| {"event": "start", "run": "pilot_Bpc_s1", "condition": "B", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.016, "vi": 1.323, "tr": 1.088, "fi": 1.04, "ru": 1.83, "el": 1.821, "hi": 2.569, "ta": 2.733, "th": 2.89, "zh": 2.791, "my": 2.85}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "parity_a", "p3_objective": "none", "freeze_in_p3": true, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_char", "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": 1.4} |
| {"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 1.5} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2674.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.153652, "cps": 20.890625, "mean_p": 0.478126, "boundary_entropy": 0.685438, "bpb": 8.041379}, "de": {"chunks": 2056.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.100131, "cps": 16.0625, "mean_p": 0.469682, "boundary_entropy": 0.685273, "bpb": 8.067143}, "vi": {"chunks": 14185.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.76946, "cps": 110.820312, "mean_p": 0.507219, "boundary_entropy": 0.688283, "bpb": 7.894113}, "tr": {"chunks": 5645.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.311827, "cps": 44.101562, "mean_p": 0.486063, "boundary_entropy": 0.686375, "bpb": 8.01701}, "fi": {"chunks": 4732.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.251555, "cps": 36.96875, "mean_p": 0.48451, "boundary_entropy": 0.686595, "bpb": 8.075294}, "ru": {"chunks": 17548.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.926016, "cps": 137.09375, "mean_p": 0.502342, "boundary_entropy": 0.690066, "bpb": 8.095633}, "el": {"chunks": 31473.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.533548, "cps": 245.882812, "mean_p": 0.517965, "boundary_entropy": 0.689704, "bpb": 7.804942}, "hi": {"chunks": 1659.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.098603, "cps": 12.960938, "mean_p": 0.476241, "boundary_entropy": 0.689492, "bpb": 8.047677}, "ta": {"chunks": 39693.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.123188, "cps": 310.101562, "mean_p": 0.514597, "boundary_entropy": 0.690542, "bpb": 8.096414}, "th": {"chunks": 44424.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.761828, "cps": 347.0625, "mean_p": 0.52078, "boundary_entropy": 0.690224, "bpb": 8.201461}, "zh": {"chunks": 12933.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.189807, "cps": 101.039062, "mean_p": 0.528309, "boundary_entropy": 0.684752, "bpb": 8.121995}, "my": {"chunks": 7048.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.371965, "cps": 55.0625, "mean_p": 0.491145, "boundary_entropy": 0.691021, "bpb": 8.112229}}, "bpb": {"en": 8.04138, "de": 8.06714, "vi": 7.89411, "tr": 8.01701, "fi": 8.07529, "ru": 8.09563, "el": 7.80494, "hi": 8.04768, "ta": 8.09641, "th": 8.20146, "zh": 8.12199, "my": 8.11223}, "gini_cpc": 0.507116, "gini_cps": 0.500863, "bpic": 2.084973, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6516785410034256, "vi": 5.0078223656304, "tr": 2.029439514461233, "fi": 1.6371767549398106, "ru": 6.026721581613623, "el": 9.980677296913099, "hi": 0.6417324941848254, "ta": 13.818190381639045, "th": 17.9746032262692, "zh": 14.25176170627287, "my": 2.4208352616119475}, "t": 3.3} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.57556, "bpb": 8.04383, "chunk_loss": 0.043969, "F": 0.34537, "G": 0.49096, "chunks": 353.66, "bytes_per_chunk": 2.895, "boundary_entropy": 0.6904, "router_noise": 0.0, "grad_norm": 7.3219, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 3.7} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2653.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.152445, "cps": 20.726562, "mean_p": 0.478021, "boundary_entropy": 0.685429, "bpb": 8.012048}, "de": {"chunks": 2040.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.099352, "cps": 15.9375, "mean_p": 0.469566, "boundary_entropy": 0.685259, "bpb": 8.041927}, "vi": {"chunks": 14105.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.765121, "cps": 110.195312, "mean_p": 0.507011, "boundary_entropy": 0.688289, "bpb": 7.873463}, "tr": {"chunks": 5608.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.309783, "cps": 43.8125, "mean_p": 0.48594, "boundary_entropy": 0.686368, "bpb": 7.996313}, "fi": {"chunks": 4696.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.249641, "cps": 36.6875, "mean_p": 0.484352, "boundary_entropy": 0.686583, "bpb": 8.050083}, "ru": {"chunks": 17483.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.922586, "cps": 136.585938, "mean_p": 0.502276, "boundary_entropy": 0.690066, "bpb": 8.083973}, "el": {"chunks": 31423.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.531111, "cps": 245.492188, "mean_p": 0.517879, "boundary_entropy": 0.68971, "bpb": 7.792293}, "hi": {"chunks": 1660.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.098663, "cps": 12.96875, "mean_p": 0.476277, "boundary_entropy": 0.689497, "bpb": 8.024787}, "ta": {"chunks": 39579.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.11709, "cps": 309.210938, "mean_p": 0.514531, "boundary_entropy": 0.690545, "bpb": 8.080319}, "th": {"chunks": 44406.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.760709, "cps": 346.921875, "mean_p": 0.520711, "boundary_entropy": 0.69023, "bpb": 8.182836}, "zh": {"chunks": 12934.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.189976, "cps": 101.046875, "mean_p": 0.528283, "boundary_entropy": 0.684755, "bpb": 8.116862}, "my": {"chunks": 7051.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.372124, "cps": 55.085938, "mean_p": 0.49115, "boundary_entropy": 0.691021, "bpb": 8.097766}}, "bpb": {"en": 8.01205, "de": 8.04193, "vi": 7.87346, "tr": 7.99631, "fi": 8.05008, "ru": 8.08397, "el": 7.79229, "hi": 8.02479, "ta": 8.08032, "th": 8.18284, "zh": 8.11686, "my": 8.09777}, "gini_cpc": 0.507764, "gini_cps": 0.501329, "bpic": 2.089878, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.651725374774326, "vi": 5.018995643969077, "tr": 2.0320964808755795, "fi": 1.6375820717293295, "ru": 6.051926061699455, "el": 10.043698575136178, "hi": 0.647202051673818, "ta": 13.88756862068677, "th": 18.109541693376254, "zh": 14.36568317639756, "my": 2.441036137468618}, "t": 5.0} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2629.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.151066, "cps": 20.539062, "mean_p": 0.477867, "boundary_entropy": 0.685416, "bpb": 7.96192}, "de": {"chunks": 2014.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.098086, "cps": 15.734375, "mean_p": 0.469405, "boundary_entropy": 0.685239, "bpb": 7.999938}, "vi": {"chunks": 13977.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.758177, "cps": 109.195312, "mean_p": 0.506724, "boundary_entropy": 0.688298, "bpb": 7.8402}, "tr": {"chunks": 5560.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.307131, "cps": 43.4375, "mean_p": 0.485758, "boundary_entropy": 0.686358, "bpb": 7.961961}, "fi": {"chunks": 4633.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.246292, "cps": 36.195312, "mean_p": 0.484106, "boundary_entropy": 0.686566, "bpb": 8.009891}, "ru": {"chunks": 17254.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.910501, "cps": 134.796875, "mean_p": 0.502048, "boundary_entropy": 0.690066, "bpb": 8.040033}, "el": {"chunks": 31369.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.52848, "cps": 245.070312, "mean_p": 0.517784, "boundary_entropy": 0.689716, "bpb": 7.7695}, "hi": {"chunks": 1670.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.099257, "cps": 13.046875, "mean_p": 0.476369, "boundary_entropy": 0.689506, "bpb": 7.983839}, "ta": {"chunks": 39396.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.107301, "cps": 307.78125, "mean_p": 0.514418, "boundary_entropy": 0.69055, "bpb": 8.042727}, "th": {"chunks": 44380.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.759092, "cps": 346.71875, "mean_p": 0.520515, "boundary_entropy": 0.690244, "bpb": 8.123365}, "zh": {"chunks": 12931.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.189468, "cps": 101.023438, "mean_p": 0.528259, "boundary_entropy": 0.684757, "bpb": 8.106778}, "my": {"chunks": 7067.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.372968, "cps": 55.210938, "mean_p": 0.491161, "boundary_entropy": 0.691021, "bpb": 8.084102}}, "bpb": {"en": 7.96192, "de": 7.99994, "vi": 7.8402, "tr": 7.96196, "fi": 8.00989, "ru": 8.04003, "el": 7.7695, "hi": 7.98384, "ta": 8.04273, "th": 8.12336, "zh": 8.10678, "my": 8.0841}, "gini_cpc": 0.50879, "gini_cps": 0.50199, "bpic": 2.09854, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6492928091689306, "vi": 5.018851633266182, "tr": 2.033095477262132, "fi": 1.6303616489457633, "ru": 6.027179330202701, "el": 10.117969435510913, "hi": 0.6570447240375805, "ta": 13.949549854926294, "th": 18.26416267799453, "zh": 14.493464234977367, "my": 2.4689099465671416}, "t": 6.6} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2559.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.147044, "cps": 19.992188, "mean_p": 0.477481, "boundary_entropy": 0.685387, "bpb": 7.791619}, "de": {"chunks": 1969.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.095894, "cps": 15.382812, "mean_p": 0.468988, "boundary_entropy": 0.685193, "bpb": 7.852585}, "vi": {"chunks": 13540.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.734472, "cps": 105.78125, "mean_p": 0.505672, "boundary_entropy": 0.688326, "bpb": 7.729913}, "tr": {"chunks": 5386.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.29752, "cps": 42.078125, "mean_p": 0.485238, "boundary_entropy": 0.686332, "bpb": 7.842913}, "fi": {"chunks": 4455.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.23683, "cps": 34.804688, "mean_p": 0.483398, "boundary_entropy": 0.686516, "bpb": 7.870478}, "ru": {"chunks": 16103.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.849763, "cps": 125.804688, "mean_p": 0.500931, "boundary_entropy": 0.690065, "bpb": 7.865208}, "el": {"chunks": 31133.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.516981, "cps": 243.226562, "mean_p": 0.517304, "boundary_entropy": 0.689748, "bpb": 7.686026}, "hi": {"chunks": 1732.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.102942, "cps": 13.53125, "mean_p": 0.476711, "boundary_entropy": 0.689543, "bpb": 7.830747}, "ta": {"chunks": 38754.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.072961, "cps": 302.765625, "mean_p": 0.513992, "boundary_entropy": 0.69057, "bpb": 7.912894}, "th": {"chunks": 44277.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.752689, "cps": 345.914062, "mean_p": 0.519897, "boundary_entropy": 0.69029, "bpb": 7.955028}, "zh": {"chunks": 12922.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.187944, "cps": 100.953125, "mean_p": 0.528242, "boundary_entropy": 0.684757, "bpb": 8.079617}, "my": {"chunks": 7116.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.375554, "cps": 55.59375, "mean_p": 0.491186, "boundary_entropy": 0.691022, "bpb": 8.001159}}, "bpb": {"en": 7.79162, "de": 7.85258, "vi": 7.72991, "tr": 7.84291, "fi": 7.87048, "ru": 7.86521, "el": 7.68603, "hi": 7.83075, "ta": 7.91289, "th": 7.95503, "zh": 8.07962, "my": 8.00116}, "gini_cpc": 0.512195, "gini_cps": 0.503663, "bpic": 2.132756, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6521494664266466, "vi": 4.99492942949961, "tr": 2.0233435624557896, "fi": 1.610607285415738, "ru": 5.7789829470408645, "el": 10.31653749183357, "hi": 0.7000783529506035, "ta": 14.097590775849879, "th": 18.720220356320702, "zh": 14.879561349774843, "my": 2.5540323724261946}, "t": 8.2} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2300.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.132161, "cps": 17.96875, "mean_p": 0.476348, "boundary_entropy": 0.685317, "bpb": 7.297508}, "de": {"chunks": 1783.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.086836, "cps": 13.929688, "mean_p": 0.467599, "boundary_entropy": 0.685052, "bpb": 7.408754}, "vi": {"chunks": 12231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.663466, "cps": 95.554688, "mean_p": 0.502575, "boundary_entropy": 0.688389, "bpb": 7.434444}, "tr": {"chunks": 4856.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.268243, "cps": 37.9375, "mean_p": 0.483426, "boundary_entropy": 0.686243, "bpb": 7.49497}, "fi": {"chunks": 3816.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.20286, "cps": 29.8125, "mean_p": 0.480874, "boundary_entropy": 0.68633, "bpb": 7.453762}, "ru": {"chunks": 12111.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.639103, "cps": 94.617188, "mean_p": 0.496817, "boundary_entropy": 0.690021, "bpb": 7.244648}, "el": {"chunks": 30501.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.486186, "cps": 238.289062, "mean_p": 0.516119, "boundary_entropy": 0.689821, "bpb": 7.51206}, "hi": {"chunks": 1893.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.112511, "cps": 14.789062, "mean_p": 0.477706, "boundary_entropy": 0.689647, "bpb": 7.379206}, "ta": {"chunks": 34877.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.865579, "cps": 272.476562, "mean_p": 0.511352, "boundary_entropy": 0.69068, "bpb": 7.49277}, "th": {"chunks": 43416.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.699161, "cps": 339.1875, "mean_p": 0.517057, "boundary_entropy": 0.690482, "bpb": 7.519925}, "zh": {"chunks": 12916.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.186929, "cps": 100.90625, "mean_p": 0.528222, "boundary_entropy": 0.684759, "bpb": 7.972107}, "my": {"chunks": 7559.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.398934, "cps": 59.054688, "mean_p": 0.491567, "boundary_entropy": 0.691038, "bpb": 7.746756}}, "bpb": {"en": 7.29751, "de": 7.40875, "vi": 7.43444, "tr": 7.49497, "fi": 7.45376, "ru": 7.24465, "el": 7.51206, "hi": 7.37921, "ta": 7.49277, "th": 7.51992, "zh": 7.97211, "my": 7.74676}, "gini_cpc": 0.522729, "gini_cps": 0.507354, "bpic": 2.280894, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6570451595416922, "vi": 5.020131672975555, "tr": 2.0296652248366236, "fi": 1.5349448172091722, "ru": 4.8357860043592975, "el": 11.245260418321756, "hi": 0.8513180179598165, "ta": 14.115944300382571, "th": 20.423258180048922, "zh": 16.547442394617118, "my": 3.018542211636424}, "t": 10.0} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2294.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.131816, "cps": 17.921875, "mean_p": 0.478687, "boundary_entropy": 0.68575, "bpb": 6.636032}, "de": {"chunks": 1724.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.083962, "cps": 13.46875, "mean_p": 0.470355, "boundary_entropy": 0.685667, "bpb": 6.756661}, "vi": {"chunks": 9744.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.52856, "cps": 76.125, "mean_p": 0.496639, "boundary_entropy": 0.688418, "bpb": 7.048676}, "tr": {"chunks": 4411.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.243661, "cps": 34.460938, "mean_p": 0.482558, "boundary_entropy": 0.686345, "bpb": 7.015109}, "fi": {"chunks": 2971.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15794, "cps": 23.210938, "mean_p": 0.478151, "boundary_entropy": 0.686223, "bpb": 6.854119}, "ru": {"chunks": 6012.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.317256, "cps": 46.96875, "mean_p": 0.487691, "boundary_entropy": 0.689678, "bpb": 6.436077}, "el": {"chunks": 23089.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.12503, "cps": 180.382812, "mean_p": 0.506376, "boundary_entropy": 0.69025, "bpb": 6.833892}, "hi": {"chunks": 3524.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.20945, "cps": 27.53125, "mean_p": 0.483426, "boundary_entropy": 0.690144, "bpb": 6.49948}, "ta": {"chunks": 21618.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.156352, "cps": 168.890625, "mean_p": 0.498108, "boundary_entropy": 0.690847, "bpb": 6.584849}, "th": {"chunks": 24798.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.541685, "cps": 193.734375, "mean_p": 0.503052, "boundary_entropy": 0.690961, "bpb": 6.461884}, "zh": {"chunks": 13088.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.216051, "cps": 102.25, "mean_p": 0.528845, "boundary_entropy": 0.684759, "bpb": 7.681133}, "my": {"chunks": 8518.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.449546, "cps": 66.546875, "mean_p": 0.492479, "boundary_entropy": 0.691083, "bpb": 6.949098}}, "bpb": {"en": 6.63603, "de": 6.75666, "vi": 7.04868, "tr": 7.01511, "fi": 6.85412, "ru": 6.43608, "el": 6.83389, "hi": 6.49948, "ta": 6.58485, "th": 6.46188, "zh": 7.68113, "my": 6.9491}, "gini_cpc": 0.49722, "gini_cps": 0.436419, "bpic": 3.151144, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6369649876985826, "vi": 4.009819651930994, "tr": 1.848490457659674, "fi": 1.1981783692664938, "ru": 2.40680255708939, "el": 8.534832164189785, "hi": 1.5889547640916093, "ta": 8.77244693122494, "th": 11.695702077482338, "zh": 16.81165814558865, "my": 3.410397225352587}, "t": 12.0} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3528.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.202724, "cps": 27.5625, "mean_p": 0.488212, "boundary_entropy": 0.686849, "bpb": 6.226399}, "de": {"chunks": 3095.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.150733, "cps": 24.179688, "mean_p": 0.483309, "boundary_entropy": 0.687365, "bpb": 6.332563}, "vi": {"chunks": 6113.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.331598, "cps": 47.757812, "mean_p": 0.489063, "boundary_entropy": 0.688305, "bpb": 6.743216}, "tr": {"chunks": 4945.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.273159, "cps": 38.632812, "mean_p": 0.487074, "boundary_entropy": 0.687059, "bpb": 6.641879}, "fi": {"chunks": 3309.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175908, "cps": 25.851562, "mean_p": 0.483023, "boundary_entropy": 0.686939, "bpb": 6.410088}, "ru": {"chunks": 2409.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.127124, "cps": 18.820312, "mean_p": 0.482052, "boundary_entropy": 0.689463, "bpb": 5.795341}, "el": {"chunks": 1349.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.065731, "cps": 10.539062, "mean_p": 0.473771, "boundary_entropy": 0.688756, "bpb": 6.226604}, "hi": {"chunks": 5062.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.300862, "cps": 39.546875, "mean_p": 0.488277, "boundary_entropy": 0.690523, "bpb": 5.768785}, "ta": {"chunks": 4454.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.238246, "cps": 34.796875, "mean_p": 0.482275, "boundary_entropy": 0.690304, "bpb": 6.128456}, "th": {"chunks": 4780.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.297171, "cps": 37.34375, "mean_p": 0.489185, "boundary_entropy": 0.690753, "bpb": 5.554456}, "zh": {"chunks": 4818.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.815781, "cps": 37.640625, "mean_p": 0.491067, "boundary_entropy": 0.686032, "bpb": 7.308084}, "my": {"chunks": 4115.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.217173, "cps": 32.148438, "mean_p": 0.489038, "boundary_entropy": 0.690968, "bpb": 5.725688}}, "bpb": {"en": 6.2264, "de": 6.33256, "vi": 6.74322, "tr": 6.64188, "fi": 6.41009, "ru": 5.79534, "el": 6.2266, "hi": 5.76879, "ta": 6.12846, "th": 5.55446, "zh": 7.30808, "my": 5.72569}, "gini_cpc": 0.31236, "gini_cps": 0.175207, "bpic": 7.99927, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.7435390632169685, "vi": 1.6357118410152236, "tr": 1.3474457478953576, "fi": 0.8677216381209865, "ru": 0.627080259544451, "el": 0.324240042945015, "hi": 1.4840981087828888, "ta": 1.175223012446487, "th": 1.46589335143445, "zh": 4.024101225319115, "my": 1.0712775584162313}, "t": 14.4} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.18169, "bpb": 6.0329, "chunk_loss": 0.029072, "F": 0.15207, "G": 0.48523, "chunks": 155.72, "bytes_per_chunk": 6.576, "boundary_entropy": 0.69152, "router_noise": 0.0, "grad_norm": 1.6407, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 15.5} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5781.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.332184, "cps": 45.164062, "mean_p": 0.496564, "boundary_entropy": 0.687365, "bpb": 5.700477}, "de": {"chunks": 5714.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.278284, "cps": 44.640625, "mean_p": 0.493163, "boundary_entropy": 0.688116, "bpb": 5.814947}, "vi": {"chunks": 4985.