| {"event": "start", "run": "emactl_noema_s2", "condition": "A", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "global", "p3_objective": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 2.2} |
| {"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": 2.2} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10933.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.628225, "cps": 85.414062, "mean_p": 0.513249, "boundary_entropy": 0.686295, "bpb": 8.157895}, "de": {"chunks": 16287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.793211, "cps": 127.242188, "mean_p": 0.524631, "boundary_entropy": 0.686391, "bpb": 8.150604}, "vi": {"chunks": 5099.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276593, "cps": 39.835938, "mean_p": 0.481643, "boundary_entropy": 0.68728, "bpb": 7.885576}, "tr": {"chunks": 8187.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.452245, "cps": 63.960938, "mean_p": 0.498002, "boundary_entropy": 0.686826, "bpb": 8.139464}, "fi": {"chunks": 13540.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.719792, "cps": 105.78125, "mean_p": 0.517169, "boundary_entropy": 0.686661, "bpb": 8.097447}, "ru": {"chunks": 33477.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.766596, "cps": 261.539062, "mean_p": 0.530376, "boundary_entropy": 0.688543, "bpb": 8.304526}, "el": {"chunks": 20696.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.00843, "cps": 161.6875, "mean_p": 0.503648, "boundary_entropy": 0.690112, "bpb": 8.265181}, "hi": {"chunks": 875.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.052006, "cps": 6.835938, "mean_p": 0.461978, "boundary_entropy": 0.68743, "bpb": 8.053074}, "ta": {"chunks": 1261.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.067451, "cps": 9.851562, "mean_p": 0.477518, "boundary_entropy": 0.690053, "bpb": 8.127442}, "th": {"chunks": 3630.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.225676, "cps": 28.359375, "mean_p": 0.484845, "boundary_entropy": 0.690398, "bpb": 7.710377}, "zh": {"chunks": 11722.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.984761, "cps": 91.578125, "mean_p": 0.520316, "boundary_entropy": 0.68562, "bpb": 8.124258}, "my": {"chunks": 1628.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085919, "cps": 12.71875, "mean_p": 0.474628, "boundary_entropy": 0.689685, "bpb": 7.688162}}, "bpb": {"en": 8.15789, "de": 8.1506, "vi": 7.88558, "tr": 8.13946, "fi": 8.09745, "ru": 8.30453, "el": 8.26518, "hi": 8.05307, "ta": 8.12744, "th": 7.71038, "zh": 8.12426, "my": 7.68816}, "gini_cpc": 0.489538, "gini_cps": 0.463551, "bpic": 3.013947, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2626223171148012, "vi": 0.440277652395235, "tr": 0.7198781817505053, "fi": 1.1457544508645294, "ru": 2.8120438593717987, "el": 1.6052044042347315, "hi": 0.08278235025738971, "ta": 0.10736788276924242, "th": 0.3592281253258146, "zh": 3.159315851384177, "my": 0.13676526035979295}, "t": 3.9} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.65891, "bpb": 8.16408, "chunk_loss": 0.065103, "F": 0.62439, "G": 0.5091, "chunks": 639.38, "bytes_per_chunk": 1.602, "boundary_entropy": 0.69066, "router_noise": 0.0, "grad_norm": 7.3918, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 4.2} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10931.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.62811, "cps": 85.398438, "mean_p": 0.513238, "boundary_entropy": 0.6863, "bpb": 8.134143}, "de": {"chunks": 16277.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.792724, "cps": 127.164062, "mean_p": 0.524625, "boundary_entropy": 0.686395, "bpb": 8.129311}, "vi": {"chunks": 5093.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276268, "cps": 39.789062, "mean_p": 0.481599, "boundary_entropy": 0.687276, "bpb": 7.872901}, "tr": {"chunks": 8214.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.453737, "cps": 64.171875, "mean_p": 0.498129, "boundary_entropy": 0.686829, "bpb": 8.124924}, "fi": {"chunks": 13526.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.719047, "cps": 105.671875, "mean_p": 0.517125, "boundary_entropy": 0.686667, "bpb": 8.076828}, "ru": {"chunks": 33476.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.766544, "cps": 261.53125, "mean_p": 0.530429, "boundary_entropy": 0.688536, "bpb": 8.267624}, "el": {"chunks": 20697.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.008478, "cps": 161.695312, "mean_p": 0.503648, "boundary_entropy": 0.690112, "bpb": 8.259746}, "hi": {"chunks": 876.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.052065, "cps": 6.84375, "mean_p": 0.462018, "boundary_entropy": 0.687436, "bpb": 8.04565}, "ta": {"chunks": 1278.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.068361, "cps": 9.984375, "mean_p": 0.477594, "boundary_entropy": 0.690061, "bpb": 8.121249}, "th": {"chunks": 3636.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.226049, "cps": 28.40625, "mean_p": 0.484847, "boundary_entropy": 0.690397, "bpb": 7.707762}, "zh": {"chunks": 11708.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.982391, "cps": 91.46875, "mean_p": 0.520277, "boundary_entropy": 0.685623, "bpb": 8.120846}, "my": {"chunks": 1616.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085286, "cps": 12.625, "mean_p": 0.474602, "boundary_entropy": 0.689682, "bpb": 7.686833}}, "bpb": {"en": 8.13414, "de": 8.12931, "vi": 7.8729, "tr": 8.12492, "fi": 8.07683, "ru": 8.26762, "el": 8.25975, "hi": 8.04565, "ta": 8.12125, "th": 7.70776, "zh": 8.12085, "my": 7.68683}, "gini_cpc": 0.489259, "gini_cps": 0.463377, "bpic": 3.014113, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2620779588703526, "vi": 0.439840038099778, "tr": 0.7223844236781818, "fi": 1.1447791884373502, "ru": 2.812474352794417, "el": 1.6055756771512673, "hi": 0.08289212231309648, "ta": 0.10883525777403733, "th": 0.35988772594999, "zh": 3.1561199251172027, "my": 0.13578200154588205}, "t": 5.5} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10928.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.627938, "cps": 85.375, "mean_p": 0.513224, "boundary_entropy": 0.68631, "bpb": 8.088349}, "de": {"chunks": 16288.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.79326, "cps": 127.25, "mean_p": 0.524618, "boundary_entropy": 0.686403, "bpb": 8.087259}, "vi": {"chunks": 5073.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.275183, "cps": 39.632812, "mean_p": 0.481502, "boundary_entropy": 0.687266, "bpb": 7.851306}, "tr": {"chunks": 8276.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.457162, "cps": 64.65625, "mean_p": 0.498371, "boundary_entropy": 0.686836, "bpb": 8.095026}, "fi": {"chunks": 13522.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.718835, "cps": 105.640625, "mean_p": 0.517052, "boundary_entropy": 0.686677, "bpb": 8.035282}, "ru": {"chunks": 33485.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.767018, "cps": 261.601562, "mean_p": 0.530578, "boundary_entropy": 0.688517, "bpb": 8.193591}, "el": {"chunks": 20686.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.007942, "cps": 161.609375, "mean_p": 0.503627, "boundary_entropy": 0.690113, "bpb": 8.25449}, "hi": {"chunks": 896.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.053254, "cps": 7.0, "mean_p": 0.462176, "boundary_entropy": 0.687462, "bpb": 8.020977}, "ta": {"chunks": 1314.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.070286, "cps": 10.265625, "mean_p": 0.477804, "boundary_entropy": 0.69008, "bpb": 8.107603}, "th": {"chunks": 3678.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.22866, "cps": 28.734375, "mean_p": 0.484932, "boundary_entropy": 0.690402, "bpb": 7.693953}, "zh": {"chunks": 11696.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.980359, "cps": 91.375, "mean_p": 0.520224, "boundary_entropy": 0.685626, "bpb": 8.113938}, "my": {"chunks": 1622.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085603, "cps": 12.671875, "mean_p": 0.4746, "boundary_entropy": 0.689682, "bpb": 7.685914}}, "bpb": {"en": 8.08835, "de": 8.08726, "vi": 7.85131, "tr": 8.09503, "fi": 8.03528, "ru": 8.19359, "el": 8.25449, "hi": 8.02098, "ta": 8.1076, "th": 7.69395, "zh": 8.11394, "my": 7.68591}, "gini_cpc": 0.488189, "gini_cps": 0.462372, "bpic": 3.010897, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2632775764528321, "vi": 0.4382330770412923, "tr": 0.7280368541845592, "fi": 1.1447548232152567, "ru": 2.8140027843558415, "el": 1.605162884066293, "hi": 0.08480791195212103, "ta": 0.11193175922187758, "th": 0.3641447840011724, "zh": 3.1537506340178663, "my": 0.1363235564793033}, "t": 7.0} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10922.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.627593, "cps": 85.328125, "mean_p": 0.513078, "boundary_entropy": 0.686344, "bpb": 7.941215}, "de": {"chunks": 16312.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.794428, "cps": 127.4375, "mean_p": 0.5245, "boundary_entropy": 0.686437, "bpb": 7.955601}, "vi": {"chunks": 5035.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.273122, "cps": 39.335938, "mean_p": 0.481292, "boundary_entropy": 0.687244, "bpb": 7.781795}, "tr": {"chunks": 8479.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.468375, "cps": 66.242188, "mean_p": 0.499113, "boundary_entropy": 0.686855, "bpb": 7.997061}, "fi": {"chunks": 13476.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.716389, "cps": 105.28125, "mean_p": 0.51673, "boundary_entropy": 0.686715, "bpb": 7.902996}, "ru": {"chunks": 33530.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.769393, "cps": 261.953125, "mean_p": 0.531283, "boundary_entropy": 0.688427, "bpb": 7.949972}, "el": {"chunks": 20456.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.996735, "cps": 159.8125, "mean_p": 0.50336, "boundary_entropy": 0.690116, "bpb": 8.206426}, "hi": {"chunks": 938.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.05575, "cps": 7.328125, "mean_p": 0.462568, "boundary_entropy": 0.687524, "bpb": 7.939638}, "ta": {"chunks": 1398.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.074779, "cps": 10.921875, "mean_p": 0.478292, "boundary_entropy": 0.690125, "bpb": 8.032249}, "th": {"chunks": 3739.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.232453, "cps": 29.210938, "mean_p": 0.484905, "boundary_entropy": 0.690396, "bpb": 7.621764}, "zh": {"chunks": 11642.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.971216, "cps": 90.953125, "mean_p": 0.519956, "boundary_entropy": 0.685644, "bpb": 8.087028}, "my": {"chunks": 1584.