| {"event": "start", "run": "exp32_local_onechunk_s1", "condition": "A", "params": 82470528, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "global", "p3_objective": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 1.9} |
| {"event": "probe_schedule", "n": 21, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 11000], "t": 1.9} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 14139.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.812446, "cps": 110.460938, "mean_p": 0.520947, "boundary_entropy": 0.686709, "bpb": 8.079001}, "de": {"chunks": 18668.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.909171, "cps": 145.84375, "mean_p": 0.525359, "boundary_entropy": 0.687245, "bpb": 8.131349}, "vi": {"chunks": 21341.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.157635, "cps": 166.726562, "mean_p": 0.523718, "boundary_entropy": 0.68791, "bpb": 8.099126}, "tr": {"chunks": 17844.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.985693, "cps": 139.40625, "mean_p": 0.529758, "boundary_entropy": 0.686399, "bpb": 8.170023}, "fi": {"chunks": 17557.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.933337, "cps": 137.164062, "mean_p": 0.526424, "boundary_entropy": 0.686837, "bpb": 8.067823}, "ru": {"chunks": 34128.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.80095, "cps": 266.625, "mean_p": 0.540375, "boundary_entropy": 0.687168, "bpb": 8.221234}, "el": {"chunks": 36914.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.798665, "cps": 288.390625, "mean_p": 0.53644, "boundary_entropy": 0.688054, "bpb": 8.074129}, "hi": {"chunks": 33657.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.000416, "cps": 262.945312, "mean_p": 0.509672, "boundary_entropy": 0.690736, "bpb": 8.173847}, "ta": {"chunks": 50614.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.707355, "cps": 395.421875, "mean_p": 0.521149, "boundary_entropy": 0.690436, "bpb": 8.094305}, "th": {"chunks": 45393.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.82207, "cps": 354.632812, "mean_p": 0.531267, "boundary_entropy": 0.689124, "bpb": 7.971664}, "zh": {"chunks": 4875.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.825432, "cps": 38.085938, "mean_p": 0.492214, "boundary_entropy": 0.686444, "bpb": 8.09271}, "my": {"chunks": 30094.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.588242, "cps": 235.109375, "mean_p": 0.506988, "boundary_entropy": 0.690851, "bpb": 8.169675}}, "bpb": {"en": 8.079, "de": 8.13135, "vi": 8.09913, "tr": 8.17002, "fi": 8.06782, "ru": 8.22123, "el": 8.07413, "hi": 8.17385, "ta": 8.0943, "th": 7.97166, "zh": 8.09271, "my": 8.16967}, "gini_cpc": 0.244844, "gini_cps": 0.27034, "bpic": 1.180051, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1190533992007132, "vi": 1.4248759281827625, "tr": 1.2132410293977671, "fi": 1.1487984777338243, "ru": 2.216700654508078, "el": 2.2138882150706616, "hi": 2.462213768688232, "ta": 3.332350057232739, "th": 3.473547534610352, "zh": 1.0159833788983057, "my": 1.9548883851595404}, "t": 5.0} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.62517, "bpb": 8.1154, "chunk_loss": 0.017276, "F": 0.00098, "G": 0.52219, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.69071, "router_noise": 0.0, "grad_norm": 4.9464, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 5.7} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 14187.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.815204, "cps": 110.835938, "mean_p": 0.52116, "boundary_entropy": 0.686692, "bpb": 8.062887}, "de": {"chunks": 18703.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.910875, "cps": 146.117188, "mean_p": 0.525549, "boundary_entropy": 0.687227, "bpb": 8.116921}, "vi": {"chunks": 21372.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.159317, "cps": 166.96875, "mean_p": 0.523898, "boundary_entropy": 0.687893, "bpb": 8.090238}, "tr": {"chunks": 17862.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.986687, "cps": 139.546875, "mean_p": 0.529885, "boundary_entropy": 0.686383, "bpb": 8.158053}, "fi": {"chunks": 17571.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.934081, "cps": 137.273438, "mean_p": 0.52653, "boundary_entropy": 0.686826, "bpb": 8.052982}, "ru": {"chunks": 34131.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.801108, "cps": 266.648438, "mean_p": 0.540536, "boundary_entropy": 0.687142, "bpb": 8.204243}, "el": {"chunks": 36915.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.798714, "cps": 288.398438, "mean_p": 0.536435, "boundary_entropy": 0.688055, "bpb": 8.066459}, "hi": {"chunks": 33597.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.99685, "cps": 262.476562, "mean_p": 0.509621, "boundary_entropy": 0.690737, "bpb": 8.14839}, "ta": {"chunks": 50613.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.707301, "cps": 395.414062, "mean_p": 0.521121, "boundary_entropy": 0.690438, "bpb": 8.07468}, "th": {"chunks": 45401.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.822568, "cps": 354.695312, "mean_p": 0.531307, "boundary_entropy": 0.68912, "bpb": 7.947023}, "zh": {"chunks": 4884.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.826956, "cps": 38.15625, "mean_p": 0.49224, "boundary_entropy": 0.686445, "bpb": 8.090749}, "my": {"chunks": 30133.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.5903, "cps": 235.414062, "mean_p": 0.50702, "boundary_entropy": 0.69085, "bpb": 8.167029}}, "bpb": {"en": 8.06289, "de": 8.11692, "vi": 8.09024, "tr": 8.15805, "fi": 8.05298, "ru": 8.20424, "el": 8.06646, "hi": 8.14839, "ta": 8.07468, "th": 7.94702, "zh": 8.09075, "my": 8.16703}, "gini_cpc": 0.244361, "gini_cps": 0.269986, "bpic": 1.179525, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1173581939390933, "vi": 1.4221178088352926, "tr": 1.210355881963792, "fi": 1.1458246126151943, "ru": 2.20939491410714, "el": 2.2064575631441645, "hi": 2.4495086527847993, "ta": 3.321009837781154, "th": 3.4624053083399384, "zh": 1.014415237465625, "my": 1.9507991019081816}, "t": 8.8} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 14285.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.820835, "cps": 111.601562, "mean_p": 0.521572, "boundary_entropy": 0.68666, "bpb": 8.02816}, "de": {"chunks": 18752.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.913262, "cps": 146.5, "mean_p": 0.525916, "boundary_entropy": 0.687191, "bpb": 8.084303}, "vi": {"chunks": 21424.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.162137, "cps": 167.375, "mean_p": 0.524245, "boundary_entropy": 0.68786, "bpb": 8.070034}, "tr": {"chunks": 17880.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.987682, "cps": 139.6875, "mean_p": 0.530114, "boundary_entropy": 0.686354, "bpb": 8.133276}, "fi": {"chunks": 17593.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.935251, "cps": 137.445312, "mean_p": 0.526751, "boundary_entropy": 0.686802, "bpb": 8.02323}, "ru": {"chunks": 34138.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.801478, "cps": 266.703125, "mean_p": 0.540847, "boundary_entropy": 0.687091, "bpb": 8.165698}, "el": {"chunks": 36915.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.798714, "cps": 288.398438, "mean_p": 0.536465, "boundary_entropy": 0.68805, "bpb": 8.053865}, "hi": {"chunks": 33568.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.995126, "cps": 262.25, "mean_p": 0.509594, "boundary_entropy": 0.690738, "bpb": 8.107136}, "ta": {"chunks": 50611.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.707194, "cps": 395.398438, "mean_p": 0.521111, "boundary_entropy": 0.690439, "bpb": 8.041022}, "th": {"chunks": 45415.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.823438, "cps": 354.804688, "mean_p": 0.531396, "boundary_entropy": 0.689109, "bpb": 7.910586}, "zh": {"chunks": 4892.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.82831, "cps": 38.21875, "mean_p": 0.492297, "boundary_entropy": 0.686447, "bpb": 8.087636}, "my": {"chunks": 30219.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.594839, "cps": 236.085938, "mean_p": 0.507075, "boundary_entropy": 0.690849, "bpb": 8.156133}}, "bpb": {"en": 8.02816, "de": 8.0843, "vi": 8.07003, "tr": 8.13328, "fi": 8.02323, "ru": 8.1657, "el": 8.05386, "hi": 8.10714, "ta": 8.04102, "th": 7.91059, "zh": 8.08764, "my": 8.15613}, "gini_cpc": 0.243666, "gini_cps": 0.269392, "bpic": 1.178356, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1126000180591507, "vi": 1.4157979957582094, "tr": 1.2032637648012325, "fi": 1.1393886652495404, "ru": 2.194687728053949, "el": 2.1913205074082094, "hi": 2.430604340321028, "ta": 3.298096239928338, "th": 3.4397123741794804, "zh": 1.0091062155545976, "my": 1.9429453629109577}, "t": 12.1} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 14585.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.838074, "cps": 113.945312, "mean_p": 0.523105, "boundary_entropy": 0.686532, "bpb": 7.903241}, "de": {"chunks": 18921.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.921492, "cps": 147.820312, "mean_p": 0.527258, "boundary_entropy": 0.687055, "bpb": 7.967939}, "vi": {"chunks": 21615.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.172498, "cps": 168.867188, "mean_p": 0.52545, "boundary_entropy": 0.687741, "bpb": 7.999373}, "tr": {"chunks": 17946.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.991327, "cps": 140.203125, "mean_p": 0.530967, "boundary_entropy": 0.686245, "bpb": 8.044735}, "fi": {"chunks": 17686.