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.27041, "cps": 38.945312, "mean_p": 0.485718, "boundary_entropy": 0.688067, "bpb": 6.193618}, "tr": {"chunks": 4483.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.247639, "cps": 35.023438, "mean_p": 0.487198, "boundary_entropy": 0.687238, "bpb": 6.160952}, "fi": {"chunks": 4469.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.237574, "cps": 34.914062, "mean_p": 0.489483, "boundary_entropy": 0.68755, "bpb": 5.889301}, "ru": {"chunks": 9947.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.524908, "cps": 77.710938, "mean_p": 0.494772, "boundary_entropy": 0.690226, "bpb": 5.366013}, "el": {"chunks": 2408.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.117332, "cps": 18.8125, "mean_p": 0.484198, "boundary_entropy": 0.68991, "bpb": 5.640015}, "hi": {"chunks": 8875.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.527489, "cps": 69.335938, "mean_p": 0.491102, "boundary_entropy": 0.690629, "bpb": 5.211598}, "ta": {"chunks": 9538.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.51019, "cps": 74.515625, "mean_p": 0.486802, "boundary_entropy": 0.690584, "bpb": 5.235529}, "th": {"chunks": 2625.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.163196, "cps": 20.507812, "mean_p": 0.48648, "boundary_entropy": 0.690623, "bpb": 5.049929}, "zh": {"chunks": 2581.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.437013, "cps": 20.164062, "mean_p": 0.484881, "boundary_entropy": 0.686169, "bpb": 6.981225}, "my": {"chunks": 24234.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.278974, "cps": 189.328125, "mean_p": 0.49916, "boundary_entropy": 0.691343, "bpb": 4.953697}}, "bpb": {"en": 5.70048, "de": 5.81495, "vi": 6.19362, "tr": 6.16095, "fi": 5.8893, "ru": 5.36601, "el": 5.64002, "hi": 5.2116, "ta": 5.23553, "th": 5.04993, "zh": 6.98122, "my": 4.9537}, "gini_cpc": 0.337288, "gini_cps": 0.371151, "bpic": 4.481329, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8377394739481859, "vi": 0.8140351751318008, "tr": 0.7454857361931226, "fi": 0.71518701710294, "ru": 1.5801708809543045, "el": 0.3532130885289483, "hi": 1.5879412832887774, "ta": 1.5358648437423015, "th": 0.49128034947966326, "zh": 1.3155753052800394, "my": 3.8501963530812593}, "t": 17.3} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.33873, "bpb": 4.81678, "chunk_loss": 0.032613, "F": 0.20432, "G": 0.48389, "chunks": 209.22, "bytes_per_chunk": 4.894, "boundary_entropy": 0.69111, "router_noise": 0.0, "grad_norm": 1.5821, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.077, "alpha": 0.03, "t": 19.3} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2502.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.143768, "cps": 19.546875, "mean_p": 0.473421, "boundary_entropy": 0.684907, "bpb": 4.488046}, "de": {"chunks": 2977.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.144986, "cps": 23.257812, "mean_p": 0.471911, "boundary_entropy": 0.685468, "bpb": 4.660432}, "vi": {"chunks": 4996.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.271006, "cps": 39.03125, "mean_p": 0.475315, "boundary_entropy": 0.684863, "bpb": 4.919781}, "tr": {"chunks": 3267.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180467, "cps": 25.523438, "mean_p": 0.470924, "boundary_entropy": 0.684533, "bpb": 4.975976}, "fi": {"chunks": 2443.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.129871, "cps": 19.085938, "mean_p": 0.473346, "boundary_entropy": 0.685551, "bpb": 4.730012}, "ru": {"chunks": 5281.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.278681, "cps": 41.257812, "mean_p": 0.478879, "boundary_entropy": 0.687825, "bpb": 3.669419}, "el": {"chunks": 5188.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.25279, "cps": 40.53125, "mean_p": 0.483068, "boundary_entropy": 0.689493, "bpb": 4.43296}, "hi": {"chunks": 3281.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.195007, "cps": 25.632812, "mean_p": 0.480899, "boundary_entropy": 0.68987, "bpb": 3.444862}, "ta": {"chunks": 8100.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.433271, "cps": 63.28125, "mean_p": 0.484676, "boundary_entropy": 0.690138, "bpb": 3.585848}, "th": {"chunks": 8856.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.550575, "cps": 69.1875, "mean_p": 0.480119, "boundary_entropy": 0.689187, "bpb": 3.324547}, "zh": {"chunks": 2084.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.352861, "cps": 16.28125, "mean_p": 0.481638, "boundary_entropy": 0.68584, "bpb": 6.317435}, "my": {"chunks": 14400.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.759975, "cps": 112.5, "mean_p": 0.49214, "boundary_entropy": 0.690927, "bpb": 3.982168}}, "bpb": {"en": 4.48805, "de": 4.66043, "vi": 4.91978, "tr": 4.97598, "fi": 4.73001, "ru": 3.66942, "el": 4.43296, "hi": 3.44486, "ta": 3.58585, "th": 3.32455, "zh": 6.31743, "my": 3.98217}, "gini_cpc": 0.310061, "gini_cps": 0.330542, "bpic": 6.055716, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0084706013983924, "vi": 1.8850206084115622, "tr": 1.2552649368104258, "fi": 0.9033340867574933, "ru": 1.9384016375290272, "el": 1.75831198803565, "hi": 1.3564005972042574, "ta": 3.0136746446651186, "th": 3.829599497275043, "zh": 2.4543759506448346, "my": 5.286106770062101}, "t": 21.6} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.92029, "bpb": 4.21309, "chunk_loss": 0.029975, "F": 0.16656, "G": 0.46125, "chunks": 170.56, "bytes_per_chunk": 6.004, "boundary_entropy": 0.68539, "router_noise": 0.0, "grad_norm": 1.5185, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.07, "alpha": 0.03, "t": 22.8} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3649.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.209676, "cps": 28.507812, "mean_p": 0.379493, "boundary_entropy": 0.552437, "bpb": 3.874269}, "de": {"chunks": 4060.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19773, "cps": 31.71875, "mean_p": 0.356731, "boundary_entropy": 0.548169, "bpb": 4.098596}, "vi": {"chunks": 7358.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.399132, "cps": 57.484375, "mean_p": 0.443599, "boundary_entropy": 0.577581, "bpb": 3.594872}, "tr": {"chunks": 5066.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.279843, "cps": 39.578125, "mean_p": 0.387769, "boundary_entropy": 0.577135, "bpb": 4.134979}, "fi": {"chunks": 3700.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.196693, "cps": 28.90625, "mean_p": 0.36117, "boundary_entropy": 0.568929, "bpb": 4.150109}, "ru": {"chunks": 8799.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.464327, "cps": 68.742188, "mean_p": 0.437452, "boundary_entropy": 0.66516, "bpb": 2.814162}, "el": {"chunks": 7093.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.345612, "cps": 55.414062, "mean_p": 0.43795, "boundary_entropy": 0.673735, "bpb": 2.990577}, "hi": {"chunks": 6249.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.371412, "cps": 48.820312, "mean_p": 0.476951, "boundary_entropy": 0.688071, "bpb": 2.451371}, "ta": {"chunks": 13511.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.722707, "cps": 105.554688, "mean_p": 0.435874, "boundary_entropy": 0.670558, "bpb": 2.375401}, "th": {"chunks": 12902.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.802114, "cps": 100.796875, "mean_p": 0.457657, "boundary_entropy": 0.68241, "bpb": 2.323845}, "zh": {"chunks": 2666.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.451405, "cps": 20.828125, "mean_p": 0.452252, "boundary_entropy": 0.676253, "bpb": 5.262051}, "my": {"chunks": 11190.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.590564, "cps": 87.421875, "mean_p": 0.380078, "boundary_entropy": 0.64103, "bpb": 2.453329}}, "bpb": {"en": 3.87427, "de": 4.0986, "vi": 3.59487, "tr": 4.13498, "fi": 4.15011, "ru": 2.81416, "el": 2.99058, "hi": 2.45137, "ta": 2.3754, "th": 2.32385, "zh": 5.26205, "my": 2.45333}, "gini_cpc": 0.253718, "gini_cps": 0.276754, "bpic": 4.449996, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9430264700860462, "vi": 1.90356143807916, "tr": 1.33464231736163, "fi": 0.9380804765478133, "ru": 2.214493246785363, "el": 1.6483118145878535, "hi": 1.7713554121567228, "ta": 3.4467698177557247, "th": 3.8254825841716977, "zh": 2.1528658028488152, "my": 2.81654677008163}, "t": 26.0} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.87761, "de": 4.09892, "vi": 3.59978, "tr": 4.14472, "fi": 4.15283, "ru": 2.79943, "el": 2.98327, "hi": 2.44712, "ta": 2.39258, "th": 2.3563, "zh": 5.27096, "my": 2.46954}, "bytes_per_chunk": {"en": 4.95, "de": 5.158, "vi": 3.227, "tr": 3.95, "fi": 5.295, "ru": 3.82, "el": 5.488, "hi": 8.242, "ta": 3.97, "th": 3.9, "zh": 6.201, "my": 4.831}, "bpb_mean": 3.38275, "t": 27.4} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.53182, "bpb": 3.65264, "chunk_loss": 0.032614, "F": 0.21576, "G": 0.39873, "chunks": 220.94, "bytes_per_chunk": 4.635, "boundary_entropy": 0.59634, "router_noise": 0.0, "grad_norm": 1.1924, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 27.4} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.19316, "bpb": 3.16406, "chunk_loss": 0.030804, "F": 0.18887, "G": 0.3521, "chunks": 193.41, "bytes_per_chunk": 5.295, "boundary_entropy": 0.52025, "router_noise": 0.0, "grad_norm": 1.469, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 29.7} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3005.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.172671, "cps": 23.476562, "mean_p": 0.272942, "boundary_entropy": 0.373989, "bpb": 3.580265}, "de": {"chunks": 3461.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.168558, "cps": 27.039062, "mean_p": 0.260167, "boundary_entropy": 0.38132, "bpb": 3.850527}, "vi": {"chunks": 7115.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.385951, "cps": 55.585938, "mean_p": 0.346056, "boundary_entropy": 0.449832, "bpb": 3.173907}, "tr": {"chunks": 4120.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.227587, "cps": 32.1875, "mean_p": 0.289343, "boundary_entropy": 0.427084, "bpb": 3.902776}, "fi": {"chunks": 3607.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19175, "cps": 28.179688, "mean_p": 0.265373, "boundary_entropy": 0.414802, "bpb": 3.92429}, "ru": {"chunks": 9552.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.504063, "cps": 74.625, "mean_p": 0.314456, "boundary_entropy": 0.510097, "bpb": 2.650328}, "el": {"chunks": 5350.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.260683, "cps": 41.796875, "mean_p": 0.372006, "boundary_entropy": 0.630176, "bpb": 2.844727}, "hi": {"chunks": 11789.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.700684, "cps": 92.101562, "mean_p": 0.465997, "boundary_entropy": 0.674487, "bpb": 2.12632}, "ta": {"chunks": 16602.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.888045, "cps": 129.703125, "mean_p": 0.453194, "boundary_entropy": 0.667509, "bpb": 1.967429}, "th": {"chunks": 14781.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.918931, "cps": 115.476562, "mean_p": 0.436909, "boundary_entropy": 0.667803, "bpb": 2.102517}, "zh": {"chunks": 2984.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.505249, "cps": 23.3125, "mean_p": 0.419983, "boundary_entropy": 0.653069, "bpb": 4.987774}, "my": {"chunks": 2480.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.130885, "cps": 19.375, "mean_p": 0.325619, "boundary_entropy": 0.603759, "bpb": 2.255813}}, "bpb": {"en": 3.58027, "de": 3.85053, "vi": 3.17391, "tr": 3.90278, "fi": 3.92429, "ru": 2.65033, "el": 2.84473, "hi": 2.12632, "ta": 1.96743, "th": 2.10252, "zh": 4.98777, "my": 2.25581}, "gini_cpc": 0.352159, "gini_cps": 0.358701, "bpic": 4.523266, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9761775958558006, "vi": 2.235174356815126, "tr": 1.318033074375755, "fi": 1.110488083980317, "ru": 2.9192060023092554, "el": 1.509706694076359, "hi": 4.057901852534038, "ta": 5.1429770210031505, "th": 5.321847133856793, "zh": 2.9260720787625094, "my": 0.7579978060631478}, "t": 30.9} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.1866, "bpb": 3.15459, "chunk_loss": 0.034413, "F": 0.2529, "G": 0.32753, "chunks": 258.97, "bytes_per_chunk": 3.954, "boundary_entropy": 0.45784, "router_noise": 0.0, "grad_norm": 0.9649, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.071, "alpha": 0.03, "t": 33.2} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.43979, "bpb": 3.51988, "chunk_loss": 0.034465, "F": 0.24164, "G": 0.3475, "chunks": 247.44, "bytes_per_chunk": 4.138, "boundary_entropy": 0.42714, "router_noise": 0.0, "grad_norm": 0.8659, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.054, "alpha": 0.03, "t": 35.9} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2991.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.171867, "cps": 23.367188, "mean_p": 0.284686, "boundary_entropy": 0.374231, "bpb": 3.19138}, "de": {"chunks": 3609.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.175766, "cps": 28.195312, "mean_p": 0.281953, "boundary_entropy": 0.395739, "bpb": 3.448875}, "vi": {"chunks": 4898.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.26569, "cps": 38.265625, "mean_p": 0.302112, "boundary_entropy": 0.385077, "bpb": 2.772838}, "tr": {"chunks": 4186.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.231232, "cps": 32.703125, "mean_p": 0.294613, "boundary_entropy": 0.401662, "bpb": 3.436372}, "fi": {"chunks": 3632.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.193079, "cps": 28.375, "mean_p": 0.281868, "boundary_entropy": 0.401486, "bpb": 3.613311}, "ru": {"chunks": 7541.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.397942, "cps": 58.914062, "mean_p": 0.269862, "boundary_entropy": 0.282063, "bpb": 2.541681}, "el": {"chunks": 8371.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.407884, "cps": 65.398438, "mean_p": 0.279634, "boundary_entropy": 0.341238, "bpb": 2.558248}, "hi": {"chunks": 5028.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.298841, "cps": 39.28125, "mean_p": 0.357215, "boundary_entropy": 0.566982, "bpb": 1.936608}, "ta": {"chunks": 8292.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.443541, "cps": 64.78125, "mean_p": 0.314749, "boundary_entropy": 0.504769, "bpb": 1.72459}, "th": {"chunks": 4544.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.282499, "cps": 35.5, "mean_p": 0.363775, "boundary_entropy": 0.615027, "bpb": 1.942939}, "zh": {"chunks": 3324.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.562817, "cps": 25.96875, "mean_p": 0.305934, "boundary_entropy": 0.489612, "bpb": 4.164096}, "my": {"chunks": 3182.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.167933, "cps": 24.859375, "mean_p": 0.356694, "boundary_entropy": 0.60593, "bpb": 2.013164}}, "bpb": {"en": 3.19138, "de": 3.44887, "vi": 2.77284, "tr": 3.43637, "fi": 3.61331, "ru": 2.54168, "el": 2.55825, "hi": 1.93661, "ta": 1.72459, "th": 1.94294, "zh": 4.1641, "my": 2.01316}, "gini_cpc": 0.223611, "gini_cps": 0.204062, "bpic": 6.439495, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0226856969877147, "vi": 1.5459069369758311, "tr": 1.3454153544985616, "fi": 1.1234187384864711, "ru": 2.315407488957638, "el": 2.373253903893623, "hi": 1.7387930805790328, "ta": 2.580723824321319, "th": 1.6437091192757631, "zh": 3.274728350306592, "my": 0.9771123588201802}, "t": 39.1} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.20097, "de": 3.44755, "vi": 2.77922, "tr": 3.44776, "fi": 3.60725, "ru": 2.52839, "el": 2.55322, "hi": 1.93435, "ta": 1.73892, "th": 1.97326, "zh": 4.17443, "my": 2.02745}, "bytes_per_chunk": {"en": 5.669, "de": 5.851, "vi": 5.16, "tr": 4.668, "fi": 5.413, "ru": 4.383, "el": 4.259, "hi": 7.365, "ta": 6.014, "th": 8.481, "zh": 3.934, "my": 16.329}, "bpb_mean": 2.7844, "t": 40.5} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.01026, "bpb": 2.9002, "chunk_loss": 0.030376, "F": 0.18158, "G": 0.29811, "chunks": 185.94, "bytes_per_chunk": 5.507, "boundary_entropy": 0.40475, "router_noise": 0.0, "grad_norm": 1.2577, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 40.5} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.07249, "bpb": 2.98997, "chunk_loss": 0.029931, "F": 0.15988, "G": 0.26561, "chunks": 163.72, "bytes_per_chunk": 6.255, "boundary_entropy": 0.37141, "router_noise": 0.0, "grad_norm": 0.9185, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 42.7} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.06174, "bpb": 2.97447, "chunk_loss": 0.030089, "F": 0.17032, "G": 0.27989, "chunks": 174.41, "bytes_per_chunk": 5.871, "boundary_entropy": 0.38507, "router_noise": 0.0, "grad_norm": 0.9493, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 44.8} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.95965, "bpb": 2.82718, "chunk_loss": 0.029643, "F": 0.15408, "G": 0.28517, "chunks": 157.78, "bytes_per_chunk": 6.49, "boundary_entropy": 0.4153, "router_noise": 0.0, "grad_norm": 0.8026, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 46.8} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2893.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166236, "cps": 22.601562, "mean_p": 0.284392, "boundary_entropy": 0.363517, "bpb": 2.913587}, "de": {"chunks": 3349.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.163103, "cps": 26.164062, "mean_p": 0.292565, "boundary_entropy": 0.405108, "bpb": 3.173822}, "vi": {"chunks": 4459.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.241877, "cps": 34.835938, "mean_p": 0.29768, "boundary_entropy": 0.380427, "bpb": 2.513935}, "tr": {"chunks": 3850.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212672, "cps": 30.078125, "mean_p": 0.311518, "boundary_entropy": 0.399462, "bpb": 3.148599}, "fi": {"chunks": 3971.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.2111, "cps": 31.023438, "mean_p": 0.347707, "boundary_entropy": 0.464396, "bpb": 3.378996}, "ru": {"chunks": 6871.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.362586, "cps": 53.679688, "mean_p": 0.242778, "boundary_entropy": 0.258415, "bpb": 2.152983}, "el": {"chunks": 7395.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.360327, "cps": 57.773438, "mean_p": 0.231106, "boundary_entropy": 0.229626, "bpb": 2.401662}, "hi": {"chunks": 6197.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.368321, "cps": 48.414062, "mean_p": 0.345946, "boundary_entropy": 0.510466, "bpb": 1.734768}, "ta": {"chunks": 9509.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.508639, "cps": 74.289062, "mean_p": 0.253296, "boundary_entropy": 0.352966, "bpb": 1.440189}, "th": {"chunks": 6099.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.379173, "cps": 47.648438, "mean_p": 0.361553, "boundary_entropy": 0.571118, "bpb": 1.794694}, "zh": {"chunks": 3252.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.550626, "cps": 25.40625, "mean_p": 0.344108, "boundary_entropy": 0.5617, "bpb": 3.861161}, "my": {"chunks": 10945.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.577634, "cps": 85.507812, "mean_p": 0.399629, "boundary_entropy": 0.598309, "bpb": 1.794384}}, "bpb": {"en": 2.91359, "de": 3.17382, "vi": 2.51393, "tr": 3.1486, "fi": 3.379, "ru": 2.15298, "el": 2.40166, "hi": 1.73477, "ta": 1.44019, "th": 1.79469, "zh": 3.86116, "my": 1.79438}, "gini_cpc": 0.229876, "gini_cps": 0.238201, "bpic": 5.579023, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9811568199027209, "vi": 1.455023531168779, "tr": 1.2793396636777847, "fi": 1.2698829440665835, "ru": 2.181154456166144, "el": 2.1675694417461098, "hi": 2.215654859888782, "ta": 3.059743807090498, "th": 2.2809368137060493, "zh": 3.312323767132243, "my": 3.4747861066228087}, "t": 49.8} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.92233, "de": 3.1728, "vi": 2.52017, "tr": 3.15878, "fi": 3.3756, "ru": 2.13865, "el": 2.39445, "hi": 1.72954, "ta": 1.45242, "th": 1.8223, "zh": 3.87175, "my": 1.80763}, "bytes_per_chunk": {"en": 5.708, "de": 6.224, "vi": 5.431, "tr": 5.106, "fi": 5.324, "ru": 5.115, "el": 5.144, "hi": 6.138, "ta": 4.993, "th": 6.917, "zh": 4.687, "my": 5.123}, "bpb_mean": 2.53053, "t": 51.1} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.86436, "bpb": 2.68971, "chunk_loss": 0.03057, "F": 0.18747, "G": 0.30574, "chunks": 191.97, "bytes_per_chunk": 5.334, "boundary_entropy": 0.41287, "router_noise": 0.0, "grad_norm": 0.6338, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 51.2} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.8879, "bpb": 2.72366, "chunk_loss": 0.030236, "F": 0.17795, "G": 0.28098, "chunks": 182.22, "bytes_per_chunk": 5.62, "boundary_entropy": 0.35879, "router_noise": 0.0, "grad_norm": 0.8201, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 53.2} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.83732, "bpb": 2.6507, "chunk_loss": 0.03008, "F": 0.17413, "G": 0.28177, "chunks": 178.31, "bytes_per_chunk": 5.743, "boundary_entropy": 0.36369, "router_noise": 0.0, "grad_norm": 0.7485, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 55.2} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.91369, "bpb": 2.76087, "chunk_loss": 0.030221, "F": 0.17548, "G": 0.28778, "chunks": 179.69, "bytes_per_chunk": 5.699, "boundary_entropy": 0.37406, "router_noise": 0.0, "grad_norm": 0.8161, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 57.2} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2885.