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.083597, "cps": 12.375, "mean_p": 0.474424, "boundary_entropy": 0.689661, "bpb": 7.683507}}, "bpb": {"en": 7.94121, "de": 7.9556, "vi": 7.7818, "tr": 7.99706, "fi": 7.903, "ru": 7.94997, "el": 8.20643, "hi": 7.93964, "ta": 8.03225, "th": 7.62176, "zh": 8.08703, "my": 7.68351}, "gini_cpc": 0.485837, "gini_cps": 0.460143, "bpic": 3.009787, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.265833991441756, "vi": 0.43518937226153825, "tr": 0.7463044510919494, "fi": 1.1414872550727257, "ru": 2.819332431117461, "el": 1.5881876620924267, "hi": 0.08883205591560636, "ta": 0.11915263470276409, "th": 0.3703875225217251, "zh": 3.140914397588094, "my": 0.1332029211917681}, "t": 8.5} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10611.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.609722, "cps": 82.898438, "mean_p": 0.5114, "boundary_entropy": 0.686513, "bpb": 7.483427}, "de": {"chunks": 16140.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.786052, "cps": 126.09375, "mean_p": 0.522868, "boundary_entropy": 0.686672, "bpb": 7.541395}, "vi": {"chunks": 4838.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.262436, "cps": 37.796875, "mean_p": 0.480523, "boundary_entropy": 0.687161, "bpb": 7.560909}, "tr": {"chunks": 9128.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.504226, "cps": 71.3125, "mean_p": 0.501699, "boundary_entropy": 0.686913, "bpb": 7.66231}, "fi": {"chunks": 13001.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.691138, "cps": 101.570312, "mean_p": 0.514409, "boundary_entropy": 0.686914, "bpb": 7.489182}, "ru": {"chunks": 33816.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.784485, "cps": 264.1875, "mean_p": 0.535777, "boundary_entropy": 0.687824, "bpb": 7.383544}, "el": {"chunks": 19650.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.957462, "cps": 153.515625, "mean_p": 0.502313, "boundary_entropy": 0.690126, "bpb": 7.973729}, "hi": {"chunks": 1041.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.061872, "cps": 8.132812, "mean_p": 0.463589, "boundary_entropy": 0.687684, "bpb": 7.68495}, "ta": {"chunks": 1665.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.089061, "cps": 13.007812, "mean_p": 0.479387, "boundary_entropy": 0.690219, "bpb": 7.809563}, "th": {"chunks": 3770.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.23438, "cps": 29.453125, "mean_p": 0.48437, "boundary_entropy": 0.690345, "bpb": 7.383205}, "zh": {"chunks": 11528.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.951913, "cps": 90.0625, "mean_p": 0.519227, "boundary_entropy": 0.685689, "bpb": 8.009399}, "my": {"chunks": 1615.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085233, "cps": 12.617188, "mean_p": 0.474685, "boundary_entropy": 0.689693, "bpb": 7.572833}}, "bpb": {"en": 7.48343, "de": 7.5414, "vi": 7.56091, "tr": 7.66231, "fi": 7.48918, "ru": 7.38354, "el": 7.97373, "hi": 7.68495, "ta": 7.80956, "th": 7.38321, "zh": 8.0094, "my": 7.57283}, "gini_cpc": 0.48134, "gini_cps": 0.45469, "bpic": 3.026592, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2891959392454035, "vi": 0.4304181016690618, "tr": 0.8269759614317663, "fi": 1.133529117454797, "ru": 2.926717646769209, "el": 1.570324892938546, "hi": 0.10147602861870751, "ta": 0.14606850153154938, "th": 0.3844041703457451, "zh": 3.2013144197363634, "my": 0.13979027404078206}, "t": 10.2} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6651.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.382175, "cps": 51.960938, "mean_p": 0.496472, "boundary_entropy": 0.686791, "bpb": 6.904381}, "de": {"chunks": 11883.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.578727, "cps": 92.835938, "mean_p": 0.50655, "boundary_entropy": 0.687732, "bpb": 6.995052}, "vi": {"chunks": 3712.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.201356, "cps": 29.0, "mean_p": 0.475641, "boundary_entropy": 0.686647, "bpb": 7.200331}, "tr": {"chunks": 9634.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.532177, "cps": 75.265625, "mean_p": 0.503078, "boundary_entropy": 0.687052, "bpb": 7.19534}, "fi": {"chunks": 8333.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.442985, "cps": 65.101562, "mean_p": 0.498573, "boundary_entropy": 0.687378, "bpb": 6.973798}, "ru": {"chunks": 34324.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.811293, "cps": 268.15625, "mean_p": 0.556795, "boundary_entropy": 0.683948, "bpb": 6.514196}, "el": {"chunks": 18484.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.900648, "cps": 144.40625, "mean_p": 0.500426, "boundary_entropy": 0.690137, "bpb": 7.390491}, "hi": {"chunks": 881.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.052363, "cps": 6.882812, "mean_p": 0.462608, "boundary_entropy": 0.68756, "bpb": 6.86139}, "ta": {"chunks": 2307.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.123402, "cps": 18.023438, "mean_p": 0.480743, "boundary_entropy": 0.690314, "bpb": 6.967024}, "th": {"chunks": 2544.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.15816, "cps": 19.875, "mean_p": 0.479518, "boundary_entropy": 0.689922, "bpb": 6.640688}, "zh": {"chunks": 10889.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.843718, "cps": 85.070312, "mean_p": 0.516179, "boundary_entropy": 0.68585, "bpb": 7.78956}, "my": {"chunks": 1803.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.095155, "cps": 14.085938, "mean_p": 0.476156, "boundary_entropy": 0.689872, "bpb": 7.157893}}, "bpb": {"en": 6.90438, "de": 6.99505, "vi": 7.20033, "tr": 7.19534, "fi": 6.9738, "ru": 6.5142, "el": 7.39049, "hi": 6.86139, "ta": 6.96702, "th": 6.64069, "zh": 7.78956, "my": 7.15789}, "gini_cpc": 0.511865, "gini_cps": 0.483203, "bpic": 3.443681, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.514296303906837, "vi": 0.5268682135320689, "tr": 1.3924937767151135, "fi": 1.1591153862579133, "ru": 4.739427141910138, "el": 2.356634802793728, "hi": 0.1370118113376241, "ta": 0.32289349615887253, "th": 0.41384071342912293, "zh": 4.82427135023655, "my": 0.2489829011413046}, "t": 12.2} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2293.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.131759, "cps": 17.914062, "mean_p": 0.481551, "boundary_entropy": 0.686311, "bpb": 6.390237}, "de": {"chunks": 3050.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.148541, "cps": 23.828125, "mean_p": 0.483972, "boundary_entropy": 0.68746, "bpb": 6.479172}, "vi": {"chunks": 3829.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.207703, "cps": 29.914062, "mean_p": 0.480437, "boundary_entropy": 0.687424, "bpb": 6.791571}, "tr": {"chunks": 7600.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.41982, "cps": 59.375, "mean_p": 0.496419, "boundary_entropy": 0.687394, "bpb": 6.743952}, "fi": {"chunks": 2139.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.11371, "cps": 16.710938, "mean_p": 0.48069, "boundary_entropy": 0.686889, "bpb": 6.485756}, "ru": {"chunks": 19204.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.013404, "cps": 150.03125, "mean_p": 0.50387, "boundary_entropy": 0.690079, "bpb": 5.908693}, "el": {"chunks": 19164.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.933782, "cps": 149.71875, "mean_p": 0.500982, "boundary_entropy": 0.690219, "bpb": 6.297452}, "hi": {"chunks": 2133.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.126776, "cps": 16.664062, "mean_p": 0.480626, "boundary_entropy": 0.689953, "bpb": 5.885484}, "ta": {"chunks": 1107.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.059214, "cps": 8.648438, "mean_p": 0.478664, "boundary_entropy": 0.690154, "bpb": 6.054824}, "th": {"chunks": 3010.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.187131, "cps": 23.515625, "mean_p": 0.481307, "boundary_entropy": 0.690111, "bpb": 5.73348}, "zh": {"chunks": 4189.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.709279, "cps": 32.726562, "mean_p": 0.486761, "boundary_entropy": 0.685658, "bpb": 7.359928}, "my": {"chunks": 3434.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.181233, "cps": 26.828125, "mean_p": 0.485155, "boundary_entropy": 0.690742, "bpb": 5.879131}}, "bpb": {"en": 6.39024, "de": 6.47917, "vi": 6.79157, "tr": 6.74395, "fi": 6.48576, "ru": 5.90869, "el": 6.29745, "hi": 5.88548, "ta": 6.05482, "th": 5.73348, "zh": 7.35993, "my": 5.87913}, "gini_cpc": 0.464241, "gini_cps": 0.47187, "bpic": 5.393819, "worst_lang": "ru", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1273726577889815, "vi": 1.5763849860384507, "tr": 3.1862739016374753, "fi": 0.863016236776586, "ru": 7.691349535755833, "el": 7.0870481352237755, "hi": 0.9621789424367333, "ta": 0.4494094670723029, "th": 1.4202522821887267, "zh": 5.383156218760287, "my": 1.3754885521841813}, "t": 14.9} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.27949, "bpb": 6.174, "chunk_loss": 0.036801, "F": 0.25571, "G": 0.48654, "chunks": 261.84, "bytes_per_chunk": 3.911, "boundary_entropy": 0.69113, "router_noise": 0.0, "grad_norm": 1.7797, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.241, "alpha": 0.03, "t": 16.3} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5231.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.30058, "cps": 40.867188, "mean_p": 0.49601, "boundary_entropy": 0.687459, "bpb": 5.771236}, "de": {"chunks": 7105.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.346028, "cps": 55.507812, "mean_p": 0.496533, "boundary_entropy": 0.688317, "bpb": 5.907513}, "vi": {"chunks": 6235.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.338215, "cps": 48.710938, "mean_p": 0.488896, "boundary_entropy": 0.688249, "bpb": 6.208939}, "tr": {"chunks": 6087.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.336243, "cps": 47.554688, "mean_p": 0.493694, "boundary_entropy": 0.687747, "bpb": 6.20941}, "fi": {"chunks": 2871.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152623, "cps": 22.429688, "mean_p": 0.488811, "boundary_entropy": 0.687767, "bpb": 5.897913}, "ru": {"chunks": 13565.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.715831, "cps": 105.976562, "mean_p": 0.497315, "boundary_entropy": 0.69017, "bpb": 5.427501}, "el": {"chunks": 10553.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.514204, "cps": 82.445312, "mean_p": 0.494748, "boundary_entropy": 0.690372, "bpb": 5.667946}, "hi": {"chunks": 21712.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.290461, "cps": 169.625, "mean_p": 0.502773, "boundary_entropy": 0.690856, "bpb": 5.147764}, "ta": {"chunks": 13082.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.699759, "cps": 102.203125, "mean_p": 0.49365, "boundary_entropy": 0.691065, "bpb": 5.23928}, "th": {"chunks": 3729.