58 749 | audio_duration float64 2.32 30 |
|---|---|---|---|---|---|---|---|---|---|
bilibili_data_1897938584_BV1pP4y1k7Jr_p168_BV1pP4y1k7Jr_p168_m4-dialogue_0764225 | [S1] So for security in general, there's, there's so many, I mean, and there's, I'm sure there's a, a, two dozen YouTube channels that could probably hook you up with like incredible, um, so maybe I, I, we can send someone and link some of those below or something. [S2] Mm-hmm. [S1] Um, I wish I could say that there wa... | 15.16 | 3.13746 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p168_BV1pP4y1k7Jr_p168_m4-dialogue_0764225.mp3 | [
4139,
3620,
6738,
3554,
5520,
3988,
11115,
7914,
7892,
5327,
994,
8177,
2995,
10637,
4074,
9168,
5595,
9943,
11610,
10152,
7579,
10538,
5449,
9282,
8532,
12033,
12205,
11870,
9356,
10185,
6039,
11190,
11924,
8045,
5459,
7963,
7866,
8333,
28... | [
0.7508398294448853,
-5.759929180145264,
-0.056198850274086,
-0.11141238361597061,
0.6000474095344543,
-0.16140349209308624,
0.4480386972427368,
0.008424490690231323,
0.04337649047374725,
-3.8917391300201416,
0.2494789958000183,
-0.008212843909859657,
-0.31358253955841064,
0.258335918188095... | 379 | 15.16 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p168_BV1pP4y1k7Jr_p168_m4-dialogue_0764228 | [S1] Excellent. Kevin, thank you so much for being here and bringing this a bit, a bit closer. I, I know more. I hope everyone else does too now. Uh, yeah, thanks. [S2] Thanks so much for having me. This has been a blast. [S1] Excellent. [S2] Super appreciate it. [S1] Bye. | 13.6 | 3.082241 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p168_BV1pP4y1k7Jr_p168_m4-dialogue_0764228.mp3 | [
11848,
9952,
12386,
12790,
12653,
9841,
9636,
993,
1416,
3239,
2670,
732,
4638,
9746,
9344,
4065,
4124,
12338,
5808,
8266,
7825,
9385,
12701,
5024,
7510,
11252,
11242,
3299,
7022,
2341,
4174,
6735,
6732,
5984,
1396,
12734,
12053,
12325,
642... | [
0.8669915795326233,
-5.761979579925537,
-0.12784768640995026,
0.01304731797426939,
0.5600253939628601,
-0.13164503872394562,
0.5026556849479675,
-0.022916769608855247,
0.024764409288764,
-3.8641905784606934,
0.060566939413547516,
-0.052647046744823456,
-0.035509686917066574,
0.094256997108... | 340 | 13.6 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005032 | [S1] Welcome everyone. Today I have with me Armin Aghajanian and I've practiced that name 10 seconds ago and I th- I think I got it down. Uh, Armin is the first author of the CM3 paper. Welcome, uh, Armin, to the channel. [S2] Thank you for having me. [S1] So I, I saw this paper and of course you have, like, uh, some b... | 29.96 | 3.467483 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005032.mp3 | [
569,
7325,
113,
8694,
11066,
5339,
9897,
12265,
8115,
8221,
11773,
11766,
3372,
2971,
11222,
10054,
627,
5430,
6007,
829,
4115,
3590,
3616,
11265,
12248,
4500,
7233,
6110,
12725,
12717,
12252,
12228,
11717,
6607,
9322,
11873,
8873,
10862,
6... | [
1.1652218103408813,
-5.334449768066406,
-0.13369393348693848,
-0.015624097548425198,
0.48509618639945984,
-0.1465269923210144,
0.31609341502189636,
-0.05134240910410881,
0.23916324973106384,
-3.867659091949463,
0.3552272617816925,
0.005989860277622938,
0.15661866962909698,
-0.0020744677167... | 749 | 29.96 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005033 | [S1] I mean, the goal here was kind of to have a single multimodal model that can do everything. [S2] Yeah. [S1] Um, image generation, image captioning, uh, image infilling to, uh, to even pure text tasks- [S2] Yeah. [S1] ... like summarization, but mostly focusing on this zero-shot setting specifically- [S2] Mm-hmm. [... | 18.32 | 3.078185 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005033.mp3 | [
9248,
5672,
1684,
3109,
9428,
12674,
11285,
8278,
3669,
3601,
4115,
5640,
8171,
10981,
6250,
11611,
7981,
10354,
10347,
8342,
4443,
10152,
10131,
11733,
5478,
2900,
108,
2870,
3922,
2897,
4258,
12130,
3506,
12030,
11413,
9071,
3107,
8520,
4... | [
0.5553148984909058,
-5.840043544769287,
-0.1398170292377472,
-0.0014071178156882524,
0.8057975172996521,
0.008617122657597065,
0.48267054557800293,
-0.004623099230229855,
-0.12895868718624115,
-3.8435122966766357,
0.3161575198173523,
-0.44937729835510254,
-0.24903933703899384,
0.1715844571... | 458 | 18.32 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005034 | [S1] ... in other modalities than text. [S2] Yeah. [S1] Um, this, this goes even further. This is multimodal, uh, there have been a lot of other approaches to multimodal. There is like this, this Ru, Rudolf, Rudolph even model, I don't know if you've seen that, that goes like image to text to image and so on. And they ... | 29.16 | 3.347572 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005034.mp3 | [
4713,
5160,
11046,
11716,
9606,
10829,
5022,
11314,
9087,
2877,
4207,
2915,
11054,
5452,
6172,
7789,
10486,
8962,
8623,
11701,
10738,
7789,
8092,
11253,
8808,
10704,
6642,
11659,
12165,
4570,
2960,
529,
2706,
11638,
12148,
10317,
6697,
11073,... | [
1.2701040506362915,
-5.414717197418213,
-0.10515940934419632,
-0.04888874664902687,
0.29527994990348816,
-0.08567271381616592,
0.3183925449848175,
-0.05983477458357811,
0.20224198698997498,
-3.8365166187286377,
0.38221025466918945,
0.16748137772083282,
0.12991736829280853,
0.04697494208812... | 729 | 29.16 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005036 | [S1] ... to be used, um, in HTML. So after DALI came out, we thought, okay, um, there are some fundamental restrictions with DALI. So the first one being the causal, uh, approach. So they train a decoder-only left-to-right model. [S2] Mm-hmm. [S1] So in some sense, you can't do things like generate the text given the i... | 29.24 | 3.190094 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005036.mp3 | [
5680,
3105,
8001,
8152,
6118,
5652,
5737,
4068,
10054,
12613,
7494,
4933,
4485,
4682,
4761,
7377,
12483,
9988,
9030,
3399,
719,
774,
4293,
3472,
3464,
1152,
5760,
4056,
9458,
9385,
9813,
9839,
4654,
92,
1627,
4123,
2729,
5657,
6822,
9268,... | [
0.4815290570259094,
-5.658483982086182,
-0.03792931139469147,
-0.04204174503684044,
0.900390625,
-0.04937825724482536,
0.530354380607605,
0.04916877672076225,
-0.0992317795753479,
-3.914426326751709,
0.2603709101676941,
-0.30559977889060974,
-0.2749859094619751,
0.20246441662311554,
-0.0... | 731 | 29.24 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005037 | [S1] Exactly. Yeah. So those were kind of the, the first weaknesses, uh, that we saw there. Uh, the approach was very clever though, right? So pretty much taking continuous data, discretizing it, and just doing sequence modeling, it seems to work very, very well. [S2] Mm-hmm. [S1] So the idea went that we could kind of... | 22.16 | 3.067929 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005037.mp3 | [
6689,
5800,
4108,
7390,
12435,
897,
4447,
4918,
3756,
9449,
674,
1117,
6184,
1539,
7299,
12742,
10717,
6134,
3885,
9377,
11996,
12597,
8877,
12589,
12532,
12573,
12525,
11445,
9963,
6233,
6048,
8144,
8043,
8049,
10477,
5270,
3738,
6765,
564... | [
0.5190719366073608,
-5.809675216674805,
-0.08339738100767136,
-0.0922091156244278,
0.8385006785392761,
-0.03293047472834587,
0.5904565453529358,
0.014024906791746616,
0.030914032831788063,
-3.8787896633148193,
0.33552995324134827,
-0.2701438069343567,
-0.2629384994506836,
0.207549393177032... | 554 | 22.16 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005039 | [S1] artificially increased the amount of images that we have available in the documents. [S2] Yeah. [S1] You actually, you look, I think we have 25 million unique images, um, | 9.8 | 3.040377 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005039.mp3 | [
6810,
9377,
5933,
7998,
2972,
3268,
5534,
5852,
8682,
8696,
948,
10086,
4234,
7081,
7648,
355,
8359,
2771,
8543,
10117,
4803,
6756,
10142,
11414,
7206,
3946,
4025,
283,
12103,
7558,
8664,
3400,
6209,
3878,
6454,
11343,
10965,
6590,
5886,
... | [
0.6738044023513794,
-5.998168468475342,
-0.16287444531917572,
0.05985132232308388,
0.8430330753326416,
0.0023553387727588415,
0.5505478978157043,
-0.04453326016664505,
-0.1308782547712326,
-3.8652780055999756,
0.166666641831398,
-0.19580289721488953,
-0.23967844247817993,
0.132769733667373... | 245 | 9.8 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005041 | [S1] This is true, but there are pros and cons to this. So encoder, decoder-only architectures, uh, they're really good for fine-tuning, but they're not so good for prompting, is at least what we noticed. Um, and also training them is a little bit more non-trivial. So decoder-only models are quite nice 'cause you get p... | 27.68 | 3.225371 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005041.mp3 | [
5672,
5145,
3202,
8117,
6115,
8144,
10668,
4050,
5512,
12241,
6632,
2964,
8025,
10955,
8353,
8054,
4198,
7880,
11198,
10525,
5589,
1109,
10602,
3570,
8094,
10421,
9494,
3456,
3294,
6166,
12579,
7152,
2877,
3389,
8742,
4363,
3553,
3657,
5968... | [
0.367492139339447,
-5.734774589538574,
-0.02196502313017845,
-0.20548945665359497,
0.989840030670166,
-0.09433220326900482,
0.5426666140556335,
0.05928289145231247,
-0.020474951714277267,
-3.862398147583008,
0.28618985414505005,
-0.283374547958374,
-0.2068447470664978,
0.22192054986953735,... | 692 | 27.68 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005042 | [S1] Or like Roberto do. It's all around that 15%. [S2] Mm-hmm. [S1] Um, so most of the times you have to go through the data multiple times. Um, for some reason they don't prompt super well. Um, | 10.32 | 2.892461 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005042.mp3 | [
4714,
8216,
3765,
2973,
5213,
2990,
7590,
9818,
6379,
174,
2899,
2838,
3108,
5482,
5423,
3020,
771,
3277,
4398,
4162,
3328,
9486,
4429,
149,
6823,
5825,
400,
8439,
7868,
9045,
3356,
2949,
12222,
12470,
8845,
4030,
7478,
6120,
6584,
5023,
... | [
0.5419936180114746,
-6.029123306274414,
-0.10970994085073471,
0.003070640144869685,
0.8190358877182007,
-0.06652595102787018,
0.6530691981315613,
0.013989835977554321,
-0.03959383815526962,
-3.8151767253875732,
0.06065863370895386,
-0.1808166354894638,
-0.21509192883968353,
0.1686427146196... | 258 | 10.32 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005043 | [S1] And the kind of the other big thing is if you want to do score-based prompting, it's kinda, it's kinda hard to do with the encoder-decoder-only architecture, right? Like if you wanna ask what's the log probability of this sequence- [S2] Yeah. [S1] ... with the masked language model, it's kinda tough to do, right? ... | 29.88 | 2.947927 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005043.mp3 | [
2608,
8225,
11117,
12196,
7693,
5540,
6756,
12194,
3571,
11698,
8986,
11942,
8082,
6950,
10869,
8053,
11932,
8148,
8532,
2580,
8480,
8670,
6615,
11931,
7273,
2897,
12708,
11860,
12061,
12759,
2896,
4419,
12645,
873,
10073,
8471,
2676,
8502,
... | [
0.5756643414497375,
-5.684328079223633,
-0.023851949721574783,
-0.11806756258010864,
0.9309369921684265,
-0.109146349132061,
0.583043098449707,
0.04692626744508743,
-0.09958599507808685,
-3.8869049549102783,
0.2996024489402771,
-0.296113520860672,
-0.25113645195961,
0.21343764662742615,
... | 747 | 29.88 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005044 | [S1] Uh, which is actually a really, really big thing to have, right? Um, and the other thing that we noticed is that depending on the setting, prompting versus fine-tuning, the size of the mask is really important. So for fine-tuning, localized information is really important. [S2] Mm-hmm. [S1] Um, so you wanna have a... | 29.88 | 2.951007 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005044.mp3 | [
3112,
4123,
11382,
3511,
4334,
3822,
741,
1124,
3123,
2805,
7307,
12672,
8670,
3927,
832,
6528,
7259,
11645,
5629,
9445,
12770,
10192,
2779,
9494,
9540,
6615,
3502,
2774,
3691,
347,
8150,
10924,
2707,
12108,
9477,
178,
339,
11166,
3230,
3... | [
0.4503988027572632,
-5.641013145446777,
0.015756880864501,
-0.12804768979549408,
0.9664720892906189,
-0.12438022345304489,
0.6204627156257629,
0.05036792531609535,
-0.04621398448944092,
-3.9272537231445312,
0.31301596760749817,
-0.2712109684944153,
-0.267521470785141,
0.2755862772464752,
... | 747 | 29.88 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005045 | [S1] Uh, so, you know, the majority of times, right, and we clip it to one, so if you get zero, it becomes one, right? [S2] Yeah. [S1] So majority of times, you're only gonna get a single mask, right? Over 50% of the time, you're only gonna get a single mask. Uh, and then you pick a, you, you uniformly, uh, sample a su... | 24.48 | 2.83761 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005045.mp3 | [
9425,
9317,
9340,
12534,
6702,
7219,
4640,
5136,
4645,
8821,
3959,
3830,
3894,
3894,
3830,