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165776, "cps": 22.539062, "mean_p": 0.275311, "boundary_entropy": 0.335136, "bpb": 2.727006}, "de": {"chunks": 3292.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160327, "cps": 25.71875, "mean_p": 0.282819, "boundary_entropy": 0.379898, "bpb": 3.031442}, "vi": {"chunks": 4528.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.24562, "cps": 35.375, "mean_p": 0.282647, "boundary_entropy": 0.345735, "bpb": 2.374767}, "tr": {"chunks": 3840.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.21212, "cps": 30.0, "mean_p": 0.290278, "boundary_entropy": 0.362983, "bpb": 2.984424}, "fi": {"chunks": 3164.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.168199, "cps": 24.71875, "mean_p": 0.303469, "boundary_entropy": 0.40982, "bpb": 3.23011}, "ru": {"chunks": 6729.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.355092, "cps": 52.570312, "mean_p": 0.234888, "boundary_entropy": 0.262845, "bpb": 1.996002}, "el": {"chunks": 6805.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.331579, "cps": 53.164062, "mean_p": 0.219463, "boundary_entropy": 0.295075, "bpb": 2.126027}, "hi": {"chunks": 7723.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.459019, "cps": 60.335938, "mean_p": 0.37877, "boundary_entropy": 0.546859, "bpb": 1.617765}, "ta": {"chunks": 9293.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.497085, "cps": 72.601562, "mean_p": 0.245305, "boundary_entropy": 0.338397, "bpb": 1.336715}, "th": {"chunks": 6943.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.431644, "cps": 54.242188, "mean_p": 0.380814, "boundary_entropy": 0.575579, "bpb": 1.59399}, "zh": {"chunks": 2895.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.490179, "cps": 22.617188, "mean_p": 0.324807, "boundary_entropy": 0.541688, "bpb": 3.705382}, "my": {"chunks": 9114.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.481001, "cps": 71.203125, "mean_p": 0.371455, "boundary_entropy": 0.570203, "bpb": 1.560895}}, "bpb": {"en": 2.72701, "de": 3.03144, "vi": 2.37477, "tr": 2.98442, "fi": 3.23011, "ru": 1.996, "el": 2.12603, "hi": 1.61776, "ta": 1.33671, "th": 1.59399, "zh": 3.70538, "my": 1.56089}, "gini_cpc": 0.221582, "gini_cps": 0.232753, "bpic": 5.710092, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9671319305837387, "vi": 1.4816361951848243, "tr": 1.2795550513904117, "fi": 1.0146187741657482, "ru": 2.1420007407983244, "el": 2.0001639112933356, "hi": 2.7689127088156447, "ta": 2.9985325690193254, "th": 2.60378069495407, "zh": 2.9568781505281176, "my": 2.9015091929055856}, "t": 60.1} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.73824, "de": 3.0305, "vi": 2.37946, "tr": 2.99571, "fi": 3.22523, "ru": 1.97969, "el": 2.11381, "hi": 1.61251, "ta": 1.34689, "th": 1.62495, "zh": 3.71715, "my": 1.57047}, "bytes_per_chunk": {"en": 5.735, "de": 6.273, "vi": 5.425, "tr": 5.227, "fi": 6.186, "ru": 5.223, "el": 5.549, "hi": 5.557, "ta": 5.49, "th": 6.632, "zh": 6.067, "my": 6.175}, "bpb_mean": 2.36122, "t": 61.4} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.64229, "bpb": 2.36932, "chunk_loss": 0.030234, "F": 0.17737, "G": 0.2754, "chunks": 181.62, "bytes_per_chunk": 5.638, "boundary_entropy": 0.34365, "router_noise": 0.0, "grad_norm": 0.7363, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 61.5} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.71272, "bpb": 2.47094, "chunk_loss": 0.029774, "F": 0.16595, "G": 0.28749, "chunks": 169.94, "bytes_per_chunk": 6.026, "boundary_entropy": 0.38245, "router_noise": 0.0, "grad_norm": 0.7279, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 63.5} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.61987, "bpb": 2.33698, "chunk_loss": 0.030293, "F": 0.18225, "G": 0.26175, "chunks": 186.62, "bytes_per_chunk": 5.487, "boundary_entropy": 0.31225, "router_noise": 0.0, "grad_norm": 0.671, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 65.5} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.71863, "bpb": 2.47946, "chunk_loss": 0.030362, "F": 0.18121, "G": 0.27163, "chunks": 185.56, "bytes_per_chunk": 5.518, "boundary_entropy": 0.34383, "router_noise": 0.0, "grad_norm": 0.7072, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 67.6} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2873.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165086, "cps": 22.445312, "mean_p": 0.247162, "boundary_entropy": 0.27908, "bpb": 2.587556}, "de": {"chunks": 3291.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160279, "cps": 25.710938, "mean_p": 0.26407, "boundary_entropy": 0.339862, "bpb": 2.911987}, "vi": {"chunks": 4319.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234283, "cps": 33.742188, "mean_p": 0.255245, "boundary_entropy": 0.291519, "bpb": 2.281252}, "tr": {"chunks": 3871.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213832, "cps": 30.242188, "mean_p": 0.266894, "boundary_entropy": 0.321115, "bpb": 2.88652}, "fi": {"chunks": 3109.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.165276, "cps": 24.289062, "mean_p": 0.273086, "boundary_entropy": 0.357387, "bpb": 3.179665}, "ru": {"chunks": 6624.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.349551, "cps": 51.75, "mean_p": 0.241017, "boundary_entropy": 0.272028, "bpb": 1.885172}, "el": {"chunks": 7297.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.355552, "cps": 57.007812, "mean_p": 0.229077, "boundary_entropy": 0.28209, "bpb": 2.008738}, "hi": {"chunks": 5909.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.351204, "cps": 46.164062, "mean_p": 0.360143, "boundary_entropy": 0.526781, "bpb": 1.557662}, "ta": {"chunks": 9513.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.508853, "cps": 74.320312, "mean_p": 0.263148, "boundary_entropy": 0.375115, "bpb": 1.25378}, "th": {"chunks": 7572.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.470749, "cps": 59.15625, "mean_p": 0.394753, "boundary_entropy": 0.584892, "bpb": 1.498537}, "zh": {"chunks": 2874.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.486624, "cps": 22.453125, "mean_p": 0.328755, "boundary_entropy": 0.541018, "bpb": 3.643526}, "my": {"chunks": 9709.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.512402, "cps": 75.851562, "mean_p": 0.400632, "boundary_entropy": 0.570057, "bpb": 1.412433}}, "bpb": {"en": 2.58756, "de": 2.91199, "vi": 2.28125, "tr": 2.88652, "fi": 3.17967, "ru": 1.88517, "el": 2.00874, "hi": 1.55766, "ta": 1.25378, "th": 1.49854, "zh": 3.64353, "my": 1.41243}, "gini_cpc": 0.227889, "gini_cps": 0.244006, "bpic": 5.731411, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9708764557165532, "vi": 1.419150832489124, "tr": 1.2952724075574893, "fi": 1.0011458070565202, "ru": 2.1173838861862837, "el": 2.1537336587668596, "hi": 2.1273914588299427, "ta": 3.082339852545306, "th": 2.851530585913397, "zh": 2.9476900493153617, "my": 3.103842097826348}, "t": 70.5} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.60037, "de": 2.91507, "vi": 2.28702, "tr": 2.89687, "fi": 3.17917, "ru": 1.8682, "el": 1.99804, "hi": 1.55379, "ta": 1.26247, "th": 1.52245, "zh": 3.65238, "my": 1.42177}, "bytes_per_chunk": {"en": 5.747, "de": 6.247, "vi": 5.611, "tr": 5.118, "fi": 6.185, "ru": 5.31, "el": 5.221, "hi": 6.839, "ta": 5.414, "th": 6.104, "zh": 6.101, "my": 5.705}, "bpb_mean": 2.26313, "t": 71.8} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.73913, "bpb": 2.50904, "chunk_loss": 0.030122, "F": 0.17529, "G": 0.28063, "chunks": 179.5, "bytes_per_chunk": 5.705, "boundary_entropy": 0.3598, "router_noise": 0.0, "grad_norm": 0.7155, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 71.8} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.52087, "bpb": 2.19414, "chunk_loss": 0.030104, "F": 0.17526, "G": 0.28142, "chunks": 179.47, "bytes_per_chunk": 5.706, "boundary_entropy": 0.3451, "router_noise": 0.0, "grad_norm": 0.7436, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 73.8} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.77216, "bpb": 2.55668, "chunk_loss": 0.030191, "F": 0.17395, "G": 0.28339, "chunks": 178.12, "bytes_per_chunk": 5.749, "boundary_entropy": 0.35904, "router_noise": 0.0, "grad_norm": 0.7831, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 75.8} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.66953, "bpb": 2.40862, "chunk_loss": 0.029977, "F": 0.16995, "G": 0.26169, "chunks": 174.03, "bytes_per_chunk": 5.884, "boundary_entropy": 0.33166, "router_noise": 0.0, "grad_norm": 0.6999, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 77.8} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.251418, "boundary_entropy": 0.283158, "bpb": 2.49812}, "de": {"chunks": 3312.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161301, "cps": 25.875, "mean_p": 0.265041, "boundary_entropy": 0.334205, "bpb": 2.794323}, "vi": {"chunks": 4359.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.236452, "cps": 34.054688, "mean_p": 0.255529, "boundary_entropy": 0.288262, "bpb": 2.224868}, "tr": {"chunks": 3902.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215544, "cps": 30.484375, "mean_p": 0.265366, "boundary_entropy": 0.30853, "bpb": 2.793896}, "fi": {"chunks": 3122.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.165967, "cps": 24.390625, "mean_p": 0.285599, "boundary_entropy": 0.368942, "bpb": 3.052971}, "ru": {"chunks": 6321.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.333562, "cps": 49.382812, "mean_p": 0.240728, "boundary_entropy": 0.280981, "bpb": 1.815067}, "el": {"chunks": 7074.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.344686, "cps": 55.265625, "mean_p": 0.224975, "boundary_entropy": 0.267008, "bpb": 1.890978}, "hi": {"chunks": 7344.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.436493, "cps": 57.375, "mean_p": 0.379269, "boundary_entropy": 0.535891, "bpb": 1.533858}, "ta": {"chunks": 9324.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.498743, "cps": 72.84375, "mean_p": 0.27683, "boundary_entropy": 0.402126, "bpb": 1.164737}, "th": {"chunks": 9197.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.571775, "cps": 71.851562, "mean_p": 0.415235, "boundary_entropy": 0.584787, "bpb": 1.422976}, "zh": {"chunks": 2969.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.502709, "cps": 23.195312, "mean_p": 0.326956, "boundary_entropy": 0.532012, "bpb": 3.574577}, "my": {"chunks": 9144.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.482584, "cps": 71.4375, "mean_p": 0.385034, "boundary_entropy": 0.567518, "bpb": 1.356536}}, "bpb": {"en": 2.49812, "de": 2.79432, "vi": 2.22487, "tr": 2.7939, "fi": 3.05297, "ru": 1.81507, "el": 1.89098, "hi": 1.53386, "ta": 1.16474, "th": 1.42298, "zh": 3.57458, "my": 1.35654}, "gini_cpc": 0.239328, "gini_cps": 0.242423, "bpic": 5.56535, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9709881940306735, "vi": 1.4233763828654848, "tr": 1.297516060381873, "fi": 0.9990725894114552, "ru": 2.0079486606144625, "el": 2.0749146586058553, "hi": 2.6275659409254417, "ta": 3.0022912731259472, "th": 3.441922994916211, "zh": 3.0261662506209643, "my": 2.905018281934128}, "t": 80.6} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.50843, "de": 2.79607, "vi": 2.23031, "tr": 2.80372, "fi": 3.05837, "ru": 1.79913, "el": 1.87811, "hi": 1.52973, "ta": 1.17453, "th": 1.44483, "zh": 3.58232, "my": 1.36587}, "bytes_per_chunk": {"en": 5.722, "de": 6.227, "vi": 5.589, "tr": 5.049, "fi": 6.205, "ru": 5.587, "el": 5.37, "hi": 5.667, "ta": 5.374, "th": 4.872, "zh": 5.752, "my": 5.918}, "bpb_mean": 2.18095, "t": 81.9} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.61209, "bpb": 2.32576, "chunk_loss": 0.030393, "F": 0.18036, "G": 0.27106, "chunks": 184.69, "bytes_per_chunk": 5.545, "boundary_entropy": 0.3301, "router_noise": 0.0, "grad_norm": 0.7016, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 82.0} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.63737, "bpb": 2.36222, "chunk_loss": 0.030331, "F": 0.17892, "G": 0.27155, "chunks": 183.22, "bytes_per_chunk": 5.589, "boundary_entropy": 0.32927, "router_noise": 0.0, "grad_norm": 0.7647, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 83.9} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.56975, "bpb": 2.26467, "chunk_loss": 0.030117, "F": 0.17239, "G": 0.26846, "chunks": 176.53, "bytes_per_chunk": 5.801, "boundary_entropy": 0.33857, "router_noise": 0.0, "grad_norm": 0.7057, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 85.9} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.71264, "bpb": 2.47082, "chunk_loss": 0.030031, "F": 0.16937, "G": 0.26788, "chunks": 173.44, "bytes_per_chunk": 5.904, "boundary_entropy": 0.33275, "router_noise": 0.0, "grad_norm": 0.7305, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 87.8} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2899.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.16658, "cps": 22.648438, "mean_p": 0.241465, "boundary_entropy": 0.26232, "bpb": 2.443563}, "de": {"chunks": 3324.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161886, "cps": 25.96875, "mean_p": 0.260722, "boundary_entropy": 0.324548, "bpb": 2.736152}, "vi": {"chunks": 4364.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.236724, "cps": 34.09375, "mean_p": 0.250118, "boundary_entropy": 0.275898, "bpb": 2.186159}, "tr": {"chunks": 3927.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216925, "cps": 30.679688, "mean_p": 0.253865, "boundary_entropy": 0.28007, "bpb": 2.729083}, "fi": {"chunks": 3077.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.163575, "cps": 24.039062, "mean_p": 0.266535, "boundary_entropy": 0.342004, "bpb": 2.994642}, "ru": {"chunks": 6201.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.32723, "cps": 48.445312, "mean_p": 0.235062, "boundary_entropy": 0.268042, "bpb": 1.741605}, "el": {"chunks": 7110.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.346441, "cps": 55.546875, "mean_p": 0.233849, "boundary_entropy": 0.285889, "bpb": 1.833287}, "hi": {"chunks": 7086.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.421159, "cps": 55.359375, "mean_p": 0.342275, "boundary_entropy": 0.503128, "bpb": 1.423506}, "ta": {"chunks": 9095.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.486494, "cps": 71.054688, "mean_p": 0.259703, "boundary_entropy": 0.375322, "bpb": 1.130688}, "th": {"chunks": 7121.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.442711, "cps": 55.632812, "mean_p": 0.385202, "boundary_entropy": 0.578215, "bpb": 1.352441}, "zh": {"chunks": 3086.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.522519, "cps": 24.109375, "mean_p": 0.311861, "boundary_entropy": 0.506199, "bpb": 3.554374}, "my": {"chunks": 7523.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.397034, "cps": 58.773438, "mean_p": 0.34245, "boundary_entropy": 0.535149, "bpb": 1.368517}}, "bpb": {"en": 2.44356, "de": 2.73615, "vi": 2.18616, "tr": 2.72908, "fi": 2.99464, "ru": 1.7416, "el": 1.83329, "hi": 1.42351, "ta": 1.13069, "th": 1.35244, "zh": 3.55437, "my": 1.36852}, "gini_cpc": 0.221281, "gini_cps": 0.214282, "bpic": 5.921358, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9718170467291591, "vi": 1.4210766541926443, "tr": 1.3022256778062185, "fi": 0.9819548458001343, "ru": 1.9643931990013843, "el": 2.079719004466455, "hi": 2.528261433593692, "ta": 2.920472618265966, "th": 2.657638003007265, "zh": 3.1367390018651413, "my": 2.3834365260148456}, "t": 90.7} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.45459, "de": 2.73842, "vi": 2.19073, "tr": 2.73546, "fi": 3.001, "ru": 1.72038, "el": 1.82291, "hi": 1.42086, "ta": 1.13866, "th": 1.37377, "zh": 3.55981, "my": 1.37754}, "bytes_per_chunk": {"en": 5.729, "de": 6.208, "vi": 5.574, "tr": 5.019, "fi": 6.226, "ru": 5.654, "el": 5.374, "hi": 5.879, "ta": 5.599, "th": 6.244, "zh": 5.8, "my": 7.254}, "bpb_mean": 2.12784, "t": 92.0} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.57008, "bpb": 2.26514, "chunk_loss": 0.030127, "F": 0.17389, "G": 0.26129, "chunks": 178.06, "bytes_per_chunk": 5.751, "boundary_entropy": 0.30384, "router_noise": 0.0, "grad_norm": 0.7842, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 92.0} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.72425, "bpb": 2.48757, "chunk_loss": 0.030184, "F": 0.17508, "G": 0.24975, "chunks": 179.28, "bytes_per_chunk": 5.712, "boundary_entropy": 0.29266, "router_noise": 0.0, "grad_norm": 0.7783, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 94.0} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.54233, "bpb": 2.22511, "chunk_loss": 0.030307, "F": 0.17801, "G": 0.2747, "chunks": 182.28, "bytes_per_chunk": 5.618, "boundary_entropy": 0.34948, "router_noise": 0.0, "grad_norm": 0.7148, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 95.9} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.60688, "bpb": 2.31824, "chunk_loss": 0.030082, "F": 0.17459, "G": 0.2624, "chunks": 178.78, "bytes_per_chunk": 5.728, "boundary_entropy": 0.32444, "router_noise": 0.0, "grad_norm": 0.7034, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 97.9} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.230462, "boundary_entropy": 0.236132, "bpb": 2.373206}, "de": {"chunks": 3311.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161253, "cps": 25.867188, "mean_p": 0.245645, "boundary_entropy": 0.29326, "bpb": 2.665528}, "vi": {"chunks": 4325.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234608, "cps": 33.789062, "mean_p": 0.242318, "boundary_entropy": 0.253808, "bpb": 2.142122}, "tr": {"chunks": 3915.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216262, "cps": 30.585938, "mean_p": 0.249114, "boundary_entropy": 0.268875, "bpb": 2.645692}, "fi": {"chunks": 3097.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.164638, "cps": 24.195312, "mean_p": 0.254763, "boundary_entropy": 0.319761, "bpb": 2.89818}, "ru": {"chunks": 6270.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.330871, "cps": 48.984375, "mean_p": 0.237866, "boundary_entropy": 0.271485, "bpb": 1.661164}, "el": {"chunks": 7043.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.343176, "cps": 55.023438, "mean_p": 0.227268, "boundary_entropy": 0.252755, "bpb": 1.789117}, "hi": {"chunks": 7650.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.454681, "cps": 59.765625, "mean_p": 0.368954, "boundary_entropy": 0.520884, "bpb": 1.389987}, "ta": {"chunks": 9817.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.525114, "cps": 76.695312, "mean_p": 0.293091, "boundary_entropy": 0.409749, "bpb": 1.153184}, "th": {"chunks": 6080.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.377992, "cps": 47.5, "mean_p": 0.38086, "boundary_entropy": 0.58696, "bpb": 1.325143}, "zh": {"chunks": 2873.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.486454, "cps": 22.445312, "mean_p": 0.295297, "boundary_entropy": 0.48052, "bpb": 3.50431}, "my": {"chunks": 7888.