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.231831, "cps": 29.132812, "mean_p": 0.474092, "boundary_entropy": 0.689124, "bpb": 5.014419}, "zh": {"chunks": 6389.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.081781, "cps": 49.914062, "mean_p": 0.498847, "boundary_entropy": 0.68693, "bpb": 6.95011}, "my": {"chunks": 16997.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.897034, "cps": 132.789062, "mean_p": 0.498273, "boundary_entropy": 0.69134, "bpb": 5.106488}}, "bpb": {"en": 5.77124, "de": 5.90751, "vi": 6.20894, "tr": 6.20941, "fi": 5.89791, "ru": 5.4275, "el": 5.66795, "hi": 5.14776, "ta": 5.23928, "th": 5.01442, "zh": 6.95011, "my": 5.10649}, "gini_cpc": 0.331804, "gini_cps": 0.318232, "bpic": 3.379663, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1512007802240432, "vi": 1.1252077533027038, "tr": 1.118644651461428, "fi": 0.5077626001316398, "ru": 2.38149669945106, "el": 1.7107025122151513, "hi": 4.293230021164829, "ta": 2.328027335546533, "th": 0.7712775996995982, "zh": 3.5989751308704783, "my": 2.9843399902345698}, "t": 18.2} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.31436, "bpb": 4.78162, "chunk_loss": 0.036621, "F": 0.26279, "G": 0.47715, "chunks": 269.09, "bytes_per_chunk": 3.805, "boundary_entropy": 0.68835, "router_noise": 0.0, "grad_norm": 1.7896, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 20.5} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4018.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.23088, "cps": 31.390625, "mean_p": 0.414038, "boundary_entropy": 0.651251, "bpb": 4.378883}, "de": {"chunks": 5392.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.262602, "cps": 42.125, "mean_p": 0.414771, "boundary_entropy": 0.653505, "bpb": 4.597219}, "vi": {"chunks": 4991.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.270735, "cps": 38.992188, "mean_p": 0.46169, "boundary_entropy": 0.680449, "bpb": 4.808207}, "tr": {"chunks": 7255.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.400762, "cps": 56.679688, "mean_p": 0.447865, "boundary_entropy": 0.666669, "bpb": 4.866776}, "fi": {"chunks": 6268.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.333209, "cps": 48.96875, "mean_p": 0.438018, "boundary_entropy": 0.665854, "bpb": 4.637474}, "ru": {"chunks": 5785.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305277, "cps": 45.195312, "mean_p": 0.486805, "boundary_entropy": 0.68811, "bpb": 3.660402}, "el": {"chunks": 8166.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.397895, "cps": 63.796875, "mean_p": 0.486801, "boundary_entropy": 0.689673, "bpb": 4.213242}, "hi": {"chunks": 17396.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.033938, "cps": 135.90625, "mean_p": 0.495507, "boundary_entropy": 0.690039, "bpb": 3.445918}, "ta": {"chunks": 11731.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.627494, "cps": 91.648438, "mean_p": 0.483461, "boundary_entropy": 0.689285, "bpb": 3.518847}, "th": {"chunks": 14217.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.883867, "cps": 111.070312, "mean_p": 0.484831, "boundary_entropy": 0.689518, "bpb": 3.27333}, "zh": {"chunks": 5796.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.981375, "cps": 45.28125, "mean_p": 0.496179, "boundary_entropy": 0.686659, "bpb": 6.272491}, "my": {"chunks": 45630.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.40817, "cps": 356.484375, "mean_p": 0.515004, "boundary_entropy": 0.690822, "bpb": 3.653501}}, "bpb": {"en": 4.37888, "de": 4.59722, "vi": 4.80821, "tr": 4.86678, "fi": 4.63747, "ru": 3.6604, "el": 4.21324, "hi": 3.44592, "ta": 3.51885, "th": 3.27333, "zh": 6.27249, "my": 3.6535}, "gini_cpc": 0.406059, "gini_cps": 0.407979, "bpic": 2.808599, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1373959150413504, "vi": 1.1726235601307542, "tr": 1.7358054720723544, "fi": 1.4432158577031184, "ru": 1.32223404967882, "el": 1.7233866254100372, "hi": 4.478251848849568, "ta": 2.717839167448008, "th": 3.8282570057371608, "zh": 4.250589077899874, "my": 10.430407608447231}, "t": 23.2} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.61956, "bpb": 3.77922, "chunk_loss": 0.03893, "F": 0.29556, "G": 0.4567, "chunks": 302.66, "bytes_per_chunk": 3.383, "boundary_entropy": 0.67539, "router_noise": 0.0, "grad_norm": 1.8901, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 24.6} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3752.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.215595, "cps": 29.3125, "mean_p": 0.414331, "boundary_entropy": 0.64915, "bpb": 3.783432}, "de": {"chunks": 4131.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201188, "cps": 32.273438, "mean_p": 0.409931, "boundary_entropy": 0.650822, "bpb": 4.054472}, "vi": {"chunks": 6234.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.338161, "cps": 48.703125, "mean_p": 0.444346, "boundary_entropy": 0.67083, "bpb": 3.516267}, "tr": {"chunks": 5588.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.308678, "cps": 43.65625, "mean_p": 0.437355, "boundary_entropy": 0.663128, "bpb": 4.143601}, "fi": {"chunks": 3865.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.205465, "cps": 30.195312, "mean_p": 0.415366, "boundary_entropy": 0.656341, "bpb": 4.099937}, "ru": {"chunks": 9075.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.478892, "cps": 70.898438, "mean_p": 0.491289, "boundary_entropy": 0.689397, "bpb": 2.820828}, "el": {"chunks": 15740.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.766944, "cps": 122.96875, "mean_p": 0.495997, "boundary_entropy": 0.690005, "bpb": 2.98911}, "hi": {"chunks": 12570.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.747103, "cps": 98.203125, "mean_p": 0.475462, "boundary_entropy": 0.684066, "bpb": 2.432483}, "ta": {"chunks": 20234.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.082321, "cps": 158.078125, "mean_p": 0.50198, "boundary_entropy": 0.686813, "bpb": 2.260437}, "th": {"chunks": 11229.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.698104, "cps": 87.726562, "mean_p": 0.491357, "boundary_entropy": 0.690395, "bpb": 2.336668}, "zh": {"chunks": 8266.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.399594, "cps": 64.578125, "mean_p": 0.421019, "boundary_entropy": 0.614882, "bpb": 4.968162}, "my": {"chunks": 11341.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.598533, "cps": 88.601562, "mean_p": 0.490715, "boundary_entropy": 0.690315, "bpb": 2.436353}}, "bpb": {"en": 3.78343, "de": 4.05447, "vi": 3.51627, "tr": 4.1436, "fi": 4.09994, "ru": 2.82083, "el": 2.98911, "hi": 2.43248, "ta": 2.26044, "th": 2.33667, "zh": 4.96816, "my": 2.43635}, "gini_cpc": 0.335261, "gini_cps": 0.291459, "bpic": 3.425851, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9331771119494885, "vi": 1.5685015293172635, "tr": 1.4317498177627734, "fi": 0.9530131765211549, "ru": 2.2212564908214296, "el": 3.557338342222938, "hi": 3.4653052398799886, "ta": 5.020160088640966, "th": 3.2380332727322134, "zh": 6.491771856454749, "my": 2.7761905071359028}, "t": 28.5} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.78588, "de": 4.0534, "vi": 3.52068, "tr": 4.15188, "fi": 4.10154, "ru": 2.80767, "el": 2.98187, "hi": 2.4277, "ta": 2.27743, "th": 2.37068, "zh": 4.97431, "my": 2.45374}, "bytes_per_chunk": {"en": 4.501, "de": 4.858, "vi": 3.843, "tr": 3.285, "fi": 4.857, "ru": 3.452, "el": 2.544, "hi": 3.718, "ta": 2.986, "th": 5.27, "zh": 1.934, "my": 4.97}, "bpb_mean": 3.32557, "t": 30.0} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.45881, "bpb": 3.54732, "chunk_loss": 0.036794, "F": 0.28021, "G": 0.44153, "chunks": 286.94, "bytes_per_chunk": 3.569, "boundary_entropy": 0.66234, "router_noise": 0.0, "grad_norm": 1.5914, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.109, "alpha": 0.03, "t": 30.1} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.3844, "bpb": 3.43996, "chunk_loss": 0.039487, "F": 0.31366, "G": 0.44388, "chunks": 321.19, "bytes_per_chunk": 3.188, "boundary_entropy": 0.65907, "router_noise": 0.0, "grad_norm": 1.8474, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.053, "alpha": 0.03, "t": 32.7} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4099.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.235534, "cps": 32.023438, "mean_p": 0.416901, "boundary_entropy": 0.64983, "bpb": 3.590377}, "de": {"chunks": 3672.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178834, "cps": 28.6875, "mean_p": 0.405909, "boundary_entropy": 0.64956, "bpb": 3.839079}, "vi": {"chunks": 5769.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.312937, "cps": 45.070312, "mean_p": 0.447056, "boundary_entropy": 0.672892, "bpb": 3.154979}, "tr": {"chunks": 4549.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.251284, "cps": 35.539062, "mean_p": 0.428906, "boundary_entropy": 0.660418, "bpb": 3.821329}, "fi": {"chunks": 2868.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152464, "cps": 22.40625, "mean_p": 0.408856, "boundary_entropy": 0.656028, "bpb": 3.888884}, "ru": {"chunks": 6405.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.337995, "cps": 50.039062, "mean_p": 0.463808, "boundary_entropy": 0.682639, "bpb": 2.641722}, "el": {"chunks": 5687.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.277104, "cps": 44.429688, "mean_p": 0.476908, "boundary_entropy": 0.688534, "bpb": 2.741809}, "hi": {"chunks": 16265.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.966716, "cps": 127.070312, "mean_p": 0.467168, "boundary_entropy": 0.680241, "bpb": 2.077241}, "ta": {"chunks": 8919.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.477079, "cps": 69.679688, "mean_p": 0.472418, "boundary_entropy": 0.678296, "bpb": 1.910939}, "th": {"chunks": 16235.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.009325, "cps": 126.835938, "mean_p": 0.477818, "boundary_entropy": 0.687204, "bpb": 2.072951}, "zh": {"chunks": 7033.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.190823, "cps": 54.945312, "mean_p": 0.400776, "boundary_entropy": 0.601892, "bpb": 4.511015}, "my": {"chunks": 26850.