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.940195, "cps": 138.171875, "mean_p": 0.527594, "boundary_entropy": 0.68671, "bpb": 7.915504}, "ru": {"chunks": 34162.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.802744, "cps": 266.890625, "mean_p": 0.542032, "boundary_entropy": 0.686897, "bpb": 8.039836}, "el": {"chunks": 36896.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.797788, "cps": 288.25, "mean_p": 0.536387, "boundary_entropy": 0.688061, "bpb": 8.009701}, "hi": {"chunks": 33502.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.991204, "cps": 261.734375, "mean_p": 0.509531, "boundary_entropy": 0.69074, "bpb": 7.990474}, "ta": {"chunks": 50609.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.707087, "cps": 395.382812, "mean_p": 0.521072, "boundary_entropy": 0.690442, "bpb": 7.93443}, "th": {"chunks": 45441.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.825054, "cps": 355.007812, "mean_p": 0.531608, "boundary_entropy": 0.689083, "bpb": 7.791261}, "zh": {"chunks": 4950.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.838131, "cps": 38.671875, "mean_p": 0.492491, "boundary_entropy": 0.686453, "bpb": 8.075842}, "my": {"chunks": 30504.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.60988, "cps": 238.3125, "mean_p": 0.507285, "boundary_entropy": 0.690842, "bpb": 8.119574}}, "bpb": {"en": 7.90324, "de": 7.96794, "vi": 7.99937, "tr": 8.04473, "fi": 7.9155, "ru": 8.03984, "el": 8.0097, "hi": 7.99047, "ta": 7.93443, "th": 7.79126, "zh": 8.07584, "my": 8.11957}, "gini_cpc": 0.241088, "gini_cps": 0.267258, "bpic": 1.174299, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0995357774280126, "vi": 1.3990388823632292, "tr": 1.1828639572843098, "fi": 1.1218516317896332, "ru": 2.151056217623376, "el": 2.145142400749038, "hi": 2.3759283963811058, "ta": 3.2301297913028026, "th": 3.3708893861661604, "zh": 1.0000677973893592, "my": 1.9209280705384328}, "t": 15.7} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 15445.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.887491, "cps": 120.664062, "mean_p": 0.528397, "boundary_entropy": 0.686016, "bpb": 7.49274}, "de": {"chunks": 19401.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.944869, "cps": 151.570312, "mean_p": 0.53184, "boundary_entropy": 0.686536, "bpb": 7.58303}, "vi": {"chunks": 22232.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.205967, "cps": 173.6875, "mean_p": 0.529734, "boundary_entropy": 0.687271, "bpb": 7.764074}, "tr": {"chunks": 18173.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.003867, "cps": 141.976562, "mean_p": 0.533788, "boundary_entropy": 0.685863, "bpb": 7.747053}, "fi": {"chunks": 17939.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.953644, "cps": 140.148438, "mean_p": 0.530404, "boundary_entropy": 0.686379, "bpb": 7.5652}, "ru": {"chunks": 34240.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.80686, "cps": 267.5, "mean_p": 0.545861, "boundary_entropy": 0.68623, "bpb": 7.634648}, "el": {"chunks": 36890.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.797495, "cps": 288.203125, "mean_p": 0.53633, "boundary_entropy": 0.688067, "bpb": 7.795918}, "hi": {"chunks": 33131.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.969153, "cps": 258.835938, "mean_p": 0.509245, "boundary_entropy": 0.690751, "bpb": 7.583094}, "ta": {"chunks": 50584.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.70575, "cps": 395.1875, "mean_p": 0.520812, "boundary_entropy": 0.690462, "bpb": 7.553574}, "th": {"chunks": 45528.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.830463, "cps": 355.6875, "mean_p": 0.532185, "boundary_entropy": 0.689015, "bpb": 7.35171}, "zh": {"chunks": 5215.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.883, "cps": 40.742188, "mean_p": 0.493445, "boundary_entropy": 0.686485, "bpb": 8.030409}, "my": {"chunks": 31551.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.665136, "cps": 246.492188, "mean_p": 0.508003, "boundary_entropy": 0.690822, "bpb": 7.940576}}, "bpb": {"en": 7.49274, "de": 7.58303, "vi": 7.76407, "tr": 7.74705, "fi": 7.5652, "ru": 7.63465, "el": 7.79592, "hi": 7.58309, "ta": 7.55357, "th": 7.35171, "zh": 8.03041, "my": 7.94058}, "gini_cpc": 0.232584, "gini_cps": 0.260676, "bpic": 1.161814, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0646525927351083, "vi": 1.3588502523851005, "tr": 1.131129379495594, "fi": 1.0745399187250166, "ru": 2.0359201900351147, "el": 2.0253683433002005, "hi": 2.218787339632507, "ta": 3.048764696719961, "th": 3.1892878245859824, "zh": 0.9949404268908266, "my": 1.8762295001302913}, "t": 20.0} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 16585.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.952997, "cps": 129.570312, "mean_p": 0.543845, "boundary_entropy": 0.683889, "bpb": 6.616212}, "de": {"chunks": 20097.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.978766, "cps": 157.007812, "mean_p": 0.545225, "boundary_entropy": 0.684549, "bpb": 6.743936}, "vi": {"chunks": 23129.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.254624, "cps": 180.695312, "mean_p": 0.542462, "boundary_entropy": 0.68542, "bpb": 7.292175}, "tr": {"chunks": 18628.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.029001, "cps": 145.53125, "mean_p": 0.541929, "boundary_entropy": 0.684612, "bpb": 7.054556}, "fi": {"chunks": 18472.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.981979, "cps": 144.3125, "mean_p": 0.539616, "boundary_entropy": 0.685069, "bpb": 6.793376}, "ru": {"chunks": 34433.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.817045, "cps": 269.007812, "mean_p": 0.556613, "boundary_entropy": 0.684096, "bpb": 6.579609}, "el": {"chunks": 36888.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.797398, "cps": 288.1875, "mean_p": 0.536004, "boundary_entropy": 0.688113, "bpb": 7.313387}, "hi": {"chunks": 32776.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.948053, "cps": 256.0625, "mean_p": 0.508868, "boundary_entropy": 0.690773, "bpb": 6.718462}, "ta": {"chunks": 50581.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.70559, "cps": 395.164062, "mean_p": 0.520702, "boundary_entropy": 0.690474, "bpb": 6.829395}, "th": {"chunks": 45759.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.844824, "cps": 357.492188, "mean_p": 0.533137, "boundary_entropy": 0.688927, "bpb": 6.571397}, "zh": {"chunks": 6581.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.114291, "cps": 51.414062, "mean_p": 0.498107, "boundary_entropy": 0.6866, "bpb": 7.936034}, "my": {"chunks": 37750.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.992295, "cps": 294.921875, "mean_p": 0.511874, "boundary_entropy": 0.690694, "bpb": 7.529463}}, "bpb": {"en": 6.61621, "de": 6.74394, "vi": 7.29218, "tr": 7.05456, "fi": 6.79338, "ru": 6.57961, "el": 7.31339, "hi": 6.71846, "ta": 6.8294, "th": 6.5714, "zh": 7.93603, "my": 7.52946}, "gini_cpc": 0.217466, "gini_cps": 0.250105, "bpic": 1.123221, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0270402633254576, "vi": 1.3165045320933437, "tr": 1.0797527583551936, "fi": 1.030411461539247, "ru": 1.9066645528790283, "el": 1.8860487254378793, "hi": 2.044134755398033, "ta": 2.839033950497583, "th": 2.9851358770190286, "zh": 1.1692492297935468, "my": 2.0905580154664123}, "t": 25.7} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17275.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.992645, "cps": 134.960938, "mean_p": 0.601583, "boundary_entropy": 0.667328, "bpb": 5.919634}, "de": {"chunks": 20724.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.009302, "cps": 161.90625, "mean_p": 0.603015, "boundary_entropy": 0.66765, "bpb": 6.089528}, "vi": {"chunks": 23939.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.298563, "cps": 187.023438, "mean_p": 0.582319, "boundary_entropy": 0.675339, "bpb": 6.918028}, "tr": {"chunks": 19513.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.077888, "cps": 152.445312, "mean_p": 0.593104, "boundary_entropy": 0.670819, "bpb": 6.481104}, "fi": {"chunks": 19332.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.027697, "cps": 151.03125, "mean_p": 0.595043, "boundary_entropy": 0.670051, "bpb": 6.137952}, "ru": {"chunks": 34545.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.822955, "cps": 269.882812, "mean_p": 0.563251, "boundary_entropy": 0.682774, "bpb": 5.668088}, "el": {"chunks": 37199.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.812552, "cps": 290.617188, "mean_p": 0.547353, "boundary_entropy": 0.686363, "bpb": 6.63278}, "hi": {"chunks": 23898.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.420386, "cps": 186.703125, "mean_p": 0.502667, "boundary_entropy": 0.690974, "bpb": 5.679103}, "ta": {"chunks": 50012.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.675154, "cps": 390.71875, "mean_p": 0.516266, "boundary_entropy": 0.690817, "bpb": 5.746383}, "th": {"chunks": 46038.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.86217, "cps": 359.671875, "mean_p": 0.532661, "boundary_entropy": 0.689123, "bpb": 5.495656}, "zh": {"chunks": 13029.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.206062, "cps": 101.789062, "mean_p": 0.523249, "boundary_entropy": 0.685846, "bpb": 7.637331}, "my": {"chunks": 52953.