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.416297, "cps": 61.625, "mean_p": 0.349326, "boundary_entropy": 0.541422, "bpb": 1.269219}}, "bpb": {"en": 2.37321, "de": 2.66553, "vi": 2.14212, "tr": 2.64569, "fi": 2.89818, "ru": 1.66116, "el": 1.78912, "hi": 1.38999, "ta": 1.15318, "th": 1.32514, "zh": 3.50431, "my": 1.26922}, "gini_cpc": 0.222683, "gini_cps": 0.231108, "bpic": 5.889825, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9706950212667754, "vi": 1.4122741123866076, "tr": 1.3018388970771484, "fi": 0.9910723284456365, "ru": 1.9917478408563014, "el": 2.0658218745491994, "hi": 2.7370478551306685, "ta": 3.161035331218085, "th": 2.27540413277053, "zh": 2.928317830257336, "my": 2.50599127382944}, "t": 100.8} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.38374, "de": 2.66817, "vi": 2.14558, "tr": 2.65192, "fi": 2.90433, "ru": 1.64168, "el": 1.77862, "hi": 1.38998, "ta": 1.16025, "th": 1.34417, "zh": 3.50807, "my": 1.27729}, "bytes_per_chunk": {"en": 5.735, "de": 6.248, "vi": 5.638, "tr": 5.037, "fi": 6.191, "ru": 5.596, "el": 5.418, "hi": 5.472, "ta": 5.221, "th": 7.455, "zh": 6.504, "my": 6.634}, "bpb_mean": 2.07115, "t": 102.0} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.57663, "bpb": 2.2746, "chunk_loss": 0.030125, "F": 0.17056, "G": 0.25469, "chunks": 174.66, "bytes_per_chunk": 5.863, "boundary_entropy": 0.32987, "router_noise": 0.0, "grad_norm": 0.7828, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 102.1} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.50022, "bpb": 2.16436, "chunk_loss": 0.030409, "F": 0.18256, "G": 0.27421, "chunks": 186.94, "bytes_per_chunk": 5.478, "boundary_entropy": 0.33368, "router_noise": 0.0, "grad_norm": 0.7028, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 104.0} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.42731, "bpb": 2.05917, "chunk_loss": 0.030001, "F": 0.16867, "G": 0.25536, "chunks": 172.72, "bytes_per_chunk": 5.929, "boundary_entropy": 0.32128, "router_noise": 0.0, "grad_norm": 0.6574, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 106.0} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.6213, "bpb": 2.33904, "chunk_loss": 0.03019, "F": 0.1759, "G": 0.25683, "chunks": 180.12, "bytes_per_chunk": 5.685, "boundary_entropy": 0.30199, "router_noise": 0.0, "grad_norm": 0.8106, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 107.9} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2897.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166466, "cps": 22.632812, "mean_p": 0.231212, "boundary_entropy": 0.236186, "bpb": 2.331455}, "de": {"chunks": 3311.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161253, "cps": 25.867188, "mean_p": 0.251801, "boundary_entropy": 0.30358, "bpb": 2.598582}, "vi": {"chunks": 4498.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.243992, "cps": 35.140625, "mean_p": 0.245975, "boundary_entropy": 0.255857, "bpb": 2.119074}, "tr": {"chunks": 3898.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215323, "cps": 30.453125, "mean_p": 0.246058, "boundary_entropy": 0.269583, "bpb": 2.630353}, "fi": {"chunks": 3064.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162883, "cps": 23.9375, "mean_p": 0.253422, "boundary_entropy": 0.317378, "bpb": 2.847679}, "ru": {"chunks": 6354.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.335303, "cps": 49.640625, "mean_p": 0.244623, "boundary_entropy": 0.274277, "bpb": 1.611162}, "el": {"chunks": 7229.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.352239, "cps": 56.476562, "mean_p": 0.23383, "boundary_entropy": 0.252783, "bpb": 1.715836}, "hi": {"chunks": 7356.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.437207, "cps": 57.46875, "mean_p": 0.370435, "boundary_entropy": 0.522437, "bpb": 1.359519}, "ta": {"chunks": 8695.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.465098, "cps": 67.929688, "mean_p": 0.300096, "boundary_entropy": 0.427053, "bpb": 1.090741}, "th": {"chunks": 8451.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.525396, "cps": 66.023438, "mean_p": 0.396864, "boundary_entropy": 0.587646, "bpb": 1.300306}, "zh": {"chunks": 3230.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.546901, "cps": 25.234375, "mean_p": 0.297364, "boundary_entropy": 0.466891, "bpb": 3.472178}, "my": {"chunks": 8782.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.463479, "cps": 68.609375, "mean_p": 0.364663, "boundary_entropy": 0.552899, "bpb": 1.248861}}, "bpb": {"en": 2.33145, "de": 2.59858, "vi": 2.11907, "tr": 2.63035, "fi": 2.84768, "ru": 1.61116, "el": 1.71584, "hi": 1.35952, "ta": 1.09074, "th": 1.30031, "zh": 3.47218, "my": 1.24886}, "gini_cpc": 0.233763, "gini_cps": 0.226315, "bpic": 5.66341, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9686846070011211, "vi": 1.4657231057183533, "tr": 1.2935014156232645, "fi": 0.9784812377146892, "ru": 2.014251154182791, "el": 2.1159870457606087, "hi": 2.626408484259733, "ta": 2.7939571540793047, "th": 3.1561865259088617, "zh": 3.2853731588933095, "my": 2.784233673628437}, "t": 110.8} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.34264, "de": 2.60184, "vi": 2.12411, "tr": 2.63818, "fi": 2.84898, "ru": 1.58941, "el": 1.70443, "hi": 1.35967, "ta": 1.09776, "th": 1.31721, "zh": 3.47653, "my": 1.2568}, "bytes_per_chunk": {"en": 5.728, "de": 6.235, "vi": 5.378, "tr": 5.049, "fi": 6.304, "ru": 5.518, "el": 5.306, "hi": 5.619, "ta": 5.86, "th": 5.302, "zh": 5.483, "my": 6.119}, "bpb_mean": 2.0298, "t": 112.1} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.57673, "bpb": 2.27473, "chunk_loss": 0.030141, "F": 0.17404, "G": 0.23846, "chunks": 178.22, "bytes_per_chunk": 5.746, "boundary_entropy": 0.25549, "router_noise": 0.0, "grad_norm": 0.902, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 112.2} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.55107, "bpb": 2.23771, "chunk_loss": 0.030013, "F": 0.17075, "G": 0.25388, "chunks": 174.84, "bytes_per_chunk": 5.857, "boundary_entropy": 0.28592, "router_noise": 0.0, "grad_norm": 0.7521, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 114.1} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.593, "bpb": 2.29821, "chunk_loss": 0.03004, "F": 0.17075, "G": 0.25431, "chunks": 174.84, "bytes_per_chunk": 5.857, "boundary_entropy": 0.3127, "router_noise": 0.0, "grad_norm": 0.6723, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 116.1} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.59825, "bpb": 2.30579, "chunk_loss": 0.030147, "F": 0.17642, "G": 0.25534, "chunks": 180.66, "bytes_per_chunk": 5.668, "boundary_entropy": 0.29608, "router_noise": 0.0, "grad_norm": 0.7003, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 118.0} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2890.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166063, "cps": 22.578125, "mean_p": 0.224576, "boundary_entropy": 0.221827, "bpb": 2.279859}, "de": {"chunks": 3302.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160814, "cps": 25.796875, "mean_p": 0.248616, "boundary_entropy": 0.298687, "bpb": 2.556372}, "vi": {"chunks": 4376.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237375, "cps": 34.1875, "mean_p": 0.239447, "boundary_entropy": 0.23825, "bpb": 2.082411}, "tr": {"chunks": 3739.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.20654, "cps": 29.210938, "mean_p": 0.230771, "boundary_entropy": 0.253866, "bpb": 2.560034}, "fi": {"chunks": 3103.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.164957, "cps": 24.242188, "mean_p": 0.250504, "boundary_entropy": 0.306924, "bpb": 2.781977}, "ru": {"chunks": 6161.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.325119, "cps": 48.132812, "mean_p": 0.238344, "boundary_entropy": 0.264358, "bpb": 1.576368}, "el": {"chunks": 7065.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.344248, "cps": 55.195312, "mean_p": 0.237206, "boundary_entropy": 0.255091, "bpb": 1.702543}, "hi": {"chunks": 6943.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.41266, "cps": 54.242188, "mean_p": 0.358726, "boundary_entropy": 0.515688, "bpb": 1.348908}, "ta": {"chunks": 8689.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.464777, "cps": 67.882812, "mean_p": 0.301854, "boundary_entropy": 0.434208, "bpb": 1.059622}, "th": {"chunks": 9293.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.577743, "cps": 72.601562, "mean_p": 0.410583, "boundary_entropy": 0.593052, "bpb": 1.252649}, "zh": {"chunks": 2844.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.481544, "cps": 22.21875, "mean_p": 0.28925, "boundary_entropy": 0.459755, "bpb": 3.452255}, "my": {"chunks": 7398.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.390437, "cps": 57.796875, "mean_p": 0.372706, "boundary_entropy": 0.580124, "bpb": 1.230826}}, "bpb": {"en": 2.27986, "de": 2.55637, "vi": 2.08241, "tr": 2.56003, "fi": 2.78198, "ru": 1.57637, "el": 1.70254, "hi": 1.34891, "ta": 1.05962, "th": 1.25265, "zh": 3.45225, "my": 1.23083}, "gini_cpc": 0.233864, "gini_cps": 0.232268, "bpic": 5.832272, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9683914340271219, "vi": 1.4294219566925033, "tr": 1.243744551315072, "fi": 0.9933359870605047, "ru": 1.957799764450247, "el": 2.072991868183338, "hi": 2.484954091836892, "ta": 2.7987918794785354, "th": 3.4790538370443356, "zh": 2.8997624842251795, "my": 2.351133168686765}, "t": 120.9} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.29027, "de": 2.56236, "vi": 2.08602, "tr": 2.56766, "fi": 2.78825, "ru": 1.5568, "el": 1.69054, "hi": 1.34601, "ta": 1.06918, "th": 1.27537, "zh": 3.45562, "my": 1.23958}, "bytes_per_chunk": {"en": 5.732, "de": 6.246, "vi": 5.567, "tr": 5.235, "fi": 6.157, "ru": 5.753, "el": 5.361, "hi": 5.965, "ta": 5.778, "th": 4.85, "zh": 6.256, "my": 7.357}, "bpb_mean": 1.99397, "t": 122.2} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.51659, "bpb": 2.18798, "chunk_loss": 0.030268, "F": 0.17508, "G": 0.2669, "chunks": 179.28, "bytes_per_chunk": 5.712, "boundary_entropy": 0.32095, "router_noise": 0.0, "grad_norm": 0.6648, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 122.3} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.39196, "bpb": 2.00818, "chunk_loss": 0.030105, "F": 0.17606, "G": 0.24621, "chunks": 180.28, "bytes_per_chunk": 5.68, "boundary_entropy": 0.27331, "router_noise": 0.0, "grad_norm": 0.6914, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 124.2} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.55867, "bpb": 2.24868, "chunk_loss": 0.030013, "F": 0.16953, "G": 0.253, "chunks": 173.59, "bytes_per_chunk": 5.899, "boundary_entropy": 0.29537, "router_noise": 0.0, "grad_norm": 0.6945, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 126.2} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.4804, "bpb": 2.13577, "chunk_loss": 0.03015, "F": 0.17502, "G": 0.2483, "chunks": 179.22, "bytes_per_chunk": 5.714, "boundary_entropy": 0.27774, "router_noise": 0.0, "grad_norm": 0.689, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 128.1} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2890.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166063, "cps": 22.578125, "mean_p": 0.223523, "boundary_entropy": 0.214694, "bpb": 2.244696}, "de": {"chunks": 3333.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162324, "cps": 26.039062, "mean_p": 0.250834, "boundary_entropy": 0.294109, "bpb": 2.509171}, "vi": {"chunks": 4437.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.240683, "cps": 34.664062, "mean_p": 0.237369, "boundary_entropy": 0.232746, "bpb": 2.053119}, "tr": {"chunks": 3842.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.21223, "cps": 30.015625, "mean_p": 0.233869, "boundary_entropy": 0.247151, "bpb": 2.510896}, "fi": {"chunks": 3137.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.166764, "cps": 24.507812, "mean_p": 0.253515, "boundary_entropy": 0.307478, "bpb": 2.739213}, "ru": {"chunks": 6234.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328971, "cps": 48.703125, "mean_p": 0.243824, "boundary_entropy": 0.262346, "bpb": 1.537464}, "el": {"chunks": 7190.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.350339, "cps": 56.171875, "mean_p": 0.235218, "boundary_entropy": 0.24689, "bpb": 1.64754}, "hi": {"chunks": 6570.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.39049, "cps": 51.328125, "mean_p": 0.324308, "boundary_entropy": 0.485376, "bpb": 1.324721}, "ta": {"chunks": 10155.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.543193, "cps": 79.335938, "mean_p": 0.309638, "boundary_entropy": 0.408052, "bpb": 1.056257}, "th": {"chunks": 6895.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.42866, "cps": 53.867188, "mean_p": 0.375392, "boundary_entropy": 0.57464, "bpb": 1.235462}, "zh": {"chunks": 2774.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.469692, "cps": 21.671875, "mean_p": 0.26978, "boundary_entropy": 0.428969, "bpb": 3.410326}, "my": {"chunks": 8230.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.434347, "cps": 64.296875, "mean_p": 0.357423, "boundary_entropy": 0.553855, "bpb": 1.200597}}, "bpb": {"en": 2.2447, "de": 2.50917, "vi": 2.05312, "tr": 2.5109, "fi": 2.73921, "ru": 1.53746, "el": 1.64754, "hi": 1.32472, "ta": 1.05626, "th": 1.23546, "zh": 3.41033, "my": 1.2006}, "gini_cpc": 0.221471, "gini_cps": 0.235223, "bpic": 5.842572, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9774829344677156, "vi": 1.4493476283922846, "tr": 1.2780065702467256, "fi": 1.0042201068027081, "ru": 1.9809971971405356, "el": 2.1096690066862283, "hi": 2.351454469734752, "ta": 3.271001442755728, "th": 2.581305951406509, "zh": 2.8283899898877105, "my": 2.6155482533511862}, "t": 131.0} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.25508, "de": 2.51562, "vi": 2.05606, "tr": 2.5177, "fi": 2.74406, "ru": 1.51678, "el": 1.63779, "hi": 1.32148, "ta": 1.06536, "th": 1.25558, "zh": 3.41382, "my": 1.20747}, "bytes_per_chunk": {"en": 5.732, "de": 6.209, "vi": 5.479, "tr": 5.153, "fi": 6.131, "ru": 5.653, "el": 5.306, "hi": 6.353, "ta": 5.102, "th": 6.416, "zh": 6.45, "my": 6.406}, "bpb_mean": 1.9589, "t": 132.3} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.48497, "bpb": 2.14235, "chunk_loss": 0.030006, "F": 0.16989, "G": 0.25083, "chunks": 173.97, "bytes_per_chunk": 5.886, "boundary_entropy": 0.3001, "router_noise": 0.0, "grad_norm": 0.6758, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 132.3} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.71705, "bpb": 2.47717, "chunk_loss": 0.030224, "F": 0.17902, "G": 0.2387, "chunks": 183.31, "bytes_per_chunk": 5.586, "boundary_entropy": 0.24916, "router_noise": 0.0, "grad_norm": 0.7825, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 134.3} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.48529, "bpb": 2.14283, "chunk_loss": 0.030201, "F": 0.17422, "G": 0.24827, "chunks": 178.41, "bytes_per_chunk": 5.74, "boundary_entropy": 0.28399, "router_noise": 0.0, "grad_norm": 0.6372, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 136.2} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.48856, "bpb": 2.14754, "chunk_loss": 0.030243, "F": 0.17554, "G": 0.25364, "chunks": 179.75, "bytes_per_chunk": 5.697, "boundary_entropy": 0.2925, "router_noise": 0.0, "grad_norm": 0.6915, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 138.2} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2892.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166178, "cps": 22.59375, "mean_p": 0.216748, "boundary_entropy": 0.197452, "bpb": 2.220078}, "de": {"chunks": 3338.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162568, "cps": 26.078125, "mean_p": 0.244318, "boundary_entropy": 0.282861, "bpb": 2.483589}, "vi": {"chunks": 4403.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.238839, "cps": 34.398438, "mean_p": 0.235961, "boundary_entropy": 0.224304, "bpb": 2.047921}, "tr": {"chunks": 3754.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.207369, "cps": 29.328125, "mean_p": 0.225045, "boundary_entropy": 0.235783, "bpb": 2.477992}, "fi": {"chunks": 3043.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161767, "cps": 23.773438, "mean_p": 0.243112, "boundary_entropy": 0.292427, "bpb": 2.700745}, "ru": {"chunks": 6025.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.317942, "cps": 47.070312, "mean_p": 0.235634, "boundary_entropy": 0.250583, "bpb": 1.522531}, "el": {"chunks": 7105.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.346197, "cps": 55.507812, "mean_p": 0.228389, "boundary_entropy": 0.226676, "bpb": 1.626459}, "hi": {"chunks": 6100.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.362556, "cps": 47.65625, "mean_p": 0.322412, "boundary_entropy": 0.492876, "bpb": 1.314605}, "ta": {"chunks": 7817.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.418133, "cps": 61.070312, "mean_p": 0.285389, "boundary_entropy": 0.414487, "bpb": 1.041828}, "th": {"chunks": 5514.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.342804, "cps": 43.078125, "mean_p": 0.3795, "boundary_entropy": 0.586745, "bpb": 1.2333}, "zh": {"chunks": 3193.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.540637, "cps": 24.945312, "mean_p": 0.283603, "boundary_entropy": 0.421643, "bpb": 3.401379}, "my": {"chunks": 7129.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.37624, "cps": 55.695312, "mean_p": 0.361621, "boundary_entropy": 0.570832, "bpb": 1.233313}}, "bpb": {"en": 2.22008, "de": 2.48359, "vi": 2.04792, "tr": 2.47799, "fi": 2.70074, "ru": 1.52253, "el": 1.62646, "hi": 1.3146, "ta": 1.04183, "th": 1.2333, "zh": 3.40138, "my": 1.23331}, "gini_cpc": 0.207369, "gini_cps": 0.193528, "bpic": 6.363156, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9782723004543037, "vi": 1.4372468817770636, "tr": 1.2478705874613754, "fi": 0.9734550456153601, "ru": 1.913258575197889, "el": 2.0832868315135955, "hi": 2.181727941403143, "ta": 2.51617286224932, "th": 2.0628684233268024, "zh": 3.2533538928693373, "my": 2.26407636054824}, "t": 141.1} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.23128, "de": 2.48957, "vi": 2.04977, "tr": 2.4822, "fi": 2.70196, "ru": 1.50072, "el": 1.616, "hi": 1.31246, "ta": 1.04885, "th": 1.24982, "zh": 3.40346, "my": 1.24015}, "bytes_per_chunk": {"en": 5.736, "de": 6.238, "vi": 5.529, "tr": 5.25, "fi": 6.249, "ru": 5.826, "el": 5.333, "hi": 6.615, "ta": 6.628, "th": 8.027, "zh": 5.516, "my": 7.401}, "bpb_mean": 1.94385, "t": 142.3} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 142.3} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "parity_a", "lm_loss": 1.50454, "bpb": 2.17059, "chunk_loss": 0.034871, "F": 0.17444, "G": 0.25323, "chunks": 178.62, "bytes_per_chunk": 5.733, "boundary_entropy": 0.29194, "router_noise": 0.0, "grad_norm": 0.7395, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 10.5454, "alpha": 0.03, "t": 142.4} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "parity_a", "lm_loss": 1.30957, "bpb": 1.88931, "chunk_loss": 0.031827, "F": 0.15503, "G": 0.20781, "chunks": 158.75, "bytes_per_chunk": 6.45, "boundary_entropy": 0.22904, "router_noise": 0.0, "grad_norm": 0.6248, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.2655, "alpha": 0.03, "t": 144.3} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "parity_a", "lm_loss": 1.41115, "bpb": 2.03586, "chunk_loss": 0.031447, "F": 0.15314, "G": 0.19571, "chunks": 156.81, "bytes_per_chunk": 6.53, "boundary_entropy": 0.20681, "router_noise": 0.0, "grad_norm": 0.6999, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.9273, "alpha": 0.03, "t": 146.3} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "parity_a", "lm_loss": 1.5766, "bpb": 2.27456, "chunk_loss": 0.030942, "F": 0.15707, "G": 0.19535, "chunks": 160.84, "bytes_per_chunk": 6.366, "boundary_entropy": 0.18459, "router_noise": 0.0, "grad_norm": 0.6555, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.9194, "alpha": 0.03, "t": 148.2} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2904.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166868, "cps": 22.6875, "mean_p": 0.200441, "boundary_entropy": 0.148896, "bpb": 2.193926}, "de": {"chunks": 3308.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161107, "cps": 25.84375, "mean_p": 0.212276, "boundary_entropy": 0.209969, "bpb": 2.449524}, "vi": {"chunks": 4240.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229997, "cps": 33.125, "mean_p": 0.204263, "boundary_entropy": 0.150252, "bpb": 2.026652}, "tr": {"chunks": 3472.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.191791, "cps": 27.125, "mean_p": 0.205827, "boundary_entropy": 0.189428, "bpb": 2.459343}, "fi": {"chunks": 3051.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162192, "cps": 23.835938, "mean_p": 0.209459, "boundary_entropy": 0.217216, "bpb": 2.672258}, "ru": {"chunks": 5871.