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.417036, "cps": 209.765625, "mean_p": 0.502825, "boundary_entropy": 0.688742, "bpb": 2.205663}}, "bpb": {"en": 3.59038, "de": 3.83908, "vi": 3.15498, "tr": 3.82133, "fi": 3.88888, "ru": 2.64172, "el": 2.74181, "hi": 2.07724, "ta": 1.91094, "th": 2.07295, "zh": 4.51101, "my": 2.20566}, "gini_cpc": 0.399543, "gini_cps": 0.375796, "bpic": 3.542016, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.7592703963643125, "vi": 1.3286286063540078, "tr": 1.06687020681593, "fi": 0.6473117127968642, "ru": 1.435013945740032, "el": 1.1764909343050538, "hi": 4.104357638781901, "ta": 2.0255216815222026, "th": 4.2852624919017455, "zh": 5.055840641382803, "my": 6.016267193793386}, "t": 34.1} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.27477, "bpb": 3.2818, "chunk_loss": 0.033796, "F": 0.23303, "G": 0.43194, "chunks": 238.62, "bytes_per_chunk": 4.291, "boundary_entropy": 0.66468, "router_noise": 0.0, "grad_norm": 1.1149, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.074, "alpha": 0.03, "t": 36.4} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.05293, "bpb": 2.96174, "chunk_loss": 0.032491, "F": 0.20999, "G": 0.42345, "chunks": 215.03, "bytes_per_chunk": 4.762, "boundary_entropy": 0.6582, "router_noise": 0.0, "grad_norm": 0.9382, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.05, "alpha": 0.03, "t": 38.9} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4623.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.265644, "cps": 36.117188, "mean_p": 0.429558, "boundary_entropy": 0.656815, "bpb": 3.195843}, "de": {"chunks": 4912.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.239225, "cps": 38.375, "mean_p": 0.434775, "boundary_entropy": 0.664428, "bpb": 3.5202}, "vi": {"chunks": 7258.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.393708, "cps": 56.703125, "mean_p": 0.455397, "boundary_entropy": 0.675802, "bpb": 2.766924}, "tr": {"chunks": 4417.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.243993, "cps": 34.507812, "mean_p": 0.44013, "boundary_entropy": 0.668982, "bpb": 3.446349}, "fi": {"chunks": 5443.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.289352, "cps": 42.523438, "mean_p": 0.449678, "boundary_entropy": 0.671692, "bpb": 3.56252}, "ru": {"chunks": 4806.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.253615, "cps": 37.546875, "mean_p": 0.411456, "boundary_entropy": 0.637188, "bpb": 2.39065}, "el": {"chunks": 5879.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.286459, "cps": 45.929688, "mean_p": 0.45004, "boundary_entropy": 0.675097, "bpb": 2.510062}, "hi": {"chunks": 7774.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.462051, "cps": 60.734375, "mean_p": 0.422841, "boundary_entropy": 0.658619, "bpb": 1.780636}, "ta": {"chunks": 10174.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.54421, "cps": 79.484375, "mean_p": 0.39704, "boundary_entropy": 0.642575, "bpb": 1.560032}, "th": {"chunks": 13560.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.843021, "cps": 105.9375, "mean_p": 0.450364, "boundary_entropy": 0.677178, "bpb": 1.911786}, "zh": {"chunks": 2689.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.4553, "cps": 21.007812, "mean_p": 0.406461, "boundary_entropy": 0.636111, "bpb": 4.142384}, "my": {"chunks": 19425.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.025174, "cps": 151.757812, "mean_p": 0.47353, "boundary_entropy": 0.676679, "bpb": 1.939128}}, "bpb": {"en": 3.19584, "de": 3.5202, "vi": 2.76692, "tr": 3.44635, "fi": 3.56252, "ru": 2.39065, "el": 2.51006, "hi": 1.78064, "ta": 1.56003, "th": 1.91179, "zh": 4.14238, "my": 1.93913}, "gini_cpc": 0.277641, "gini_cps": 0.30009, "bpic": 4.219228, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9005465727081897, "vi": 1.4820881970027457, "tr": 0.9184955881735642, "fi": 1.089247765375106, "ru": 0.9547172705932307, "el": 1.0783576951365528, "hi": 1.7393608480628062, "ta": 2.048643973869966, "th": 3.1735025457916723, "zh": 1.7139477819606452, "my": 3.859205260971075}, "t": 42.3} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.20626, "de": 3.51886, "vi": 2.77111, "tr": 3.4557, "fi": 3.55498, "ru": 2.37556, "el": 2.50304, "hi": 1.77446, "ta": 1.57782, "th": 1.94226, "zh": 4.14905, "my": 1.95443}, "bytes_per_chunk": {"en": 3.681, "de": 4.2, "vi": 3.285, "tr": 4.209, "fi": 3.316, "ru": 7.469, "el": 6.439, "hi": 6.681, "ta": 4.938, "th": 3.44, "zh": 4.909, "my": 2.741}, "bpb_mean": 2.73196, "t": 43.7} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.20925, "bpb": 3.18727, "chunk_loss": 0.033732, "F": 0.23059, "G": 0.42455, "chunks": 236.12, "bytes_per_chunk": 4.337, "boundary_entropy": 0.65601, "router_noise": 0.0, "grad_norm": 1.0104, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.096, "alpha": 0.03, "t": 43.7} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.17034, "bpb": 3.13113, "chunk_loss": 0.032234, "F": 0.2045, "G": 0.4242, "chunks": 209.41, "bytes_per_chunk": 4.89, "boundary_entropy": 0.65947, "router_noise": 0.0, "grad_norm": 1.0488, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 46.1} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.05017, "bpb": 2.95777, "chunk_loss": 0.034904, "F": 0.2518, "G": 0.4313, "chunks": 257.84, "bytes_per_chunk": 3.971, "boundary_entropy": 0.6639, "router_noise": 0.0, "grad_norm": 0.8168, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 48.4} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.93026, "bpb": 2.78478, "chunk_loss": 0.033177, "F": 0.22198, "G": 0.4276, "chunks": 227.31, "bytes_per_chunk": 4.505, "boundary_entropy": 0.66113, "router_noise": 0.0, "grad_norm": 0.8102, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 50.7} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4324.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.248463, "cps": 33.78125, "mean_p": 0.439535, "boundary_entropy": 0.663434, "bpb": 2.884743}, "de": {"chunks": 4919.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.239566, "cps": 38.429688, "mean_p": 0.442578, "boundary_entropy": 0.668406, "bpb": 3.194928}, "vi": {"chunks": 5316.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.288365, "cps": 41.53125, "mean_p": 0.448395, "boundary_entropy": 0.674673, "bpb": 2.568255}, "tr": {"chunks": 5245.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.289731, "cps": 40.976562, "mean_p": 0.450779, "boundary_entropy": 0.670882, "bpb": 3.114931}, "fi": {"chunks": 3455.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183669, "cps": 26.992188, "mean_p": 0.432525, "boundary_entropy": 0.668157, "bpb": 3.363056}, "ru": {"chunks": 3951.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.208496, "cps": 30.867188, "mean_p": 0.419031, "boundary_entropy": 0.650762, "bpb": 2.158886}, "el": {"chunks": 8251.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.402037, "cps": 64.460938, "mean_p": 0.437432, "boundary_entropy": 0.668799, "bpb": 2.222094}, "hi": {"chunks": 8192.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.486895, "cps": 64.0, "mean_p": 0.428903, "boundary_entropy": 0.659994, "bpb": 1.656197}, "ta": {"chunks": 12347.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.660444, "cps": 96.460938, "mean_p": 0.429436, "boundary_entropy": 0.652602, "bpb": 1.405889}, "th": {"chunks": 17283.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.074479, "cps": 135.023438, "mean_p": 0.426783, "boundary_entropy": 0.665921, "bpb": 1.755019}, "zh": {"chunks": 1913.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.323908, "cps": 14.945312, "mean_p": 0.410354, "boundary_entropy": 0.647938, "bpb": 3.971288}, "my": {"chunks": 21875.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.154475, "cps": 170.898438, "mean_p": 0.484442, "boundary_entropy": 0.685511, "bpb": 1.680762}}, "bpb": {"en": 2.88474, "de": 3.19493, "vi": 2.56826, "tr": 3.11493, "fi": 3.36306, "ru": 2.15889, "el": 2.22209, "hi": 1.6562, "ta": 1.40589, "th": 1.75502, "zh": 3.97129, "my": 1.68076}, "gini_cpc": 0.344485, "gini_cps": 0.371147, "bpic": 3.953611, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9641905048226106, "vi": 1.1605938508854396, "tr": 1.1660935040077738, "fi": 0.7392215229629012, "ru": 0.8391435297620936, "el": 1.6180955997880224, "hi": 1.9596265083922897, "ta": 2.658118962091453, "th": 4.324505956302333, "zh": 1.303646858131698, "my": 4.646469818952381}, "t": 54.0} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.8941, "de": 3.19584, "vi": 2.57331, "tr": 3.12768, "fi": 3.35541, "ru": 2.14203, "el": 2.2117, "hi": 1.6501, "ta": 1.41963, "th": 1.78411, "zh": 3.98362, "my": 1.69366}, "bytes_per_chunk": {"en": 3.957, "de": 4.198, "vi": 4.462, "tr": 3.668, "fi": 5.769, "ru": 8.666, "el": 4.461, "hi": 5.899, "ta": 3.643, "th": 2.724, "zh": 7.604, "my": 2.458}, "bpb_mean": 2.5026, "t": 55.4} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.91644, "bpb": 2.76483, "chunk_loss": 0.033103, "F": 0.21884, "G": 0.43152, "chunks": 224.09, "bytes_per_chunk": 4.57, "boundary_entropy": 0.66563, "router_noise": 0.0, "grad_norm": 1.0762, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.094, "alpha": 0.03, "t": 55.4} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.94391, "bpb": 2.80447, "chunk_loss": 0.032905, "F": 0.21542, "G": 0.4335, "chunks": 220.59, "bytes_per_chunk": 4.642, "boundary_entropy": 0.6659, "router_noise": 0.0, "grad_norm": 0.7213, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 57.8} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.78488, "bpb": 2.57504, "chunk_loss": 0.033976, "F": 0.23431, "G": 0.4349, "chunks": 239.94, "bytes_per_chunk": 4.268, "boundary_entropy": 0.66512, "router_noise": 0.0, "grad_norm": 0.8955, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 60.1} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.86731, "bpb": 2.69396, "chunk_loss": 0.033389, "F": 0.22449, "G": 0.43423, "chunks": 229.88, "bytes_per_chunk": 4.455, "boundary_entropy": 0.66554, "router_noise": 0.0, "grad_norm": 0.7548, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 62.4} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4103.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.235764, "cps": 32.054688, "mean_p": 0.438408, "boundary_entropy": 0.663697, "bpb": 2.690794}, "de": {"chunks": 4676.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.227731, "cps": 36.53125, "mean_p": 0.443856, "boundary_entropy": 0.670568, "bpb": 3.026229}, "vi": {"chunks": 5625.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.305126, "cps": 43.945312, "mean_p": 0.449601, "boundary_entropy": 0.67459, "bpb": 2.362892}, "tr": {"chunks": 4120.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.227587, "cps": 32.1875, "mean_p": 0.44911, "boundary_entropy": 0.673439, "bpb": 3.005791}, "fi": {"chunks": 3459.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183882, "cps": 27.023438, "mean_p": 0.442092, "boundary_entropy": 0.672738, "bpb": 3.214705}, "ru": {"chunks": 4628.