3830,
3614,
3110,
4699,
4178,
3873,
9096,
3035,
10487,
7982,
1238,
4707,
12548,
11973,
8020,
5495,
6367,
3584,
2989,
5659,
10477,
11566,
10945,
6603,... | [
0.6583912372589111,
-5.705014228820801,
-0.07360173761844635,
-0.08180833607912064,
0.8375846147537231,
-0.15486684441566467,
0.6048232913017273,
0.04324870929121971,
-0.03798147663474083,
-3.946392297744751,
0.3444843590259552,
-0.24288900196552277,
-0.367144912481308,
0.13587257266044617... | 612 | 24.48 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005046 | [S1] Uh, of sequences super, super easily. Um, so we're kind of going all in on this objective. And so we have some follow-up work looking at, um, causal masked, uh, scaling loss for text. [S2] Yeah. [S1] So this is some ongoing work that we have now. Um, so we're pushing heavily on this. Um, so the general argument th... | 26.12 | 2.891961 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005046.mp3 | [
3112,
6672,
9463,
2998,
814,
3886,
3950,
3950,
3958,
3894,
3893,
3749,
3108,
3091,
9778,
7208,
8365,
6439,
3057,
10217,
8680,
8144,
11221,
5062,
11937,
3754,
5422,
9230,
3929,
3472,
797,
3806,
336,
9048,
9048,
9168,
8640,
10581,
6277,
880... | [
0.5831874012947083,
-5.628437519073486,
0.0006664996035397053,
-0.11938658356666565,
0.8361992835998535,
-0.11809232085943222,
0.6856356263160706,
0.0494314469397068,
-0.04052959382534027,
-3.87125563621521,
0.34149158000946045,
-0.29919612407684326,
-0.2709292769432068,
0.2259798496961593... | 653 | 26.12 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005047 | [S1] Yeah, I mean, I was, I was, it is intuitively a good trade-off. So I think here you make the case if, if I interpret this correctly, that this word nationalist right here is really important to fill in this mask. And, and if it, if it were just sort of left to right, it would be very difficult to fill this in. Yet... | 27.28 | 3.383163 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005047.mp3 | [
9740,
3617,
6869,
9092,
10666,
11132,
1901,
9323,
11767,
4886,
12085,
6246,
11684,
12163,
6668,
4112,
6171,
12630,
4705,
8757,
8702,
11629,
2733,
10484,
10613,
3395,
5456,
2953,
3392,
11289,
9753,
9249,
12159,
7102,
7782,
5294,
10547,
11502,
... | [
1.3475185632705688,
-5.303053379058838,
-0.08278574794530869,
-0.057306885719299316,
0.32496553659439087,
-0.1512376368045807,
0.2973371148109436,
-0.00833735242486,
0.16124358773231506,
-3.85955548286438,
0.45666518807411194,
0.10522022098302841,
0.15831851959228516,
-0.03193601593375206,... | 682 | 27.28 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005048 | [S1] XLNet or so. [S2] Mm-hmm. [S1] They were saying, "Well, we just train on all possible paths," right, "of decoding." Like all possible sequence of masking out tokens. And it was, it was never really satisfying because I always thought, "Well, there is something to left, to right." However, sometimes, as you say, it... | 26.88 | 3.414309 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005048.mp3 | [
6184,
6704,
8741,
12270,
12774,
9897,
9259,
11740,
8690,
6105,
5577,
7267,
12790,
11246,
3567,
3383,
2782,
3404,
8717,
6638,
7151,
7159,
7449,
11360,
11776,
1902,
3326,
386,
4499,
9201,
8600,
468,
4023,
253,
45,
2566,
1173,
8873,
7270,
73... | [
1.3980648517608643,
-5.334873676300049,
-0.14898784458637238,
-0.019883234053850174,
0.47033676505088806,
-0.002542436821386218,
0.24097613990306854,
-0.008950452320277691,
0.2345191240310669,
-3.7959377765655518,
0.3486000895500183,
0.1445368081331253,
0.16480445861816406,
-0.063200131058... | 672 | 26.88 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005049 | [S1] Yeah, like specifically in this example, right? Like in the zero-shot prompting case, right? Like let's say we want to tag nationalist with some entity link, right? Um, if it appears beforehand in the sequence, there's no way to prompt the language model to generate like an entity link before the entity appears, r... | 17.64 | 2.986214 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005049.mp3 | [
7268,
8309,
12588,
12078,
7660,
6746,
12693,
10705,
10665,
7934,
5469,
9718,
7442,
9882,
12525,
6548,
8089,
3913,
10928,
8182,
6634,
8136,
8155,
8988,
5616,
8484,
11817,
3737,
7963,
12490,
12134,
6157,
3043,
10677,
3049,
8064,
926,
9820,
43... | [
0.44087308645248413,
-5.9464287757873535,
-0.12498395889997482,
-0.04570232704281807,
0.940559983253479,
-0.046013250946998596,
0.544094979763031,
0.028351375833153725,
-0.061456892639398575,
-3.830165386199951,
0.20383496582508087,
-0.2808398902416229,
-0.2794525623321533,
0.1697081178426... | 441 | 17.64 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005050 | [S1] So that was kind of another reason that we had because like I said, like HTML data is very localized, right? Like in Wikipedia, this, this A tag which represents the entity link always appears before the entity. [S2] Yeah. [S1] Either, uh, we have the option of, you know, training two models, right? One left to ri... | 17.76 | 2.997638 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005050.mp3 | [
7413,
3248,
6425,
8665,
3034,
11062,
10477,
7983,
2783,
3223,
3163,
3163,
9249,
8369,
4161,
12,
3104,
6144,
8685,
11771,
7796,
7988,
6042,
1425,
12138,
5564,
12083,
8403,
12527,
2849,
4829,
12589,
8205,
8404,
8117,
10556,
6411,
8630,
3340,
... | [
0.5568743348121643,
-5.871746063232422,
-0.1662430614233017,
-0.08632943779230118,
0.8830399513244629,
-0.045867934823036194,
0.5201193690299988,
-0.00924619846045971,
-0.18953171372413635,
-3.851850748062134,
0.2141604721546173,
-0.2607576549053192,
-0.2729645073413849,
0.1916399300098419... | 444 | 17.76 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005051 | [S1] Um, um, or you can kind of do this kind of clever rotation of the document. [S2] Mm-hmm. [S1] Um, you said. Uh, yeah, the XLNet approach is definitely interesting, uh, which is, you know, having different permutations of the source document. But, uh, like you said, I think there's a lot of inductive biased, um, fo... | 22.92 | 3.012765 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005051.mp3 | [
6184,
7768,
9319,
9662,
5502,
6519,
4406,
4398,
4398,
3749,
8790,
6157,
9505,
9378,
9838,
7212,
7204,
9766,
4702,
9843,
4688,
6176,
3625,
3624,
4619,
6903,
3959,
4270,
3829,
822,
3829,
3685,
3100,
3099,
4626,
7193,
10778,
8318,
7925,
8098... | [
0.7017578482627869,
-5.690310001373291,
-0.11321047693490982,
-0.09269697964191437,
0.8659124970436096,
-0.08294415473937988,
0.6176546216011047,
0.03933802619576454,
0.07287397235631943,
-3.8938138484954834,
0.25063207745552063,
-0.20939697325229645,
-0.24789029359817505,
0.09522468596696... | 573 | 22.92 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005052 | [S1] Is there, just, just for my understanding, is there a reason behind these arrows? Like why do the arrows, like, are like double arrows, and there's a line, and there's like a double arrow again? Like, is, is, does that have a specific meaning? And here the arrows are only here? [S2] Yeah, so arrows pretty much, uh... | 23.84 | 3.322117 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005052.mp3 | [
5665,
5752,
8746,
12189,
6915,
5473,
642,
2770,
12203,
11673,
11698,
8027,
8075,
5590,
4901,
12003,
8746,
8361,
11826,
6193,
11340,
7344,
7386,
5728,
5737,
3082,
4652,
2078,
2657,
6353,
8346,
3170,
3617,
8780,
12482,
7169,
4108,
10113,
1071... | [
1.139747977256775,
-5.48531436920166,
-0.07617650181055069,
-0.056376535445451736,
0.5105111002922058,
-0.17371025681495667,
0.3731830418109894,
-0.05364483222365379,
0.08014220744371414,
-3.895420789718628,
0.2821034789085388,
-0.06860675662755966,
-0.021288588643074036,
0.061314865946769... | 596 | 23.84 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005053 | [S1] ... go like this. Okay, I see, I see. [S2] Yeah. [S1] 'Cause I was, I was like, okay, is, is there some meaning? But yes, there is. And this shows that in the mask language model object, if you only actually generate very small number of tokens, and you, you wouldn't even get like a loss for the other tokens. | 17.16 | 3.334201 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005053.mp3 | [
9131,
10549,
8429,
7854,
10398,
3758,
1187,
6178,
12411,
10920,
8240,
8248,
7217,
7707,
3221,
5469,
11774,
7436,
11482,
1755,
8916,
3931,
11181,
7007,
4462,
4957,
9201,
3048,
2897,
2889,
3336,
3904,
4633,
9454,
10605,
3228,
11547,
4594,
972... | [
1.5161118507385254,
-5.391561985015869,
-0.2617987394332886,
-0.02983153611421585,
0.3638658821582794,
-0.22876127064228058,
0.2940629720687866,
-0.051951996982097626,
0.19411225616931915,
-3.871201276779175,
0.39528387784957886,
0.08201995491981506,
0.12567339837551117,
-0.124344803392887... | 429 | 17.16 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005054 | [S1] Uh, yeah. So there's a couple of ways to approach this. So the, the very first thing is that modeling, and I think we mentioned this quickly in the paper, but modeling image tokens versus text tokens, it's quite different actually. So for like text usually follows, like textual tokens follow like a Zipfian distrib... | 25.6 | 3.142817 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005054.mp3 | [
3104,
4641,
8949,
6455,
6382,
8438,
7990,
7983,
6391,
6236,
5724,
6699,
7323,
9891,
12442,
8426,
11931,
9882,
9897,
5674,
6153,
6185,
5153,
9298,
6192,
6868,
5680,
7388,
5216,
3617,
3609,
2793,
5641,
7372,
12610,
11212,
11683,
11125,
3957,
... | [
0.5301077961921692,
-5.709949016571045,
-0.14261822402477264,
-0.07400034368038177,
0.9070348739624023,
-0.08009347319602966,
0.5856414437294006,
-0.013717480935156345,
-0.04881018027663231,
-3.858583450317383,
0.30508291721343994,
-0.3479403555393219,
-0.3015248775482178,
0.19020749628543... | 640 | 25.6 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005055 | [S1] ... you know, images would be being optimized for, but text kind of stayed flat. So we don't really have explanations for why this is happening. Um, I think there needs to be future, like, scaling laws looking at multimodal sequence modeling. [S2] Mm-hmm. [S1] And when I say multimodal, I'm not just talking about,... | 23.24 | 3.348522 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005055.mp3 | [
6185,
5193,
1180,
6805,
12547,
7303,
8541,
3830,
3165,
6754,
9825,
8361,
7904,
9248,
5169,
9755,
11741,
11602,
7714,
10467,
11740,
8770,
10177,
8663,
6559,
1353,
3912,
1809,
3180,
5293,
8991,
8200,
6240,
7074,
10054,
4357,
5665,
4068,
12678... | [
0.6812966465950012,
-5.687626838684082,
-0.14964663982391357,
-0.008826239965856075,
0.8995286822319031,
-0.06994454562664032,
0.5125292539596558,
0.03266770765185356,
-0.16639961302280426,
-3.8617911338806152,
0.2819024324417114,
-0.24768505990505219,
-0.21201950311660767,
0.1805310547351... | 581 | 23.24 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005056 | [S1] Uh, yeah, it seems appropriate that an image could be expressed in something like a sequence of tokens. It's- [S2] Mm-hmm. [S1] It's just a bit... I'm not too big of a fan of how this is currently done because the tokens, they also... Like, they al- already, they seem to be a bit localized in the image and so on. ... | 29.88 | 3.412359 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005056.mp3 | [
3688,
2083,
11887,
9663,
957,
4470,
4470,
3958,
3958,
3518,
6006,
3164,
3169,
8801,
11827,
8353,
3624,
9749,
7365,
7299,
9857,
9929,
9606,
11243,
11689,
11637,
6645,
4772,
12637,
7435,
10256,
3554,
3024,
12118,
12674,
11908,
7708,
10909,
34... | [
1.5987828969955444,
-5.187533855438232,
-0.066188283264637,
-0.027952272444963455,
0.43298572301864624,
-0.19841349124908447,
0.2675727903842926,
0.0006297763902693987,
0.3052566647529602,
-3.872225284576416,
0.5745472311973572,
0.07498157769441605,
0.14353755116462708,
-0.1158081516623497... | 747 | 29.88 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005057 | [S1] Um, it took a little bit of hand-holding to work, especially the, the 13 billion parameter model, took a little bit of hand-holding to work. So a lot of the times the pathologies we see is, are things like gradient underflow or overflow. [S2] Mm-hmm. [S1] Uh, gradient explosions happen, although they're more, they... | 29.88 | 3.120978 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005057.mp3 | [
11508,
10676,
8043,
11051,
10540,
10611,
11484,
8294,
8798,
12571,
12572,
9948,
12508,
12508,
12500,
12435,
9736,
10889,
5738,
544,
3097,
12436,
11628,
10004,
8392,
10704,
6122,
10540,
11875,
6562,
10546,
7780,
2716,
11125,
10987,
10461,
3091,
... | [
0.6296629905700684,
-5.5373101234436035,
-0.06742647290229797,
-0.0707489475607872,
0.954764187335968,
-0.10964055359363556,
0.5392454862594604,
0.09172938019037247,
-0.020251629874110222,
-3.8815107345581055,
0.3165019154548645,
-0.35131698846817017,
-0.20431803166866302,