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.309815, "cps": 45.867188, "mean_p": 0.195853, "boundary_entropy": 0.16835, "bpb": 1.494689}, "el": {"chunks": 6883.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.33538, "cps": 53.773438, "mean_p": 0.201634, "boundary_entropy": 0.177057, "bpb": 1.606334}, "hi": {"chunks": 4028.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.239406, "cps": 31.46875, "mean_p": 0.129368, "boundary_entropy": 0.216569, "bpb": 1.326297}, "ta": {"chunks": 4757.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.254453, "cps": 37.164062, "mean_p": 0.11999, "boundary_entropy": 0.186407, "bpb": 1.090125}, "th": {"chunks": 3052.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.189742, "cps": 23.84375, "mean_p": 0.139709, "boundary_entropy": 0.289306, "bpb": 1.239699}, "zh": {"chunks": 1573.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.266339, "cps": 12.289062, "mean_p": 0.144088, "boundary_entropy": 0.261213, "bpb": 3.380143}, "my": {"chunks": 3728.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.196749, "cps": 29.125, "mean_p": 0.140611, "boundary_entropy": 0.300452, "bpb": 1.217217}}, "bpb": {"en": 2.19393, "de": 2.44952, "vi": 2.02665, "tr": 2.45934, "fi": 2.67226, "ru": 1.49469, "el": 1.60633, "hi": 1.3263, "ta": 1.09012, "th": 1.2397, "zh": 3.38014, "my": 1.21722}, "gini_cpc": 0.136891, "gini_cps": 0.187429, "bpic": 8.188726, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9654740423901376, "vi": 1.3783205230797508, "tr": 1.1493615733294742, "fi": 0.9719811271846831, "ru": 1.8566514206589766, "el": 2.009853639315374, "hi": 1.4347026389792836, "ta": 1.5248783207601802, "th": 1.137079872680517, "zh": 1.596109888249238, "my": 1.1790712118595128}, "t": 151.1} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.20483, "de": 2.45867, "vi": 2.02805, "tr": 2.46454, "fi": 2.67708, "ru": 1.47058, "el": 1.59741, "hi": 1.32302, "ta": 1.09769, "th": 1.26103, "zh": 3.38318, "my": 1.22647}, "bytes_per_chunk": {"en": 5.739, "de": 6.258, "vi": 5.699, "tr": 5.605, "fi": 6.27, "ru": 6.008, "el": 5.515, "hi": 10.04, "ta": 10.339, "th": 14.707, "zh": 11.056, "my": 13.739}, "bpb_mean": 1.93271, "t": 152.3} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "parity_a", "lm_loss": 1.67616, "bpb": 2.41819, "chunk_loss": 0.030817, "F": 0.14502, "G": 0.18178, "chunks": 148.5, "bytes_per_chunk": 6.896, "boundary_entropy": 0.19156, "router_noise": 0.0, "grad_norm": 0.7007, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.082, "n_target_mean": 9.9946, "alpha": 0.03, "t": 152.3} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "parity_a", "lm_loss": 1.43381, "bpb": 2.06856, "chunk_loss": 0.030759, "F": 0.14752, "G": 0.17975, "chunks": 151.06, "bytes_per_chunk": 6.779, "boundary_entropy": 0.17176, "router_noise": 0.0, "grad_norm": 0.6527, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.7257, "alpha": 0.03, "t": 154.3} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "parity_a", "lm_loss": 1.36872, "bpb": 1.97464, "chunk_loss": 0.031002, "F": 0.14734, "G": 0.17484, "chunks": 150.88, "bytes_per_chunk": 6.787, "boundary_entropy": 0.16597, "router_noise": 0.0, "grad_norm": 0.6849, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.29, "alpha": 0.03, "t": 156.2} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "parity_a", "lm_loss": 1.34166, "bpb": 1.93561, "chunk_loss": 0.03079, "F": 0.1543, "G": 0.18341, "chunks": 158.0, "bytes_per_chunk": 6.481, "boundary_entropy": 0.16237, "router_noise": 0.0, "grad_norm": 0.69, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.1792, "alpha": 0.03, "t": 158.1} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2900.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166638, "cps": 22.65625, "mean_p": 0.196315, "boundary_entropy": 0.133385, "bpb": 2.170767}, "de": {"chunks": 3331.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162227, "cps": 26.023438, "mean_p": 0.202206, "boundary_entropy": 0.17915, "bpb": 2.429769}, "vi": {"chunks": 4233.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229618, "cps": 33.070312, "mean_p": 0.194918, "boundary_entropy": 0.127654, "bpb": 2.005628}, "tr": {"chunks": 3478.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.192123, "cps": 27.171875, "mean_p": 0.202187, "boundary_entropy": 0.175092, "bpb": 2.423284}, "fi": {"chunks": 3059.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162618, "cps": 23.898438, "mean_p": 0.200801, "boundary_entropy": 0.19199, "bpb": 2.639857}, "ru": {"chunks": 5771.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.304538, "cps": 45.085938, "mean_p": 0.184587, "boundary_entropy": 0.151776, "bpb": 1.466722}, "el": {"chunks": 6830.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332797, "cps": 53.359375, "mean_p": 0.189832, "boundary_entropy": 0.142977, "bpb": 1.568668}, "hi": {"chunks": 2963.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.176107, "cps": 23.148438, "mean_p": 0.090772, "boundary_entropy": 0.191687, "bpb": 1.32543}, "ta": {"chunks": 4446.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.237818, "cps": 34.734375, "mean_p": 0.097475, "boundary_entropy": 0.167662, "bpb": 1.052186}, "th": {"chunks": 2984.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.185514, "cps": 23.3125, "mean_p": 0.125435, "boundary_entropy": 0.254219, "bpb": 1.220106}, "zh": {"chunks": 1344.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.227565, "cps": 10.5, "mean_p": 0.124129, "boundary_entropy": 0.232347, "bpb": 3.359424}, "my": {"chunks": 3744.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.197593, "cps": 29.25, "mean_p": 0.112655, "boundary_entropy": 0.224132, "bpb": 1.197534}}, "bpb": {"en": 2.17077, "de": 2.42977, "vi": 2.00563, "tr": 2.42328, "fi": 2.63986, "ru": 1.46672, "el": 1.56867, "hi": 1.32543, "ta": 1.05219, "th": 1.22011, "zh": 3.35942, "my": 1.19753}, "gini_cpc": 0.131092, "gini_cps": 0.19411, "bpic": 8.512765, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9735277778244271, "vi": 1.377942987009343, "tr": 1.1529358631737547, "fi": 0.9758739292306958, "ru": 1.8275445910290238, "el": 1.9971283690123947, "hi": 1.0568240815699133, "ta": 1.4271516079350002, "th": 1.1132786382686803, "zh": 1.3656265399301704, "my": 1.1857648882968268}, "t": 161.0} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.18391, "de": 2.43734, "vi": 2.00689, "tr": 2.42761, "fi": 2.64831, "ru": 1.44509, "el": 1.55833, "hi": 1.32164, "ta": 1.05944, "th": 1.23706, "zh": 3.3633, "my": 1.20636}, "bytes_per_chunk": {"en": 5.736, "de": 6.241, "vi": 5.701, "tr": 5.523, "fi": 6.236, "ru": 6.109, "el": 5.531, "hi": 11.367, "ta": 11.025, "th": 14.727, "zh": 13.098, "my": 12.906}, "bpb_mean": 1.90794, "t": 162.3} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "parity_a", "lm_loss": 1.46483, "bpb": 2.11331, "chunk_loss": 0.031043, "F": 0.15891, "G": 0.18386, "chunks": 162.72, "bytes_per_chunk": 6.293, "boundary_entropy": 0.15386, "router_noise": 0.0, "grad_norm": 0.7448, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.2423, "alpha": 0.03, "t": 162.3} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "parity_a", "lm_loss": 1.40669, "bpb": 2.02943, "chunk_loss": 0.030761, "F": 0.14746, "G": 0.17163, "chunks": 151.0, "bytes_per_chunk": 6.781, "boundary_entropy": 0.15067, "router_noise": 0.0, "grad_norm": 0.7207, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.9644, "alpha": 0.03, "t": 164.3} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "parity_a", "lm_loss": 1.48414, "bpb": 2.14116, "chunk_loss": 0.030741, "F": 0.14978, "G": 0.17456, "chunks": 153.38, "bytes_per_chunk": 6.676, "boundary_entropy": 0.14925, "router_noise": 0.0, "grad_norm": 0.6968, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.7181, "alpha": 0.03, "t": 166.2} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "parity_a", "lm_loss": 1.3706, "bpb": 1.97736, "chunk_loss": 0.030884, "F": 0.1467, "G": 0.17116, "chunks": 150.22, "bytes_per_chunk": 6.817, "boundary_entropy": 0.1521, "router_noise": 0.0, "grad_norm": 0.6893, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.4663, "alpha": 0.03, "t": 168.2} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2886.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165833, "cps": 22.546875, "mean_p": 0.190471, "boundary_entropy": 0.115748, "bpb": 2.147138}, "de": {"chunks": 3327.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162032, "cps": 25.992188, "mean_p": 0.198533, "boundary_entropy": 0.167403, "bpb": 2.393333}, "vi": {"chunks": 4223.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229075, "cps": 32.992188, "mean_p": 0.190765, "boundary_entropy": 0.107339, "bpb": 2.002559}, "tr": {"chunks": 3206.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177098, "cps": 25.046875, "mean_p": 0.195023, "boundary_entropy": 0.160925, "bpb": 2.398323}, "fi": {"chunks": 3070.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.163202, "cps": 23.984375, "mean_p": 0.195699, "boundary_entropy": 0.1764, "bpb": 2.582161}, "ru": {"chunks": 5702.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.300897, "cps": 44.546875, "mean_p": 0.181848, "boundary_entropy": 0.137588, "bpb": 1.449283}, "el": {"chunks": 6730.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.327925, "cps": 52.578125, "mean_p": 0.187555, "boundary_entropy": 0.1435, "bpb": 1.545019}, "hi": {"chunks": 2383.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.141634, "cps": 18.617188, "mean_p": 0.087481, "boundary_entropy": 0.185669, "bpb": 1.293212}, "ta": {"chunks": 4028.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.215459, "cps": 31.46875, "mean_p": 0.090582, "boundary_entropy": 0.155396, "bpb": 1.050618}, "th": {"chunks": 2815.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.175008, "cps": 21.992188, "mean_p": 0.110245, "boundary_entropy": 0.224184, "bpb": 1.2018}, "zh": {"chunks": 1345.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.227735, "cps": 10.507812, "mean_p": 0.119455, "boundary_entropy": 0.205685, "bpb": 3.345635}, "my": {"chunks": 2943.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.15532, "cps": 22.992188, "mean_p": 0.098563, "boundary_entropy": 0.205763, "bpb": 1.183803}}, "bpb": {"en": 2.14714, "de": 2.39333, "vi": 2.00256, "tr": 2.39832, "fi": 2.58216, "ru": 1.44928, "el": 1.54502, "hi": 1.29321, "ta": 1.05062, "th": 1.2018, "zh": 3.34563, "my": 1.1838}, "gini_cpc": 0.146864, "gini_cps": 0.206995, "bpic": 8.996695, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9770756430523635, "vi": 1.3813563641879345, "tr": 1.0679250162761182, "fi": 0.9841341118253563, "ru": 1.8144532882791455, "el": 1.9774340699644017, "hi": 0.8540764813275957, "ta": 1.2992471888433375, "th": 1.0553223293353848, "zh": 1.3732722211387975, "my": 0.93660113443204}, "t": 171.0} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.15956, "de": 2.40275, "vi": 2.00289, "tr": 2.40154, "fi": 2.58861, "ru": 1.42698, "el": 1.53567, "hi": 1.29123, "ta": 1.05725, "th": 1.21914, "zh": 3.34934, "my": 1.19343}, "bytes_per_chunk": {"en": 5.738, "de": 6.24, "vi": 5.707, "tr": 6.144, "fi": 6.214, "ru": 6.215, "el": 5.588, "hi": 12.098, "ta": 12.221, "th": 15.582, "zh": 13.232, "my": 15.465}, "bpb_mean": 1.8857, "t": 172.3} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "parity_a", "lm_loss": 1.48561, "bpb": 2.14329, "chunk_loss": 0.030412, "F": 0.14056, "G": 0.1708, "chunks": 143.94, "bytes_per_chunk": 7.114, "boundary_entropy": 0.15511, "router_noise": 0.0, "grad_norm": 0.7354, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.7667, "alpha": 0.03, "t": 172.3} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "parity_a", "lm_loss": 1.38999, "bpb": 2.00533, "chunk_loss": 0.030332, "F": 0.14664, "G": 0.17396, "chunks": 150.16, "bytes_per_chunk": 6.82, "boundary_entropy": 0.143, "router_noise": 0.0, "grad_norm": 0.7205, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.8476, "alpha": 0.03, "t": 174.3} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "parity_a", "lm_loss": 1.41743, "bpb": 2.04492, "chunk_loss": 0.030432, "F": 0.15778, "G": 0.18097, "chunks": 161.56, "bytes_per_chunk": 6.338, "boundary_entropy": 0.12951, "router_noise": 0.0, "grad_norm": 0.6859, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 7.9333, "alpha": 0.03, "t": 176.2} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "parity_a", "lm_loss": 1.37052, "bpb": 1.97724, "chunk_loss": 0.030545, "F": 0.14053, "G": 0.16971, "chunks": 143.91, "bytes_per_chunk": 7.116, "boundary_entropy": 0.15242, "router_noise": 0.0, "grad_norm": 0.7073, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.8477, "alpha": 0.03, "t": 178.2} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2884.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165719, "cps": 22.53125, "mean_p": 0.187741, "boundary_entropy": 0.106791, "bpb": 2.141882}, "de": {"chunks": 3297.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160571, "cps": 25.757812, "mean_p": 0.192476, "boundary_entropy": 0.154551, "bpb": 2.373305}, "vi": {"chunks": 4205.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228099, "cps": 32.851562, "mean_p": 0.188829, "boundary_entropy": 0.108766, "bpb": 1.985352}, "tr": {"chunks": 3065.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169309, "cps": 23.945312, "mean_p": 0.188585, "boundary_entropy": 0.152579, "bpb": 2.385386}, "fi": {"chunks": 3006.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.1598, "cps": 23.484375, "mean_p": 0.186937, "boundary_entropy": 0.16091, "bpb": 2.568674}, "ru": {"chunks": 5725.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.302111, "cps": 44.726562, "mean_p": 0.17682, "boundary_entropy": 0.126828, "bpb": 1.427375}, "el": {"chunks": 6747.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.328753, "cps": 52.710938, "mean_p": 0.183855, "boundary_entropy": 0.137105, "bpb": 1.511929}, "hi": {"chunks": 1518.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.090223, "cps": 11.859375, "mean_p": 0.077602, "boundary_entropy": 0.169187, "bpb": 1.288772}, "ta": {"chunks": 3155.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.168762, "cps": 24.648438, "mean_p": 0.085837, "boundary_entropy": 0.149131, "bpb": 1.049415}, "th": {"chunks": 2600.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.161641, "cps": 20.3125, "mean_p": 0.10408, "boundary_entropy": 0.21332, "bpb": 1.190855}, "zh": {"chunks": 1372.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.232306, "cps": 10.71875, "mean_p": 0.114512, "boundary_entropy": 0.191442, "bpb": 3.315322}, "my": {"chunks": 1937.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.102227, "cps": 15.132812, "mean_p": 0.091409, "boundary_entropy": 0.2117, "bpb": 1.18047}}, "bpb": {"en": 2.14188, "de": 2.3733, "vi": 1.98535, "tr": 2.38539, "fi": 2.56867, "ru": 1.42738, "el": 1.51193, "hi": 1.28877, "ta": 1.04942, "th": 1.19086, "zh": 3.31532, "my": 1.18047}, "gini_cpc": 0.194388, "gini_cps": 0.247233, "bpic": 9.71327, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.968936696267765, "vi": 1.3764223701598788, "tr": 1.0216656490908895, "fi": 0.9642862119730901, "ru": 1.8230355633300275, "el": 1.9838038517456449, "hi": 0.5444344057390985, "ta": 1.0183633433847006, "th": 0.9753963966566314, "zh": 1.4018112237349545, "my": 0.6168720722485395}, "t": 181.0} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.15205, "de": 2.3798, "vi": 1.9878, "tr": 2.38742, "fi": 2.57679, "ru": 1.40602, "el": 1.50305, "hi": 1.28692, "ta": 1.05672, "th": 1.20941, "zh": 3.31663, "my": 1.18944}, "bytes_per_chunk": {"en": 5.74, "de": 6.275, "vi": 5.72, "tr": 6.346, "fi": 6.328, "ru": 6.231, "el": 5.607, "hi": 14.882, "ta": 13.806, "th": 16.865, "zh": 12.848, "my": 22.567}, "bpb_mean": 1.87101, "t": 182.3} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "parity_a", "lm_loss": 1.3246, "bpb": 1.91099, "chunk_loss": 0.030712, "F": 0.14575, "G": 0.16609, "chunks": 149.25, "bytes_per_chunk": 6.861, "boundary_entropy": 0.12679, "router_noise": 0.0, "grad_norm": 0.7467, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.6071, "alpha": 0.03, "t": 182.4} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "parity_a", "lm_loss": 1.55938, "bpb": 2.2497, "chunk_loss": 0.030403, "F": 0.14481, "G": 0.16869, "chunks": 148.28, "bytes_per_chunk": 6.906, "boundary_entropy": 0.13445, "router_noise": 0.0, "grad_norm": 0.7096, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.2757, "alpha": 0.03, "t": 184.3} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "parity_a", "lm_loss": 1.51234, "bpb": 2.18184, "chunk_loss": 0.030317, "F": 0.15427, "G": 0.17986, "chunks": 157.97, "bytes_per_chunk": 6.482, "boundary_entropy": 0.13254, "router_noise": 0.0, "grad_norm": 0.7242, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.1473, "alpha": 0.03, "t": 186.2} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "parity_a", "lm_loss": 1.33484, "bpb": 1.92577, "chunk_loss": 0.031027, "F": 0.14233, "G": 0.16144, "chunks": 145.75, "bytes_per_chunk": 7.026, "boundary_entropy": 0.14031, "router_noise": 0.0, "grad_norm": 0.6351, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.8728, "alpha": 0.03, "t": 188.2} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.188126, "boundary_entropy": 0.105196, "bpb": 2.116959}, "de": {"chunks": 3312.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161301, "cps": 25.875, "mean_p": 0.190263, "boundary_entropy": 0.144943, "bpb": 2.353447}, "vi": {"chunks": 4231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229509, "cps": 33.054688, "mean_p": 0.188395, "boundary_entropy": 0.105587, "bpb": 1.952682}, "tr": {"chunks": 3200.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176766, "cps": 25.0, "mean_p": 0.191264, "boundary_entropy": 0.149336, "bpb": 2.361009}, "fi": {"chunks": 3011.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160066, "cps": 23.523438, "mean_p": 0.185141, "boundary_entropy": 0.150485, "bpb": 2.555513}, "ru": {"chunks": 5660.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.298681, "cps": 44.21875, "mean_p": 0.177246, "boundary_entropy": 0.125885, "bpb": 1.412211}, "el": {"chunks": 6626.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.322857, "cps": 51.765625, "mean_p": 0.183506, "boundary_entropy": 0.135038, "bpb": 1.494211}, "hi": {"chunks": 1474.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.087608, "cps": 11.515625, "mean_p": 0.081488, "boundary_entropy": 0.180176, "bpb": 1.272783}, "ta": {"chunks": 3240.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.173308, "cps": 25.3125, "mean_p": 0.087175, "boundary_entropy": 0.150559, "bpb": 1.038918}, "th": {"chunks": 2605.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.161952, "cps": 20.351562, "mean_p": 0.11072, "boundary_entropy": 0.228237, "bpb": 1.175443}, "zh": {"chunks": 1451.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.245682, "cps": 11.335938, "mean_p": 0.119094, "boundary_entropy": 0.187276, "bpb": 3.296522}, "my": {"chunks": 2892.