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.244222, "cps": 36.15625, "mean_p": 0.430632, "boundary_entropy": 0.658776, "bpb": 2.024693}, "el": {"chunks": 6274.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.305706, "cps": 49.015625, "mean_p": 0.444897, "boundary_entropy": 0.674772, "bpb": 2.126051}, "hi": {"chunks": 19163.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.13896, "cps": 149.710938, "mean_p": 0.453591, "boundary_entropy": 0.667041, "bpb": 1.568171}, "ta": {"chunks": 12553.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.671463, "cps": 98.070312, "mean_p": 0.418297, "boundary_entropy": 0.647402, "bpb": 1.343311}, "th": {"chunks": 15966.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.992602, "cps": 124.734375, "mean_p": 0.42009, "boundary_entropy": 0.660347, "bpb": 1.629769}, "zh": {"chunks": 3188.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.53979, "cps": 24.90625, "mean_p": 0.426811, "boundary_entropy": 0.658236, "bpb": 3.840238}, "my": {"chunks": 21359.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.127243, "cps": 166.867188, "mean_p": 0.481074, "boundary_entropy": 0.68348, "bpb": 1.504298}}, "bpb": {"en": 2.69079, "de": 3.02623, "vi": 2.36289, "tr": 3.00579, "fi": 3.21471, "ru": 2.02469, "el": 2.12605, "hi": 1.56817, "ta": 1.34331, "th": 1.62977, "zh": 3.84024, "my": 1.5043}, "gini_cpc": 0.363203, "gini_cps": 0.374443, "bpic": 3.651093, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9659278734995338, "vi": 1.2942017658792995, "tr": 0.9653154736775881, "fi": 0.7799401574467302, "ru": 1.0358735387003655, "el": 1.29666047386792, "hi": 4.830933173602904, "ta": 2.8480306746894635, "th": 4.210150908177227, "zh": 2.289535982484218, "my": 4.781235582422661}, "t": 65.6} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.69944, "de": 3.03061, "vi": 2.3685, "tr": 3.0168, "fi": 3.21183, "ru": 2.00829, "el": 2.11196, "hi": 1.56293, "ta": 1.35548, "th": 1.65611, "zh": 3.84987, "my": 1.51407}, "bytes_per_chunk": {"en": 4.218, "de": 4.26, "vi": 4.328, "tr": 4.631, "fi": 5.544, "ru": 6.955, "el": 6.039, "hi": 2.716, "ta": 4.322, "th": 2.873, "zh": 4.527, "my": 2.514}, "bpb_mean": 2.36549, "t": 67.0} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.85467, "bpb": 2.67572, "chunk_loss": 0.034903, "F": 0.24973, "G": 0.43487, "chunks": 255.72, "bytes_per_chunk": 4.004, "boundary_entropy": 0.66656, "router_noise": 0.0, "grad_norm": 0.803, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 67.0} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.65289, "bpb": 2.38462, "chunk_loss": 0.034792, "F": 0.24863, "G": 0.43773, "chunks": 254.59, "bytes_per_chunk": 4.022, "boundary_entropy": 0.66783, "router_noise": 0.0, "grad_norm": 0.7567, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 69.4} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.73314, "bpb": 2.5004, "chunk_loss": 0.033056, "F": 0.21991, "G": 0.43599, "chunks": 225.19, "bytes_per_chunk": 4.547, "boundary_entropy": 0.66815, "router_noise": 0.0, "grad_norm": 0.7981, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 71.6} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.89103, "bpb": 2.72818, "chunk_loss": 0.033039, "F": 0.21707, "G": 0.43647, "chunks": 222.28, "bytes_per_chunk": 4.607, "boundary_entropy": 0.66822, "router_noise": 0.0, "grad_norm": 0.8848, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 73.9} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4701.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.270126, "cps": 36.726562, "mean_p": 0.444732, "boundary_entropy": 0.664615, "bpb": 2.554994}, "de": {"chunks": 4521.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.220182, "cps": 35.320312, "mean_p": 0.444201, "boundary_entropy": 0.670902, "bpb": 2.89501}, "vi": {"chunks": 7742.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.419962, "cps": 60.484375, "mean_p": 0.458987, "boundary_entropy": 0.675531, "bpb": 2.271005}, "tr": {"chunks": 5424.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.299619, "cps": 42.375, "mean_p": 0.455651, "boundary_entropy": 0.672133, "bpb": 2.872704}, "fi": {"chunks": 3873.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.20589, "cps": 30.257812, "mean_p": 0.442029, "boundary_entropy": 0.67188, "bpb": 3.122131}, "ru": {"chunks": 9118.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.481161, "cps": 71.234375, "mean_p": 0.439859, "boundary_entropy": 0.660761, "bpb": 1.915617}, "el": {"chunks": 9998.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.487161, "cps": 78.109375, "mean_p": 0.454139, "boundary_entropy": 0.676585, "bpb": 2.006132}, "hi": {"chunks": 11131.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.661575, "cps": 86.960938, "mean_p": 0.439857, "boundary_entropy": 0.663824, "bpb": 1.531019}, "ta": {"chunks": 9429.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.504359, "cps": 73.664062, "mean_p": 0.424664, "boundary_entropy": 0.650246, "bpb": 1.325648}, "th": {"chunks": 4467.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.277712, "cps": 34.898438, "mean_p": 0.419175, "boundary_entropy": 0.663281, "bpb": 1.586718}, "zh": {"chunks": 2655.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.449543, "cps": 20.742188, "mean_p": 0.426563, "boundary_entropy": 0.66179, "bpb": 3.703664}, "my": {"chunks": 4933.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.260344, "cps": 38.539062, "mean_p": 0.464608, "boundary_entropy": 0.685199, "bpb": 1.409347}}, "bpb": {"en": 2.55499, "de": 2.89501, "vi": 2.271, "tr": 2.8727, "fi": 3.12213, "ru": 1.91562, "el": 2.00613, "hi": 1.53102, "ta": 1.32565, "th": 1.58672, "zh": 3.70366, "my": 1.40935}, "gini_cpc": 0.199238, "gini_cps": 0.230466, "bpic": 4.920774, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8151095263990654, "vi": 1.5546903183004361, "tr": 1.1091824732198765, "fi": 0.762200944422904, "ru": 1.7812473964165263, "el": 1.8034585132915921, "hi": 2.449136432986817, "ta": 1.8671277568423215, "th": 1.0280843691151764, "zh": 1.6641978315455215, "my": 0.9637882080524283}, "t": 77.1} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.56338, "de": 2.89556, "vi": 2.27734, "tr": 2.88516, "fi": 3.11963, "ru": 1.89676, "el": 1.99516, "hi": 1.52705, "ta": 1.33396, "th": 1.61047, "zh": 3.71575, "my": 1.41801}, "bytes_per_chunk": {"en": 3.666, "de": 4.5, "vi": 3.161, "tr": 3.53, "fi": 4.784, "ru": 3.782, "el": 3.683, "hi": 4.949, "ta": 5.356, "th": 10.567, "zh": 5.556, "my": 10.634}, "bpb_mean": 2.26985, "t": 78.4} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.56085, "bpb": 2.25183, "chunk_loss": 0.03401, "F": 0.23276, "G": 0.43588, "chunks": 238.34, "bytes_per_chunk": 4.296, "boundary_entropy": 0.66721, "router_noise": 0.0, "grad_norm": 0.776, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 78.5} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.77788, "bpb": 2.56493, "chunk_loss": 0.033767, "F": 0.22897, "G": 0.43942, "chunks": 234.47, "bytes_per_chunk": 4.367, "boundary_entropy": 0.66973, "router_noise": 0.0, "grad_norm": 0.8645, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 80.8} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.60539, "bpb": 2.31608, "chunk_loss": 0.033822, "F": 0.23056, "G": 0.44044, "chunks": 236.09, "bytes_per_chunk": 4.337, "boundary_entropy": 0.66961, "router_noise": 0.0, "grad_norm": 0.8186, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 83.1} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.62166, "bpb": 2.33956, "chunk_loss": 0.033554, "F": 0.22617, "G": 0.43658, "chunks": 231.59, "bytes_per_chunk": 4.422, "boundary_entropy": 0.66914, "router_noise": 0.0, "grad_norm": 0.8408, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 85.4} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4105.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.235879, "cps": 32.070312, "mean_p": 0.442642, "boundary_entropy": 0.664261, "bpb": 2.491302}, "de": {"chunks": 4936.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.240394, "cps": 38.5625, "mean_p": 0.4499, "boundary_entropy": 0.671836, "bpb": 2.80705}, "vi": {"chunks": 6218.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.337293, "cps": 48.578125, "mean_p": 0.452644, "boundary_entropy": 0.674746, "bpb": 2.234409}, "tr": {"chunks": 4684.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.258742, "cps": 36.59375, "mean_p": 0.456411, "boundary_entropy": 0.673492, "bpb": 2.80254}, "fi": {"chunks": 3821.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.203126, "cps": 29.851562, "mean_p": 0.45094, "boundary_entropy": 0.674954, "bpb": 3.050613}, "ru": {"chunks": 8562.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.451821, "cps": 66.890625, "mean_p": 0.438147, "boundary_entropy": 0.662756, "bpb": 1.839138}, "el": {"chunks": 6541.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.318716, "cps": 51.101562, "mean_p": 0.444287, "boundary_entropy": 0.676251, "bpb": 1.941267}, "hi": {"chunks": 10287.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.611412, "cps": 80.367188, "mean_p": 0.443532, "boundary_entropy": 0.667155, "bpb": 1.475832}, "ta": {"chunks": 8307.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.444343, "cps": 64.898438, "mean_p": 0.429751, "boundary_entropy": 0.656759, "bpb": 1.200449}, "th": {"chunks": 3607.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.224246, "cps": 28.179688, "mean_p": 0.423035, "boundary_entropy": 0.666863, "bpb": 1.477179}, "zh": {"chunks": 2924.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.49509, "cps": 22.84375, "mean_p": 0.432421, "boundary_entropy": 0.665415, "bpb": 3.648647}, "my": {"chunks": 7853.