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.794649, "cps": 413.695312, "mean_p": 0.528349, "boundary_entropy": 0.689607, "bpb": 6.661498}}, "bpb": {"en": 5.91963, "de": 6.08953, "vi": 6.91803, "tr": 6.4811, "fi": 6.13795, "ru": 5.66809, "el": 6.63278, "hi": 5.6791, "ta": 5.74638, "th": 5.49566, "zh": 7.63733, "my": 6.6615}, "gini_cpc": 0.221314, "gini_cps": 0.244258, "bpic": 1.070647, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.016780574815748, "vi": 1.3081842826342172, "tr": 1.085874306798699, "fi": 1.035311330986019, "ru": 1.8364624249204817, "el": 1.8259819666516541, "hi": 1.4309107553495053, "ta": 2.694975473835649, "th": 2.8833771169533056, "zh": 2.2224075592384893, "my": 2.8153556034383995}, "t": 34.8} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17393.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.999425, "cps": 135.882812, "mean_p": 0.792709, "boundary_entropy": 0.504482, "bpb": 5.49021}, "de": {"chunks": 20836.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.014757, "cps": 162.78125, "mean_p": 0.79224, "boundary_entropy": 0.50628, "bpb": 5.649542}, "vi": {"chunks": 24208.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.313154, "cps": 189.125, "mean_p": 0.736482, "boundary_entropy": 0.569658, "bpb": 6.444588}, "tr": {"chunks": 19723.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.089488, "cps": 154.085938, "mean_p": 0.78105, "boundary_entropy": 0.519423, "bpb": 6.070392}, "fi": {"chunks": 19557.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.039658, "cps": 152.789062, "mean_p": 0.781507, "boundary_entropy": 0.518301, "bpb": 5.690374}, "ru": {"chunks": 34556.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823536, "cps": 269.96875, "mean_p": 0.646095, "boundary_entropy": 0.648115, "bpb": 4.924962}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.589503, "boundary_entropy": 0.675121, "bpb": 5.321338}, "hi": {"chunks": 43165.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.565527, "cps": 337.226562, "mean_p": 0.563649, "boundary_entropy": 0.683228, "bpb": 4.709763}, "ta": {"chunks": 50904.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.722867, "cps": 397.6875, "mean_p": 0.561859, "boundary_entropy": 0.683948, "bpb": 4.95329}, "th": {"chunks": 46084.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.86503, "cps": 360.03125, "mean_p": 0.619119, "boundary_entropy": 0.662949, "bpb": 4.806862}, "zh": {"chunks": 16167.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.737386, "cps": 126.304688, "mean_p": 0.642628, "boundary_entropy": 0.646513, "bpb": 6.844093}, "my": {"chunks": 53719.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.835075, "cps": 419.679688, "mean_p": 0.548429, "boundary_entropy": 0.686804, "bpb": 4.835418}}, "bpb": {"en": 5.49021, "de": 5.64954, "vi": 6.44459, "tr": 6.07039, "fi": 5.69037, "ru": 4.92496, "el": 5.32134, "hi": 4.70976, "ta": 4.95329, "th": 4.80686, "zh": 6.84409, "my": 4.83542}, "gini_cpc": 0.221721, "gini_cps": 0.234026, "bpic": 1.000534, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.015340161301107, "vi": 1.313909316136225, "tr": 1.0901143246637082, "fi": 1.0402553919883355, "ru": 1.8245840508123017, "el": 1.816762886083451, "hi": 2.5670025233460985, "ta": 2.724432572691774, "th": 2.8666767619936353, "zh": 2.7389595527812265, "my": 2.836704951112105}, "t": 49.6} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.35844, "bpb": 4.8452, "chunk_loss": 0.008231, "F": 0.00098, "G": 0.77494, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.51058, "router_noise": 0.0, "grad_norm": 3.2101, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.591, "alpha": 0.03, "t": 64.7} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.844234, "boundary_entropy": 0.424369, "bpb": 4.351306}, "de": {"chunks": 20861.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.015974, "cps": 162.976562, "mean_p": 0.84576, "boundary_entropy": 0.418082, "bpb": 4.611573}, "vi": {"chunks": 24255.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.315704, "cps": 189.492188, "mean_p": 0.713193, "boundary_entropy": 0.588087, "bpb": 4.800908}, "tr": {"chunks": 19732.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.089985, "cps": 154.15625, "mean_p": 0.789806, "boundary_entropy": 0.49188, "bpb": 4.802107}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.825895, "boundary_entropy": 0.44854, "bpb": 4.606451}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.751498, "boundary_entropy": 0.556119, "bpb": 3.239892}, "el": {"chunks": 37257.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815378, "cps": 291.070312, "mean_p": 0.676449, "boundary_entropy": 0.624105, "bpb": 3.715803}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.673838, "boundary_entropy": 0.626948, "bpb": 2.805365}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.701645, "boundary_entropy": 0.605617, "bpb": 2.876841}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.701878, "boundary_entropy": 0.604991, "bpb": 2.819389}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.856756, "boundary_entropy": 0.40555, "bpb": 6.211469}, "my": {"chunks": 53724.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.835339, "cps": 419.71875, "mean_p": 0.671352, "boundary_entropy": 0.61942, "bpb": 2.880282}}, "bpb": {"en": 4.35131, "de": 4.61157, "vi": 4.80091, "tr": 4.80211, "fi": 4.60645, "ru": 3.23989, "el": 3.7158, "hi": 2.80536, "ta": 2.87684, "th": 2.81939, "zh": 6.21147, "my": 2.88028}, "gini_cpc": 0.221536, "gini_cps": 0.233853, "bpic": 1.000146, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0150993505006827, "vi": 1.3145707689388406, "tr": 1.0890464140692686, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.81381450508406, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8328970903633786}, "t": 77.3} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.36349, "bpb": 3.40979, "chunk_loss": 0.003484, "F": 0.00098, "G": 0.90758, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.28622, "router_noise": 0.0, "grad_norm": 2.3996, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.41, "alpha": 0.03, "t": 105.7} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.96102, "boundary_entropy": 0.158849, "bpb": 3.485933}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.956232, "boundary_entropy": 0.169639, "bpb": 3.740562}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.928086, "boundary_entropy": 0.243182, "bpb": 3.078655}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.947626, "boundary_entropy": 0.186874, "bpb": 3.769505}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.961937, "boundary_entropy": 0.152777, "bpb": 3.808157}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.983363, "boundary_entropy": 0.080154, "bpb": 2.613296}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.961985, "boundary_entropy": 0.159766, "bpb": 2.772602}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.964613, "boundary_entropy": 0.151595, "bpb": 2.057358}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.963148, "boundary_entropy": 0.156615, "bpb": 1.916238}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.967684, "boundary_entropy": 0.141384, "bpb": 2.092931}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.941247, "boundary_entropy": 0.218339, "bpb": 4.505707}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.939727, "boundary_entropy": 0.225401, "bpb": 2.176433}}, "bpb": {"en": 3.48593, "de": 3.74056, "vi": 3.07866, "tr": 3.76951, "fi": 3.80816, "ru": 2.6133, "el": 2.7726, "hi": 2.05736, "ta": 1.91624, "th": 2.09293, "zh": 4.50571, "my": 2.17643}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 129.0} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.11691, "bpb": 3.05406, "chunk_loss": 0.000935, "F": 0.00098, "G": 0.97878, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.0976, "router_noise": 0.0, "grad_norm": 1.5426, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.402, "alpha": 0.03, "t": 145.9} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 1.99125, "bpb": 2.87277, "chunk_loss": 0.00041, "F": 0.00098, "G": 0.99346, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.03872, "router_noise": 0.0, "grad_norm": 1.5343, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.336, "alpha": 0.03, "t": 179.5} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.79429, "bpb": 2.58861, "chunk_loss": 0.000267, "F": 0.00098, "G": 0.99744, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.0176, "router_noise": 0.0, "grad_norm": 1.1028, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.315, "alpha": 0.03, "t": 211.0} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.79797, "bpb": 2.59393, "chunk_loss": 0.000227, "F": 0.00098, "G": 0.99858, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.01059, "router_noise": 0.0, "grad_norm": 1.0183, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.296, "alpha": 0.03, "t": 240.6} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.77183, "bpb": 2.55622, "chunk_loss": 0.000208, "F": 0.00098, "G": 0.99909, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00722, "router_noise": 0.0, "grad_norm": 0.9532, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.281, "alpha": 0.03, "t": 268.8} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.62301, "bpb": 2.34151, "chunk_loss": 0.000196, "F": 0.00098, "G": 0.99944, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00467, "router_noise": 