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.152628, "cps": 22.59375, "mean_p": 0.088761, "boundary_entropy": 0.177917, "bpb": 1.165481}}, "bpb": {"en": 2.11696, "de": 2.35345, "vi": 1.95268, "tr": 2.36101, "fi": 2.55551, "ru": 1.41221, "el": 1.49421, "hi": 1.27278, "ta": 1.03892, "th": 1.17544, "zh": 3.29652, "my": 1.16548}, "gini_cpc": 0.174557, "gini_cps": 0.224421, "bpic": 9.454364, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9709881940306735, "vi": 1.381579599886182, "tr": 1.064082878837005, "fi": 0.9635514307232196, "ru": 1.7979733300233918, "el": 1.9435092632064457, "hi": 0.5273736651585106, "ta": 1.043267237765773, "th": 0.9749058825439523, "zh": 1.478938103621091, "my": 0.9187787479607937}, "t": 191.0} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.12894, "de": 2.36268, "vi": 1.95443, "tr": 2.36555, "fi": 2.56176, "ru": 1.39097, "el": 1.48554, "hi": 1.27073, "ta": 1.04588, "th": 1.1932, "zh": 3.29342, "my": 1.17365}, "bytes_per_chunk": {"en": 5.736, "de": 6.259, "vi": 5.7, "tr": 6.135, "fi": 6.323, "ru": 6.269, "el": 5.702, "hi": 15.862, "ta": 14.142, "th": 17.172, "zh": 12.429, "my": 15.888}, "bpb_mean": 1.85223, "t": 192.3} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "parity_a", "lm_loss": 1.37042, "bpb": 1.9771, "chunk_loss": 0.030493, "F": 0.14386, "G": 0.16685, "chunks": 147.31, "bytes_per_chunk": 6.951, "boundary_entropy": 0.13502, "router_noise": 0.0, "grad_norm": 0.6858, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.6708, "alpha": 0.03, "t": 192.4} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "parity_a", "lm_loss": 1.36988, "bpb": 1.97632, "chunk_loss": 0.030584, "F": 0.14096, "G": 0.1655, "chunks": 144.34, "bytes_per_chunk": 7.094, "boundary_entropy": 0.13503, "router_noise": 0.0, "grad_norm": 0.7567, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.9964, "alpha": 0.03, "t": 194.3} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "parity_a", "lm_loss": 1.44213, "bpb": 2.08056, "chunk_loss": 0.030902, "F": 0.14511, "G": 0.16506, "chunks": 148.59, "bytes_per_chunk": 6.891, "boundary_entropy": 0.1331, "router_noise": 0.0, "grad_norm": 0.7145, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.4167, "alpha": 0.03, "t": 196.2} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "parity_a", "lm_loss": 1.40112, "bpb": 2.02138, "chunk_loss": 0.030401, "F": 0.14566, "G": 0.16976, "chunks": 149.16, "bytes_per_chunk": 6.865, "boundary_entropy": 0.12974, "router_noise": 0.0, "grad_norm": 0.6952, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.2873, "alpha": 0.03, "t": 198.2} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.184756, "boundary_entropy": 0.095216, "bpb": 2.101586}, "de": {"chunks": 3296.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160522, "cps": 25.75, "mean_p": 0.187737, "boundary_entropy": 0.139355, "bpb": 2.331155}, "vi": {"chunks": 4204.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228044, "cps": 32.84375, "mean_p": 0.183647, "boundary_entropy": 0.089449, "bpb": 1.947969}, "tr": {"chunks": 3047.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168315, "cps": 23.804688, "mean_p": 0.184142, "boundary_entropy": 0.139063, "bpb": 2.344235}, "fi": {"chunks": 3000.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159481, "cps": 23.4375, "mean_p": 0.179868, "boundary_entropy": 0.139124, "bpb": 2.52305}, "ru": {"chunks": 5643.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.297784, "cps": 44.085938, "mean_p": 0.174391, "boundary_entropy": 0.117329, "bpb": 1.409126}, "el": {"chunks": 6650.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.324027, "cps": 51.953125, "mean_p": 0.182169, "boundary_entropy": 0.125343, "bpb": 1.473493}, "hi": {"chunks": 1387.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.082437, "cps": 10.835938, "mean_p": 0.075863, "boundary_entropy": 0.165247, "bpb": 1.262582}, "ta": {"chunks": 3243.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.173469, "cps": 25.335938, "mean_p": 0.084867, "boundary_entropy": 0.142989, "bpb": 1.01602}, "th": {"chunks": 2649.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.164688, "cps": 20.695312, "mean_p": 0.107538, "boundary_entropy": 0.212396, "bpb": 1.15537}, "zh": {"chunks": 1319.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.223332, "cps": 10.304688, "mean_p": 0.109996, "boundary_entropy": 0.175868, "bpb": 3.274054}, "my": {"chunks": 2948.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.155584, "cps": 23.03125, "mean_p": 0.090475, "boundary_entropy": 0.17844, "bpb": 1.166108}}, "bpb": {"en": 2.10159, "de": 2.33116, "vi": 1.94797, "tr": 2.34424, "fi": 2.52305, "ru": 1.40913, "el": 1.47349, "hi": 1.26258, "ta": 1.01602, "th": 1.15537, "zh": 3.27405, "my": 1.16611}, "gini_cpc": 0.172198, "gini_cps": 0.229283, "bpic": 9.529486, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9676362554817475, "vi": 1.374665083482168, "tr": 1.0146102352659638, "fi": 0.961361459737398, "ru": 1.7950567049269173, "el": 1.9532513649263312, "hi": 0.49693401429063316, "ta": 1.0456800333281053, "th": 0.9927461908662191, "zh": 1.3462592156462092, "my": 0.9378673984161272}, "t": 201.1} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.11256, "de": 2.33806, "vi": 1.9486, "tr": 2.3503, "fi": 2.53115, "ru": 1.38708, "el": 1.4642, "hi": 1.25937, "ta": 1.02421, "th": 1.17508, "zh": 3.2736, "my": 1.17481}, "bytes_per_chunk": {"en": 5.742, "de": 6.286, "vi": 5.716, "tr": 6.363, "fi": 6.326, "ru": 6.29, "el": 5.676, "hi": 15.928, "ta": 13.443, "th": 16.617, "zh": 13.079, "my": 15.781}, "bpb_mean": 1.83658, "t": 202.3} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "parity_a", "lm_loss": 1.45844, "bpb": 2.10408, "chunk_loss": 0.031324, "F": 0.15829, "G": 0.17464, "chunks": 162.09, "bytes_per_chunk": 6.317, "boundary_entropy": 0.12446, "router_noise": 0.0, "grad_norm": 1.3891, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.7077, "alpha": 0.03, "t": 202.4} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "parity_a", "lm_loss": 1.32161, "bpb": 1.90668, "chunk_loss": 0.03072, "F": 0.15601, "G": 0.1729, "chunks": 159.75, "bytes_per_chunk": 6.41, "boundary_entropy": 0.11168, "router_noise": 0.0, "grad_norm": 0.6845, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.9757, "alpha": 0.03, "t": 204.3} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "parity_a", "lm_loss": 1.58566, "bpb": 2.28762, "chunk_loss": 0.030421, "F": 0.14111, "G": 0.1608, "chunks": 144.5, "bytes_per_chunk": 7.087, "boundary_entropy": 0.12247, "router_noise": 0.0, "grad_norm": 0.7386, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.6591, "alpha": 0.03, "t": 206.3} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "parity_a", "lm_loss": 1.30721, "bpb": 1.8859, "chunk_loss": 0.030532, "F": 0.14413, "G": 0.16326, "chunks": 147.59, "bytes_per_chunk": 6.938, "boundary_entropy": 0.1215, "router_noise": 0.0, "grad_norm": 0.6829, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.7851, "alpha": 0.03, "t": 208.2} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.184241, "boundary_entropy": 0.091697, "bpb": 2.085321}, "de": {"chunks": 3301.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160766, "cps": 25.789062, "mean_p": 0.186293, "boundary_entropy": 0.131754, "bpb": 2.313464}, "vi": {"chunks": 4188.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227177, "cps": 32.71875, "mean_p": 0.1833, "boundary_entropy": 0.084332, "bpb": 1.937292}, "tr": {"chunks": 3046.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168259, "cps": 23.796875, "mean_p": 0.181338, "boundary_entropy": 0.130166, "bpb": 2.321428}, "fi": {"chunks": 3009.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15996, "cps": 23.507812, "mean_p": 0.181954, "boundary_entropy": 0.140243, "bpb": 2.527471}, "ru": {"chunks": 5610.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.296042, "cps": 43.828125, "mean_p": 0.169707, "boundary_entropy": 0.120523, "bpb": 1.39017}, "el": {"chunks": 6617.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.322419, "cps": 51.695312, "mean_p": 0.180033, "boundary_entropy": 0.126125, "bpb": 1.457136}, "hi": {"chunks": 3421.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.203328, "cps": 26.726562, "mean_p": 0.086899, "boundary_entropy": 0.164053, "bpb": 1.243723}, "ta": {"chunks": 3825.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.2046, "cps": 29.882812, "mean_p": 0.082721, "boundary_entropy": 0.136272, "bpb": 1.012687}, "th": {"chunks": 2561.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.159217, "cps": 20.007812, "mean_p": 0.09425, "boundary_entropy": 0.18729, "bpb": 1.150841}, "zh": {"chunks": 1294.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.219099, "cps": 10.109375, "mean_p": 0.105907, "boundary_entropy": 0.160118, "bpb": 3.273989}, "my": {"chunks": 2377.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.125449, "cps": 18.570312, "mean_p": 0.089157, "boundary_entropy": 0.187489, "bpb": 1.159197}}, "bpb": {"en": 2.08532, "de": 2.31346, "vi": 1.93729, "tr": 2.32143, "fi": 2.52747, "ru": 1.39017, "el": 1.45714, "hi": 1.24372, "ta": 1.01269, "th": 1.15084, "zh": 3.27399, "my": 1.1592}, "gini_cpc": 0.148928, "gini_cps": 0.208066, "bpic": 9.107717, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9684332578324515, "vi": 1.3684852141100288, "tr": 1.013575083709584, "fi": 0.9635780151833346, "ru": 1.7833238806828922, "el": 1.942213049826983, "hi": 1.2248265157020763, "ta": 1.232487571018289, "th": 0.9591026513805326, "zh": 1.319828226148743, "my": 0.7556877531796592}, "t": 211.1} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.09738, "de": 2.32185, "vi": 1.93702, "tr": 2.32732, "fi": 2.53373, "ru": 1.36823, "el": 1.4466, "hi": 1.24106, "ta": 1.02095, "th": 1.16961, "zh": 3.27262, "my": 1.16727}, "bytes_per_chunk": {"en": 5.739, "de": 6.272, "vi": 5.714, "tr": 6.365, "fi": 6.312, "ru": 6.338, "el": 5.701, "hi": 10.733, "ta": 12.641, "th": 17.444, "zh": 13.985, "my": 20.895}, "bpb_mean": 1.8253, "t": 212.4} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "parity_a", "lm_loss": 1.37226, "bpb": 1.97976, "chunk_loss": 0.030442, "F": 0.15314, "G": 0.17094, "chunks": 156.81, "bytes_per_chunk": 6.53, "boundary_entropy": 0.11491, "router_noise": 0.0, "grad_norm": 0.711, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 8.4667, "alpha": 0.03, "t": 212.4} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "parity_a", "lm_loss": 1.48906, "bpb": 2.14825, "chunk_loss": 0.030425, "F": 0.13077, "G": 0.15153, "chunks": 133.91, "bytes_per_chunk": 7.647, "boundary_entropy": 0.12887, "router_noise": 0.0, "grad_norm": 0.6775, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 11.14, "alpha": 0.03, "t": 214.3} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "parity_a", "lm_loss": 1.49694, "bpb": 2.15962, "chunk_loss": 0.030469, "F": 0.13345, "G": 0.15499, "chunks": 136.66, "bytes_per_chunk": 7.493, "boundary_entropy": 0.12849, "router_noise": 0.0, "grad_norm": 0.7089, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.5674, "alpha": 0.03, "t": 216.3} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "parity_a", "lm_loss": 1.42055, "bpb": 2.04942, "chunk_loss": 0.030522, "F": 0.14355, "G": 0.16001, "chunks": 147.0, "bytes_per_chunk": 6.966, "boundary_entropy": 0.11789, "router_noise": 0.0, "grad_norm": 0.7131, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.7537, "alpha": 0.03, "t": 218.2} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.183747, "boundary_entropy": 0.089635, "bpb": 2.073236}, "de": {"chunks": 3304.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160912, "cps": 25.8125, "mean_p": 0.186841, "boundary_entropy": 0.131979, "bpb": 2.298369}, "vi": {"chunks": 4216.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228695, "cps": 32.9375, "mean_p": 0.183016, "boundary_entropy": 0.085979, "bpb": 1.922784}, "tr": {"chunks": 3099.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.171187, "cps": 24.210938, "mean_p": 0.18433, "boundary_entropy": 0.129427, "bpb": 2.297714}, "fi": {"chunks": 3024.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160757, "cps": 23.625, "mean_p": 0.180738, "boundary_entropy": 0.131188, "bpb": 2.48274}, "ru": {"chunks": 5610.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.296042, "cps": 43.828125, "mean_p": 0.17088, "boundary_entropy": 0.113807, "bpb": 1.372294}, "el": {"chunks": 6658.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.324417, "cps": 52.015625, "mean_p": 0.181893, "boundary_entropy": 0.119422, "bpb": 1.434461}, "hi": {"chunks": 3285.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.195245, "cps": 25.664062, "mean_p": 0.0861, "boundary_entropy": 0.161624, "bpb": 1.236405}, "ta": {"chunks": 4147.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.221824, "cps": 32.398438, "mean_p": 0.086915, "boundary_entropy": 0.133045, "bpb": 1.017608}, "th": {"chunks": 2824.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.175567, "cps": 22.0625, "mean_p": 0.103796, "boundary_entropy": 0.199816, "bpb": 1.13998}, "zh": {"chunks": 1368.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.231629, "cps": 10.6875, "mean_p": 0.10938, "boundary_entropy": 0.163371, "bpb": 3.240679}, "my": {"chunks": 2801.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.147826, "cps": 21.882812, "mean_p": 0.092329, "boundary_entropy": 0.186094, "bpb": 1.148922}}, "bpb": {"en": 2.07324, "de": 2.29837, "vi": 1.92278, "tr": 2.29771, "fi": 2.48274, "ru": 1.37229, "el": 1.43446, "hi": 1.23641, "ta": 1.01761, "th": 1.13998, "zh": 3.24068, "my": 1.14892}, "gini_cpc": 0.139124, "gini_cps": 0.193828, "bpic": 8.878887, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.969649019826607, "vi": 1.3781116102319522, "tr": 1.031568224638851, "fi": 0.9687168069722863, "ru": 1.783941375101411, "el": 1.9549239937949487, "hi": 1.1765414504039875, "ta": 1.3367047679657187, "th": 1.0579631924239046, "zh": 1.3957884042989288, "my": 0.8907927301546088}, "t": 221.1} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.08683, "de": 2.30625, "vi": 1.9235, "tr": 2.3039, "fi": 2.49016, "ru": 1.35118, "el": 1.42626, "hi": 1.23389, "ta": 1.024, "th": 1.15782, "zh": 3.23807, "my": 1.15703}, "bytes_per_chunk": {"en": 5.74, "de": 6.266, "vi": 5.71, "tr": 6.259, "fi": 6.295, "ru": 6.328, "el": 5.672, "hi": 10.996, "ta": 11.827, "th": 15.546, "zh": 13.289, "my": 16.867}, "bpb_mean": 1.80824, "t": 222.4} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "parity_a", "lm_loss": 1.41114, "bpb": 2.03585, "chunk_loss": 0.030665, "F": 0.14127, "G": 0.15947, "chunks": 144.66, "bytes_per_chunk": 7.079, "boundary_entropy": 0.12272, "router_noise": 0.0, "grad_norm": 0.7691, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 10.4895, "alpha": 0.03, "t": 222.4} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "parity_a", "lm_loss": 1.30075, "bpb": 1.87658, "chunk_loss": 0.030647, "F": 0.1441, "G": 0.1622, "chunks": 147.56, "bytes_per_chunk": 6.939, "boundary_entropy": 0.11301, "router_noise": 0.0, "grad_norm": 0.7284, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.696, "alpha": 0.03, "t": 224.4} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "parity_a", "lm_loss": 1.44905, "bpb": 2.09053, "chunk_loss": 0.030451, "F": 0.14374, "G": 0.16163, "chunks": 147.19, "bytes_per_chunk": 6.957, "boundary_entropy": 0.11569, "router_noise": 0.0, "grad_norm": 0.717, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.4908, "alpha": 0.03, "t": 226.3} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "parity_a", "lm_loss": 1.23619, "bpb": 1.78345, "chunk_loss": 0.030774, "F": 0.14532, "G": 0.16062, "chunks": 148.81, "bytes_per_chunk": 6.881, "boundary_entropy": 0.12028, "router_noise": 0.0, "grad_norm": 0.6916, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.3128, "alpha": 0.03, "t": 228.3} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2886.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165833, "cps": 22.546875, "mean_p": 0.183336, "boundary_entropy": 0.087425, "bpb": 2.060022}, "de": {"chunks": 3299.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160668, "cps": 25.773438, "mean_p": 0.185143, "boundary_entropy": 0.127172, "bpb": 2.288534}, "vi": {"chunks": 4218.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228804, "cps": 32.953125, "mean_p": 0.182667, "boundary_entropy": 0.090399, "bpb": 1.91214}, "tr": {"chunks": 3081.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170193, "cps": 24.070312, "mean_p": 0.182215, "boundary_entropy": 0.127493, "bpb": 2.289708}, "fi": {"chunks": 3009.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15996, "cps": 23.507812, "mean_p": 0.177679, "boundary_entropy": 0.129386, "bpb": 2.462147}, "ru": {"chunks": 5626.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.296887, "cps": 43.953125, "mean_p": 0.171953, "boundary_entropy": 0.115176, "bpb": 1.359009}, "el": {"chunks": 6673.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325147, "cps": 52.132812, "mean_p": 0.182348, "boundary_entropy": 0.113783, "bpb": 1.418329}, "hi": {"chunks": 2035.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.120951, "cps": 15.898438, "mean_p": 0.080935, "boundary_entropy": 0.166857, "bpb": 1.230427}, "ta": {"chunks": 3608.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.192993, "cps": 28.1875, "mean_p": 0.082364, "boundary_entropy": 0.133821, "bpb": 0.99944}, "th": {"chunks": 2718.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.168977, "cps": 21.234375, "mean_p": 0.099039, "boundary_entropy": 0.192079, "bpb": 1.127911}, "zh": {"chunks": 1279.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.216559, "cps": 9.992188, "mean_p": 0.10285, "boundary_entropy": 0.15536, "bpb": 3.227484}, "my": {"chunks": 2650.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.139856, "cps": 20.703125, "mean_p": 0.097878, "boundary_entropy": 0.202746, "bpb": 1.139875}}, "bpb": {"en": 2.06002, "de": 2.28853, "vi": 1.91214, "tr": 2.28971, "fi": 2.46215, "ru": 1.35901, "el": 1.41833, "hi": 1.23043, "ta": 0.99944, "th": 1.12791, "zh": 3.22748, "my": 1.13987}, "gini_cpc": 0.159914, "gini_cps": 0.219865, "bpic": 9.341829, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9688525838382166, "vi": 1.379720848720035, "tr": 1.026287266109395, "fi": 0.9645796555317581, "ru": 1.7902690634616756, "el": 1.9606861142455354, "hi": 0.7293519259343925, "ta": 1.163774542538918, "th": 1.0189577588396364, "zh": 1.3058848853803138, "my": 0.8433547421831146}, "t": 231.1} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.07298, "de": 2.29702, "vi": 1.91479, "tr": 2.29392, "fi": 2.46976, "ru": 1.33765, "el": 1.40849, "hi": 1.22821, "ta": 1.00682, "th": 1.14526, "zh": 3.22666, "my": 1.14875}, "bytes_per_chunk": {"en": 5.74, "de": 6.277, "vi": 5.707, "tr": 6.299, "fi": 6.316, "ru": 6.311, "el": 5.667, "hi": 13.065, "ta": 12.7, "th": 16.483, "zh": 14.185, "my": 17.28}, "bpb_mean": 1.79586, "t": 232.4} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584, "t": 232.4} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "none", "lm_loss": 1.25148, "bpb": 1.8055, "chunk_loss": 0.0, "F": 0.15186, "G": 0.16823, "chunks": 155.5, "bytes_per_chunk": 6.585, "boundary_entropy": 0.11372, "router_noise": 0.0, "grad_norm": 0.7049, "lr": 9.12e-05, "router_trainable": false, "sec_per_step": 0.084, "alpha": 0.0, "t": 232.4} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "none", "lm_loss": 1.27809, "bpb": 1.8439, "chunk_loss": 0.0, "F": 0.15427, "G": 0.16765, "chunks": 157.97, "bytes_per_chunk": 6.482, "boundary_entropy": 0.10204, "router_noise": 0.0, "grad_norm": 0.7128, "lr": 8.81e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 234.3} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "none", "lm_loss": 1.36288, "bpb": 1.96623, "chunk_loss": 0.0, "F": 0.14032, "G": 0.15953, "chunks": 143.69, "bytes_per_chunk": 7.127, "boundary_entropy": 0.11882, "router_noise": 0.0, "grad_norm": 0.7329, "lr": 8.5e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 236.2} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "none", "lm_loss": 1.32609, "bpb": 1.91314, "chunk_loss": 0.0, "F": 0.14874, "G": 0.16339, "chunks": 152.31, "bytes_per_chunk": 6.723, "boundary_entropy": 0.10989, "router_noise": 0.0, "grad_norm": 0.7108, "lr": 8.21e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 238.2} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.183964, "boundary_entropy": 0.090233, "bpb": 2.050702}, "de": {"chunks": 3303.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160863, "cps": 25.804688, "mean_p": 0.186806, "boundary_entropy": 0.132514, "bpb": 2.274676}, "vi": {"chunks": 4230.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229455, "cps": 33.046875, "mean_p": 0.183383, "boundary_entropy": 0.089306, "bpb": 1.906278}, "tr": {"chunks": 3047.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168315, "cps": 23.804688, "mean_p": 0.180419, "boundary_entropy": 0.128488, "bpb": 2.272499}, "fi": {"chunks": 3009.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15996, "cps": 23.507812, "mean_p": 0.179146, "boundary_entropy": 0.131428, "bpb": 2.445118}, "ru": {"chunks": 5649.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.2981, "cps": 44.132812, "mean_p": 0.175236, "boundary_entropy": 0.111764, "bpb": 1.351973}, "el": {"chunks": 6685.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325732, "cps": 52.226562, "mean_p": 0.184062, "boundary_entropy": 0.11287, "bpb": 1.41307}, "hi": {"chunks": 3647.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.216761, "cps": 28.492188, "mean_p": 0.08945, "boundary_entropy": 0.164888, "bpb": 1.220038}, "ta": {"chunks": 4382.