0.17886973917484... | 747 | 29.88 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005058 | [S1] So the surprising thing is it kind of just worked out of the box. Apart from having to tune, I think we tuned, tuned like learning rate, um, we had to tune weight t-k- [S2] Mm-hmm. [S1] Uh, and batch size. Apart from tuning those things, it just worked almost straight out of the box. Um, and what you said is actua... | 22.12 | 3.085381 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005058.mp3 | [
6176,
2337,
8481,
8648,
3036,
11061,
7982,
7918,
7902,
5271,
5254,
3220,
8866,
12441,
9881,
11891,
7265,
5666,
5289,
3609,
3624,
5145,
8082,
10686,
6121,
8144,
8999,
10538,
4081,
5612,
10681,
9985,
3538,
11986,
9245,
5768,
11755,
11457,
117... | [
0.7639814019203186,
-5.65359354019165,
-0.13365598022937775,
-0.08160878717899323,
0.8842290639877319,
-0.10396628826856613,
0.5884767770767212,
0.07535403966903687,
-0.045985300093889236,
-3.886378526687622,
0.2744047939777374,
-0.41015079617500305,
-0.21050845086574554,
0.083352133631706... | 553 | 22.12 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005059 | [S1] The good news is once, once CM3 is released, we're going to release the, the checkpoint that we used for this model. Um, uh, I think the model that we have now is continuing training, so we'll release that one too. So, uh, people will be able to play around with both. [S2] Excellent. [S1] Uh, but one thing I'd lik... | 22.12 | 3.108213 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005059.mp3 | [
5672,
5153,
8073,
8622,
9307,
9627,
10659,
11475,
10891,
12099,
11078,
6096,
5587,
10659,
5956,
3016,
3923,
2731,
5365,
10677,
11564,
10756,
3482,
8153,
8154,
1939,
11885,
3692,
2675,
8053,
11500,
11459,
8169,
6104,
6040,
5568,
12172,
11651,
... | [
0.7214276790618896,
-5.7504143714904785,
-0.15831363201141357,
-0.011624723672866821,
0.7879557013511658,
-0.10316849499940872,
0.6476016044616699,
0.07379555702209473,
-0.0665346160531044,
-3.854581594467163,
0.3270426392555237,
-0.3116005063056946,
-0.21867817640304565,
0.176747635006904... | 553 | 22.12 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005060 | seems to be that scale plays a slightly larger role, um, in multi-modal than it does in text. [S1] Mm-hmm. [S2] Um, so I, I think the qua- the quantitative thing that we saw is that if you looked at the data efficiency jumps between like, uh, I'm forgetting the exact numbers, but like, like let's make them up, like the... | 26.24 | 2.9587 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005060.mp3 | [
5680,
6675,
3561,
10720,
5520,
4048,
12612,
12483,
7773,
3622,
8426,
5089,
8080,
5504,
8216,
5520,
6617,
8966,
5155,
8800,
6621,
12164,
12549,
12677,
7237,
9,
2980,
6079,
4716,
7872,
5617,
5600,
8152,
5008,
11866,
5019,
11668,
11740,
7804,
... | [
0.7911475896835327,
-5.646492004394531,
-0.14265969395637512,
-0.12898625433444977,
0.8690744042396545,
-0.14197003841400146,
0.5763978958129883,
0.05096158757805824,
-0.13841423392295837,
-3.8719751834869385,
0.35708436369895935,
-0.31276199221611023,
-0.2483028918504715,
0.11180908977985... | 656 | 26.24 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005061 | [S1] And the data efficiency there, well, let's say it was like the larger model was five times more efficient in terms of data. So, uh, in order to reach the same perplexity, it would need five times less data. Uh, using these same exact models, we saw that in the multimodal case, it was 10x. So there was all, almost ... | 29.36 | 3.141938 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005061.mp3 | [
2600,
5152,
10981,
11758,
9861,
8781,
12749,
4822,
3161,
9771,
5787,
6874,
4635,
524,
18,
5312,
10741,
9430,
8256,
8667,
11217,
11665,
8465,
8545,
11108,
3562,
6640,
8629,
11605,
12697,
4562,
9302,
8971,
6040,
5003,
7558,
3075,
2963,
11767,... | [
0.747488260269165,
-5.61313009262085,
-0.0506150908768177,
-0.0882011130452156,
0.8534215688705444,
-0.14760227501392365,
0.5148462653160095,
0.0481768399477005,
-0.1430002599954605,
-3.898984432220459,
0.36811134219169617,
-0.3566192388534546,
-0.27860477566719055,
0.14393950998783112,
... | 734 | 29.36 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005062 | [S1] It's just really, really tedious. So one of the main things is, um, you know, whenever you have a ton of nodes that you're running, there's infrastructure issues that pop up, right? [S2] Yeah. [S1] So like if one GPU goes down, right, then, then all of training is paused, right? So infrastructure issues are kind o... | 22.48 | 2.972791 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005062.mp3 | [
6184,
4646,
10069,
11728,
10184,
12160,
8155,
8484,
6040,
3904,
3652,
619,
347,
10653,
10412,
2837,
12118,
6340,
171,
2900,
10982,
2644,
3486,
7556,
11856,
8080,
12256,
4891,
10054,
4357,
3852,
4292,
4997,
1391,
1379,
6304,
2888,
4242,
6705... | [
0.6559805274009705,
-5.723817348480225,
-0.16170595586299896,
-0.049537088721990585,
0.8207542896270752,
-0.042625848203897476,
0.5271332263946533,
0.021287858486175537,
-0.1011388748884201,
-3.851170301437378,
0.281141996383667,
-0.3661023676395416,
-0.2043776959180832,
0.1429226100444793... | 562 | 22.48 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005063 | [S1] Yeah, because of the computers, one model. So it, it really comes down to intuition. Um, so both Mike Lewis and Namung Goyo, who are on the paper, have trained these really, really big models before. Um, so they had a ton of great intuition about how to get things to work, um, in terms of these very large models. ... | 21.36 | 3.064457 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005063.mp3 | [
9911,
6825,
4108,
12613,
11211,
11252,
8183,
1462,
6437,
4379,
9377,
9317,
5558,
8702,
2324,
2595,
5416,
11724,
9932,
12641,
6643,
10611,
2944,
5586,
11428,
2769,
11694,
6242,
12634,
4587,
11295,
8866,
7592,
9156,
8015,
6404,
3502,
6358,
34... | [
0.7409189939498901,
-5.765048503875732,
-0.18782426416873932,
-0.0693642646074295,
0.829537034034729,
-0.09320662915706635,
0.6646050214767456,
-0.0038552256301045418,
-0.013655648566782475,
-3.8209526538848877,
0.2910911738872528,
-0.23902243375778198,
-0.16891388595104218,
0.132843479514... | 534 | 21.36 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005064 | [S1] Yeah. Um, yeah, that's a great question. So I, I think at the beginning of the project the, the push was really to have a single model, uh, that can do any image task in the zero shot case. [S2] Mm-hmm. [S1] Um, and so kind of the story that we built around it is, can we describe all the tasks that we're intereste... | 20.24 | 3.050452 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005064.mp3 | [
10840,
9936,
7381,
12101,
11175,
6007,
6438,
6370,
9370,
12580,
12596,
8429,
12565,
9972,
12509,
9910,
12014,
12398,
4694,
9842,
6769,
8729,
6454,
10621,
10549,
10613,
11054,
11062,
11054,
8438,
8798,
5661,
3109,
6746,
9313,
12384,
11818,
623... | [
0.5921516418457031,
-5.811474323272705,
-0.15524977445602417,
-0.029538581147789955,
0.8115456104278564,
-0.07711499184370041,
0.7324839234352112,
-0.0057883006520569324,
-0.009338914416730404,
-3.8098347187042236,
0.20814239978790283,
-0.2833087742328644,
-0.20876343548297882,
0.216498211... | 506 | 20.24 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005066 | [S1] So some of them didn't work. Some of them only worked at scale, and we can kind of go, go through this. Um, specifically, like one thing is that, like, the captioning only worked at scale. [S2] Okay. [S1] So the 13 billion model was the only model that could caption well. [S2] And the captioning, you go mainly wit... | 19.56 | 3.024697 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005066.mp3 | [
7274,
5744,
4754,
6032,
5595,
7995,
10470,
3221,
3307,
7666,
8152,
5603,
8121,
11370,
7851,
8364,
5656,
3019,
11185,
8732,
6677,
5184,
8660,
11164,
8259,
11117,
8726,
8807,
2733,
5309,
5541,
5572,
11036,
9313,
12450,
4458,
2355,
11625,
8144... | [
0.9237847924232483,
-5.627383232116699,
-0.18362149596214294,
-0.04513449966907501,
0.6892415881156921,
-0.08378496766090393,
0.602705180644989,
-0.023922458291053772,
-0.018701517954468727,
-3.8874757289886475,
0.28114771842956543,
-0.050910621881484985,
-0.18081773817539215,
0.0927539393... | 489 | 19.56 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005067 | [S1] No. But like the figure that you're on now, I think is kind of interesting. So we can kind of get unconditional image generation by just asking the model to generate a sequence of tokens after the image tag. [S2] Yeah. [S1] So we saw one interesting behavior is that the model, for some reason, almost always wanted... | 29.52 | 3.1289 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005067.mp3 | [
8280,
6875,
9354,
9939,
6421,
6454,
5999,
3183,
6747,
9372,
10022,
12532,
12654,
10020,
7201,
10904,
2870,
5339,
10749,
10004,
9322,
8082,
9638,
8680,
3569,
10664,
9482,
8994,
8091,
6210,
8100,
11707,
4364,
9548,
5469,
4253,
8446,
10870,
86... | [
0.4780593812465668,
-5.656517028808594,
0.007854422554373741,
-0.11560782790184021,
0.8826935887336731,
-0.11308593302965164,
0.5160194039344788,
0.037720683962106705,
-0.09160512685775757,
-3.8814265727996826,
0.2963036298751831,
-0.29134756326675415,
-0.27484840154647827,
0.2774980366230... | 738 | 29.52 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005068 | [S1] When you say it wanted to, it, that's just what it did. [S2] Yeah. [S1] Like when you, when you sampled, did you, like, I mean, this, when you say it wanted to, it could also be that in the internet, humans most of the time write alt first and then the source. [S2] Yeah, so we actually looked into this. So, uh, | 18.56 | 3.31566 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005068.mp3 | [
9454,
4406,
9445,
3893,
6453,
3308,
11252,
9365,
11604,
9099,
6428,
6049,
2952,
12212,
11689,
12676,
4821,
12701,
4429,
3624,
3620,
2669,
2667,
2667,
2862,
3319,
7870,
8500,
11412,
10614,
11940,
10848,
6705,
5377,
7080,
11629,
10395,
5254,
... | [
1.2690702676773071,
-5.63243293762207,
-0.09413961321115494,
-0.15101736783981323,
0.49613821506500244,
-0.1835859715938568,
0.38476502895355225,
-0.04621761292219162,
0.23730815947055817,
-3.7856569290161133,
0.2802961766719818,
-0.004244206007570028,
0.044994745403528214,
0.0129019860178... | 464 | 18.56 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005069 | [S1] A lot of text does have alt, um, but it's around like, I want to say like 70% to 80% mark, if I recall correctly. So it wouldn't explain why the model almost always wants to generate alt text. Now the, the theory that we kind of have is that without alt text, you have much higher perplexities for images. [S2] Mm-h... | 29.52 | 3.290174 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005069.mp3 | [
2600,
11801,
11126,
8116,
5213,
3290,
5404,
7989,
10545,
12010,
5450,
11131,
3298,
2345,
8384,
11168,
7156,
11770,
12124,
11355,
913,
1354,
524,
5377,
11191,
8053,
4577,
5584,
3936,
9532,
11261,
9015,
2914,
4329,
9257,
8759,
2870,
3382,
786... | [
0.5922963619232178,
-5.648042678833008,
-0.027443496510386467,
-0.17250250279903412,
0.8413140177726746,
-0.12771651148796082,
0.5736762285232544,
0.06957423686981201,
-0.1475413739681244,
-3.882505416870117,
0.3568275570869446,
-0.31710201501846313,
-0.2605026066303253,
0.2639823853969574... | 738 | 29.52 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005070 | [S1] Yeah, that's true. There, there isn't, for VQ-V again, there isn't something explicit, but the, I think the way that the layers are constructed, you do still get some implicit dependencies across the- [S2] Yeah. [S1] ... tokens. And so I think this is what the Transformer's kind of pulling apart here. [S2] Yeah. [... | 25.84 | 2.928408 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005070.mp3 | [
5672,
9360,
9995,
10178,
10713,
10664,
11252,
2998,
6498,
10217,
6962,
9387,
9378,
3658,
7210,
1026,
3586,
8181,
11645,
6938,
465,
4553,
7632,
12753,
3043,
2901,
404,
798,
3790,
3097,
2962,
11189,
10724,
5581,
4308,
9887,
5184,
6125,
8684,
... | [
0.6663565635681152,
-5.6926655769348145,
-0.13549385964870453,
-0.09141099452972412,
0.8959706425666809,
-0.10389921069145203,
0.5755966901779175,
0.024735908955335617,
-0.0937647596001625,
-3.8897695541381836,
0.26194116473197937,
-0.26350513100624084,
-0.2429581731557846,
0.1384037584066... | 646 | 25.84 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005071 | [S1] Um, so being able to generate trivially, um, being able to compute log probabilities, I think tokens are probably the easiest way to go. [S2] Yeah. [S1] Uh, and one thing is you can naturally increase the resolution of tokens images just by increasing how many tokens you use per image. [S2] Mm-hmm. [S1] So in some... | 19.84 | 2.818317 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005071.mp3 | [