0.0, "grad_norm": 0.8297, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.267, "alpha": 0.03, "t": 295.5} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.56088, "bpb": 2.25188, "chunk_loss": 0.000192, "F": 0.00098, "G": 0.99955, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.0039, "router_noise": 0.0, "grad_norm": 0.8445, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.257, "alpha": 0.03, "t": 321.2} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999691, "boundary_entropy": 0.002774, "bpb": 2.337446}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999627, "boundary_entropy": 0.003284, "bpb": 2.616441}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.999741, "boundary_entropy": 0.002378, "bpb": 2.126672}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.999624, "boundary_entropy": 0.003313, "bpb": 2.632175}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999728, "boundary_entropy": 0.002485, "bpb": 2.853465}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999689, "boundary_entropy": 0.0028, "bpb": 1.687933}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.999689, "boundary_entropy": 0.002805, "bpb": 1.781484}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.999809, "boundary_entropy": 0.001816, "bpb": 1.383445}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.999816, "boundary_entropy": 0.001752, "bpb": 1.135823}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.999743, "boundary_entropy": 0.002359, "bpb": 1.333656}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999708, "boundary_entropy": 0.002649, "bpb": 3.487845}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.999766, "boundary_entropy": 0.002169, "bpb": 1.264966}}, "bpb": {"en": 2.33745, "de": 2.61644, "vi": 2.12667, "tr": 2.63218, "fi": 2.85346, "ru": 1.68793, "el": 1.78148, "hi": 1.38344, "ta": 1.13582, "th": 1.33366, "zh": 3.48784, "my": 1.26497}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 349.4} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.35022, "de": 2.60846, "vi": 2.12863, "tr": 2.62582, "fi": 2.86252, "ru": 1.68962, "el": 1.77559, "hi": 1.39087, "ta": 1.13629, "th": 1.33199, "zh": 3.49268, "my": 1.27656}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 2.05577, "t": 353.8} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.54094, "bpb": 2.22311, "chunk_loss": 0.000186, "F": 0.00098, "G": 0.99971, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00263, "router_noise": 0.0, "grad_norm": 0.616, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.328, "alpha": 0.03, "t": 354.0} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.45932, "bpb": 2.10536, "chunk_loss": 0.000184, "F": 0.00098, "G": 0.99977, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00215, "router_noise": 0.0, "grad_norm": 0.7457, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.245, "alpha": 0.03, "t": 378.5} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.51443, "bpb": 2.18486, "chunk_loss": 0.000183, "F": 0.00098, "G": 0.99981, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00183, "router_noise": 0.0, "grad_norm": 0.7065, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 402.6} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.39111, "bpb": 2.00695, "chunk_loss": 0.000181, "F": 0.00098, "G": 0.99985, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00147, "router_noise": 0.0, "grad_norm": 0.6968, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 426.3} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.35004, "bpb": 1.94769, "chunk_loss": 0.00018, "F": 0.00098, "G": 0.99987, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00128, "router_noise": 0.0, "grad_norm": 0.6194, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.233, "alpha": 0.03, "t": 449.6} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.43174, "bpb": 2.06557, "chunk_loss": 0.00018, "F": 0.00098, "G": 0.99988, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.0012, "router_noise": 0.0, "grad_norm": 0.6733, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.228, "alpha": 0.03, "t": 472.4} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.46639, "bpb": 2.11556, "chunk_loss": 0.00018, "F": 0.00098, "G": 0.99989, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00111, "router_noise": 0.0, "grad_norm": 0.6062, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 496.4} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.49947, "bpb": 2.16327, "chunk_loss": 0.00018, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00106, "router_noise": 0.0, "grad_norm": 0.5939, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.225, "alpha": 0.03, "t": 518.9} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.30218, "bpb": 1.87865, "chunk_loss": 0.00018, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00106, "router_noise": 0.0, "grad_norm": 0.5982, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 542.8} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.28612, "bpb": 1.85548, "chunk_loss": 0.00018, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00106, "router_noise": 0.0, "grad_norm": 0.5672, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.226, "alpha": 0.03, "t": 565.4} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999897, "boundary_entropy": 0.001046, "bpb": 2.116976}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999891, "boundary_entropy": 0.001099, "bpb": 2.322506}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001017, "bpb": 1.971024}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.999895, "boundary_entropy": 0.001061, "bpb": 2.295064}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999897, "boundary_entropy": 0.001046, "bpb": 2.503548}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999897, "boundary_entropy": 0.001052, "bpb": 1.426688}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.999897, "boundary_entropy": 0.001051, "bpb": 1.49101}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.00102, "bpb": 1.20643}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001024, "bpb": 1.010469}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001026, "bpb": 1.14879}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999899, "boundary_entropy": 0.001026, "bpb": 3.163672}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.999899, "boundary_entropy": 0.001036, "bpb": 1.146913}}, "bpb": {"en": 2.11698, "de": 2.32251, "vi": 1.97102, "tr": 2.29506, "fi": 2.50355, "ru": 1.42669, "el": 1.49101, "hi": 1.20643, "ta": 1.01047, "th": 1.14879, "zh": 3.16367, "my": 1.14691}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 591.4} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.1257, "de": 2.31769, "vi": 1.97831, "tr": 2.28767, "fi": 2.51738, "ru": 1.4261, "el": 1.48798, "hi": 1.21536, "ta": 1.0105, "th": 1.14188, "zh": 3.17256, "my": 1.15549}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.81972, "t": 595.7} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.38983, "bpb": 2.00509, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00105, "router_noise": 0.0, "grad_norm": 0.5635, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.306, "alpha": 0.03, "t": 596.0} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.40731, "bpb": 2.03032, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00105, "router_noise": 0.0, "grad_norm": 0.571, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 620.1} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.27151, "bpb": 1.8344, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.5299, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.223, "alpha": 0.03, "t": 642.4} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.41503, "bpb": 2.04146, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.5447, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 666.4} |
| {"event": "train", "step": 2400, "phase": 1, "objective": "global", "lm_loss": 1.32171, "bpb": 1.90682, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.5859, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 690.4} |
| {"event": "train", "step": 2500, "phase": 1, "objective": "global", "lm_loss": 1.29138, "bpb": 1.86306, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.5546, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.226, "alpha": 0.03, "t": 713.0} |
| {"event": "train", "step": 2600, "phase": 1, "objective": "global", "lm_loss": 1.3887, "bpb": 2.00347, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.5293, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 737.0} |
| {"event": "train", "step": 2700, "phase": 1, "objective": "global", "lm_loss": 1.24983, "bpb": 1.80312, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.532, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.242, "alpha": 0.03, "t": 761.2} |
| {"event": "train", "step": 2800, "phase": 1, "objective": "global", "lm_loss": 1.32003, "bpb": 1.90441, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.4924, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.224, "alpha": 0.03, "t": 783.6} |
| {"event": "train", "step": 2900, "phase": 1, "objective": "global", "lm_loss": 1.38969, "bpb": 2.0049, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.5764, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 807.5} |