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.41445, "cps": 61.351562, "mean_p": 0.469937, "boundary_entropy": 0.686294, "bpb": 1.373711}}, "bpb": {"en": 2.4913, "de": 2.80705, "vi": 2.23441, "tr": 2.80254, "fi": 3.05061, "ru": 1.83914, "el": 1.94127, "hi": 1.47583, "ta": 1.20045, "th": 1.47718, "zh": 3.64865, "my": 1.37371}, "gini_cpc": 0.195682, "gini_cps": 0.211779, "bpic": 5.341791, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.019139660092636, "vi": 1.4299423691959139, "tr": 1.0969259070592332, "fi": 0.8611446602853916, "ru": 1.915477116201042, "el": 1.3511832807171633, "hi": 2.592057466250884, "ta": 1.8837779093135234, "th": 0.9506836738596001, "zh": 2.0989151600820484, "my": 1.7570461963082757}, "t": 88.7} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.49746, "de": 2.80639, "vi": 2.23951, "tr": 2.81272, "fi": 3.05176, "ru": 1.82048, "el": 1.92898, "hi": 1.47177, "ta": 1.21144, "th": 1.50046, "zh": 3.65531, "my": 1.38243}, "bytes_per_chunk": {"en": 4.421, "de": 4.046, "vi": 3.971, "tr": 4.129, "fi": 5.04, "ru": 4.039, "el": 5.276, "hi": 5.027, "ta": 6.487, "th": 11.693, "zh": 5.471, "my": 6.743}, "bpb_mean": 2.19823, "t": 89.9} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.70824, "bpb": 2.46447, "chunk_loss": 0.032809, "F": 0.21432, "G": 0.43816, "chunks": 219.47, "bytes_per_chunk": 4.666, "boundary_entropy": 0.67002, "router_noise": 0.0, "grad_norm": 0.8105, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 90.0} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.61584, "bpb": 2.33116, "chunk_loss": 0.033167, "F": 0.22049, "G": 0.43895, "chunks": 225.78, "bytes_per_chunk": 4.535, "boundary_entropy": 0.6711, "router_noise": 0.0, "grad_norm": 0.9951, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 92.2} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.72173, "bpb": 2.48393, "chunk_loss": 0.033799, "F": 0.23035, "G": 0.43866, "chunks": 235.88, "bytes_per_chunk": 4.341, "boundary_entropy": 0.66906, "router_noise": 0.0, "grad_norm": 0.7868, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 94.5} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.5164, "bpb": 2.18771, "chunk_loss": 0.033735, "F": 0.22998, "G": 0.43871, "chunks": 235.5, "bytes_per_chunk": 4.348, "boundary_entropy": 0.6693, "router_noise": 0.0, "grad_norm": 0.7654, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 96.7} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4730.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.271792, "cps": 36.953125, "mean_p": 0.446286, "boundary_entropy": 0.665254, "bpb": 2.419445}, "de": {"chunks": 5185.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.25252, "cps": 40.507812, "mean_p": 0.452345, "boundary_entropy": 0.671955, "bpb": 2.756551}, "vi": {"chunks": 5520.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.29943, "cps": 43.125, "mean_p": 0.449232, "boundary_entropy": 0.674018, "bpb": 2.174434}, "tr": {"chunks": 5165.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.285312, "cps": 40.351562, "mean_p": 0.458758, "boundary_entropy": 0.673345, "bpb": 2.715253}, "fi": {"chunks": 4288.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.227952, "cps": 33.5, "mean_p": 0.454253, "boundary_entropy": 0.675261, "bpb": 2.989795}, "ru": {"chunks": 7391.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.390026, "cps": 57.742188, "mean_p": 0.440932, "boundary_entropy": 0.666154, "bpb": 1.810393}, "el": {"chunks": 7020.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.342055, "cps": 54.84375, "mean_p": 0.448172, "boundary_entropy": 0.677415, "bpb": 1.9177}, "hi": {"chunks": 8814.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.523863, "cps": 68.859375, "mean_p": 0.442511, "boundary_entropy": 0.667328, "bpb": 1.43912}, "ta": {"chunks": 10913.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.583739, "cps": 85.257812, "mean_p": 0.438545, "boundary_entropy": 0.658888, "bpb": 1.162418}, "th": {"chunks": 7531.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.4682, "cps": 58.835938, "mean_p": 0.431575, "boundary_entropy": 0.66814, "bpb": 1.417487}, "zh": {"chunks": 2246.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.380291, "cps": 17.546875, "mean_p": 0.427927, "boundary_entropy": 0.665272, "bpb": 3.596684}, "my": {"chunks": 18039.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.952027, "cps": 140.929688, "mean_p": 0.466699, "boundary_entropy": 0.684164, "bpb": 1.333711}}, "bpb": {"en": 2.41944, "de": 2.75655, "vi": 2.17443, "tr": 2.71525, "fi": 2.98979, "ru": 1.81039, "el": 1.9177, "hi": 1.43912, "ta": 1.16242, "th": 1.41749, "zh": 3.59668, "my": 1.33371}, "gini_cpc": 0.233309, "gini_cps": 0.270495, "bpic": 4.419302, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9290933102171732, "vi": 1.1016887515761626, "tr": 1.0497424357847083, "fi": 0.8386985123971029, "ru": 1.435016740392822, "el": 1.2585174638797971, "hi": 1.9274403777224194, "ta": 2.14773979996008, "th": 1.722640136536805, "zh": 1.3991983642248649, "my": 3.5027735527208996}, "t": 99.8} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.42702, "de": 2.76028, "vi": 2.1806, "tr": 2.72786, "fi": 2.99217, "ru": 1.7924, "el": 1.90924, "hi": 1.43594, "ta": 1.16991, "th": 1.43815, "zh": 3.60384, "my": 1.34475}, "bytes_per_chunk": {"en": 3.75, "de": 3.853, "vi": 4.613, "tr": 3.791, "fi": 4.628, "ru": 4.643, "el": 5.389, "hi": 5.784, "ta": 4.051, "th": 6.291, "zh": 6.735, "my": 3.321}, "bpb_mean": 2.14851, "t": 101.2} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.71727, "bpb": 2.47749, "chunk_loss": 0.032699, "F": 0.21115, "G": 0.43816, "chunks": 216.22, "bytes_per_chunk": 4.736, "boundary_entropy": 0.67045, "router_noise": 0.0, "grad_norm": 0.8107, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 101.2} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.71434, "bpb": 2.47327, "chunk_loss": 0.032624, "F": 0.2103, "G": 0.43847, "chunks": 215.34, "bytes_per_chunk": 4.755, "boundary_entropy": 0.6711, "router_noise": 0.0, "grad_norm": 0.7485, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 103.4} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.68427, "bpb": 2.42988, "chunk_loss": 0.034947, "F": 0.24881, "G": 0.44419, "chunks": 254.78, "bytes_per_chunk": 4.019, "boundary_entropy": 0.6714, "router_noise": 0.0, "grad_norm": 0.8944, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 105.6} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.57525, "bpb": 2.2726, "chunk_loss": 0.035741, "F": 0.26172, "G": 0.44545, "chunks": 268.0, "bytes_per_chunk": 3.821, "boundary_entropy": 0.67172, "router_noise": 0.0, "grad_norm": 0.8749, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 107.9} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5117.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.29403, "cps": 39.976562, "mean_p": 0.446949, "boundary_entropy": 0.663615, "bpb": 2.351049}, "de": {"chunks": 5155.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.251059, "cps": 40.273438, "mean_p": 0.44875, "boundary_entropy": 0.670668, "bpb": 2.668827}, "vi": {"chunks": 6915.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.375102, "cps": 54.023438, "mean_p": 0.454985, "boundary_entropy": 0.674332, "bpb": 2.127762}, "tr": {"chunks": 5172.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.285699, "cps": 40.40625, "mean_p": 0.456414, "boundary_entropy": 0.67194, "bpb": 2.644989}, "fi": {"chunks": 5166.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.274627, "cps": 40.359375, "mean_p": 0.458556, "boundary_entropy": 0.675298, "bpb": 2.894612}, "ru": {"chunks": 7356.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.388179, "cps": 57.46875, "mean_p": 0.442128, "boundary_entropy": 0.667552, "bpb": 1.746868}, "el": {"chunks": 6395.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.311602, "cps": 49.960938, "mean_p": 0.445792, "boundary_entropy": 0.676174, "bpb": 1.82221}, "hi": {"chunks": 16638.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.988886, "cps": 129.984375, "mean_p": 0.457226, "boundary_entropy": 0.668346, "bpb": 1.399161}, "ta": {"chunks": 10369.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.55464, "cps": 81.007812, "mean_p": 0.436758, "boundary_entropy": 0.659666, "bpb": 1.124449}, "th": {"chunks": 11758.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.730992, "cps": 91.859375, "mean_p": 0.433344, "boundary_entropy": 0.670166, "bpb": 1.354248}, "zh": {"chunks": 2860.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.484253, "cps": 22.34375, "mean_p": 0.442769, "boundary_entropy": 0.67195, "bpb": 3.521134}, "my": {"chunks": 17890.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.944163, "cps": 139.765625, "mean_p": 0.459728, "boundary_entropy": 0.68247, "bpb": 1.2965}}, "bpb": {"en": 2.35105, "de": 2.66883, "vi": 2.12776, "tr": 2.64499, "fi": 2.89461, "ru": 1.74687, "el": 1.82221, "hi": 1.39916, "ta": 1.12445, "th": 1.35425, "zh": 3.52113, "my": 1.2965}, "gini_cpc": 0.270445, "gini_cps": 0.287994, "bpic": 3.807691, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8538566510678801, "vi": 1.2757269956253927, "tr": 0.9716652428947848, "fi": 0.9340093453193153, "ru": 1.3202045022463793, "el": 1.0597621561041002, "hi": 3.363215921290248, "ta": 1.8863405824958346, "th": 2.4861143127107925, "zh": 1.6469533339761535, "my": 3.211113583687006}, "t": 111.0} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.36045, "de": 2.67398, "vi": 2.13179, "tr": 2.65327, "fi": 2.90032, "ru": 1.72658, "el": 1.81255, "hi": 1.39574, "ta": 1.13196, "th": 1.37432, "zh": 3.53195, "my": 1.30692}, "bytes_per_chunk": {"en": 3.346, "de": 3.947, "vi": 3.65, "tr": 3.778, "fi": 3.553, "ru": 4.582, "el": 5.662, "hi": 2.674, "ta": 4.845, "th": 3.751, "zh": 5.365, "my": 3.115}, "bpb_mean": 2.08332, "t": 112.4} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.79253, "bpb": 2.58607, "chunk_loss": 0.035692, "F": 0.26013, "G": 0.44651, "chunks": 266.38, "bytes_per_chunk": 3.844, "boundary_entropy": 0.67322, "router_noise": 0.0, "grad_norm": 0.8362, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 112.5} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.57771, "bpb": 2.27615, "chunk_loss": 0.034228, "F": 0.23602, "G": 0.44149, "chunks": 241.69, "bytes_per_chunk": 4.237, "boundary_entropy": 0.67137, "router_noise": 0.0, "grad_norm": 0.7847, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 114.6} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.56354, "bpb": 2.25571, "chunk_loss": 0.033145, "F": 0.22025, "G": 0.43718, "chunks": 225.53, "bytes_per_chunk": 4.54, "boundary_entropy": 0.66993, "router_noise": 0.0, "grad_norm": 0.7339, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 116.8} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.52821, "bpb": 2.20475, "chunk_loss": 0.034344, "F": 0.23895, "G": 0.44255, "chunks": 244.69, "bytes_per_chunk": 4.185, "boundary_entropy": 0.67103, "router_noise": 0.0, "grad_norm": 0.9255, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 119.0} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4231.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.243119, "cps": 33.054688, "mean_p": 0.442546, "boundary_entropy": 0.662906, "bpb": 2.323319}, "de": {"chunks": 4771.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.232358, "cps": 37.273438, "mean_p": 0.445849, "boundary_entropy": 0.670337, "bpb": 2.625657}, "vi": {"chunks": 4251.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230594, "cps": 33.210938, "mean_p": 0.440154, "boundary_entropy": 0.671022, "bpb": 2.11059}, "tr": {"chunks": 5195.