6697,
5680,
4099,
6903,
3375,
8558,
3446,
3894,
4341,
3766,
6758,
3109,
1571,
3609,
4106,
3945,
10688,
5548,
10911,
8225,
6628,
11715,
12171,
9293,
12629,
12773,
6236,
5366,
3155,
8072,
5620,
8790,
8328,
12673,
10734,
12718,
10316,
11623,
3... | [
0.8314253091812134,
-5.754639625549316,
-0.15594181418418884,
-0.07305227220058441,
0.8860535025596619,
-0.11599288135766983,
0.38688671588897705,
0.010230038315057755,
-0.08689290285110474,
-3.8577325344085693,
0.3705455958843231,
-0.3403589129447937,
-0.2932743728160858,
0.11051873117685... | 496 | 19.84 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005072 | [S1] Yeah, I mean, yeah, down to probably, probably you could at some point get more, more tokens than pixels. I wouldn't know what, what that would mean, but, um, I guess the resolution isn't even limited by the resolution of the image itself. [S2] Yeah. [S1] Uh, so there's, there's this, this interesting thing you ca... | 27.56 | 3.370621 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005072.mp3 | [
2592,
4163,
10060,
11653,
6647,
1958,
7470,
4029,
6182,
12125,
12613,
7692,
5649,
3610,
3104,
2052,
12677,
11239,
4542,
1397,
4909,
9449,
9811,
12410,
6176,
3632,
9216,
3566,
7101,
8958,
8366,
11860,
8288,
6536,
7643,
5895,
1164,
8297,
7922... | [
1.4333769083023071,
-5.3358845710754395,
-0.10644093155860901,
-0.04977979511022568,
0.38016873598098755,
-0.2321617752313614,
0.37844616174697876,
-0.06387689709663391,
0.22387456893920898,
-3.812920093536377,
0.4950195550918579,
0.12312746047973633,
0.1636083722114563,
-0.093692615628242... | 689 | 27.56 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005073 | [S1] Yeah, so actually because of our objective, 'cause we sampled the number of masks, right? [S2] Yeah. [S1] Um, you can actually mask out like five, six, seven masks. [S2] Yeah. [S1] And it should still work. Um, I don't think there was any specific reason that we stuck to masking out a single thing. [S2] Mm-hmm. [S... | 18.08 | 2.813773 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005073.mp3 | [
12597,
8438,
7595,
6819,
11667,
10648,
11573,
1444,
6616,
5595,
7910,
9966,
10747,
9525,
3507,
12761,
2851,
11161,
7237,
35,
10656,
11726,
7314,
12641,
4084,
10546,
5898,
448,
10612,
8859,
874,
5879,
10997,
8053,
5525,
5518,
3029,
3348,
937... | [
0.8239614367485046,
-5.689973831176758,
-0.1650405079126358,
0.009456761181354523,
0.8786248564720154,
-0.06041525676846504,
0.646399199962616,
0.015166476368904114,
0.03946761414408684,
-3.865266799926758,
0.2029205560684204,
-0.2596071660518646,
-0.19275911152362823,
0.10252983868122101,... | 452 | 18.08 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005074 | [S1] Um, then you also give the ground truth, which is here on the right, and then there's one model that does infilling unconditional, so just looking at the image, and then there's one model that does it conditionally, and the conditional is, uh, conditioned with this thing right here as the, the alt text. So the und... | 22.64 | 3.350773 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005074.mp3 | [
3746,
4176,
4645,
7343,
4470,
1526,
1389,
1973,
950,
1469,
4398,
3100,
558,
5670,
4643,
8737,
8376,
6176,
10376,
7280,
7254,
2132,
3336,
12727,
12790,
11756,
11763,
11859,
10892,
10830,
10582,
7462,
9313,
11826,
8368,
7189,
7236,
12483,
583... | [
1.2791224718093872,
-5.4186320304870605,
-0.14843668043613434,
0.0346260741353035,
0.37835562229156494,
-0.20579296350479126,
0.3913874924182892,
-0.05760342255234718,
0.30010661482810974,
-3.7998158931732178,
0.4830188453197479,
0.02013026364147663,
0.10787133127450943,
-0.085736215114593... | 566 | 22.64 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005075 | [S1] I, I'm not, I'm not sure the unconditionality has something much to do with it because there is no, this doesn't look like natural. You know, you know what I mean, a little bit? Like- [S2] Yeah, a little bit. [S1] This, this, this shouldn't be like, just because it's not conditioned on it. If it's not conditioned ... | 28.84 | 3.223518 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005075.mp3 | [
9387,
5681,
10862,
8695,
7573,
6752,
11819,
11262,
7335,
7789,
8716,
8050,
10417,
4260,
7200,
9768,
11126,
9654,
7717,
8916,
10492,
10994,
7273,
11010,
12698,
12754,
7632,
8047,
10612,
5459,
5451,
2883,
5517,
5510,
3022,
6356,
7849,
11361,
... | [
1.3141064643859863,
-5.299694061279297,
-0.08993315696716309,
-0.04245516285300255,
0.454461932182312,
-0.16674794256687164,
0.3258766531944275,
0.00840977393090725,
0.2363186478614807,
-3.8238980770111084,
0.3899213671684265,
0.16861768066883087,
0.15273086726665497,
0.010320127941668034,... | 721 | 28.84 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005076 | [S1] Yeah. So, so one theory that we kind of have here, um, is that the, the model needs to understand the continue, continuation of the, the horizontal lines, right? [S2] Mm-hmm. [S1] And that re- that requires some semantic understanding that this is, for example, uh, a bench, right? And actually, if you look at the,... | 23.88 | 2.909264 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005076.mp3 | [
6184,
5217,
4819,
9989,
3447,
3830,
5794,
12393,
6305,
12389,
12619,
10220,
8437,
12589,
12716,
12661,
12526,
6753,
7074,
11224,
5611,
7978,
7909,
5270,
3234,
6177,
6311,
6641,
11208,
2988,
10413,
2652,
5235,
7995,
12019,
8720,
3616,
8138,
... | [
0.6113797426223755,
-5.743223190307617,
-0.12515336275100708,
-0.0442841537296772,
0.8713556528091431,
-0.06609911471605301,
0.5538712739944458,
-0.0023277997970581055,
-0.05974908173084259,
-3.8113696575164795,
0.3483738303184509,
-0.3771849274635315,
-0.29915162920951843,
0.1723812073469... | 597 | 23.88 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005077 | [S1] the top of the bench. So I think the model has a tough time understanding the, the high level semantic content of the image, which is fixed by feeding in text. [S2] Yeah. [S1] Uh, now I think of course if you have, I think if you have a larger model that's trained for longer with a higher resolution, uh, this prob... | 26.8 | 3.185842 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005077.mp3 | [
3609,
6810,
4133,
3778,
4512,
8958,
2942,
6692,
5865,
108,
3609,
5457,
8807,
3112,
10602,
11638,
12139,
9300,
8913,
12681,
7449,
9321,
12013,
9397,
9331,
9453,
7853,
9902,
9830,
8285,
9820,
7338,
5649,
4009,
6096,
3501,
10542,
7918,
5847,
... | [
0.6393207311630249,
-5.696508884429932,
-0.08787187933921814,
-0.09511761367321014,
0.8811278939247131,
-0.116745226085186,
0.5645191073417664,
0.022349920123815536,
-0.10996787250041962,
-3.838660955429077,
0.26181545853614807,
-0.3358478248119354,
-0.2591032087802887,
0.23158352077007294... | 670 | 26.8 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005078 | [S1] Just if you change the tokens even a little bit, the, the blurring aspect happens very, very quickly with VQ-VAE again. Compared to, for example, the VQ-VAE from DALI, um, which requires more tokens, so 1024 tokens versus the 256 we use here. Um, but, uh, it's more direct in some sense. [S2] Yeah. [S1] Um, so, yea... | 26 | 3.097037 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005078.mp3 | [
9384,
3170,
11649,
10707,
10731,
8673,
5512,
4754,
12701,
3418,
9704,
9093,
8454,
11465,
12746,
9177,
11762,
12114,
8907,
10192,
6273,
8174,
6676,
8392,
10152,
9150,
7973,
3149,
12449,
9054,
8215,
6411,
2896,
6691,
12685,
12044,
5418,
11931,
... | [
0.762718915939331,
-5.6942548751831055,
-0.126667782664299,
-0.06870976090431213,
0.8650137186050415,
-0.15679915249347687,
0.4691796898841858,
-0.009896313771605492,
-0.11425244063138962,
-3.876173973083496,
0.3323832154273987,
-0.32683059573173523,
-0.2753857970237732,
0.1433420330286026... | 650 | 26 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005079 | [S1] The, in the case that it doesn't generate a car at all, it just generates mountains, right? Just because of what landscapes are easier to generate. Um, the other thing that we saw kind of tough compared to Dali is, you know, the data that we used only came from Wikipedia or Common Crawl News. So none of it was fic... | 28.08 | 2.929623 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005079.mp3 | [
6177,
5744,
3589,
531,
7808,
3446,
3375,
3887,
1254,
1766,
3886,
1254,
9321,
6688,
12134,
11277,
8332,
7159,
11937,
12768,
7660,
12254,
7629,
3408,
5997,
4838,
3486,
3083,
7872,
10726,
8118,
6485,
3520,
8926,
8198,
12681,
9173,
8694,
7325,
... | [
0.6055261492729187,
-5.681286811828613,
-0.014551146887242794,
-0.11382073163986206,
0.8887355327606201,
-0.09788358956575394,
0.4546700716018677,
0.04003608226776123,
-0.14931969344615936,
-3.897603750228882,
0.24035589396953583,
-0.3351925015449524,
-0.23812147974967957,
0.20077732205390... | 702 | 28.08 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005080 | [S1] Um, like really fantasy based prompts. [S2] Yeah. [S1] Uh, so that's kind of one downside. And actually this is one criticism I have, uh, of the evaluation that we did for the FID matrix, which is a way to measure, uh, you know, the, the quality of images, which is, uh, we actually took the table from Glide, um, f... | 24.64 | 3.159062 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005080.mp3 | [
4129,
1555,
6838,
4030,
3446,
1405,
4861,
3613,
5669,
102,
358,
4821,
9825,
4322,
164,
3469,
11045,
2598,
350,
7430,
7557,
1234,
9834,
2650,
8680,
9136,
6584,
12766,
10222,
9902,
8913,
3024,
7493,
2378,
833,
12614,
7493,
3105,
10665,
1009... | [
0.5437418818473816,
-5.791359901428223,
-0.1590699404478073,
-0.07180386036634445,
0.8716330528259277,
-0.03036629967391491,
0.4862011969089508,
0.02295839414000511,
-0.0342797227203846,
-3.8253071308135986,
0.32316917181015015,
-0.4600207507610321,
-0.21504844725131989,
0.2057060450315475... | 616 | 24.64 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005081 | [S1] almost all non-fiction. [S2] Yeah. [S1] Uh, like non-fantasy images. So this is really sh- like, it's under-representing Dali. So I think if you casted a wider net here and had something that included a wider array, uh, a, a bigger distribution of images, I think Dali's results here would be, uh, much, much strong... | 21.52 | 3.071398 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005081.mp3 | [
5680,
5680,
9261,
7933,
10843,
10421,
8438,
6560,
6464,
1187,
11837,
5366,
7990,
8438,
5878,
8375,
5149,
5269,
203,
3722,
6753,
6739,
8263,
7685,
8718,
11742,
8447,
5885,
7278,
11916,
4528,
9201,
6640,
4079,
2326,
12451,
5025,
12744,
5016,
... | [
0.8619099855422974,
-5.739057540893555,
-0.18915438652038574,
-0.0886913612484932,
0.8959051370620728,
-0.08354726433753967,
0.482243150472641,
0.06283774226903915,
-0.04598671570420265,
-3.872945785522461,
0.27835020422935486,
-0.3874365985393524,
-0.23730885982513428,
0.13125380873680115... | 538 | 21.52 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005082 | [S1] Mm-hmm. [S2] Yeah. [S1] You, you did, you did discuss a little bit. You also said you ha- you, you saw subsampled, uh, web data in, and, and you cited some concerns as well. Um, but there is also quality issue with sort of the, the wider you cast the net. | 19.16 | 3.33269 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005082.mp3 | [
8795,
8854,
9427,
11478,
2068,
7186,
7849,
5650,
3749,
6445,
6380,
3829,
5868,
9451,
57,
5216,
9832,
8754,
7200,
5664,
7192,
7912,
8744,
7249,
6160,
9770,
6232,
7201,
7193,
7866,
7256,
6176,
5288,
4641,
5225,
12675,
12653,
6892,
7324,
700... | [
1.354225754737854,
-5.529987335205078,
-0.24342480301856995,
-0.028116751462221146,
0.3844878375530243,
-0.20657660067081451,
0.40890225768089294,
-0.04576203227043152,
0.30996057391166687,
-3.832313060760498,
0.3780074715614319,
-0.01071259193122387,
0.1866087168455124,
-0.072264254093170... | 479 | 19.16 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005083 | [S1] I think at the beginning we had some ethical concerns of like, like I said, we have very weak alignment, so you can prompt with anything, right? [S2] Yeah. [S1] We had some ethical concerns about images that you can generate if you were just trained on all of Common Chrome. [S2] Mm-hmm. [S1] Um, so we tried to thi... | 26 | 2.836699 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005083.mp3 | [
8714,
9266,
11260,
12627,
5704,
5601,
11601,
11857,
9059,
11644,
3104,
8081,
11565,
2602,
8682,
9478,
11993,
11677,
11683,
8339,
12051,
11285,
628,
3869,
10002,
7466,
11700,
11565,
8256,
3434,
5520,
5806,
8231,
6748,
9251,
12278,
10743,
9372,... | [
0.776684045791626,
-5.657994270324707,
-0.11586405336856842,
-0.07203765213489532,
0.841436505317688,
-0.08237715065479279,
0.490559458732605,
0.05345109850168228,
0.018902411684393883,
-3.8975963592529297,