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.234394, "cps": 34.234375, "mean_p": 0.091991, "boundary_entropy": 0.135459, "bpb": 0.98641}, "th": {"chunks": 2824.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.175567, "cps": 22.0625, "mean_p": 0.104101, "boundary_entropy": 0.195104, "bpb": 1.117777}, "zh": {"chunks": 1505.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.254826, "cps": 11.757812, "mean_p": 0.118047, "boundary_entropy": 0.164489, "bpb": 3.214117}, "my": {"chunks": 3623.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.191208, "cps": 28.304688, "mean_p": 0.108471, "boundary_entropy": 0.20538, "bpb": 1.12819}}, "bpb": {"en": 2.0507, "de": 2.27468, "vi": 1.90628, "tr": 2.2725, "fi": 2.44512, "ru": 1.35197, "el": 1.41307, "hi": 1.22004, "ta": 0.98641, "th": 1.11778, "zh": 3.21412, "my": 1.12819}, "gini_cpc": 0.135564, "gini_cps": 0.188087, "bpic": 8.568261, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9696913082088021, "vi": 1.383166818061268, "tr": 1.0146102352659638, "fi": 0.9642455441166102, "ru": 1.7969653244962178, "el": 1.9635316352680487, "hi": 1.3066462509862575, "ta": 1.412941691657033, "th": 1.0583296500589665, "zh": 1.5361031990504512, "my": 1.1526097640643247}, "t": 241.0} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.06443, "de": 2.28347, "vi": 1.90685, "tr": 2.27654, "fi": 2.45228, "ru": 1.33059, "el": 1.40379, "hi": 1.21762, "ta": 0.99335, "th": 1.13539, "zh": 3.21033, "my": 1.13666}, "bytes_per_chunk": {"en": 5.74, "de": 6.273, "vi": 5.703, "tr": 6.357, "fi": 6.302, "ru": 6.282, "el": 5.64, "hi": 10.514, "ta": 11.31, "th": 15.884, "zh": 12.063, "my": 13.205}, "bpb_mean": 1.78427, "t": 242.3} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "none", "lm_loss": 1.32877, "bpb": 1.91701, "chunk_loss": 0.0, "F": 0.14807, "G": 0.16441, "chunks": 151.62, "bytes_per_chunk": 6.754, "boundary_entropy": 0.1191, "router_noise": 0.0, "grad_norm": 0.7289, "lr": 7.92e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 242.3} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "none", "lm_loss": 1.16668, "bpb": 1.68317, "chunk_loss": 0.0, "F": 0.14142, "G": 0.16316, "chunks": 144.81, "bytes_per_chunk": 7.071, "boundary_entropy": 0.12823, "router_noise": 0.0, "grad_norm": 0.7901, "lr": 7.63e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 244.2} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "none", "lm_loss": 1.29099, "bpb": 1.8625, "chunk_loss": 0.0, "F": 0.1524, "G": 0.16914, "chunks": 156.06, "bytes_per_chunk": 6.561, "boundary_entropy": 0.11252, "router_noise": 0.0, "grad_norm": 0.718, "lr": 7.36e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 246.1} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "none", "lm_loss": 1.29022, "bpb": 1.86139, "chunk_loss": 0.0, "F": 0.1586, "G": 0.17548, "chunks": 162.41, "bytes_per_chunk": 6.305, "boundary_entropy": 0.10787, "router_noise": 0.0, "grad_norm": 0.6712, "lr": 7.09e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 248.1} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.183509, "boundary_entropy": 0.087431, "bpb": 2.039342}, "de": {"chunks": 3292.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160327, "cps": 25.71875, "mean_p": 0.184783, "boundary_entropy": 0.125797, "bpb": 2.266387}, "vi": {"chunks": 4216.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228695, "cps": 32.9375, "mean_p": 0.182951, "boundary_entropy": 0.086696, "bpb": 1.897473}, "tr": {"chunks": 3054.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168701, "cps": 23.859375, "mean_p": 0.180384, "boundary_entropy": 0.125355, "bpb": 2.263499}, "fi": {"chunks": 3008.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159906, "cps": 23.5, "mean_p": 0.177989, "boundary_entropy": 0.127667, "bpb": 2.436554}, "ru": {"chunks": 5655.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.298417, "cps": 44.179688, "mean_p": 0.175862, "boundary_entropy": 0.113216, "bpb": 1.34131}, "el": {"chunks": 6693.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326122, "cps": 52.289062, "mean_p": 0.184096, "boundary_entropy": 0.112354, "bpb": 1.392126}, "hi": {"chunks": 3957.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.235186, "cps": 30.914062, "mean_p": 0.097155, "boundary_entropy": 0.169486, "bpb": 1.208292}, "ta": {"chunks": 4081.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.218294, "cps": 31.882812, "mean_p": 0.087789, "boundary_entropy": 0.136353, "bpb": 0.985746}, "th": {"chunks": 2786.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.173205, "cps": 21.765625, "mean_p": 0.103485, "boundary_entropy": 0.196708, "bpb": 1.114193}, "zh": {"chunks": 1582.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.267863, "cps": 12.359375, "mean_p": 0.123093, "boundary_entropy": 0.166065, "bpb": 3.203192}, "my": {"chunks": 2377.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.125449, "cps": 18.570312, "mean_p": 0.094944, "boundary_entropy": 0.196031, "bpb": 1.122207}}, "bpb": {"en": 2.03934, "de": 2.26639, "vi": 1.89747, "tr": 2.2635, "fi": 2.43655, "ru": 1.34131, "el": 1.39213, "hi": 1.20829, "ta": 0.98575, "th": 1.11419, "zh": 3.20319, "my": 1.12221}, "gini_cpc": 0.157509, "gini_cps": 0.200338, "bpic": 8.80474, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9664619396377162, "vi": 1.3785889609802142, "tr": 1.0169411416154428, "fi": 0.9639250902966978, "ru": 1.798873944065518, "el": 1.9658814113461558, "hi": 1.4177129737188432, "ta": 1.3158865914313902, "th": 1.0440886703485412, "zh": 1.6146945255134972, "my": 0.7562112639196521}, "t": 250.9} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.05184, "de": 2.27408, "vi": 1.89788, "tr": 2.26732, "fi": 2.4452, "ru": 1.32086, "el": 1.38446, "hi": 1.20471, "ta": 0.99278, "th": 1.13183, "zh": 3.19927, "my": 1.13132}, "bytes_per_chunk": {"en": 5.739, "de": 6.283, "vi": 5.71, "tr": 6.354, "fi": 6.314, "ru": 6.273, "el": 5.635, "hi": 10.108, "ta": 12.22, "th": 16.1, "zh": 11.415, "my": 18.997}, "bpb_mean": 1.77513, "t": 252.2} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "none", "lm_loss": 1.45634, "bpb": 2.10105, "chunk_loss": 0.0, "F": 0.13651, "G": 0.1594, "chunks": 139.78, "bytes_per_chunk": 7.326, "boundary_entropy": 0.12916, "router_noise": 0.0, "grad_norm": 0.6819, "lr": 6.82e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 252.2} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "none", "lm_loss": 1.34264, "bpb": 1.93702, "chunk_loss": 0.0, "F": 0.13943, "G": 0.15765, "chunks": 142.78, "bytes_per_chunk": 7.172, "boundary_entropy": 0.1253, "router_noise": 0.0, "grad_norm": 0.7393, "lr": 6.57e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 254.1} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "none", "lm_loss": 1.39673, "bpb": 2.01506, "chunk_loss": 0.0, "F": 0.15027, "G": 0.17024, "chunks": 153.88, "bytes_per_chunk": 6.655, "boundary_entropy": 0.11909, "router_noise": 0.0, "grad_norm": 0.7358, "lr": 6.32e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 256.0} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "none", "lm_loss": 1.36767, "bpb": 1.97313, "chunk_loss": 0.0, "F": 0.15063, "G": 0.16929, "chunks": 154.25, "bytes_per_chunk": 6.639, "boundary_entropy": 0.11513, "router_noise": 0.0, "grad_norm": 0.6913, "lr": 6.08e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 257.9} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.184385, "boundary_entropy": 0.090865, "bpb": 2.031533}, "de": {"chunks": 3288.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160132, "cps": 25.6875, "mean_p": 0.184128, "boundary_entropy": 0.125599, "bpb": 2.254401}, "vi": {"chunks": 4204.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228044, "cps": 32.84375, "mean_p": 0.184009, "boundary_entropy": 0.086126, "bpb": 1.889612}, "tr": {"chunks": 3024.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167044, "cps": 23.625, "mean_p": 0.179077, "boundary_entropy": 0.126963, "bpb": 2.245087}, "fi": {"chunks": 3017.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160385, "cps": 23.570312, "mean_p": 0.180595, "boundary_entropy": 0.133552, "bpb": 2.415772}, "ru": {"chunks": 5645.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.297889, "cps": 44.101562, "mean_p": 0.175913, "boundary_entropy": 0.110962, "bpb": 1.331342}, "el": {"chunks": 6684.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325683, "cps": 52.21875, "mean_p": 0.185647, "boundary_entropy": 0.113254, "bpb": 1.3827}, "hi": {"chunks": 4006.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.238098, "cps": 31.296875, "mean_p": 0.100366, "boundary_entropy": 0.167393, "bpb": 1.19435}, "ta": {"chunks": 4274.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.228617, "cps": 33.390625, "mean_p": 0.093172, "boundary_entropy": 0.136243, "bpb": 0.977901}, "th": {"chunks": 2872.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.178551, "cps": 22.4375, "mean_p": 0.103841, "boundary_entropy": 0.191993, "bpb": 1.105129}, "zh": {"chunks": 1648.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.279038, "cps": 12.875, "mean_p": 0.12814, "boundary_entropy": 0.170132, "bpb": 3.179834}, "my": {"chunks": 2853.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.15057, "cps": 22.289062, "mean_p": 0.101317, "boundary_entropy": 0.204715, "bpb": 1.115496}}, "bpb": {"en": 2.03153, "de": 2.2544, "vi": 1.88961, "tr": 2.24509, "fi": 2.41577, "ru": 1.33134, "el": 1.3827, "hi": 1.19435, "ta": 0.9779, "th": 1.10513, "zh": 3.17983, "my": 1.1155}, "gini_cpc": 0.14832, "gini_cps": 0.187968, "bpic": 8.643327, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9652876237936849, "vi": 1.374665083482168, "tr": 1.0069515429748193, "fi": 0.96680917467591, "ru": 1.7956929114500175, "el": 1.9632379132582856, "hi": 1.4352686815056068, "ta": 1.3781179347654402, "th": 1.0763182560089772, "zh": 1.68205851962468, "my": 0.9076443988063809}, "t": 260.8} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.04423, "de": 2.2626, "vi": 1.88832, "tr": 2.2501, "fi": 2.4247, "ru": 1.30987, "el": 1.37377, "hi": 1.19087, "ta": 0.98618, "th": 1.12321, "zh": 3.17789, "my": 1.12426}, "bytes_per_chunk": {"en": 5.739, "de": 6.294, "vi": 5.706, "tr": 6.381, "fi": 6.289, "ru": 6.279, "el": 5.639, "hi": 10.074, "ta": 11.6, "th": 15.52, "zh": 10.936, "my": 16.804}, "bpb_mean": 1.763, "t": 262.1} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "none", "lm_loss": 1.25072, "bpb": 1.80441, "chunk_loss": 0.0, "F": 0.15393, "G": 0.17291, "chunks": 157.62, "bytes_per_chunk": 6.496, "boundary_entropy": 0.11698, "router_noise": 0.0, "grad_norm": 0.7065, "lr": 5.85e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 262.1} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "none", "lm_loss": 1.33386, "bpb": 1.92435, "chunk_loss": 0.0, "F": 0.14847, "G": 0.16588, "chunks": 152.03, "bytes_per_chunk": 6.735, "boundary_entropy": 0.11972, "router_noise": 0.0, "grad_norm": 0.7313, "lr": 5.62e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 264.0} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "none", "lm_loss": 1.37435, "bpb": 1.98276, "chunk_loss": 0.0, "F": 0.15054, "G": 0.16885, "chunks": 154.16, "bytes_per_chunk": 6.643, "boundary_entropy": 0.11913, "router_noise": 0.0, "grad_norm": 0.7384, "lr": 5.41e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 265.9} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "none", "lm_loss": 1.27149, "bpb": 1.83437, "chunk_loss": 0.0, "F": 0.1344, "G": 0.15557, "chunks": 137.62, "bytes_per_chunk": 7.441, "boundary_entropy": 0.13623, "router_noise": 0.0, "grad_norm": 0.7457, "lr": 5.2e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 267.8} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.183554, "boundary_entropy": 0.088081, "bpb": 2.021865}, "de": {"chunks": 3302.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160814, "cps": 25.796875, "mean_p": 0.185635, "boundary_entropy": 0.128954, "bpb": 2.241289}, "vi": {"chunks": 4202.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227936, "cps": 32.828125, "mean_p": 0.183506, "boundary_entropy": 0.086822, "bpb": 1.881037}, "tr": {"chunks": 3070.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169585, "cps": 23.984375, "mean_p": 0.182085, "boundary_entropy": 0.129377, "bpb": 2.23528}, "fi": {"chunks": 3008.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159906, "cps": 23.5, "mean_p": 0.177334, "boundary_entropy": 0.126995, "bpb": 2.401817}, "ru": {"chunks": 5626.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.296887, "cps": 43.953125, "mean_p": 0.175915, "boundary_entropy": 0.114028, "bpb": 1.324819}, "el": {"chunks": 6711.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326999, "cps": 52.429688, "mean_p": 0.187499, "boundary_entropy": 0.112802, "bpb": 1.374726}, "hi": {"chunks": 3956.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.235126, "cps": 30.90625, "mean_p": 0.099658, "boundary_entropy": 0.174529, "bpb": 1.193076}, "ta": {"chunks": 4356.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.233003, "cps": 34.03125, "mean_p": 0.096173, "boundary_entropy": 0.141177, "bpb": 0.97288}, "th": {"chunks": 2852.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177308, "cps": 22.28125, "mean_p": 0.109119, "boundary_entropy": 0.203331, "bpb": 1.101437}, "zh": {"chunks": 1593.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.269726, "cps": 12.445312, "mean_p": 0.123827, "boundary_entropy": 0.170586, "bpb": 3.17837}, "my": {"chunks": 3388.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.178805, "cps": 26.46875, "mean_p": 0.102767, "boundary_entropy": 0.197534, "bpb": 1.109839}}, "bpb": {"en": 2.02186, "de": 2.24129, "vi": 1.88104, "tr": 2.23528, "fi": 2.40182, "ru": 1.32482, "el": 1.37473, "hi": 1.19308, "ta": 0.97288, "th": 1.10144, "zh": 3.17837, "my": 1.10984}, "gini_cpc": 0.13877, "gini_cps": 0.185854, "bpic": 8.537573, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9690620652141212, "vi": 1.3735353382814666, "tr": 1.0219149563218044, "fi": 0.9635913212211102, "ru": 1.7890292649412727, "el": 1.9704858699846655, "hi": 1.4168639201820927, "ta": 1.4040718517623996, "th": 1.0684529124621018, "zh": 1.6253588655323052, "my": 1.0774743912045035}, "t": 270.6} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.03473, "de": 2.25022, "vi": 1.88151, "tr": 2.24007, "fi": 2.4106, "ru": 1.30388, "el": 1.36541, "hi": 1.19042, "ta": 0.97983, "th": 1.11801, "zh": 3.17605, "my": 1.11882}, "bytes_per_chunk": {"en": 5.739, "de": 6.265, "vi": 5.71, "tr": 6.312, "fi": 6.31, "ru": 6.292, "el": 5.628, "hi": 10.166, "ta": 11.411, "th": 15.662, "zh": 11.506, "my": 14.194}, "bpb_mean": 1.7558, "t": 271.9} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "none", "lm_loss": 1.46209, "bpb": 2.10936, "chunk_loss": 0.0, "F": 0.15671, "G": 0.17449, "chunks": 160.47, "bytes_per_chunk": 6.381, "boundary_entropy": 0.10684, "router_noise": 0.0, "grad_norm": 0.7478, "lr": 5e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 272.0} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "none", "lm_loss": 1.27387, "bpb": 1.83781, "chunk_loss": 0.0, "F": 0.13727, "G": 0.15675, "chunks": 140.56, "bytes_per_chunk": 7.285, "boundary_entropy": 0.13166, "router_noise": 0.0, "grad_norm": 0.7045, "lr": 4.81e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 273.9} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "none", "lm_loss": 1.4463, "bpb": 2.08657, "chunk_loss": 0.0, "F": 0.14886, "G": 0.16732, "chunks": 152.44, "bytes_per_chunk": 6.718, "boundary_entropy": 0.1178, "router_noise": 0.0, "grad_norm": 0.7695, "lr": 4.63e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 275.8} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "none", "lm_loss": 1.23494, "bpb": 1.78164, "chunk_loss": 0.0, "F": 0.13757, "G": 0.15523, "chunks": 140.88, "bytes_per_chunk": 7.269, "boundary_entropy": 0.1276, "router_noise": 0.0, "grad_norm": 0.6644, "lr": 4.45e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 277.7} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.183707, "boundary_entropy": 0.088706, "bpb": 2.017029}, "de": {"chunks": 3298.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160619, "cps": 25.765625, "mean_p": 0.18541, "boundary_entropy": 0.12858, "bpb": 2.235396}, "vi": {"chunks": 4204.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228044, "cps": 32.84375, "mean_p": 0.183804, "boundary_entropy": 0.086546, "bpb": 1.87572}, "tr": {"chunks": 3055.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168757, "cps": 23.867188, "mean_p": 0.181898, "boundary_entropy": 0.128972, "bpb": 2.227274}, "fi": {"chunks": 3012.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160119, "cps": 23.53125, "mean_p": 0.178592, "boundary_entropy": 0.129241, "bpb": 2.396285}, "ru": {"chunks": 5646.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.297942, "cps": 44.109375, "mean_p": 0.176679, "boundary_entropy": 0.114112, "bpb": 1.318421}, "el": {"chunks": 6710.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.32695, "cps": 52.421875, "mean_p": 0.188021, "boundary_entropy": 0.111405, "bpb": 1.364129}, "hi": {"chunks": 3976.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.236315, "cps": 31.0625, "mean_p": 0.09913, "boundary_entropy": 0.168584, "bpb": 1.187383}, "ta": {"chunks": 4273.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.228564, "cps": 33.382812, "mean_p": 0.093692, "boundary_entropy": 0.136811, "bpb": 0.9672}, "th": {"chunks": 2857.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177619, "cps": 22.320312, "mean_p": 0.108311, "boundary_entropy": 0.204943, "bpb": 1.095286}, "zh": {"chunks": 1617.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.273789, "cps": 12.632812, "mean_p": 0.127187, "boundary_entropy": 0.174433, "bpb": 3.166451}, "my": {"chunks": 3417.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.180336, "cps": 26.695312, "mean_p": 0.104549, "boundary_entropy": 0.199632, "bpb": 1.106218}}, "bpb": {"en": 2.01703, "de": 2.2354, "vi": 1.87572, "tr": 2.22727, "fi": 2.39629, "ru": 1.31842, "el": 1.36413, "hi": 1.18738, "ta": 0.9672, "th": 1.09529, "zh": 3.16645, "my": 1.10622}, "gini_cpc": 0.139628, "gini_cps": 0.184767, "bpic": 8.537573, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9682234134037631, "vi": 1.374665083482168, "tr": 1.0172741282367967, "fi": 0.9652069055763477, "ru": 1.7960110147115678, "el": 1.9708746855121329, "hi": 1.4245202889830983, "ta": 1.3777954925719995, "th": 1.070696816649599, "zh": 1.6504178557239733, "my": 1.0870735754368748}, "t": 280.5} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.02942, "de": 2.24543, "vi": 1.87677, "tr": 2.23316, "fi": 2.40478, "ru": 1.29733, "el": 1.3539, "hi": 1.1836, "ta": 0.97543, "th": 1.11163, "zh": 3.16384, "my": 1.11527}, "bytes_per_chunk": {"en": 5.74, "de": 6.276, "vi": 5.708, "tr": 6.328, "fi": 6.301, "ru": 6.282, "el": 5.613, "hi": 10.118, "ta": 11.603, "th": 15.688, "zh": 11.214, "my": 13.801}, "bpb_mean": 1.74921, "t": 281.8} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "none", "lm_loss": 1.34436, "bpb": 1.9395, "chunk_loss": 0.0, "F": 0.15195, "G": 0.1712, "chunks": 155.59, "bytes_per_chunk": 6.581, "boundary_entropy": 0.1172, "router_noise": 0.0, "grad_norm": 0.7567, "lr": 4.29e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 281.8} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "none", "lm_loss": 1.31028, "bpb": 1.89034, "chunk_loss": 0.0, "F": 0.14902, "G": 0.16834, "chunks": 152.59, "bytes_per_chunk": 6.711, "boundary_entropy": 0.11641, "router_noise": 0.0, "grad_norm": 0.7599, "lr": 4.14e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 283.7} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "none", "lm_loss": 1.2783, "bpb": 1.8442, "chunk_loss": 0.0, "F": 0.14258, "G": 0.16233, "chunks": 146.0, "bytes_per_chunk": 7.014, "boundary_entropy": 0.12247, "router_noise": 0.0, "grad_norm": 0.7016, "lr": 3.99e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 285.6} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "none", "lm_loss": 1.35438, "bpb": 1.95396, "chunk_loss": 0.0, "F": 0.14963, "G": 0.16894, "chunks": 153.22, "bytes_per_chunk": 6.683, "boundary_entropy": 0.11316, "router_noise": 0.0, "grad_norm": 0.7235, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 287.6} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.184194, "boundary_entropy": 0.089809, "bpb": 2.010622}, "de": {"chunks": 3291.