| {"event": "probe", "step": 3000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 2.022004}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.9999, "boundary_entropy": 0.001023, "bpb": 2.183322}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.875349}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 2.171962}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 2.346761}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.9999, "boundary_entropy": 0.001021, "bpb": 1.330148}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.9999, "boundary_entropy": 0.001025, "bpb": 1.376647}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.00102, "bpb": 1.127361}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.948447}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001021, "bpb": 1.076589}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.9999, "boundary_entropy": 0.00102, "bpb": 3.001654}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.001025, "bpb": 1.075343}}, "bpb": {"en": 2.022, "de": 2.18332, "vi": 1.87535, "tr": 2.17196, "fi": 2.34676, "ru": 1.33015, "el": 1.37665, "hi": 1.12736, "ta": 0.94845, "th": 1.07659, "zh": 3.00165, "my": 1.07534}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 833.6} |
| {"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 2.03021, "de": 2.17521, "vi": 1.88252, "tr": 2.16513, "fi": 2.36106, "ru": 1.33064, "el": 1.37474, "hi": 1.13555, "ta": 0.94922, "th": 1.07186, "zh": 3.00902, "my": 1.08453}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.71414, "t": 838.0} |
| {"event": "train", "step": 3000, "phase": 1, "objective": "global", "lm_loss": 1.41923, "bpb": 2.04752, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00103, "router_noise": 0.0, "grad_norm": 0.5218, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.307, "alpha": 0.03, "t": 838.3} |
| {"event": "train", "step": 3100, "phase": 1, "objective": "global", "lm_loss": 1.30139, "bpb": 1.87751, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5117, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.231, "alpha": 0.03, "t": 861.4} |
| {"event": "train", "step": 3200, "phase": 1, "objective": "global", "lm_loss": 1.33554, "bpb": 1.92678, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5295, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.262, "alpha": 0.03, "t": 887.6} |
| {"event": "train", "step": 3300, "phase": 1, "objective": "global", "lm_loss": 1.20171, "bpb": 1.7337, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4834, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.264, "alpha": 0.03, "t": 914.0} |
| {"event": "train", "step": 3400, "phase": 1, "objective": "global", "lm_loss": 1.25822, "bpb": 1.81523, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4842, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.259, "alpha": 0.03, "t": 939.9} |
| {"event": "train", "step": 3500, "phase": 1, "objective": "global", "lm_loss": 1.24244, "bpb": 1.79247, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4853, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.254, "alpha": 0.03, "t": 965.4} |
| {"event": "train", "step": 3600, "phase": 1, "objective": "global", "lm_loss": 1.30403, "bpb": 1.88131, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5141, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.262, "alpha": 0.03, "t": 991.5} |
| {"event": "train", "step": 3700, "phase": 1, "objective": "global", "lm_loss": 1.21813, "bpb": 1.75739, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5081, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.25, "alpha": 0.03, "t": 1016.5} |
| {"event": "train", "step": 3800, "phase": 1, "objective": "global", "lm_loss": 1.36754, "bpb": 1.97294, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5275, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.257, "alpha": 0.03, "t": 1042.2} |
| {"event": "train", "step": 3900, "phase": 1, "objective": "global", "lm_loss": 1.21832, "bpb": 1.75766, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4835, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.259, "alpha": 0.03, "t": 1068.1} |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.966428}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.9999, "boundary_entropy": 0.001017, "bpb": 2.121711}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.81251}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 2.08359}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 2.249151}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 1.279434}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.9999, "boundary_entropy": 0.00102, "bpb": 1.308139}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 1.06912}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.903684}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 1.026402}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.893337}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.00102, "bpb": 1.0325}}, "bpb": {"en": 1.96643, "de": 2.12171, "vi": 1.81251, "tr": 2.08359, "fi": 2.24915, "ru": 1.27943, "el": 1.30814, "hi": 1.06912, "ta": 0.90368, "th": 1.0264, "zh": 2.89334, "my": 1.0325}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 1095.0} |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.97253, "de": 2.11462, "vi": 1.82054, "tr": 2.07723, "fi": 2.26324, "ru": 1.28066, "el": 1.3051, "hi": 1.07782, "ta": 0.90496, "th": 1.02359, "zh": 2.89973, "my": 1.03987}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.64832, "t": 1099.3} |
| {"event": "train", "step": 4000, "phase": 1, "objective": "global", "lm_loss": 1.233, "bpb": 1.77884, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4636, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.315, "alpha": 0.03, "t": 1099.6} |
| {"event": "train", "step": 4100, "phase": 1, "objective": "global", "lm_loss": 1.29947, "bpb": 1.87474, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4702, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.258, "alpha": 0.03, "t": 1125.4} |
| {"event": "train", "step": 4200, "phase": 1, "objective": "global", "lm_loss": 1.10899, "bpb": 1.59994, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4638, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.256, "alpha": 0.03, "t": 1151.1} |
| {"event": "train", "step": 4300, "phase": 1, "objective": "global", "lm_loss": 1.28261, "bpb": 1.85041, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4607, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.244, "alpha": 0.03, "t": 1175.5} |
| {"event": "phase_transition", "step": 4400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 1200.9} |
| {"event": "train", "step": 4400, "phase": 2, "objective": "global", "lm_loss": 1.30307, "bpb": 1.87993, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5109, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.257, "alpha": 0.03, "t": 1201.2} |
| {"event": "train", "step": 4500, "phase": 2, "objective": "global", "lm_loss": 1.28875, "bpb": 1.85927, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4916, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.257, "alpha": 0.03, "t": 1226.9} |
| {"event": "train", "step": 4600, "phase": 2, "objective": "global", "lm_loss": 1.21937, "bpb": 1.75918, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.476, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.242, "alpha": 0.03, "t": 1251.1} |
| {"event": "train", "step": 4700, "phase": 2, "objective": "global", "lm_loss": 1.24005, "bpb": 1.78901, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4748, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.252, "alpha": 0.03, "t": 1276.3} |
| {"event": "train", "step": 4800, "phase": 2, "objective": "global", "lm_loss": 1.25066, "bpb": 1.80433, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4511, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1300.3} |
| {"event": "train", "step": 4900, "phase": 2, "objective": "global", "lm_loss": 1.23538, "bpb": 1.78227, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.488, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1324.2} |
| {"event": "probe", "step": 5000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.92215}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.9999, "boundary_entropy": 0.001017, "bpb": 2.064368}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.746123}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 2.028858}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 2.177474}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.242231}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 1.256636}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 1.034896}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.880242}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001021, "bpb": 0.996348}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 2.822298}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.00102, "bpb": 1.002155}}, "bpb": {"en": 1.92215, "de": 2.06437, "vi": 1.74612, "tr": 2.02886, "fi": 2.17747, "ru": 1.24223, "el": 1.25664, "hi": 1.0349, "ta": 0.88024, "th": 0.99635, "zh": 2.8223, "my": 1.00216}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 1350.3} |
| {"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.92733, "de": 2.05868, "vi": 1.75602, "tr": 2.02112, "fi": 2.19186, "ru": 1.24325, "el": 1.2538, "hi": 1.04308, "ta": 0.88312, "th": 0.99445, "zh": 2.831, "my": 1.0106}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.60119, "t": 1354.8} |