0.36310824751853943,
-0.3720209002494812,
-0.3535950183868408,
0.10398752987384796,... | 650 | 26 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005085 | [S1] ... from Convo. So you could probably do something like this here. [S2] Yeah. [S1] Um, I, I questioned the efficacy just because very large models, they only need to see a data point a couple times in order to pick it up. Um, so, uh, I think there's like some very fundamental engineering work that, that's being do... | 29.28 | 2.864815 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005085.mp3 | [
3748,
9314,
6734,
548,
3797,
1858,
4009,
164,
5661,
7340,
9514,
3829,
3765,
3742,
3686,
8802,
8939,
562,
6307,
8868,
6228,
8926,
12540,
12077,
12589,
11509,
12644,
10083,
9435,
6616,
3357,
9026,
8517,
12129,
5486,
8426,
3050,
8041,
7866,
... | [
0.7790442705154419,
-5.628880023956299,
-0.059953946620225906,
-0.10613711178302765,
0.7764776945114136,
-0.11056230962276459,
0.5010915398597717,
0.028427235782146454,
-0.008226552978157997,
-3.926746368408203,
0.2952783405780792,
-0.28693583607673645,
-0.3286925256252289,
0.1271604150533... | 732 | 29.28 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005086 | [S1] Sort of the, let's say diversity makes, is, is probably the best, so you can always choose which one you wanna, you wanna use. I don't know. I'm sorry, this is just a rant by now. [S2] [LAUGHS] [S1] Um, you, you do have some- [S2] Yeah. [S1] Sorry, go ahead. | 14.56 | 3.371556 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005086.mp3 | [
4716,
6712,
4362,
3408,
8502,
8043,
3397,
5516,
11062,
5404,
3449,
4842,
5312,
7078,
12037,
9477,
10054,
9670,
6812,
12451,
12377,
9779,
5849,
9753,
7728,
9776,
3112,
10859,
10121,
9956,
5804,
7272,
4713,
7754,
1243,
557,
660,
5999,
12719,
... | [
1.5207031965255737,
-5.4955010414123535,
-0.155094712972641,
0.0015473091043531895,
0.5042995810508728,
-0.3227607309818268,
0.33524778485298157,
-0.0525294654071331,
0.3158743977546692,
-3.867234706878662,
0.2412707358598709,
0.09879428148269653,
0.08441774547100067,
-0.1803397834300995,
... | 364 | 14.56 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005087 | [S1] I was going to say, uh, with respect to what you're saying, there's, the solution doesn't necessarily have to lie on the language model side. [S2] Yeah. [S1] Um, so one thing is you can think of language modeling as just pure density estimation over tokens, right? So if you're doing that, like, of course you're go... | 26.56 | 2.814624 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005087.mp3 | [
9249,
11940,
11764,
6859,
4307,
10029,
7605,
7404,
9469,
3893,
9845,
8125,
5813,
5801,
9891,
11517,
11259,
5356,
11574,
3464,
6363,
11428,
11477,
5601,
3024,
11758,
5062,
9315,
9562,
9835,
7998,
6134,
3365,
2726,
3227,
3171,
2661,
603,
3107... | [
0.5404999256134033,
-5.731481075286865,
-0.04953303560614586,
-0.21874170005321503,
0.982938289642334,
-0.03795509785413742,
0.5497425198554993,
0.045219603925943375,
-0.11003222316503525,
-3.8567380905151367,
0.2426614910364151,
-0.3487958312034607,
-0.21398967504501343,
0.219850838184356... | 664 | 26.56 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005088 | [S1] Uh, select in some sense the mode of the slice of the density. [S2] Yeah. [S1] Right? Um, and so something probably similarly can be done here. Like a great example is like, take Codex, for example, right? I think in the Codex paper, what they do is they generate a ton of samples and then they re-rank the samples | 19.52 | 3.07124 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005088.mp3 | [
3617,
10841,
2278,
8623,
3439,
6510,
3895,
3886,
3950,
3886,
3886,
3813,
3747,
7329,
12377,
12377,
9753,
5288,
5169,
4178,
6505,
8160,
11224,
3035,
8375,
2658,
5478,
10748,
12573,
9377,
3857,
11883,
7454,
10821,
6129,
5537,
5384,
5604,
1055... | [
0.5509515404701233,
-5.8523993492126465,
-0.09048181772232056,
-0.1299087107181549,
0.8809938430786133,
-0.0544285923242569,
0.6361330151557922,
0.040567632764577866,
-0.12839213013648987,
-3.878086805343628,
0.3014317452907562,
-0.2674492299556732,
-0.22910857200622559,
0.2076898068189621... | 488 | 19.52 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005089 | [S1] Uh, in terms of perplexity, so average log probability. And then they take the mode, so essentially the exact mode of that, uh, density estimation, right? [S2] Yeah. [S1] So one thing to argue is that, you know, you could, you could train language models that do pure density estimation over all the text that we ha... | 17.56 | 2.871121 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005089.mp3 | [
6697,
5144,
4191,
6838,
10678,
5494,
6007,
2991,
3439,
6510,
3942,
6817,
9321,
8812,
10071,
10254,
10888,
11216,
6642,
8045,
8533,
8230,
6430,
3009,
333,
8559,
5487,
8358,
91,
2865,
4257,
6705,
10730,
3562,
3733,
10601,
7088,
2788,
8126,
... | [
0.552497148513794,
-5.87379789352417,
-0.10516513139009476,
-0.07830794155597687,
0.9094569683074951,
-0.061423324048519135,
0.5350652933120728,
0.0032347189262509346,
-0.1094798818230629,
-3.838300943374634,
0.2663555145263672,
-0.3072481155395508,
-0.1974966824054718,
0.13748612999916077... | 439 | 17.56 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005091 | [S1] Entity linking, right? So in some sense, the genre objective was a subset of our much more general objective. [S2] Yeah. [S1] Right? Uh, and it's not too surprising we beat out genre just because our models are bigger, um, in the, in our fine-tuned case. But the really, really cool thing, I think, was that we can ... | 24.64 | 3.213307 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005091.mp3 | [
6184,
8225,
12021,
7150,
7260,
6993,
4049,
4642,
8528,
10010,
10117,
3823,
2595,
2708,
4901,
4693,
9762,
10137,
2325,
7242,
3612,
4122,
4625,
1124,
2844,
2948,
9007,
7503,
8265,
8857,
9309,
10038,
10485,
12085,
12470,
7340,
6297,
8144,
5422... | [
0.5653201937675476,
-5.729944229125977,
-0.07255721837282181,
-0.11711633205413818,
0.975304901599884,
-0.014536376111209393,
0.5771563649177551,
0.05875702202320099,
-0.043911244720220566,
-3.9247357845306396,
0.3319615125656128,
-0.295087605714798,
-0.3032214641571045,
0.1845033466815948... | 616 | 24.64 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005092 | [S1] You want to disambiguate this entity. You can place a mask there with this A tag, right? Uh, and then our model will fill in what it thinks the disambiguation is. [S2] Yeah. [S1] So that's kind of cool. Uh, I couldn't find any, like, zero-shot baselines like this, so I think this is kind of the first paper to do t... | 20.68 | 3.153219 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005092.mp3 | [
6184,
5168,
4314,
12124,
7901,
8061,
8740,
8781,
2752,
6110,
4066,
8144,
9179,
8734,
8490,
12238,
7233,
9041,
5838,
5486,
12758,
5720,
1363,
4918,
5520,
4766,
12219,
10772,
6024,
1491,
5712,
6993,
12677,
6866,
12652,
12661,
12525,
8301,
933... | [
0.6877406239509583,
-5.738762378692627,
-0.13726769387722015,
-0.01686912775039673,
0.777477502822876,
-0.08713610470294952,
0.5451517701148987,
0.00037001632153987885,
-0.1197715550661087,
-3.90496826171875,
0.2855823040008545,
-0.20252691209316254,
-0.2051299661397934,
0.1241445094347000... | 517 | 20.68 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005094 | [S1] Good, actually, at least from a semantic level. [S2] Mm-hmm. [S1] So one problem is that we don't actually generate in the style of, uh, I think MS Coco here. Um, so we didn't report like blue four, uh, numbers or like the standard numbers. But if you look at the semantic, uh, uh, similarity using BERT score, the,... | 26.84 | 3.252657 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005094.mp3 | [
2593,
3089,
3093,
2580,
11811,
11730,
8102,
423,
1502,
4293,
5705,
3723,
7185,
8362,
8744,
12455,
6590,
3516,
4845,
11939,
7640,
12680,
1875,
2643,
861,
2379,
4690,
12726,
9213,
5814,
5395,
12683,
12100,
2960,
4945,
6137,
8094,
10790,
8054,... | [
0.5378593802452087,
-5.66707181930542,
0.022643469274044037,
-0.18654721975326538,
0.9024410843849182,
-0.015990998595952988,
0.6323264241218567,
0.038084354251623154,
-0.03882395103573799,
-3.9159910678863525,
0.2807256877422333,
-0.33595043420791626,
-0.23141376674175262,
0.2915445566177... | 671 | 26.84 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005095 | [S1] Uh, and so you can kind of see the style here is weird. It tries to explicitly state what type of airplane it is. [S2] Yeah. [S1] Uh, uh, but that's kind of a interesting behavior. Uh, so I think definitely at scale, uh, you know, you could get a single model that I think could be competitive with MS Coco. | 18.08 | 2.973956 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005095.mp3 | [
4267,
5648,
7351,
3902,
6454,
6007,
4278,
3382,
1334,
3838,
6765,
5724,
9377,
12385,
12393,
8736,
5673,
3099,
7867,
7209,
6177,
7721,
7187,
7415,
8205,
3930,
11168,
3536,
10613,
7980,
7918,
2775,
1178,
4707,
12675,
10644,
8859,
12712,
9373,... | [
0.6619564890861511,
-5.805587291717529,
-0.1128150075674057,
-0.09805795550346375,
0.8805098533630371,
-0.046635452657938004,
0.615429162979126,
0.043862245976924896,
-0.10980731248855591,
-3.840017795562744,
0.19881325960159302,
-0.2857528626918793,
-0.22588233649730682,
0.215276852250099... | 452 | 18.08 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005096 | [S1] Um, from, you know, our, I think our 100 million parameter model to our 13 billion parameter model, around the 60 billion mark is where we'll see grounding in this setup. [S2] Okay. [S1] It's kind of a linear plot. So our expectation is that if you scale this up to 60 billion, that you should be able to achieve, I... | 25 | 2.903306 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005096.mp3 | [
10009,
6177,
9460,
6582,
3446,
5999,
6447,
3894,
3446,
3894,
3830,
4206,
5662,
6175,
3092,
3610,
3611,
4618,
6658,
3609,
8785,
5146,
3561,
8682,
9201,
4089,
5526,
5550,
6006,
6006,
8495,
6455,
6766,
8222,
8286,
4686,
9313,
5715,
9861,
121... | [
0.5353163480758667,
-5.654455184936523,
-0.04814765229821205,
-0.04542998969554901,
0.8561930060386658,
-0.13069485127925873,
0.48964881896972656,
-0.0067140161991119385,
-0.15465286374092102,
-3.931480646133423,
0.2787112593650818,
-0.272274911403656,
-0.2508125305175781,
0.18628089129924... | 625 | 25 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005097 | [S1] You can actually make these predictions, which is, which is, you know, it's, it's cool. Like, I'm amazed by this. [S2] Yeah. I definitely don't think we're going to be like an order of magnitude off, right? [S1] Yeah. [S2] So, so I think with the 100 billion parameter, 100 billion and 175 billion like 2-3 size, we... | 25.68 | 3.164018 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005097.mp3 | [
3090,
9801,
7380,
9931,
8966,
1568,
5666,
12485,
5446,
4201,
12761,
9709,
8774,
4959,
12093,
9213,
9900,
10155,
10184,
3218,
12117,
7237,
3109,
3118,
6182,
9702,
12781,
11729,
11659,
9827,
10146,
2273,
17,
467,
12566,
12677,
4947,
8073,
419... | [
0.8953903913497925,
-5.533348560333252,
-0.04351796954870224,
-0.09412535279989243,
0.7386606335639954,
-0.1772143840789795,
0.47738316655158997,
-0.02903137356042862,
0.011160250753164291,
-3.9532644748687744,
0.32654038071632385,
-0.24817147850990295,
-0.19892679154872894,
0.073106169700... | 642 | 25.68 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005098 | Um, so, yeah, it's interesting to see what the next, you know, step-wise changes in behavior will be if you scale this up. Um- [S1] With respect to the HTML, right- [S2] Mm-hmm. [S1] ... uh, that you use, which is, I- I thought it was, it was pretty cool because it is data that is, you know, so available, and your argu... | 25.04 | 3.011238 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005098.mp3 | [
7413,
3680,
4718,
8574,
5494,
6007,
3895,
6455,
8806,
6307,
9378,
9299,
7334,
9902,
4252,
4144,
3793,
8608,
4048,
8046,
8046,
2862,
5343,
5775,
3212,
6298,
11930,
7270,
4827,
9809,
6657,
3616,
2720,
1569,
3681,
4384,
7653,
11231,
10671,
8... | [
1.1455217599868774,
-5.5781660079956055,
-0.0843309536576271,
-0.13460680842399597,
0.5460092425346375,
-0.08362966775894165,
0.35195329785346985,
-0.055667776614427567,
0.062145158648490906,
-3.85332989692688,
0.4076891243457794,
-0.029649117961525917,
0.038054950535297394,
0.043059155344... | 626 | 25.04 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005099 | [S1] Yeah. So, so in some sense, do, do we want to model every single token? So in, in the case that you have infinite compute? Sure. [S2] Mm-hmm. [S1] Right. Um, but here, there's kind of a min-max problem that you have to solve, right? Which is you want to kind of, you want to maximize the amount of semantic informat... | 24.32 | 3.260757 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005099.mp3 | [