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160279, "cps": 25.710938, "mean_p": 0.185732, "boundary_entropy": 0.129705, "bpb": 2.226431}, "vi": {"chunks": 4203.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22799, "cps": 32.835938, "mean_p": 0.1832, "boundary_entropy": 0.085748, "bpb": 1.870747}, "tr": {"chunks": 3053.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168646, "cps": 23.851562, "mean_p": 0.181844, "boundary_entropy": 0.129559, "bpb": 2.219773}, "fi": {"chunks": 3008.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159906, "cps": 23.5, "mean_p": 0.177551, "boundary_entropy": 0.128202, "bpb": 2.389036}, "ru": {"chunks": 5682.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.299842, "cps": 44.390625, "mean_p": 0.180032, "boundary_entropy": 0.11629, "bpb": 1.312717}, "el": {"chunks": 6702.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.32656, "cps": 52.359375, "mean_p": 0.187771, "boundary_entropy": 0.112078, "bpb": 1.355996}, "hi": {"chunks": 4002.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.23786, "cps": 31.265625, "mean_p": 0.103828, "boundary_entropy": 0.170203, "bpb": 1.182187}, "ta": {"chunks": 4242.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.226906, "cps": 33.140625, "mean_p": 0.096728, "boundary_entropy": 0.140308, "bpb": 0.966807}, "th": {"chunks": 2880.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.179049, "cps": 22.5, "mean_p": 0.110574, "boundary_entropy": 0.207325, "bpb": 1.093054}, "zh": {"chunks": 1596.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.270234, "cps": 12.46875, "mean_p": 0.126999, "boundary_entropy": 0.172151, "bpb": 3.161785}, "my": {"chunks": 3185.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.168092, "cps": 24.882812, "mean_p": 0.103808, "boundary_entropy": 0.202603, "bpb": 1.101228}}, "bpb": {"en": 2.01062, "de": 2.22643, "vi": 1.87075, "tr": 2.21977, "fi": 2.38904, "ru": 1.31272, "el": 1.356, "hi": 1.18219, "ta": 0.96681, "th": 1.09305, "zh": 3.16179, "my": 1.10123}, "gini_cpc": 0.141321, "gini_cps": 0.185921, "bpic": 8.579755, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9661683606767084, "vi": 1.3743380936906642, "tr": 1.0166081549940886, "fi": 0.9639250902966978, "ru": 1.8074627321273693, "el": 1.9685249094340258, "hi": 1.4338355625026056, "ta": 1.3677997845753385, "th": 1.0793163570006459, "zh": 1.6289838575976878, "my": 1.0132658290214942}, "t": 290.4} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.02306, "de": 2.23562, "vi": 1.87132, "tr": 2.22437, "fi": 2.39634, "ru": 1.29183, "el": 1.34657, "hi": 1.17851, "ta": 0.9745, "th": 1.11027, "zh": 3.15808, "my": 1.11081}, "bytes_per_chunk": {"en": 5.737, "de": 6.288, "vi": 5.708, "tr": 6.356, "fi": 6.304, "ru": 6.247, "el": 5.627, "hi": 10.077, "ta": 11.654, "th": 15.415, "zh": 11.205, "my": 15.066}, "bpb_mean": 1.74344, "t": 291.7} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "none", "lm_loss": 1.22728, "bpb": 1.7706, "chunk_loss": 0.0, "F": 0.14377, "G": 0.16411, "chunks": 147.22, "bytes_per_chunk": 6.956, "boundary_entropy": 0.12522, "router_noise": 0.0, "grad_norm": 0.7071, "lr": 3.73e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 291.7} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "none", "lm_loss": 1.28043, "bpb": 1.84727, "chunk_loss": 0.0, "F": 0.14029, "G": 0.16054, "chunks": 143.66, "bytes_per_chunk": 7.128, "boundary_entropy": 0.1323, "router_noise": 0.0, "grad_norm": 0.7345, "lr": 3.62e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 293.6} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "none", "lm_loss": 1.4543, "bpb": 2.09811, "chunk_loss": 0.0, "F": 0.15674, "G": 0.17686, "chunks": 160.5, "bytes_per_chunk": 6.38, "boundary_entropy": 0.11694, "router_noise": 0.0, "grad_norm": 0.7477, "lr": 3.51e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 295.5} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "none", "lm_loss": 1.30142, "bpb": 1.87755, "chunk_loss": 0.0, "F": 0.15146, "G": 0.17068, "chunks": 155.09, "bytes_per_chunk": 6.602, "boundary_entropy": 0.11453, "router_noise": 0.0, "grad_norm": 0.7445, "lr": 3.41e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 297.4} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.184642, "boundary_entropy": 0.09118, "bpb": 2.004998}, "de": {"chunks": 3291.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160279, "cps": 25.710938, "mean_p": 0.185399, "boundary_entropy": 0.12846, "bpb": 2.220705}, "vi": {"chunks": 4199.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227773, "cps": 32.804688, "mean_p": 0.184151, "boundary_entropy": 0.089216, "bpb": 1.865999}, "tr": {"chunks": 3050.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16848, "cps": 23.828125, "mean_p": 0.181888, "boundary_entropy": 0.130235, "bpb": 2.213951}, "fi": {"chunks": 3016.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160332, "cps": 23.5625, "mean_p": 0.17851, "boundary_entropy": 0.128658, "bpb": 2.380648}, "ru": {"chunks": 5672.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.299314, "cps": 44.3125, "mean_p": 0.180809, "boundary_entropy": 0.117481, "bpb": 1.307035}, "el": {"chunks": 6700.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326463, "cps": 52.34375, "mean_p": 0.188251, "boundary_entropy": 0.111062, "bpb": 1.349833}, "hi": {"chunks": 3996.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.237504, "cps": 31.21875, "mean_p": 0.102469, "boundary_entropy": 0.168413, "bpb": 1.175757}, "ta": {"chunks": 4341.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.232201, "cps": 33.914062, "mean_p": 0.097942, "boundary_entropy": 0.140984, "bpb": 0.960791}, "th": {"chunks": 2910.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.180914, "cps": 22.734375, "mean_p": 0.111896, "boundary_entropy": 0.208218, "bpb": 1.085908}, "zh": {"chunks": 1618.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.273959, "cps": 12.640625, "mean_p": 0.129091, "boundary_entropy": 0.176705, "bpb": 3.150014}, "my": {"chunks": 3041.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.160492, "cps": 23.757812, "mean_p": 0.103523, "boundary_entropy": 0.202805, "bpb": 1.101993}}, "bpb": {"en": 2.005, "de": 2.22071, "vi": 1.866, "tr": 2.21395, "fi": 2.38065, "ru": 1.30704, "el": 1.34983, "hi": 1.17576, "ta": 0.96079, "th": 1.08591, "zh": 3.15001, "my": 1.10199}, "gini_cpc": 0.143573, "gini_cps": 0.186308, "bpic": 8.581289, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9654995006139346, "vi": 1.3720796117593148, "tr": 1.0149061081136677, "fi": 0.9658196390139373, "ru": 1.803032629453363, "el": 1.9665750995678986, "hi": 1.4306947549679911, "ta": 1.3987525609909526, "th": 1.0898042622090394, "zh": 1.6502952626805765, "my": 0.9667843741772586}, "t": 300.3} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.019, "de": 2.23088, "vi": 1.86538, "tr": 2.21911, "fi": 2.38806, "ru": 1.28627, "el": 1.33974, "hi": 1.17263, "ta": 0.96848, "th": 1.10448, "zh": 3.14649, "my": 1.11098}, "bytes_per_chunk": {"en": 5.737, "de": 6.293, "vi": 5.707, "tr": 6.348, "fi": 6.287, "ru": 6.259, "el": 5.633, "hi": 10.095, "ta": 11.438, "th": 15.363, "zh": 11.303, "my": 15.722}, "bpb_mean": 1.73763, "t": 301.6} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "none", "lm_loss": 1.31222, "bpb": 1.89313, "chunk_loss": 0.0, "F": 0.15158, "G": 0.17242, "chunks": 155.22, "bytes_per_chunk": 6.597, "boundary_entropy": 0.11779, "router_noise": 0.0, "grad_norm": 0.7324, "lr": 3.33e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 301.6} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "none", "lm_loss": 1.21562, "bpb": 1.75377, "chunk_loss": 0.0, "F": 0.15094, "G": 0.17086, "chunks": 154.56, "bytes_per_chunk": 6.625, "boundary_entropy": 0.12116, "router_noise": 0.0, "grad_norm": 0.749, "lr": 3.25e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 303.5} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "none", "lm_loss": 1.27719, "bpb": 1.8426, "chunk_loss": 0.0, "F": 0.15714, "G": 0.17609, "chunks": 160.91, "bytes_per_chunk": 6.364, "boundary_entropy": 0.1115, "router_noise": 0.0, "grad_norm": 0.7391, "lr": 3.18e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 305.4} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "none", "lm_loss": 1.3016, "bpb": 1.8778, "chunk_loss": 0.0, "F": 0.14392, "G": 0.16484, "chunks": 147.38, "bytes_per_chunk": 6.948, "boundary_entropy": 0.12383, "router_noise": 0.0, "grad_norm": 0.7503, "lr": 3.13e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 307.3} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.184698, "boundary_entropy": 0.091491, "bpb": 2.003676}, "de": {"chunks": 3294.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160425, "cps": 25.734375, "mean_p": 0.186891, "boundary_entropy": 0.131899, "bpb": 2.214081}, "vi": {"chunks": 4202.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227936, "cps": 32.828125, "mean_p": 0.18457, "boundary_entropy": 0.087815, "bpb": 1.862532}, "tr": {"chunks": 3039.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167873, "cps": 23.742188, "mean_p": 0.18103, "boundary_entropy": 0.129031, "bpb": 2.208046}, "fi": {"chunks": 3006.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.1598, "cps": 23.484375, "mean_p": 0.178674, "boundary_entropy": 0.13026, "bpb": 2.375446}, "ru": {"chunks": 5669.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.299156, "cps": 44.289062, "mean_p": 0.179573, "boundary_entropy": 0.116157, "bpb": 1.303352}, "el": {"chunks": 6710.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.32695, "cps": 52.421875, "mean_p": 0.190254, "boundary_entropy": 0.114497, "bpb": 1.348612}, "hi": {"chunks": 3991.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.237207, "cps": 31.179688, "mean_p": 0.103331, "boundary_entropy": 0.166936, "bpb": 1.174381}, "ta": {"chunks": 4474.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.239315, "cps": 34.953125, "mean_p": 0.100542, "boundary_entropy": 0.140405, "bpb": 0.958894}, "th": {"chunks": 3005.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.18682, "cps": 23.476562, "mean_p": 0.113465, "boundary_entropy": 0.208566, "bpb": 1.084076}, "zh": {"chunks": 1599.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.270742, "cps": 12.492188, "mean_p": 0.12827, "boundary_entropy": 0.173763, "bpb": 3.148816}, "my": {"chunks": 2697.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.142337, "cps": 21.070312, "mean_p": 0.101789, "boundary_entropy": 0.204832, "bpb": 1.099044}}, "bpb": {"en": 2.00368, "de": 2.21408, "vi": 1.86253, "tr": 2.20805, "fi": 2.37545, "ru": 1.30335, "el": 1.34861, "hi": 1.17438, "ta": 0.95889, "th": 1.08408, "zh": 3.14882, "my": 1.09904}, "gini_cpc": 0.150677, "gini_cps": 0.193061, "bpic": 8.609974, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9667142467641778, "vi": 1.3735353382814666, "tr": 1.0115959453622032, "fi": 0.9629506355022132, "ru": 1.8027029688859, "el": 1.970192249679199, "hi": 1.4293993694253622, "ta": 1.4421068560112433, "th": 1.1257717398136802, "zh": 1.6314807444985286, "my": 0.8577179554541163}, "t": 310.1} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.01709, "de": 2.22512, "vi": 1.86233, "tr": 2.21275, "fi": 2.38348, "ru": 1.28234, "el": 1.33686, "hi": 1.17111, "ta": 0.96703, "th": 1.10134, "zh": 3.14447, "my": 1.10822}, "bytes_per_chunk": {"en": 5.738, "de": 6.279, "vi": 5.707, "tr": 6.35, "fi": 6.31, "ru": 6.258, "el": 5.623, "hi": 10.075, "ta": 11.074, "th": 14.995, "zh": 11.365, "my": 17.077}, "bpb_mean": 1.73435, "t": 311.4} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "none", "lm_loss": 1.34776, "bpb": 1.94441, "chunk_loss": 0.0, "F": 0.14362, "G": 0.16585, "chunks": 147.06, "bytes_per_chunk": 6.963, "boundary_entropy": 0.12648, "router_noise": 0.0, "grad_norm": 0.7387, "lr": 3.08e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 311.5} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "none", "lm_loss": 1.30037, "bpb": 1.87604, "chunk_loss": 0.0, "F": 0.14474, "G": 0.16574, "chunks": 148.22, "bytes_per_chunk": 6.909, "boundary_entropy": 0.12748, "router_noise": 0.0, "grad_norm": 0.7264, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 313.4} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "none", "lm_loss": 1.21472, "bpb": 1.75247, "chunk_loss": 0.0, "F": 0.1373, "G": 0.16214, "chunks": 140.59, "bytes_per_chunk": 7.283, "boundary_entropy": 0.12896, "router_noise": 0.0, "grad_norm": 0.7358, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 315.3} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "none", "lm_loss": 1.23395, "bpb": 1.78022, "chunk_loss": 0.0, "F": 0.142, "G": 0.16294, "chunks": 145.41, "bytes_per_chunk": 7.042, "boundary_entropy": 0.12225, "router_noise": 0.0, "grad_norm": 0.7593, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 317.2} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.184505, "boundary_entropy": 0.091087, "bpb": 2.001056}, "de": {"chunks": 3282.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15984, "cps": 25.640625, "mean_p": 0.18559, "boundary_entropy": 0.129535, "bpb": 2.211994}, "vi": {"chunks": 4204.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228044, "cps": 32.84375, "mean_p": 0.184244, "boundary_entropy": 0.087565, "bpb": 1.860585}, "tr": {"chunks": 3048.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16837, "cps": 23.8125, "mean_p": 0.181173, "boundary_entropy": 0.128722, "bpb": 2.206169}, "fi": {"chunks": 3012.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160119, "cps": 23.53125, "mean_p": 0.178967, "boundary_entropy": 0.129067, "bpb": 2.374406}, "ru": {"chunks": 5683.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.299894, "cps": 44.398438, "mean_p": 0.179483, "boundary_entropy": 0.113638, "bpb": 1.301023}, "el": {"chunks": 6731.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.327973, "cps": 52.585938, "mean_p": 0.190763, "boundary_entropy": 0.112705, "bpb": 1.343464}, "hi": {"chunks": 3993.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.237325, "cps": 31.195312, "mean_p": 0.102903, "boundary_entropy": 0.166223, "bpb": 1.171806}, "ta": {"chunks": 4241.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.226852, "cps": 33.132812, "mean_p": 0.096903, "boundary_entropy": 0.141992, "bpb": 0.959092}, "th": {"chunks": 2857.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177619, "cps": 22.320312, "mean_p": 0.109559, "boundary_entropy": 0.206358, "bpb": 1.082583}, "zh": {"chunks": 1699.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.287674, "cps": 13.273438, "mean_p": 0.133945, "boundary_entropy": 0.178156, "bpb": 3.142019}, "my": {"chunks": 2380.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.125607, "cps": 18.59375, "mean_p": 0.094372, "boundary_entropy": 0.196496, "bpb": 1.101421}}, "bpb": {"en": 2.00106, "de": 2.21199, "vi": 1.86058, "tr": 2.20617, "fi": 2.37441, "ru": 1.30102, "el": 1.34346, "hi": 1.17181, "ta": 0.95909, "th": 1.08258, "zh": 3.14202, "my": 1.10142}, "gini_cpc": 0.161287, "gini_cps": 0.197895, "bpic": 8.718927, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.963526150027638, "vi": 1.374665083482168, "tr": 1.014943221887318, "fi": 0.9652069055763477, "ru": 1.8077808353889193, "el": 1.9770428477171633, "hi": 1.430611044745853, "ta": 1.367477342381898, "th": 1.070696816649599, "zh": 1.73411251507423, "my": 0.757165674433644}, "t": 320.0} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.01431, "de": 2.22119, "vi": 1.86032, "tr": 2.21118, "fi": 2.38305, "ru": 1.28115, "el": 1.33288, "hi": 1.16799, "ta": 0.96709, "th": 1.10009, "zh": 3.13681, "my": 1.11032}, "bytes_per_chunk": {"en": 5.737, "de": 6.293, "vi": 5.707, "tr": 6.35, "fi": 6.292, "ru": 6.246, "el": 5.6, "hi": 10.107, "ta": 11.685, "th": 15.873, "zh": 10.475, "my": 19.93}, "bpb_mean": 1.7322, "t": 321.3} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.184505, "boundary_entropy": 0.091087, "bpb": 2.001056}, "de": {"chunks": 3282.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15984, "cps": 25.640625, "mean_p": 0.18559, "boundary_entropy": 0.129535, "bpb": 2.211994}, "vi": {"chunks": 4204.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228044, "cps": 32.84375, "mean_p": 0.184244, "boundary_entropy": 0.087565, "bpb": 1.860585}, "tr": {"chunks": 3048.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16837, "cps": 23.8125, "mean_p": 0.181173, "boundary_entropy": 0.128722, "bpb": 2.206169}, "fi": {"chunks": 3012.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160119, "cps": 23.53125, "mean_p": 0.178967, "boundary_entropy": 0.129067, "bpb": 2.374406}, "ru": {"chunks": 5683.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.299894, "cps": 44.398438, "mean_p": 0.179483, "boundary_entropy": 0.113638, "bpb": 1.301023}, "el": {"chunks": 6731.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.327973, "cps": 52.585938, "mean_p": 0.190763, "boundary_entropy": 0.112705, "bpb": 1.343464}, "hi": {"chunks": 3993.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.237325, "cps": 31.195312, "mean_p": 0.102903, "boundary_entropy": 0.166223, "bpb": 1.171806}, "ta": {"chunks": 4241.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.226852, "cps": 33.132812, "mean_p": 0.096903, "boundary_entropy": 0.141992, "bpb": 0.959092}, "th": {"chunks": 2857.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177619, "cps": 22.320312, "mean_p": 0.109559, "boundary_entropy": 0.206358, "bpb": 1.082583}, "zh": {"chunks": 1699.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.287674, "cps": 13.273438, "mean_p": 0.133945, "boundary_entropy": 0.178156, "bpb": 3.142019}, "my": {"chunks": 2380.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.125607, "cps": 18.59375, "mean_p": 0.094372, "boundary_entropy": 0.196496, "bpb": 1.101421}}, "bpb": {"en": 2.00106, "de": 2.21199, "vi": 1.86058, "tr": 2.20617, "fi": 2.37441, "ru": 1.30102, "el": 1.34346, "hi": 1.17181, "ta": 0.95909, "th": 1.08258, "zh": 3.14202, "my": 1.10142}, "gini_cpc": 0.161287, "gini_cps": 0.197895, "bpic": 8.718927, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.963526150027638, "vi": 1.374665083482168, "tr": 1.014943221887318, "fi": 0.9652069055763477, "ru": 1.8077808353889193, "el": 1.9770428477171633, "hi": 1.430611044745853, "ta": 1.367477342381898, "th": 1.070696816649599, "zh": 1.73411251507423, "my": 0.757165674433644}, "t": 322.3} |
| {"event": "final", "run": "pilot_Bpc_s1", "condition": "B", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.16589093834396368, "cps": 22.5546875, "bpb": 2.022277036500003, "bytes_per_chunk": 5.737448019260232}, "de": {"cpc": 0.1598402571470316, "cps": 25.640625, "bpb": 2.212262530285491, "bytes_per_chunk": 6.28281085226728}, "vi": {"cpc": 0.22804448060754, "cps": 32.84375, "bpb": 1.8635332016418173, "bytes_per_chunk": 5.69333680827035}, "tr": {"cpc": 0.16836988344473292, "cps": 23.8125, "bpb": 2.20278255634921, "bytes_per_chunk": 6.390327141533811}, "fi": {"cpc": 0.16011907926213387, "cps": 23.53125, "bpb": 2.3973121972336267, "bytes_per_chunk": 6.275441074378187}, "ru": {"cpc": 0.29989445910290236, "cps": 44.3984375, "bpb": 1.2820044161675894, "bytes_per_chunk": 6.228695662504603}, "el": {"cpc": 0.3279734931540223, "cps": 52.5859375, "bpb": 1.3292369306063332, "bytes_per_chunk": 5.610958904109589}, "hi": {"cpc": 0.2373254086181278, "cps": 31.1953125, "bpb": 1.1773838762501472, "bytes_per_chunk": 10.146855041610218}, "ta": {"cpc": 0.22685209949184273, "cps": 33.1328125, "bpb": 0.9680200115514627, "bytes_per_chunk": 11.766152741309275}, "th": {"cpc": 0.1776188995958968, "cps": 22.3203125, "bpb": 1.0968234512144357, "bytes_per_chunk": 15.853406307641135}, "zh": {"cpc": 0.2876735523196749, "cps": 13.2734375, "bpb": 3.1441993133138864, "bytes_per_chunk": 10.574799814437565}, "my": {"cpc": 0.1256069242136373, "cps": 18.59375, "bpb": 1.1201699240864111, "bytes_per_chunk": 19.878971714567378}}, "gini_cpc": 0.1612873717171932, "gini_cps": 0.1978951314264943, "bpic": 8.718926778290205, "seed": 1, "mean_boundary_entropy": 0.12611162662506104, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584}], "t": 324.7} |
|
|