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.286969, "cps": 40.585938, "mean_p": 0.457405, "boundary_entropy": 0.672592, "bpb": 2.594063}, "fi": {"chunks": 4318.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.229547, "cps": 33.734375, "mean_p": 0.449766, "boundary_entropy": 0.673353, "bpb": 2.834582}, "ru": {"chunks": 7860.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.414776, "cps": 61.40625, "mean_p": 0.440165, "boundary_entropy": 0.667123, "bpb": 1.704451}, "el": {"chunks": 6606.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.321883, "cps": 51.609375, "mean_p": 0.44462, "boundary_entropy": 0.675281, "bpb": 1.780977}, "hi": {"chunks": 8685.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.516196, "cps": 67.851562, "mean_p": 0.438512, "boundary_entropy": 0.666997, "bpb": 1.378008}, "ta": {"chunks": 7994.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.427601, "cps": 62.453125, "mean_p": 0.429984, "boundary_entropy": 0.657377, "bpb": 1.119168}, "th": {"chunks": 16734.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.040348, "cps": 130.734375, "mean_p": 0.435699, "boundary_entropy": 0.66749, "bpb": 1.316412}, "zh": {"chunks": 2608.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.441585, "cps": 20.375, "mean_p": 0.439301, "boundary_entropy": 0.670922, "bpb": 3.516191}, "my": {"chunks": 3730.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.196855, "cps": 29.140625, "mean_p": 0.453839, "boundary_entropy": 0.683681, "bpb": 1.2683}}, "bpb": {"en": 2.32332, "de": 2.62566, "vi": 2.11059, "tr": 2.59406, "fi": 2.83458, "ru": 1.70445, "el": 1.78098, "hi": 1.37801, "ta": 1.11917, "th": 1.31641, "zh": 3.51619, "my": 1.2683}, "gini_cpc": 0.270294, "gini_cps": 0.272414, "bpic": 4.985269, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9557363504394694, "vi": 0.9484819224373396, "tr": 1.1803643801789527, "fi": 0.944173592290455, "ru": 1.7060604945654918, "el": 1.3239720565081774, "hi": 2.1232241472496747, "ta": 1.7588134143301877, "th": 4.279172503496588, "zh": 1.8163320205315403, "my": 0.8097044957084901}, "t": 122.2} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.33334, "de": 2.63115, "vi": 2.11724, "tr": 2.60089, "fi": 2.84118, "ru": 1.68435, "el": 1.77275, "hi": 1.37488, "ta": 1.12662, "th": 1.33785, "zh": 3.52375, "my": 1.27752}, "bytes_per_chunk": {"en": 4.154, "de": 4.142, "vi": 6.21, "tr": 3.781, "fi": 4.314, "ru": 4.437, "el": 5.432, "hi": 5.304, "ta": 6.25, "th": 2.764, "zh": 5.732, "my": 14.019}, "bpb_mean": 2.05179, "t": 123.5} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.46179, "bpb": 2.10891, "chunk_loss": 0.032636, "F": 0.21167, "G": 0.44014, "chunks": 216.75, "bytes_per_chunk": 4.724, "boundary_entropy": 0.67131, "router_noise": 0.0, "grad_norm": 0.7108, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 123.5} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.43975, "bpb": 2.07712, "chunk_loss": 0.031786, "F": 0.1958, "G": 0.44065, "chunks": 200.5, "bytes_per_chunk": 5.107, "boundary_entropy": 0.6714, "router_noise": 0.0, "grad_norm": 0.7222, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 125.7} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.54984, "bpb": 2.23594, "chunk_loss": 0.033226, "F": 0.22089, "G": 0.43999, "chunks": 226.19, "bytes_per_chunk": 4.527, "boundary_entropy": 0.66946, "router_noise": 0.0, "grad_norm": 0.6943, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 127.9} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.62295, "bpb": 2.34142, "chunk_loss": 0.033086, "F": 0.21774, "G": 0.44155, "chunks": 222.97, "bytes_per_chunk": 4.593, "boundary_entropy": 0.67152, "router_noise": 0.0, "grad_norm": 0.7765, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 130.1} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4577.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.263001, "cps": 35.757812, "mean_p": 0.442911, "boundary_entropy": 0.662247, "bpb": 2.281551}, "de": {"chunks": 5184.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.252472, "cps": 40.5, "mean_p": 0.446105, "boundary_entropy": 0.669053, "bpb": 2.577465}, "vi": {"chunks": 5520.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.29943, "cps": 43.125, "mean_p": 0.444433, "boundary_entropy": 0.671777, "bpb": 2.085821}, "tr": {"chunks": 5316.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.293653, "cps": 41.53125, "mean_p": 0.458372, "boundary_entropy": 0.672382, "bpb": 2.549594}, "fi": {"chunks": 4589.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.243953, "cps": 35.851562, "mean_p": 0.452641, "boundary_entropy": 0.673531, "bpb": 2.820323}, "ru": {"chunks": 6214.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327916, "cps": 48.546875, "mean_p": 0.437905, "boundary_entropy": 0.666488, "bpb": 1.682134}, "el": {"chunks": 7186.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.350144, "cps": 56.140625, "mean_p": 0.443189, "boundary_entropy": 0.674264, "bpb": 1.751116}, "hi": {"chunks": 9914.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.589242, "cps": 77.453125, "mean_p": 0.442005, "boundary_entropy": 0.666664, "bpb": 1.348331}, "ta": {"chunks": 12428.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.664777, "cps": 97.09375, "mean_p": 0.445569, "boundary_entropy": 0.66343, "bpb": 1.112393}, "th": {"chunks": 15530.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.965496, "cps": 121.328125, "mean_p": 0.43239, "boundary_entropy": 0.66711, "bpb": 1.306535}, "zh": {"chunks": 2704.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.457839, "cps": 21.125, "mean_p": 0.44428, "boundary_entropy": 0.673298, "bpb": 3.470454}, "my": {"chunks": 6431.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.339403, "cps": 50.242188, "mean_p": 0.450297, "boundary_entropy": 0.681227, "bpb": 1.262787}}, "bpb": {"en": 2.28155, "de": 2.57746, "vi": 2.08582, "tr": 2.54959, "fi": 2.82032, "ru": 1.68213, "el": 1.75112, "hi": 1.34833, "ta": 1.11239, "th": 1.30653, "zh": 3.47045, "my": 1.26279}, "gini_cpc": 0.246269, "gini_cps": 0.256124, "bpic": 4.48379, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9599658717209525, "vi": 1.138516013754697, "tr": 1.1165485929680825, "fi": 0.927575741149786, "ru": 1.2468242554979785, "el": 1.331341823804774, "hi": 2.2404592508378185, "ta": 2.5276619033631125, "th": 3.6710778826896573, "zh": 1.740830360965741, "my": 1.290500987742911}, "t": 133.2} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.28867, "de": 2.58299, "vi": 2.08982, "tr": 2.55564, "fi": 2.82218, "ru": 1.66141, "el": 1.74447, "hi": 1.34539, "ta": 1.11984, "th": 1.32763, "zh": 3.47671, "my": 1.27139}, "bytes_per_chunk": {"en": 3.811, "de": 3.88, "vi": 4.446, "tr": 3.636, "fi": 4.146, "ru": 5.436, "el": 5.196, "hi": 4.461, "ta": 3.894, "th": 3.018, "zh": 5.844, "my": 10.317}, "bpb_mean": 2.02384, "t": 134.5} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.42769, "bpb": 2.05972, "chunk_loss": 0.033503, "F": 0.22562, "G": 0.44204, "chunks": 231.03, "bytes_per_chunk": 4.432, "boundary_entropy": 0.67057, "router_noise": 0.0, "grad_norm": 1.1488, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 134.6} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.60148, "bpb": 2.31044, "chunk_loss": 0.032964, "F": 0.21704, "G": 0.43883, "chunks": 222.25, "bytes_per_chunk": 4.607, "boundary_entropy": 0.67034, "router_noise": 0.0, "grad_norm": 0.8068, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 136.8} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.64875, "bpb": 2.37864, "chunk_loss": 0.03385, "F": 0.22968, "G": 0.44549, "chunks": 235.19, "bytes_per_chunk": 4.354, "boundary_entropy": 0.67355, "router_noise": 0.0, "grad_norm": 0.8223, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 139.0} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.38202, "bpb": 1.99384, "chunk_loss": 0.032767, "F": 0.21329, "G": 0.43902, "chunks": 218.41, "bytes_per_chunk": 4.689, "boundary_entropy": 0.66956, "router_noise": 0.0, "grad_norm": 0.746, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 141.1} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4245.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.243923, "cps": 33.164062, "mean_p": 0.44047, "boundary_entropy": 0.661586, "bpb": 2.243203}, "de": {"chunks": 4707.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.229241, "cps": 36.773438, "mean_p": 0.443948, "boundary_entropy": 0.668427, "bpb": 2.535685}, "vi": {"chunks": 4338.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.235313, "cps": 33.890625, "mean_p": 0.445061, "boundary_entropy": 0.673406, "bpb": 2.063195}, "tr": {"chunks": 5048.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.278849, "cps": 39.4375, "mean_p": 0.457182, "boundary_entropy": 0.671834, "bpb": 2.512729}, "fi": {"chunks": 4529.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.240763, "cps": 35.382812, "mean_p": 0.45172, "boundary_entropy": 0.673033, "bpb": 2.738008}, "ru": {"chunks": 6987.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.368707, "cps": 54.585938, "mean_p": 0.439224, "boundary_entropy": 0.666862, "bpb": 1.648402}, "el": {"chunks": 6380.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.310871, "cps": 49.84375, "mean_p": 0.444761, "boundary_entropy": 0.675694, "bpb": 1.732019}, "hi": {"chunks": 10420.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.619316, "cps": 81.40625, "mean_p": 0.447801, "boundary_entropy": 0.667713, "bpb": 1.337653}, "ta": {"chunks": 13832.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.739877, "cps": 108.0625, "mean_p": 0.443121, "boundary_entropy": 0.663141, "bpb": 1.085849}, "th": {"chunks": 16737.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.040535, "cps": 130.757812, "mean_p": 0.43492, "boundary_entropy": 0.665523, "bpb": 1.261981}, "zh": {"chunks": 2809.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.475618, "cps": 21.945312, "mean_p": 0.445315, "boundary_entropy": 0.673645, "bpb": 3.468585}, "my": {"chunks": 9268.