6184,
5673,
6667,
4814,
9605,
11245,
8183,
1462,
3811,
9378,
9371,
9901,
8301,
9453,
6821,
9906,
6696,
4306,
8168,
8673,
6608,
10662,
7990,
10542,
10534,
7902,
5342,
662,
3226,
8801,
9322,
8219,
4262,
6640,
8584,
3547,
10550,
7918,
3222,
... | [
0.5741584300994873,
-5.736083984375,
-0.1057298481464386,
-0.030532846227288246,
0.8990451097488403,
0.020806441083550453,
0.5845118165016174,
0.04744845628738403,
-0.18413224816322327,
-3.867203950881958,
0.2711511254310608,
-0.29120686650276184,
-0.20789115130901337,
0.19763624668121338,... | 608 | 24.32 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005101 | [S1] ... the document structure is described in the minimal set of tokens. [S2] Mm-hmm. [S1] Um, so maybe that's, you know, th- that's a pure engineering project as well. Um- | 8.8 | 3.09608 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005101.mp3 | [
6162,
4619,
3597,
5697,
8172,
12141,
5184,
5588,
7925,
8051,
9389,
8984,
4324,
10790,
11574,
2458,
7040,
3993,
2901,
8053,
5860,
10008,
10129,
5534,
8837,
3907,
1300,
6168,
3010,
5934,
8664,
4953,
9451,
405,
11254,
7527,
3164,
5852,
9311,
... | [
0.8165577054023743,
-5.8433003425598145,
-0.20970144867897034,
-0.035951171070337296,
0.6994810104370117,
-0.00976146012544632,
0.7343937158584595,
-0.05591447651386261,
-0.10539602488279343,
-3.9584991931915283,
0.03257292881608009,
-0.1876157820224762,
-0.24116572737693787,
0.09688514471... | 220 | 8.8 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005102 | [S1] When you, when you think of HTML and the DOM, it is a tree, right? [S2] Mm-hmm. [S1] Which is different from a linear sequence. Uh, do you, do you think there is, do you think there's value in treating the tree as a tree? Do you think it's mainly a limitation of the models we have? They go, let's say, like, | 22.36 | 3.392074 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005102.mp3 | [
3112,
7704,
621,
11253,
9812,
11533,
12035,
8462,
1617,
6697,
1133,
10740,
9732,
11533,
11076,
3933,
10144,
3880,
11697,
11473,
9762,
11417,
8565,
10934,
936,
2334,
12605,
11750,
11725,
7235,
11585,
10185,
4865,
11088,
10133,
8670,
7399,
8558... | [
1.365696907043457,
-5.461874008178711,
-0.18791699409484863,
-0.07458525151014328,
0.48125478625297546,
-0.11823632568120956,
0.3086365759372711,
-0.053443703800439835,
0.1630043387413025,
-3.8053536415100098,
0.4696930944919586,
0.08169235289096832,
0.06827466934919357,
-0.089421145617961... | 559 | 22.36 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005103 | [S1] Yeah. So one thing about transformers is it seems that they can learn the inductive bias of the data fairly well. [S2] Mm-hmm. [S1] And it's not necessarily encoded. Um, so my argument to this is that usually for these large scale runs, the best thing is just to keep it as simple as possible. [S2] Mm-hmm. [S1] Mos... | 26.08 | 3.02013 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005103.mp3 | [
6184,
3121,
4610,
4245,
10058,
10626,
10657,
11242,
6071,
6518,
7012,
4458,
4458,
5106,
6122,
10713,
8144,
8171,
10546,
7915,
7900,
10453,
3286,
3737,
9306,
9370,
9370,
12377,
12377,
8728,
1048,
7169,
9770,
4648,
6161,
3616,
6673,
2720,
125... | [
0.6643928289413452,
-5.644629001617432,
-0.04753477871417999,
-0.12482832372188568,
0.8305262327194214,
-0.016311343759298325,
0.4736465513706207,
0.0510065071284771,
0.008662700653076172,
-3.8618321418762207,
0.34627875685691833,
-0.39495471119880676,
-0.24365273118019104,
0.2045432478189... | 652 | 26.08 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005104 | [S1] Uh, structure. [S2] Yeah. [S1] So this isn't in the paper, but we looked at attention scores and, and then you can, you can see very clearly that the model knows what are like boundaries between HTML elements, um, for example. But I, but again, there's also a ton of work to be done as well. So like some exciting w... | 23.12 | 3.153943 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005104.mp3 | [
6177,
5744,
6162,
4406,
3886,
4342,
3830,
886,
821,
3765,
3100,
3620,
3156,
6689,
5728,
3113,
9956,
8176,
9680,
10057,
9441,
354,
5550,
2934,
9386,
9442,
9680,
10201,
1515,
5590,
3533,
6427,
9954,
9889,
8802,
4126,
5676,
3620,
4299,
9678,... | [
0.6569706201553345,
-5.619163513183594,
-0.06565150618553162,
-0.1559051126241684,
0.8027578592300415,
-0.10961243510246277,
0.511383056640625,
0.026004040613770485,
-0.15418249368667603,
-3.9316766262054443,
0.2696574330329895,
-0.3304748833179474,
-0.18825821578502655,
0.1309171915054321... | 578 | 23.12 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005105 | [S1] That work is really clever, right? 'Cause it introduces an explicit inductive bias that the further away a token is, the probably less likely that you are to look at it. And it gets rid of the need for, you know, positional, uh, representations. [S2] Yeah. [S1] So you can imagine, like, an extension of Alibi here ... | 17.64 | 3.203776 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005105.mp3 | [
6681,
3089,
1555,
8017,
11771,
8122,
5740,
2803,
8027,
5741,
11481,
4068,
3908,
1472,
2794,
8084,
10925,
5468,
5078,
11931,
12778,
3297,
8043,
10672,
9370,
2995,
2918,
275,
282,
3495,
11765,
12107,
9376,
12577,
6078,
6404,
897,
9127,
7363,
... | [
0.6354101896286011,
-5.778890132904053,
-0.16059240698814392,
-0.06948000192642212,
0.8439001441001892,
-0.01453481800854206,
0.47808194160461426,
-0.045139823108911514,
-0.07747700810432434,
-3.8773815631866455,
0.366366982460022,
-0.24217908084392548,
-0.22961652278900146,
0.146486580371... | 441 | 17.64 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005107 | [S1] So, this is all stuff that needs to be done in the future as well. Uh, but that being said, I think if you have enough compute, these models can learn anything. It mostly becomes an efficiency angle- [S2] Yeah. [S1] ... uh, going forward. [S2] So, a- about this paper, so what, what I have a bit of a trouble with i... | 28.6 | 2.888895 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005107.mp3 | [
2600,
5665,
4891,
6552,
3536,
6007,
5423,
5342,
3214,
3732,
6810,
12572,
9444,
4706,
3601,
6340,
1090,
2972,
8556,
5592,
3026,
10461,
2944,
3532,
6755,
10805,
5438,
7926,
10998,
5879,
5879,
2606,
5286,
2838,
723,
6298,
9754,
3346,
8168,
8... | [
1.065171241760254,
-5.512720584869385,
-0.11135951429605484,
-0.015655873343348503,
0.5396607518196106,
-0.07767954468727112,
0.3935372233390808,
-0.08045855164527893,
0.039674706757068634,
-3.854276657104492,
0.25036972761154175,
-0.04074300825595856,
0.02725587598979473,
0.10765967518091... | 715 | 28.6 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005109 | [S1] I think both Google and Codex have similar sizes. [S2] Mm-hmm. [S1] While being able to have a bidirectional, uh, bidirectional option. [S2] Yeah. [S1] So, um, there are a couple teams within Facebook that are trying out this objective with some success. Um, so, uh, there will be future work about this. [S2] Excel... | 19.76 | 3.01203 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005109.mp3 | [
6217,
4130,
7415,
3073,
7060,
12366,
9765,
2634,
19,
3097,
6193,
10528,
2871,
7982,
9796,
1298,
8289,
9323,
6555,
7917,
10397,
3214,
11931,
3359,
10412,
10342,
7710,
7774,
7782,
5326,
147,
7267,
7406,
10094,
12702,
12711,
10159,
7534,
8789,... | [
0.8227458000183105,
-5.72928524017334,
-0.17533573508262634,
-0.06289814412593842,
0.7253779768943787,
0.015758901834487915,
0.6482430100440979,
-0.024681303650140762,
-0.12339013069868088,
-3.8842244148254395,
0.3485275208950043,
-0.3002658486366272,
-0.16704711318016052,
0.06641860306262... | 494 | 19.76 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005110 | [S1] ... Speech tokens, right? [S2] Mm-hmm. [S1] Um, very simple. I think they use K-means. I might be wrong, though. Uh, uh, just to find discrete tokens for speech. So imagine how you have a single model that has video, images, you know, text. [S2] Yeah. [S1] Uh, right. Speech, everything kind of put into one, right?... | 25.68 | 3.154738 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005110.mp3 | [
8088,
4249,
9195,
11650,
6915,
11585,
12747,
4808,
11600,
4596,
2862,
661,
12458,
5020,
4198,
6157,
4301,
1984,
4241,
620,
389,
11630,
12687,
6800,
3212,
8907,
6492,
3733,
1163,
9762,
5730,
7180,
8371,
4676,
1045,
1172,
6760,
5665,
8745,
... | [
0.953518271446228,
-5.608548164367676,
-0.13336127996444702,
0.036005470901727676,
0.7650227546691895,
-0.13604699075222015,
0.4859637916088104,
0.009496111422777176,
-0.03198367357254028,
-3.9043350219726562,
0.3956448435783386,
-0.3507716953754425,
-0.27314308285713196,
0.075426630675792... | 642 | 25.68 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005112 | [S1] Is it just 'cause we don't have, like, bidirectional, like, masks? Is that one? Is it because we only mask for, like, causal models in upper triangular matrix? Is there something more fundamental there? I think kind of peeling that apart and figuring out what's going on there is kind of important too. Um, but I th... | 28.68 | 3.068643 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005112.mp3 | [
5672,
8804,
10006,
5504,
10716,
12565,
11520,
11713,
8549,
5584,
9379,
9194,
3340,
898,
5363,
12676,
8907,
7997,
8350,
7474,
12213,
11134,
3172,
3361,
2653,
3446,
7581,
9931,
11875,
5083,
11873,
9818,
5345,
4641,
3617,
7233,
5352,
4128,
102... | [
0.7815670967102051,
-5.566966533660889,
-0.017502833157777786,
-0.05448583886027336,
0.7984830141067505,
-0.052847735583782196,
0.5760988593101501,
0.021310575306415558,
-0.04271530732512474,
-3.8599984645843506,
0.34944602847099304,
-0.3337513208389282,
-0.2743844985961914,
0.211369454860... | 717 | 28.68 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005113 | [S1] ... So working on. [S2] It seems like it, yeah. Um, is there anything else about the paper or the research direction you want to shout out? You want people to know that we haven't mentioned so far? [S1] Yeah, I mean, we'll be releasing all this code really, really soon. Um, we're just waiting on some internal appr... | 28.32 | 3.331201 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005113.mp3 | [
3172,
5080,
6592,
6391,
1189,
44,
3125,
3421,
3413,
11874,
6754,
515,
6158,
6357,
3814,
3830,
9390,
3748,
9386,
9825,
6680,
6152,
4620,
2652,
7190,
6678,
9772,
7190,
4173,
2659,
6801,
8284,
4890,
5201,
336,
1812,
12612,
8853,
8957,
3914,
... | [
1.1760562658309937,
-5.422362804412842,
-0.038247838616371155,
-0.07319270074367523,
0.5774392485618591,
-0.20787201821804047,
0.4421811103820801,
-0.016926752403378487,
0.193706214427948,
-3.9227957725524902,
0.3589385151863098,
-0.14553046226501465,
-0.15620839595794678,
0.05418905988335... | 708 | 28.32 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005114 | [S1] Nice. What does it, what does it take to, like, just to forward propagate? What's, like, the minimal configuration? [S2] Um, so with the recent Deep Speed stuff that was released for inference, I'm not really sure, 'cause I think they said that you can use one GPU for, like, a 6.7 billion model. [S1] Yeah. [S2] So... | 22.24 | 3.295453 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005114.mp3 | [
9504,
3661,
8405,
3958,
9140,
4397,
10134,
10139,
3425,
3812,
3418,
3282,
5713,
6305,
8296,
3104,
3803,
6216,
12645,
12603,
8957,
4852,
9434,
6884,
174,
10740,
9825,
8128,
10546,
2826,
2962,
12180,
7169,
3291,
3830,
6956,
3308,
10612,
9874,... | [
0.9448638558387756,
-5.707254886627197,
-0.07434029877185822,
-0.1463700234889984,
0.5685898661613464,
-0.11460467427968979,
0.45863375067710876,
0.0007776094134896994,
0.07413630187511444,
-3.9117443561553955,
0.37723323702812195,
-0.1061607375741005,
-0.18074746429920197,
0.1126175224781... | 556 | 22.24 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005115 | [S1] But if, like, without that, just give us a ballpark, you know, what, what, what would it, what would it be like forward propping through this model? [S2] Yeah. So, so one thing is you could do it on a CPU if you have a strong enough CPU. [S1] Yeah. [S2] Uh, but, but for, for inference, I think what I used was four... | 22.4 | 3.256148 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005115.mp3 | [
4715,
1562,
10474,
10623,
1252,
4642,
6430,
9436,
12629,
8053,
5941,
5996,
4342,
9436,
5980,
10747,
11757,
9810,
10211,
3292,
1122,
3676,
3291,
3163,
5217,
3156,
2858,
3688,
2578,
6213,
6691,
3604,
5214,
8212,
6185,
6155,
5150,
6162,
3097,
... | [
1.0848783254623413,
-5.597113132476807,
-0.11961248517036438,
-0.13196709752082825,
0.5782998204231262,
-0.11550962924957275,
0.43514004349708557,
-0.022369232028722763,
0.005797355901449919,
-3.8531646728515625,
0.25610801577568054,
-0.08360420912504196,
-0.06403204798698425,
0.1092138960... | 560 | 22.4 |
bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005116 | [S1] Cool. Excellent. Well, Armin, thank you so much for being here. This was really cool. Um, really value the, like, also the kind of behind the scenes in insights we got here. And- [S2] Yeah. [S1] ... I hope to see you again very soon with even, like, CM4. [S2] [LAUGHS] Yeah, thank you for having me. [S1] Excellent. | 20.44 | 3.392068 | 24,000 | audio/en/bilibili_data_1897938584_BV1pP4y1k7Jr_p169_BV1pP4y1k7Jr_p169_m4-dialogue_1005116.mp3 | [
5665,
12699,
2931,
7833,
10392,
2653,
3746,
12385,
11361,
9306,
9778,
9992,
4208,
12315,
11765,
11619,
11370,
9707,
8664,
5568,
10484,
5341,
886,
6174,
9738,
8272,
4050,
4124,
1132,
3310,
12150,
8375,
3230,
6038,
4971,
9385,
9313,
7849,
720... | [
1.504079818725586,
-5.347790718078613,
-0.18520720303058624,
-0.07411409169435501,
0.444141685962677,
-0.19969849288463593,
0.3341980576515198,
-0.04926002770662308,
0.21607713401317596,
-3.836867570877075,
0.41316235065460205,
0.03981507197022438,
-0.013373060151934624,
-0.069020733237266... | 511 | 20.44 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722462 | [S1] Exactly. I'm not surprised that he did something like that here, but, uh, we're gonna have to see if Shrek can actually handle himself in a macro game here. Again, Solki really forced to react entirely to what his opponent's doing. Can we point out how high up we are, Artosis? [S2] People have no idea. [S1] The fa... | 27.16 | 2.953597 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722462.mp3 | [
6067,
7654,
4893,
7004,
1491,
3923,
11767,
11245,
10741,
9565,
8686,
5558,
5934,
3878,
7007,
4503,
1942,
1439,
1318,
7285,
9077,
12125,
2723,
9397,
9205,
5220,
11467,
10674,
8058,
7202,
5457,
8145,
11638,
2601,
10482,
356,
5999,
9014,
7540,... | [
0.5941882729530334,
-5.453787326812744,
0.12069356441497803,
-0.2630542814731598,
0.5408663749694824,
-0.2603806257247925,
0.46188199520111084,
-0.0822848230600357,
-0.1668166220188141,
-3.9626123905181885,
0.20871387422084808,
0.11274435371160507,
0.17228825390338898,
0.09633652120828629,... | 679 | 27.16 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722465 | [S1] Which is actually why it was a really smart way to open up considering the map we had and the fact that this is a best of seven and Starcraft one, that can take quite a while. I think it was an intelligent move by him because it still masks what else, uh, Sharp has in store. Silky has not been able to get a feel f... | 28.76 | 2.949562 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722465.mp3 | [
4829,
1374,
7987,
11980,
12160,
10715,
9044,
6617,
8684,
10672,
11184,
9891,
7076,
9664,
9092,
8900,
116,
11133,
9082,
12713,
11733,
9494,
2741,
11174,
6592,
5607,
748,
8148,
7835,
8092,
10117,
7136,
8664,
9153,
6173,
10614,
11124,
8166,
16... | [
0.46745485067367554,
-5.442844390869141,
0.1556340754032135,
-0.26001909375190735,
0.5414820313453674,
-0.2123761922121048,
0.3392694890499115,
-0.0836121216416359,
-0.16435129940509796,
-3.9752118587493896,
0.16322310268878937,
0.10900667309761047,
0.2316664159297943,
0.12418297678232193,... | 719 | 28.76 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722466 | [S1] Uh, we'll see. Loser picks, by the way. [S2] Yes, yes. Maybe something like Aztec coming up. You can't really do that eBay block. Uh, but, you know, there's, there's a huge variety of strategies that each side could do here. [S1] Right. [S2] So, who prepared well for these maps? It's such a big deal in a best of f... | 20.8 | 2.932808 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722466.mp3 | [
7397,
4023,
2934,
5550,
4014,
8695,
3951,
3878,
3236,
10365,
10525,
6113,
3886,
4560,
11117,
11667,
7559,
1302,
8866,
9884,
8820,
7268,
7267,
8795,
1109,
5468,
10523,
11030,
1929,
4060,
8108,
644,
8362,
4531,
12196,
11108,
2031,
4705,
11493... | [
0.16766847670078278,
-5.6549248695373535,
0.10288908332586288,
-0.11935625970363617,
0.5179528594017029,
-0.3122020363807678,
0.32915693521499634,
-0.090701162815094,
-0.15493568778038025,
-3.9524030685424805,
0.06266447901725769,
-0.011625941842794418,
0.31024324893951416,
0.0898034721612... | 520 | 20.8 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722471 | [S1] There are so many people in these- [S2] It's so cool to have remastered back here. Um, and just see how much it's thriving in Korea. It is killing it out here. [S1] Absolutely. I mean- [S2] It's, it's so cool to just see that we still see new strategies and even a little bit of the old, like we saw in game one, th... | 20.04 | 2.838413 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722471.mp3 | [
4618,
3869,
3374,
325,
5042,
9121,
8146,
8145,
8144,
5056,
11622,
10998,
7982,
4118,
11061,
7301,
5005,
10053,
3674,
6560,
10062,
10182,
3795,
3825,
3175,
797,
7197,
3730,
7373,
4230,
4522,
6105,
4040,
10610,
7907,
7315,
3929,
2978,
10354,
... | [
0.43051913380622864,
-5.649112224578857,
0.08414746075868607,
-0.033544644713401794,
0.5063705444335938,
-0.1529918760061264,
0.3416031002998352,
-0.07531318813562393,
-0.24474568665027618,
-3.8151588439941406,
0.16342106461524963,
0.25938481092453003,
0.2658153772354126,
0.131821408867836... | 501 | 20.04 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722474 | [S1] Now, Command Center first is a little bit risky. For those of you guys that are watching at home and saying, "When I try Command Center first, I always lose." Yeah, you're supposed to. Um, but, you know, when you're prepping in a series like this, if you can just squeeze out a CC right away, and the Zerg was going... | 28.76 | 2.929536 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722474.mp3 | [
12441,
3780,
9702,
11237,
11252,
10551,
5238,
1626,
4154,
8731,
11939,
9315,
12193,
7022,
7771,
12770,
11528,
10608,
9754,
2897,
5568,
11691,
6404,
8559,
2950,
3420,
3697,
3376,
5998,
10605,
8086,
5445,
8143,
4445,
9176,
4994,
8777,
3984,
6... | [
0.19820035994052887,
-5.45549201965332,
0.21553944051265717,
-0.15972748398780823,
0.5753318071365356,
-0.1898409128189087,
0.16758407652378082,
-0.06713113933801651,
-0.2863144874572754,
-3.944300889968872,
0.13026216626167297,
0.22821152210235596,
0.29735448956489563,
0.1401519477367401,... | 719 | 28.76 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722476 | [S1] Well, let me, let me tell you why I think that this is a brilliant one for Sharp on this map. [S2] Please. [S1] He wins map one. So he counter chooses. It's not like you're choosing a map, you're like, "On this map, I go quick spawning pool," and that's how I'm gonna beat him. [S2] Right. [S1] No, not likely, righ... | 16.4 | 3.074605 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722476.mp3 | [
6875,
4331,
8431,
5467,
11635,
8155,
5357,
12250,
7492,
5720,
9370,
8275,
9501,
12005,
6234,
2709,
6309,
8347,
9498,
9435,
9834,
652,
8179,
12443,
7915,
12041,
9745,
9112,
4532,
11197,
7971,
9929,
11660,
94,
5438,
6014,
10149,
6615,
1469,
... | [
0.29677432775497437,
-5.68985652923584,
0.13686327636241913,
-0.05572324991226196,
0.38309207558631897,
-0.17338259518146515,
0.28693363070487976,
-0.13010142743587494,
-0.1377047300338745,
-3.927976131439209,
0.13049346208572388,
0.19311757385730743,
0.24244502186775208,
0.165587618947029... | 410 | 16.4 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722478 | [S1] Uh, we're gonna have to see, uh, what Sharp has when theselings get down here, where he's positioned. Marines are quite brittle against Zerglings, unless they're positioned well. And then if you have enough of them and they're wedged in these odd spots where thelings can't get surrounded- [S2] Yeah. [S1] ... then ... | 26.8 | 2.846885 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722478.mp3 | [
4202,
27,
6902,
3447,
3684,
2876,
780,
9507,
10675,
12013,
1121,
4010,
3863,
7088,
9120,
8640,
3549,
11604,
11604,
10062,
12614,
9094,
8518,
6550,
4438,
9828,
9894,
6958,
3438,
1390,
3886,
5918,
1117,
2665,
5674,
1122,
5301,
9150,
2073,
2... | [
0.31265589594841003,
-5.522141933441162,
0.18179406225681305,
-0.2281891256570816,
0.5545830130577087,
-0.20067112147808075,
0.3107374906539917,
-0.06167791411280632,
-0.2225964516401291,
-3.9811551570892334,
0.18975414335727692,
0.06864188611507416,
0.18112486600875854,
0.1333931833505630... | 670 | 26.8 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722493 | [S1] And that allows that hatchery to, to not be destroyed. Had that not gotten up, there's a very, very high chance that, uh, in a base race, Zerg just loses there. [S2] Mm-hmm. [S1] Stim Marines are gonna kill stuff way faster than mutas and lurkers trying to destroy all those terrain buildings. [S2] Terran buildings... | 16.32 | 3.032176 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722493.mp3 | [
9363,
9769,
12591,
12278,
12647,
10829,
6214,
337,
12719,
12790,
11248,
12154,
7946,
11068,
8237,
2652,
9143,
12734,
12725,
11699,
10348,
11038,
3971,
3392,
3218,
6134,
12270,
6647,
11711,
7542,
3201,
3178,
6194,
3294,
4905,
7604,
12780,
1163... | [
0.5325060486793518,
-5.620867729187012,
-0.056848347187042236,
-0.10364147275686264,
0.6071645617485046,
-0.18628527224063873,
0.41577446460723877,
-0.07268019020557404,
0.025012699887156487,
-4.070735454559326,
0.270595908164978,
0.09797345846891403,
0.22850088775157928,
-0.05340101569890... | 408 | 16.32 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722494 | [S1] Yeah, yeah. So, yeah, that could have gone the other way. So again, kudos to Sharp. He was super decisive and that is awesome to see. But I think this should have put the fear of God into him because- [S2] Yeah. [S1] ... Solki is looking so strong. Like, yeah, Sharp won game one. He dictated that whole game. I thi... | 21.52 | 2.872863 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722494.mp3 | [
8925,
1959,
7135,
3967,
3381,
7407,
6890,
12444,
8796,
12645,
6899,
6308,
9508,
9323,
1685,
7518,
1967,
3318,
6947,
7387,
9771,
12500,
11883,
7242,
3369,
2953,
3556,
11181,
10605,
5423,
5783,
1237,
5738,
5668,
9366,
9573,
8364,
6756,
8348,
... | [
0.4372956454753876,
-5.547576427459717,
0.09135223925113678,
-0.1842593252658844,
0.43234890699386597,
-0.2821720242500305,
0.3368433117866516,
-0.09941856563091278,
-0.33102652430534363,
-3.991565465927124,
0.14925990998744965,
0.2589882016181946,
0.32938483357429504,
0.05165363475680351,... | 538 | 21.52 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722513 | [S1] and behind he has a sunken barracks production. [S2] Yeah, it's very important. [S1] It, it is incredibly important because Solki, first off, he scouted this, so he went three hatch. You go two hatch against tack or weirdness, uh, against this type of play, you actually want three hatch. But now he's moving out wi... | 22.24 | 2.87915 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722513.mp3 | [
9755,
9840,
12086,
9190,
12357,
6731,
2994,
6071,
3829,
10033,
4413,
12094,
11573,
12212,
12683,
8708,
10133,
12677,
4444,
12791,
11767,
6348,
384,
2870,
3356,
8688,
6536,
6133,
12726,
9756,
10154,
4967,
5698,
5874,
5039,
7605,
397,
1619,
9... | [
-0.01364364754408598,
-5.649439811706543,
0.17958153784275055,
-0.029706675559282303,
0.5230919718742371,
-0.2313353270292282,
0.15650026500225067,
-0.11997874826192856,
-0.43895015120506287,
-3.860229730606079,
-0.04097861796617508,
0.28363746404647827,
0.39489492774009705,
0.165256470441... | 556 | 22.24 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722533 | [S1] ... right, that Solki would have a three-one lead and somehow Sharp would manage to win every single game, uh, from there on out. So I don't know, I don't know what Sharp's supposed to do here. We don't know what the map, um, that's been chosen is just yet, but this is gonna be possibly the most important game in ... | 28.64 | 2.94064 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722533.mp3 | [
12043,
4043,
6039,
8260,
823,
3022,
4015,
6583,
12775,
12164,
11221,
4747,
6354,
6747,
859,
10095,
4519,
4447,
6272,
9362,
9428,
9428,
9364,
8867,
3924,
6624,
3027,
10365,
7893,
12121,
10139,
10957,
2870,
9475,
10172,
11253,
3236,
2998,
675... | [
0.42883116006851196,
-5.451328277587891,
0.18202634155750275,
-0.2629266679286957,
0.48988401889801025,
-0.2814437448978424,
0.40542861819267273,
-0.02546270377933979,
-0.2117350846529007,
-4.001719951629639,
0.16921034455299377,
0.08857928216457367,
0.06577567011117935,
0.1085555553436279... | 716 | 28.64 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722534 | [S1] ... but three-one. [S2] ... best of three from there on out, so. [S1] Yeah. So, Sharp needs his ultimate play here, but the thing is, he already chose Fighting Spirit, right? That was his first map choice. [S2] Interesting that he picked a super basic map. Sometimes you see the players pick one of the, you know, t... | 21.88 | 3.00259 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722534.mp3 | [