| {"event": "train", "step": 5000, "phase": 2, "objective": "global", "lm_loss": 1.18445, "bpb": 1.70879, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4926, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.308, "alpha": 0.03, "t": 1355.0} |
| {"event": "train", "step": 5100, "phase": 2, "objective": "global", "lm_loss": 1.17367, "bpb": 1.69324, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4716, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1378.9} |
| {"event": "train", "step": 5200, "phase": 2, "objective": "global", "lm_loss": 1.29278, "bpb": 1.86509, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.478, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1402.8} |
| {"event": "train", "step": 5300, "phase": 2, "objective": "global", "lm_loss": 1.19762, "bpb": 1.72781, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4816, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 1426.9} |
| {"event": "train", "step": 5400, "phase": 2, "objective": "global", "lm_loss": 1.14844, "bpb": 1.65685, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4957, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 1451.0} |
| {"event": "train", "step": 5500, "phase": 2, "objective": "global", "lm_loss": 1.15675, "bpb": 1.66884, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4448, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 1475.1} |
| {"event": "train", "step": 5600, "phase": 2, "objective": "global", "lm_loss": 1.18423, "bpb": 1.70849, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4726, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 1499.2} |
| {"event": "train", "step": 5700, "phase": 2, "objective": "global", "lm_loss": 1.27265, "bpb": 1.83604, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4766, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 1523.2} |
| {"event": "train", "step": 5800, "phase": 2, "objective": "global", "lm_loss": 1.27734, "bpb": 1.84281, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5223, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 1547.1} |
| {"event": "train", "step": 5900, "phase": 2, "objective": "global", "lm_loss": 1.22439, "bpb": 1.76642, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5507, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 1571.2} |
| {"event": "probe", "step": 6000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.8745}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 2.01368}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.695619}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001015, "bpb": 1.954403}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 2.118228}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.21529}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 1.228583}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 1.016072}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.999901, "boundary_entropy": 0.001019, "bpb": 0.858126}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.973235}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.780877}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.975416}}, "bpb": {"en": 1.8745, "de": 2.01368, "vi": 1.69562, "tr": 1.9544, "fi": 2.11823, "ru": 1.21529, "el": 1.22858, "hi": 1.01607, "ta": 0.85813, "th": 0.97323, "zh": 2.78088, "my": 0.97542}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 1597.3} |
| {"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.87852, "de": 2.00871, "vi": 1.70859, "tr": 1.94902, "fi": 2.13045, "ru": 1.21478, "el": 1.22753, "hi": 1.02242, "ta": 0.86051, "th": 0.97131, "zh": 2.78945, "my": 0.9841}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.56212, "t": 1601.8} |
| {"event": "train", "step": 6000, "phase": 2, "objective": "global", "lm_loss": 1.23058, "bpb": 1.77536, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.462, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.308, "alpha": 0.03, "t": 1602.0} |
| {"event": "train", "step": 6100, "phase": 2, "objective": "global", "lm_loss": 1.18611, "bpb": 1.7112, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5045, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 1626.0} |
| {"event": "train", "step": 6200, "phase": 2, "objective": "global", "lm_loss": 1.12423, "bpb": 1.62192, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4767, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1649.9} |
| {"event": "train", "step": 6300, "phase": 2, "objective": "global", "lm_loss": 1.2112, "bpb": 1.74739, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4805, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1673.8} |
| {"event": "train", "step": 6400, "phase": 2, "objective": "global", "lm_loss": 1.1493, "bpb": 1.65809, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4549, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1697.6} |
| {"event": "train", "step": 6500, "phase": 2, "objective": "global", "lm_loss": 1.13812, "bpb": 1.64196, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4702, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 1721.3} |
| {"event": "train", "step": 6600, "phase": 2, "objective": "global", "lm_loss": 1.18768, "bpb": 1.71345, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4756, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 1745.3} |
| {"event": "train", "step": 6700, "phase": 2, "objective": "global", "lm_loss": 1.2692, "bpb": 1.83107, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5102, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1769.3} |
| {"event": "train", "step": 6800, "phase": 2, "objective": "global", "lm_loss": 1.12543, "bpb": 1.62366, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4818, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.238, "alpha": 0.03, "t": 1793.1} |
| {"event": "train", "step": 6900, "phase": 2, "objective": "global", "lm_loss": 1.19418, "bpb": 1.72284, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4851, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.236, "alpha": 0.03, "t": 1816.7} |
| {"event": "probe", "step": 7000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.844961}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999901, "boundary_entropy": 0.001016, "bpb": 1.972751}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.654689}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001015, "bpb": 1.906082}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 2.068323}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.193944}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.200078}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 0.987065}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.843187}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.951527}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.736183}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.999901, "boundary_entropy": 0.001019, "bpb": 0.942952}}, "bpb": {"en": 1.84496, "de": 1.97275, "vi": 1.65469, "tr": 1.90608, "fi": 2.06832, "ru": 1.19394, "el": 1.20008, "hi": 0.98706, "ta": 0.84319, "th": 0.95153, "zh": 2.73618, "my": 0.94295}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 1842.7} |
| {"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.84656, "de": 1.96912, "vi": 1.66676, "tr": 1.90053, "fi": 2.08179, "ru": 1.19385, "el": 1.19782, "hi": 0.99369, "ta": 0.84667, "th": 0.94967, "zh": 2.74643, "my": 0.9523}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.52877, "t": 1847.3} |
| {"event": "train", "step": 7000, "phase": 2, "objective": "global", "lm_loss": 1.19521, "bpb": 1.72432, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4872, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.308, "alpha": 0.03, "t": 1847.6} |
| {"event": "train", "step": 7100, "phase": 2, "objective": "global", "lm_loss": 1.0933, "bpb": 1.5773, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.463, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 1871.3} |
| {"event": "train", "step": 7200, "phase": 2, "objective": "global", "lm_loss": 1.15416, "bpb": 1.6651, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5024, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 1895.0} |
| {"event": "train", "step": 7300, "phase": 2, "objective": "global", "lm_loss": 1.19528, "bpb": 1.72442, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4901, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.236, "alpha": 0.03, "t": 1918.6} |
| {"event": "train", "step": 7400, "phase": 2, "objective": "global", "lm_loss": 1.14649, "bpb": 1.65403, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4889, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.236, "alpha": 0.03, "t": 1942.1} |
| {"event": "train", "step": 7500, "phase": 2, "objective": "global", "lm_loss": 1.19178, "bpb": 1.71938, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4679, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.236, "alpha": 0.03, "t": 1965.8} |
| {"event": "train", "step": 7600, "phase": 2, "objective": "global", "lm_loss": 1.13993, "bpb": 1.64457, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4794, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 1989.7} |
| {"event": "phase_transition", "step": 7700, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 2012.9} |