3498,
4316,
1902,
6555,
10620,
8986,
5933,
5507,
12173,
6405,
5557,
11253,
9006,
8117,
12004,
11092,
8935,
811,
8336,
3004,
3236,
2949,
12174,
7623,
3809,
6074,
726,
8230,
3356,
5038,
3493,
918,
3837,
4342,
8917,
5039,
4982,
4406,
702,
38... | [
0.626207172870636,
-5.509791374206543,
0.0011054184287786484,
-0.14979933202266693,
0.5761592388153076,
-0.11459394544363022,
0.31920045614242554,
-0.11288071423768997,
-0.06914866715669632,
-3.967236280441284,
0.15465882420539856,
0.24096998572349548,
0.1994689255952835,
0.026113169267773... | 547 | 21.88 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722535 | [S1] I think that there's a good chance. I mean, if you go into something like Circuit Breakers and you're starting to... The thing is, it's not like Benzene... I don't feel like it's necessarily super sharp favorite or something, but... [S2] Right. [S1] At least he can have a very good plan there, right? Whereas somet... | 27.96 | 2.845337 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722535.mp3 | [
9882,
4478,
2517,
4884,
5865,
1458,
12702,
11868,
10010,
6506,
11043,
8272,
6131,
11042,
5451,
6428,
3520,
10917,
8738,
6557,
10532,
3096,
10179,
10195,
2409,
9972,
11545,
11344,
8857,
11491,
9527,
7205,
3425,
6088,
860,
7606,
3102,
12629,
... | [
0.4133116900920868,
-5.4956536293029785,
0.13187244534492493,
-0.23592133820056915,
0.5461583733558655,
-0.21820595860481262,
0.2903214395046234,
-0.12060992419719696,
-0.21443785727024078,
-4.035321235656738,
0.1410248577594757,
0.21666894853115082,
0.2375868260860443,
0.07351825386285782... | 699 | 27.96 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722541 | [S1] No, this one is. [S2] This is actually- [S1] We already explained it. [S2] The distance is gonna be so big between the two of them. I mean, you, two, two. I think Sharp still has a shot. Three, one. I mean, against Solki, who is the epitome of stable Zerg. | 16.16 | 2.884682 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722541.mp3 | [
11932,
10893,
12783,
8042,
10545,
10548,
3327,
2719,
4812,
3152,
10222,
11693,
6056,
4480,
3747,
5301,
6526,
9750,
7062,
7637,
2403,
1604,
850,
6041,
6026,
1482,
5074,
6756,
9867,
1309,
7598,
3977,
337,
12198,
11542,
1345,
3943,
10223,
1124... | [
0.25772684812545776,
-5.775623798370361,
0.09686864167451859,
-0.06098697707056999,
0.4951719045639038,
-0.15747876465320587,
0.4107334017753601,
-0.10082672536373138,
-0.3901287317276001,
-3.9379732608795166,
0.053896669298410416,
0.19410285353660583,
0.15889926254749298,
0.14420609176158... | 404 | 16.16 |
bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722546 | [S1] It's kind of interesting. [S2] Yeah, let's wait. What's going on here? [S1] This is actually really weird, guys. [S2] He's not mining gas. [S1] We are floating at 176 gas, which doesn't point to anything, really. [S2] But I feel like he'd make a spire here. Like, is he just trying to get his third up real fast? [S... | 25.48 | 2.887832 | 24,000 | audio/en/bilibili_data_1904149_BV1At411v72m_p2_BV1At411v72m_p2_m4-dialogue_0722546.mp3 | [
4259,
8739,
786,
4378,
11427,
1908,
4918,
6277,
2351,
803,
5022,
533,
8833,
12776,
813,
454,
1901,
4918,
1897,
6033,
6912,
7086,
12434,
9372,
12059,
12067,
9507,
12003,
8364,
9883,
12441,
9364,
7268,
7443,
11980,
6867,
9836,
12595,
9492,
... | [
0.2794584631919861,
-5.568016529083252,
0.12603892385959625,
-0.16675631701946259,
0.6180981993675232,
-0.16077758371829987,
0.2362687587738037,
-0.13624584674835205,
-0.2504780888557434,
-3.9150431156158447,
0.09902293980121613,
0.133882075548172,
0.23658069968223572,
0.08011444658041,
... | 637 | 25.48 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042385 | [S1] I don't think we live yet. [S2] You just said the broadcast is live on my screen. [S1] Oh, well, I can't even see none of that. All right, so we live and direct. [S2] In seconds, bro. We're 10 seconds into the show. | 11.76 | 3.044344 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042385.mp3 | [
3746,
3668,
4701,
3703,
3190,
6678,
11776,
3574,
10014,
9813,
635,
6926,
6788,
2717,
8670,
11581,
11510,
6454,
9911,
218,
4308,
7111,
3375,
6966,
4314,
9883,
12499,
9875,
12500,
12571,
10011,
9500,
10012,
9948,
8931,
12507,
9946,
12443,
125... | [
1.4024889469146729,
-5.5113630294799805,
-0.33354952931404114,
0.052684199064970016,
0.4295353889465332,
-0.004049305338412523,
0.6159186959266663,
-0.0717848613858223,
0.046472493559122086,
-3.9875450134277344,
0.32131484150886536,
0.29271677136421204,
-0.205180361866951,
-0.0908473283052... | 294 | 11.76 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042386 | [S1] I like that. I like that. Right away. Right away. I pre- I appreciate that. I'm sure the fans appreciate us sitting down chopping it up, man. It's- [S2] For sure. [S1] Again, we gonna, we gonna just jump right into it, have some casual conversations, man. A lot of people wanna hear what you have to say. Um, I'm wi... | 21.48 | 3.393797 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042386.mp3 | [
9818,
11958,
8190,
6519,
5284,
8027,
8122,
8121,
12250,
11986,
7329,
5523,
11189,
3575,
4894,
2982,
11197,
8868,
7891,
10684,
11708,
7445,
7321,
997,
5102,
3755,
2909,
8109,
8123,
8898,
11173,
2652,
5365,
12279,
5093,
3924,
12051,
9947,
114... | [
1.500266432762146,
-5.258368492126465,
-0.30497562885284424,
-0.14483830332756042,
0.33637866377830505,
-0.03135235235095024,
0.6571282148361206,
-0.00803691428154707,
0.08965881913900375,
-4.065943717956543,
0.42386215925216675,
0.19185592234134674,
-0.2655370533466339,
-0.135320082306861... | 537 | 21.48 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042387 | [S1] You know, I landed where I landed now. Like, as far as not just my career, but like my mentality and how I approached it. [S2] Do you think those, do you think those core values that, that you was built off of are lost today on some people? [S1] I mean, I don't, I can't expect everybody to have that. You know what... | 28.2 | 3.350965 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042387.mp3 | [
7194,
8352,
4746,
7493,
1028,
9055,
814,
2869,
5629,
3238,
587,
5972,
12270,
12772,
10143,
3598,
5955,
7583,
109,
2620,
3830,
3038,
478,
2998,
6447,
2635,
3285,
7599,
11686,
10079,
6659,
3030,
2390,
3139,
11030,
11757,
12193,
12204,
3887,
... | [
1.3866856098175049,
-5.180866241455078,
-0.15829774737358093,
-0.02878906950354576,
0.572469174861908,
-0.21173575520515442,
0.37487876415252686,
0.0372820682823658,
-0.015954354777932167,
-3.9706461429595947,
0.38916125893592834,
0.13719697296619415,
-0.06541123986244202,
-0.0291016940027... | 705 | 28.2 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042388 | [S1] So how do you, so you, and, and because you, you, when your story speaks to a very specific group of people. [S2] Yeah. [S1] You know, kids, uh, younger generation who, um, had to take the long road to get to success. [S2] Right. [S1] Had to overcome so many people saying, "You can't do this and you can't do that,... | 27.28 | 3.207995 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042388.mp3 | [
10010,
12450,
3939,
10208,
987,
10549,
8043,
5222,
3879,
4979,
9149,
9149,
8121,
12004,
8155,
8923,
10114,
10579,
7827,
10444,
5837,
3860,
9939,
12507,
12115,
12578,
9764,
4452,
6040,
5493,
7982,
4227,
12610,
10634,
7825,
7746,
5188,
2700,
... | [
1.145275592803955,
-5.330851078033447,
-0.19324515759944916,
-0.24699878692626953,
0.4749446511268616,
-0.015728307887911797,
0.43834438920021057,
0.0024948790669441223,
-0.1303074210882187,
-4.0678391456604,
0.5406015515327454,
0.3022092580795288,
-0.25863340497016907,
-0.0309053026139736... | 682 | 27.28 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042389 | [S1] on your mindset, your work ethic, and what you had to go through to be able to do what you are doing today. And a lot of people might look at it, they look at us and say, "Oh, that shit is, it happened. Oh, they, they superstars. They, they, you know, they're in the league, they're getting this much money, this, t... | 28.2 | 3.301891 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042389.mp3 | [
6681,
5657,
12407,
5367,
7791,
7855,
8303,
5807,
6383,
4391,
4455,
4438,
6149,
6726,
7235,
5290,
12354,
9745,
2578,
12497,
3595,
5673,
4739,
12741,
10596,
8108,
3741,
5670,
10926,
8574,
11575,
6519,
10094,
7431,
2017,
9192,
1936,
7014,
1272... | [
1.2783739566802979,
-5.230589866638184,
-0.12478605657815933,
-0.2713713049888611,
0.3876449465751648,
0.010446939617395401,
0.437965989112854,
0.06984291225671768,
-0.11523713916540146,
-4.061929702758789,
0.7519879937171936,
0.2208717167377472,
-0.2548117935657501,
0.030392274260520935,
... | 705 | 28.2 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042390 | [S1] Right. You, you went to, you know, high school, didn't really have that many schools, only you went to a small school, but you embraced that and you, you made that part of your motivation to, to what you are today. [S2] Yeah. [S1] So, talk to those people about what that, like, what that mindset was. [S2] I mean, ... | 23.4 | 3.249312 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042390.mp3 | [
5657,
9837,
1132,
3029,
4477,
7101,
12703,
6556,
356,
6930,
12229,
10452,
5254,
4235,
9994,
8517,
10437,
5764,
5785,
8810,
4193,
9307,
4202,
4098,
3235,
7988,
11494,
7178,
8080,
5558,
3367,
4245,
3664,
9868,
7948,
10765,
2943,
2806,
1827,
... | [
1.4268934726715088,
-5.313533782958984,
-0.21571803092956543,
-0.039596039801836014,
0.44230949878692627,
-0.04757123067975044,
0.5154119729995728,
0.04883540794253349,
-0.1451103240251541,
-3.9855387210845947,
0.516624391078949,
0.29921525716781616,
-0.13420481979846954,
0.005287264473736... | 585 | 23.4 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042391 | [S1] My dad, after games, would just be like, "Man, they ain't... These dudes ain't nobody." You know what I'm saying? [S2] Like, that's a fact. [S1] They really... These dudes ain't nobody. I don't care who you playing against. They, you know what I'm saying? They probably scared or, you know, it's just hype. My dad w... | 26.76 | 3.109517 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042391.mp3 | [
5665,
5754,
2571,
6239,
8054,
5565,
12790,
2562,
5184,
9631,
11756,
11241,
11757,
11694,
12719,
3778,
12590,
12789,
9078,
2667,
8792,
924,
2894,
3609,
9939,
7583,
12662,
12076,
12629,
12565,
9438,
6238,
3794,
238,
3862,
6145,
9601,
8543,
39... | [
1.5433162450790405,
-5.0996222496032715,
-0.2662953734397888,
0.06920693814754486,
0.4616803228855133,
-0.1462610512971878,
0.4050191342830658,
-0.056880660355091095,
-0.04088101163506508,
-3.9934208393096924,
0.4635699987411499,
0.20954711735248566,
-0.14350803196430206,
0.097128935158252... | 669 | 26.76 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042392 | [S1] I took on that same mentality. And when I was at Weber, once I started to, you know, I started working with Phil Beckner, who I still train with to this day. He was like a grad assistant. We started working out and I started to just see them, I started to feel the results and I started to get better and then took ... | 28.04 | 3.284027 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042392.mp3 | [
4714,
6201,
10926,
6141,
6141,
8188,
9756,
5392,
3408,
9069,
10669,
7323,
3811,
6327,
8246,
11622,
8710,
10223,
12790,
10095,
4368,
7968,
770,
8999,
12734,
12701,
9997,
3605,
5669,
12207,
12446,
6849,
2960,
5045,
12725,
12778,
11643,
3100,
... | [
1.400766134262085,
-5.170370101928711,
-0.23706349730491638,
0.012092736549675465,
0.5263747572898865,
-0.050137314945459366,
0.4481668174266815,
-0.0021307747811079025,
-0.0723748579621315,
-4.016578197479248,
0.40955638885498047,
0.1966991424560547,
-0.06468803435564041,
0.04499357193708... | 701 | 28.04 |
bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042393 | [S1] The confidence is built in. I'd have been- [S2] Right. [S1] I'd have been through it for years and years and years, and in worse situations. So, I mean, that's what it takes. [S2] And, and- | 9.92 | 3.147455 | 24,000 | audio/en/bilibili_data_19103351_BV18k4y1176d_BV18k4y1176d_m4-dialogue_0042393.mp3 | [
7201,
3602,
5906,
1255,
11362,
10219,
8695,
8052,
8178,
11367,
3178,
2922,
7302,
5590,
12079,
9557,
328,
3302,
3778,
82,
3091,
8490,
12246,
10541,
7902,
6346,
8224,
476,
12267,
12121,
11604,
7591,
2342,
9805,
5724,
9266,
7208,
9792,
5146,
... | [
1.3420453071594238,
-5.54221773147583,
-0.22971811890602112,
0.06624488532543182,
0.4615028500556946,
-0.1633937805891037,
0.4302873909473419,
-0.10362023115158081,
-0.1393570452928543,
-4.0294928550720215,
0.30273672938346863,
0.18777482211589813,
-0.1314750760793686,
-0.07123175263404846... | 248 | 9.92 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.