| {"event": "train", "step": 7700, "phase": 3, "objective": "global", "lm_loss": 1.14516, "bpb": 1.65211, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5037, "lr": 9.06e-05, "router_trainable": true, "sec_per_step": 0.235, "alpha": 0.03, "t": 2013.1} |
| {"event": "train", "step": 7800, "phase": 3, "objective": "global", "lm_loss": 1.1766, "bpb": 1.69747, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4838, "lr": 8.73e-05, "router_trainable": true, "sec_per_step": 0.238, "alpha": 0.03, "t": 2036.9} |
| {"event": "train", "step": 7900, "phase": 3, "objective": "global", "lm_loss": 1.07972, "bpb": 1.55771, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4717, "lr": 8.4e-05, "router_trainable": true, "sec_per_step": 0.236, "alpha": 0.03, "t": 2060.5} |
| {"event": "probe", "step": 8000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.81948}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.942991}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.626989}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001015, "bpb": 1.870388}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 2.031455}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.174319}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 1.175456}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 0.969656}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.825764}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.935232}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.703654}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.929638}}, "bpb": {"en": 1.81948, "de": 1.94299, "vi": 1.62699, "tr": 1.87039, "fi": 2.03146, "ru": 1.17432, "el": 1.17546, "hi": 0.96966, "ta": 0.82576, "th": 0.93523, "zh": 2.70365, "my": 0.92964}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 2086.0} |
| {"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.82089, "de": 1.93963, "vi": 1.64081, "tr": 1.86438, "fi": 2.04352, "ru": 1.17442, "el": 1.17249, "hi": 0.97591, "ta": 0.82937, "th": 0.93338, "zh": 2.71467, "my": 0.93936}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.50407, "t": 2090.5} |
| {"event": "train", "step": 8000, "phase": 3, "objective": "global", "lm_loss": 1.1013, "bpb": 1.58884, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4746, "lr": 8.08e-05, "router_trainable": true, "sec_per_step": 0.302, "alpha": 0.03, "t": 2090.7} |
| {"event": "train", "step": 8100, "phase": 3, "objective": "global", "lm_loss": 1.16767, "bpb": 1.6846, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5122, "lr": 7.77e-05, "router_trainable": true, "sec_per_step": 0.234, "alpha": 0.03, "t": 2114.1} |
| {"event": "train", "step": 8200, "phase": 3, "objective": "global", "lm_loss": 1.16888, "bpb": 1.68634, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4853, "lr": 7.47e-05, "router_trainable": true, "sec_per_step": 0.234, "alpha": 0.03, "t": 2137.5} |
| {"event": "train", "step": 8300, "phase": 3, "objective": "global", "lm_loss": 1.11216, "bpb": 1.60451, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.494, "lr": 7.17e-05, "router_trainable": true, "sec_per_step": 0.233, "alpha": 0.03, "t": 2160.8} |
| {"event": "train", "step": 8400, "phase": 3, "objective": "global", "lm_loss": 1.26345, "bpb": 1.82278, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5151, "lr": 6.88e-05, "router_trainable": true, "sec_per_step": 0.234, "alpha": 0.03, "t": 2184.2} |
| {"event": "train", "step": 8500, "phase": 3, "objective": "global", "lm_loss": 1.14922, "bpb": 1.65797, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4759, "lr": 6.6e-05, "router_trainable": true, "sec_per_step": 0.233, "alpha": 0.03, "t": 2207.5} |
| {"event": "train", "step": 8600, "phase": 3, "objective": "global", "lm_loss": 1.17826, "bpb": 1.69988, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.507, "lr": 6.33e-05, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 2231.2} |
| {"event": "train", "step": 8700, "phase": 3, "objective": "global", "lm_loss": 1.13118, "bpb": 1.63195, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5211, "lr": 6.07e-05, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 2255.3} |
| {"event": "train", "step": 8800, "phase": 3, "objective": "global", "lm_loss": 1.20178, "bpb": 1.7338, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5182, "lr": 5.82e-05, "router_trainable": true, "sec_per_step": 0.247, "alpha": 0.03, "t": 2280.0} |
| {"event": "train", "step": 8900, "phase": 3, "objective": "global", "lm_loss": 1.1411, "bpb": 1.64625, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4896, "lr": 5.58e-05, "router_trainable": true, "sec_per_step": 0.25, "alpha": 0.03, "t": 2305.0} |
| {"event": "probe", "step": 9000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.804012}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.923246}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.601953}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001015, "bpb": 1.847999}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 2.000333}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.161966}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.156591}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 0.95585}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.809456}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.92408}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.683507}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.913693}}, "bpb": {"en": 1.80401, "de": 1.92325, "vi": 1.60195, "tr": 1.848, "fi": 2.00033, "ru": 1.16197, "el": 1.15659, "hi": 0.95585, "ta": 0.80946, "th": 0.92408, "zh": 2.68351, "my": 0.91369}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 2331.9} |
| {"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.8053, "de": 1.92009, "vi": 1.61588, "tr": 1.84256, "fi": 2.0127, "ru": 1.16167, "el": 1.15373, "hi": 0.96254, "ta": 0.81427, "th": 0.92232, "zh": 2.69372, "my": 0.92158}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.48553, "t": 2336.4} |
| {"event": "train", "step": 9000, "phase": 3, "objective": "global", "lm_loss": 1.19679, "bpb": 1.7266, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5449, "lr": 5.35e-05, "router_trainable": true, "sec_per_step": 0.316, "alpha": 0.03, "t": 2336.7} |
| {"event": "train", "step": 9100, "phase": 3, "objective": "global", "lm_loss": 1.16413, "bpb": 1.67948, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5181, "lr": 5.12e-05, "router_trainable": true, "sec_per_step": 0.231, "alpha": 0.03, "t": 2359.8} |
| {"event": "train", "step": 9200, "phase": 3, "objective": "global", "lm_loss": 1.08484, "bpb": 1.56509, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4841, "lr": 4.91e-05, "router_trainable": true, "sec_per_step": 0.23, "alpha": 0.03, "t": 2382.8} |
| {"event": "train", "step": 9300, "phase": 3, "objective": "global", "lm_loss": 1.12458, "bpb": 1.62242, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5053, "lr": 4.71e-05, "router_trainable": true, "sec_per_step": 0.235, "alpha": 0.03, "t": 2406.3} |
| {"event": "train", "step": 9400, "phase": 3, "objective": "global", "lm_loss": 1.0756, "bpb": 1.55177, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.494, "lr": 4.52e-05, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 2430.0} |
| {"event": "train", "step": 9500, "phase": 3, "objective": "global", "lm_loss": 1.18167, "bpb": 1.70478, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5541, "lr": 4.34e-05, "router_trainable": true, "sec_per_step": 0.238, "alpha": 0.03, "t": 2453.8} |
| {"event": "train", "step": 9600, "phase": 3, "objective": "global", "lm_loss": 1.12171, "bpb": 1.61829, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.501, "lr": 4.17e-05, "router_trainable": true, "sec_per_step": 0.242, "alpha": 0.03, "t": 2478.1} |
| {"event": "train", "step": 9700, "phase": 3, "objective": "global", "lm_loss": 1.14122, "bpb": 1.64644, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5308, "lr": 4.01e-05, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 2502.0} |
| {"event": "train", "step": 9800, "phase": 3, "objective": "global", "lm_loss": 1.05797, "bpb": 1.52632, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4918, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 2526.1} |
| {"event": "train", "step": 9900, "phase": 3, "objective": "global", "lm_loss": 1.12687, "bpb": 1.62573, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.4964, "lr": 3.72e-05, "router_trainable": true, "sec_per_step": 0.239, "alpha": 0.03, "t": 2550.0} |
| {"event": "probe", "step": 10000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.791307}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.909317}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.588669}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001015, "bpb": 1.830332}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.981202}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.15043}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.14478}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 0.946861}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.802265}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.914736}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.665265}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.901919}}, "bpb": {"en": 1.79131, "de": 1.90932, "vi": 1.58867, "tr": 1.83033, "fi": 1.9812, "ru": 1.15043, "el": 1.14478, "hi": 0.94686, "ta": 0.80227, "th": 0.91474, "zh": 2.66526, "my": 0.90192}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 2576.0} |
| {"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.79201, "de": 1.90613, "vi": 1.60282, "tr": 1.8248, "fi": 1.99412, "ru": 1.15038, "el": 1.14194, "hi": 0.95292, "ta": 0.80682, "th": 0.91327, "zh": 2.67559, "my": 0.91048}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.47261, "t": 2580.5} |
| {"event": "train", "step": 10000, "phase": 3, "objective": "global", "lm_loss": 1.15153, "bpb": 1.66131, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5021, "lr": 3.6e-05, "router_trainable": true, "sec_per_step": 0.307, "alpha": 0.03, "t": 2580.7} |
| {"event": "train", "step": 10100, "phase": 3, "objective": "global", "lm_loss": 1.21302, "bpb": 1.75002, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5082, "lr": 3.49e-05, "router_trainable": true, "sec_per_step": 0.238, "alpha": 0.03, "t": 2604.5} |
| {"event": "train", "step": 10200, "phase": 3, "objective": "global", "lm_loss": 1.12947, "bpb": 1.62948, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5243, "lr": 3.38e-05, "router_trainable": true, "sec_per_step": 0.236, "alpha": 0.03, "t": 2628.1} |
| {"event": "train", "step": 10300, "phase": 3, "objective": "global", "lm_loss": 1.11819, "bpb": 1.6132, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5113, "lr": 3.3e-05, "router_trainable": true, "sec_per_step": 0.232, "alpha": 0.03, "t": 2651.3} |
| {"event": "train", "step": 10400, "phase": 3, "objective": "global", "lm_loss": 1.11317, "bpb": 1.60596, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.509, "lr": 3.22e-05, "router_trainable": true, "sec_per_step": 0.233, "alpha": 0.03, "t": 2674.6} |
| {"event": "train", "step": 10500, "phase": 3, "objective": "global", "lm_loss": 1.07527, "bpb": 1.55128, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5048, "lr": 3.15e-05, "router_trainable": true, "sec_per_step": 0.232, "alpha": 0.03, "t": 2697.8} |
| {"event": "train", "step": 10600, "phase": 3, "objective": "global", "lm_loss": 1.17118, "bpb": 1.68965, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.509, "lr": 3.1e-05, "router_trainable": true, "sec_per_step": 0.231, "alpha": 0.03, "t": 2720.9} |
| {"event": "train", "step": 10700, "phase": 3, "objective": "global", "lm_loss": 1.14087, "bpb": 1.64592, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5157, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.233, "alpha": 0.03, "t": 2744.2} |
| {"event": "train", "step": 10800, "phase": 3, "objective": "global", "lm_loss": 1.1198, "bpb": 1.61552, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5239, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.23, "alpha": 0.03, "t": 2767.2} |
| {"event": "train", "step": 10900, "phase": 3, "objective": "global", "lm_loss": 1.18856, "bpb": 1.71473, "chunk_loss": 0.000179, "F": 0.00098, "G": 0.9999, "chunks": 1.0, "bytes_per_chunk": 1024.0, "boundary_entropy": 0.00102, "router_noise": 0.0, "grad_norm": 0.5138, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.229, "alpha": 0.03, "t": 2790.1} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.785442}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.900472}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.57901}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001015, "bpb": 1.822283}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.971093}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.145529}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.136687}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 0.941021}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.797711}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.908699}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.652856}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.894206}}, "bpb": {"en": 1.78544, "de": 1.90047, "vi": 1.57901, "tr": 1.82228, "fi": 1.97109, "ru": 1.14553, "el": 1.13669, "hi": 0.94102, "ta": 0.79771, "th": 0.9087, "zh": 2.65286, "my": 0.89421}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 2815.4} |
| {"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.78621, "de": 1.89772, "vi": 1.59361, "tr": 1.81608, "fi": 1.98427, "ru": 1.14528, "el": 1.13436, "hi": 0.94736, "ta": 0.8021, "th": 0.90697, "zh": 2.66203, "my": 0.9029}, "bytes_per_chunk": {"en": 1024.0, "de": 1024.0, "vi": 1024.0, "tr": 1024.0, "fi": 1024.0, "ru": 1024.0, "el": 1024.0, "hi": 1024.0, "ta": 1024.0, "th": 1024.0, "zh": 1024.0, "my": 1024.0}, "bpb_mean": 1.46491, "t": 2819.9} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 17418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 1.000862, "cps": 136.078125, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 1.785442}, "de": {"chunks": 20862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 1.016023, "cps": 162.984375, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.900472}, "vi": {"chunks": 24274.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.316734, "cps": 189.640625, "mean_p": 0.9999, "boundary_entropy": 0.001016, "bpb": 1.57901}, "tr": {"chunks": 19742.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 1.090537, "cps": 154.234375, "mean_p": 0.9999, "boundary_entropy": 0.001015, "bpb": 1.822283}, "fi": {"chunks": 19572.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 1.040455, "cps": 152.90625, "mean_p": 0.999901, "boundary_entropy": 0.001015, "bpb": 1.971093}, "ru": {"chunks": 34560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.823747, "cps": 270.0, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.145529}, "el": {"chunks": 37264.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.815719, "cps": 291.125, "mean_p": 0.999901, "boundary_entropy": 0.001018, "bpb": 1.136687}, "hi": {"chunks": 43173.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.566003, "cps": 337.289062, "mean_p": 0.9999, "boundary_entropy": 0.001018, "bpb": 0.941021}, "ta": {"chunks": 50915.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.723455, "cps": 397.773438, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.797711}, "th": {"chunks": 46085.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.865092, "cps": 360.039062, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.908699}, "zh": {"chunks": 16173.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.738402, "cps": 126.351562, "mean_p": 0.999901, "boundary_entropy": 0.001014, "bpb": 2.652856}, "my": {"chunks": 53743.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.836342, "cps": 419.867188, "mean_p": 0.9999, "boundary_entropy": 0.001019, "bpb": 0.894206}}, "bpb": {"en": 1.78544, "de": 1.90047, "vi": 1.57901, "tr": 1.82228, "fi": 1.97109, "ru": 1.14553, "el": 1.13669, "hi": 0.94102, "ta": 0.79771, "th": 0.9087, "zh": 2.65286, "my": 0.89421}, "gini_cpc": 0.221516, "gini_cps": 0.233853, "bpic": 1.0, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.01514801064883, "vi": 1.3156005295906583, "tr": 1.0895983329898389, "fi": 1.0395590357977051, "ru": 1.8221761311426539, "el": 1.8141552920914839, "hi": 2.563793186225859, "ta": 2.721110089193083, "th": 2.8626243461390897, "zh": 2.7360433739796948, "my": 2.8338989711748765}, "t": 2822.3} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/exp32_local_onechunk_s1/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/exp32_local_onechunk_s1/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/exp32_local_onechunk_s1/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/exp32_local_onechunk_s1/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/exp32_local_onechunk_s1/th0.70"], "t": 2837.0} |
| {"event": "final", "run": "exp32_local_onechunk_s1", "condition": "A", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 1.000861920358559, "cps": 136.078125, "bpb": 1.7801580655464224, "bytes_per_chunk": 1024.0}, "de": {"cpc": 1.0160229873861588, "cps": 162.984375, "bpb": 1.8993159442270804, "bytes_per_chunk": 1024.0}, "vi": {"cpc": 1.3167344724708434, "cps": 189.640625, "bpb": 1.5974955500390096, "bytes_per_chunk": 1024.0}, "tr": {"cpc": 1.0905374799756946, "cps": 154.234375, "bpb": 1.8139317497999687, "bytes_per_chunk": 1024.0}, "fi": {"cpc": 1.040455052894583, "cps": 152.90625, "bpb": 1.9888535841660229, "bytes_per_chunk": 1024.0}, "ru": {"cpc": 1.8237467018469657, "cps": 270.0, "bpb": 1.1453247085408835, "bytes_per_chunk": 1024.0}, "el": {"cpc": 1.8157189494713248, "cps": 291.125, "bpb": 1.1316046474579413, "bytes_per_chunk": 1024.0}, "hi": {"cpc": 2.566002971768202, "cps": 337.2890625, "bpb": 0.9489036454579403, "bytes_per_chunk": 1024.0}, "ta": {"cpc": 2.7234554693768387, "cps": 397.7734375, "bpb": 0.8009907835009732, "bytes_per_chunk": 1024.0}, "th": {"cpc": 2.8650917003419334, "cps": 360.0390625, "bpb": 0.9089351568692311, "bytes_per_chunk": 1024.0}, "zh": {"cpc": 2.7384016254656283, "cps": 126.3515625, "bpb": 2.6648868792850267, "bytes_per_chunk": 1024.0}, "my": {"cpc": 2.8363415663922313, "cps": 419.8671875, "bpb": 0.9014855832367469, "bytes_per_chunk": 1024.0}}, "gini_cpc": 0.22151585453568945, "gini_cps": 0.2338532044751218, "bpic": 1.0, "seed": 1, "mean_boundary_entropy": 0.0010205036960542202, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 7700, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 2846.6} |
|
|