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.489128, "cps": 72.40625, "mean_p": 0.45128, "boundary_entropy": 0.681173, "bpb": 1.222074}}, "bpb": {"en": 2.2432, "de": 2.53569, "vi": 2.06319, "tr": 2.51273, "fi": 2.73801, "ru": 1.6484, "el": 1.73202, "hi": 1.33765, "ta": 1.08585, "th": 1.26198, "zh": 3.46858, "my": 1.22207}, "gini_cpc": 0.284346, "gini_cps": 0.292956, "bpic": 4.29766, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9398060073596946, "vi": 0.9647012279798853, "tr": 1.1431815861689754, "fi": 0.9870447954198664, "ru": 1.5115689232404412, "el": 1.2744601463255307, "hi": 2.5389787828351507, "ta": 3.0332341464290535, "th": 4.265824424353396, "zh": 1.9498658009642227, "my": 2.005252537911084}, "t": 144.3} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.25212, "de": 2.54016, "vi": 2.06623, "tr": 2.51848, "fi": 2.74607, "ru": 1.62776, "el": 1.72497, "hi": 1.33582, "ta": 1.0931, "th": 1.27925, "zh": 3.47123, "my": 1.2318}, "bytes_per_chunk": {"en": 4.112, "de": 4.307, "vi": 5.725, "tr": 3.803, "fi": 4.103, "ru": 4.973, "el": 5.426, "hi": 4.593, "ta": 3.48, "th": 2.803, "zh": 5.354, "my": 7.279}, "bpb_mean": 1.99058, "t": 145.6} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.29736, "bpb": 1.87169, "chunk_loss": 0.034122, "F": 0.23584, "G": 0.44227, "chunks": 241.5, "bytes_per_chunk": 4.24, "boundary_entropy": 0.66984, "router_noise": 0.0, "grad_norm": 0.759, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 145.7} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.53633, "bpb": 2.21645, "chunk_loss": 0.034098, "F": 0.2359, "G": 0.44057, "chunks": 241.56, "bytes_per_chunk": 4.239, "boundary_entropy": 0.6704, "router_noise": 0.0, "grad_norm": 0.7299, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 147.8} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.48773, "bpb": 2.14634, "chunk_loss": 0.032157, "F": 0.20374, "G": 0.43802, "chunks": 208.62, "bytes_per_chunk": 4.908, "boundary_entropy": 0.66961, "router_noise": 0.0, "grad_norm": 0.7718, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 150.0} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.46281, "bpb": 2.11039, "chunk_loss": 0.033756, "F": 0.2298, "G": 0.44176, "chunks": 235.31, "bytes_per_chunk": 4.352, "boundary_entropy": 0.67153, "router_noise": 0.0, "grad_norm": 0.8207, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 152.2} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4498.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.258461, "cps": 35.140625, "mean_p": 0.441226, "boundary_entropy": 0.661573, "bpb": 2.2206}, "de": {"chunks": 5063.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.246579, "cps": 39.554688, "mean_p": 0.447069, "boundary_entropy": 0.66915, "bpb": 2.490375}, "vi": {"chunks": 5512.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.298996, "cps": 43.0625, "mean_p": 0.447095, "boundary_entropy": 0.673007, "bpb": 2.045373}, "tr": {"chunks": 5406.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.298625, "cps": 42.234375, "mean_p": 0.459074, "boundary_entropy": 0.672213, "bpb": 2.475489}, "fi": {"chunks": 4565.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.242677, "cps": 35.664062, "mean_p": 0.453036, "boundary_entropy": 0.673812, "bpb": 2.700669}, "ru": {"chunks": 8644.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.456148, "cps": 67.53125, "mean_p": 0.444869, "boundary_entropy": 0.66872, "bpb": 1.609512}, "el": {"chunks": 8332.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.405984, "cps": 65.09375, "mean_p": 0.446868, "boundary_entropy": 0.674305, "bpb": 1.705396}, "hi": {"chunks": 11926.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.708826, "cps": 93.171875, "mean_p": 0.450917, "boundary_entropy": 0.670047, "bpb": 1.310106}, "ta": {"chunks": 12249.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.655202, "cps": 95.695312, "mean_p": 0.444482, "boundary_entropy": 0.665468, "bpb": 1.066777}, "th": {"chunks": 7728.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.480448, "cps": 60.375, "mean_p": 0.435886, "boundary_entropy": 0.670949, "bpb": 1.245142}, "zh": {"chunks": 2909.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.49255, "cps": 22.726562, "mean_p": 0.447366, "boundary_entropy": 0.674383, "bpb": 3.420955}, "my": {"chunks": 19171.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.011769, "cps": 149.773438, "mean_p": 0.464494, "boundary_entropy": 0.681883, "bpb": 1.21012}}, "bpb": {"en": 2.2206, "de": 2.49038, "vi": 2.04537, "tr": 2.47549, "fi": 2.70067, "ru": 1.60951, "el": 1.7054, "hi": 1.31011, "ta": 1.06678, "th": 1.24514, "zh": 3.42096, "my": 1.21012}, "gini_cpc": 0.254391, "gini_cps": 0.28564, "bpic": 3.997594, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9540259531246829, "vi": 1.1568331789085537, "tr": 1.1553941360252005, "fi": 0.9389307609978825, "ru": 1.7648598086983247, "el": 1.5707717617386705, "hi": 2.742485886083099, "ta": 2.535010918644297, "th": 1.858877271188673, "zh": 1.9057018154736942, "my": 3.9145879978745035}, "t": 155.3} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.22983, "de": 2.49874, "vi": 2.04899, "tr": 2.48267, "fi": 2.71096, "ru": 1.58922, "el": 1.69604, "hi": 1.30969, "ta": 1.07545, "th": 1.26369, "zh": 3.42469, "my": 1.21899}, "bytes_per_chunk": {"en": 3.887, "de": 3.873, "vi": 4.491, "tr": 3.605, "fi": 4.093, "ru": 3.823, "el": 4.343, "hi": 3.752, "ta": 3.918, "th": 5.932, "zh": 5.319, "my": 2.797}, "bpb_mean": 1.96241, "t": 156.7} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 156.7} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "global", "lm_loss": 1.32268, "bpb": 1.90822, "chunk_loss": 0.035732, "F": 0.26047, "G": 0.44608, "chunks": 266.72, "bytes_per_chunk": 3.839, "boundary_entropy": 0.67254, "router_noise": 0.0, "grad_norm": 0.7804, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 156.7} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "global", "lm_loss": 1.48098, "bpb": 2.1366, "chunk_loss": 0.033622, "F": 0.22702, "G": 0.44254, "chunks": 232.47, "bytes_per_chunk": 4.405, "boundary_entropy": 0.67113, "router_noise": 0.0, "grad_norm": 0.8354, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 158.9} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "global", "lm_loss": 1.46867, "bpb": 2.11884, "chunk_loss": 0.033196, "F": 0.22058, "G": 0.44127, "chunks": 225.88, "bytes_per_chunk": 4.533, "boundary_entropy": 0.67095, "router_noise": 0.0, "grad_norm": 0.7189, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 161.0} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "global", "lm_loss": 1.5367, "bpb": 2.21699, "chunk_loss": 0.033945, "F": 0.23318, "G": 0.43963, "chunks": 238.78, "bytes_per_chunk": 4.288, "boundary_entropy": 0.66867, "router_noise": 0.0, "grad_norm": 0.847, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 163.2} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4229.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.243004, "cps": 33.039062, "mean_p": 0.4384, "boundary_entropy": 0.660639, "bpb": 2.203661}, "de": {"chunks": 4873.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.237325, "cps": 38.070312, "mean_p": 0.446082, "boundary_entropy": 0.669387, "bpb": 2.480142}, "vi": {"chunks": 5142.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.278926, "cps": 40.171875, "mean_p": 0.445426, "boundary_entropy": 0.672644, "bpb": 2.044913}, "tr": {"chunks": 5323.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.29404, "cps": 41.585938, "mean_p": 0.456888, "boundary_entropy": 0.671421, "bpb": 2.426406}, "fi": {"chunks": 4345.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.230982, "cps": 33.945312, "mean_p": 0.449348, "boundary_entropy": 0.672449, "bpb": 2.662891}, "ru": {"chunks": 6459.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.340844, "cps": 50.460938, "mean_p": 0.438724, "boundary_entropy": 0.668185, "bpb": 1.589015}, "el": {"chunks": 7436.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.362325, "cps": 58.09375, "mean_p": 0.441412, "boundary_entropy": 0.673249, "bpb": 1.65916}, "hi": {"chunks": 8728.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.518752, "cps": 68.1875, "mean_p": 0.440297, "boundary_entropy": 0.667204, "bpb": 1.307949}, "ta": {"chunks": 6924.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.370366, "cps": 54.09375, "mean_p": 0.427918, "boundary_entropy": 0.660655, "bpb": 1.062458}, "th": {"chunks": 15956.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.99198, "cps": 124.65625, "mean_p": 0.43241, "boundary_entropy": 0.666057, "bpb": 1.2394}, "zh": {"chunks": 2415.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.408906, "cps": 18.867188, "mean_p": 0.447101, "boundary_entropy": 0.675055, "bpb": 3.408702}, "my": {"chunks": 9381.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.495092, "cps": 73.289062, "mean_p": 0.463416, "boundary_entropy": 0.684853, "bpb": 1.19008}}, "bpb": {"en": 2.20366, "de": 2.48014, "vi": 2.04491, "tr": 2.42641, "fi": 2.66289, "ru": 1.58902, "el": 1.65916, "hi": 1.30795, "ta": 1.06246, "th": 1.2394, "zh": 3.4087, "my": 1.19008}, "gini_cpc": 0.236925, "gini_cps": 0.251707, "bpic": 4.725727, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9766308511878006, "vi": 1.1478241696250135, "tr": 1.2100194458845732, "fi": 0.9505267258901093, "ru": 1.402628003276784, "el": 1.4910251567985704, "hi": 2.1347454654885936, "ta": 1.5241160086770804, "th": 4.082154121375616, "zh": 1.6827133005243256, "my": 2.0373807335609504}, "t": 166.3} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.21173, "de": 2.48608, "vi": 2.04705, "tr": 2.4364, "fi": 2.672, "ru": 1.56823, "el": 1.65127, "hi": 1.30577, "ta": 1.07033, "th": 1.25848, "zh": 3.41006, "my": 1.19908}, "bytes_per_chunk": {"en": 4.133, "de": 4.216, "vi": 4.711, "tr": 3.62, "fi": 4.411, "ru": 5.218, "el": 4.834, "hi": 5.308, "ta": 7.133, "th": 2.928, "zh": 6.592, "my": 6.127}, "bpb_mean": 1.94304, "t": 167.6} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "global", "lm_loss": 1.51018, "bpb": 2.17873, "chunk_loss": 0.033983, "F": 0.23431, "G": 0.43976, "chunks": 239.94, "bytes_per_chunk": 4.268, "boundary_entropy": 0.67007, "router_noise": 0.0, "grad_norm": 0.8277, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 167.7} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "global", "lm_loss": 1.45181, "bpb": 2.09452, "chunk_loss": 0.032807, "F": 0.21417, "G": 0.43914, "chunks": 219.31, "bytes_per_chunk": 4.669, "boundary_entropy": 0.67012, "router_noise": 0.0, "grad_norm": 0.7924, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 169.9} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "global", "lm_loss": 1.50361, "bpb": 2.16925, "chunk_loss": 0.034016, "F": 0.23444, "G": 0.44093, "chunks": 240.06, "bytes_per_chunk": 4.266, "boundary_entropy": 0.67009, "router_noise": 0.0, "grad_norm": 0.8809, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 172.0} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "global", "lm_loss": 1.53329, "bpb": 2.21206, "chunk_loss": 0.034523, "F": 0.24286, "G": 0.44177, "chunks": 248.69, "bytes_per_chunk": 4.118, "boundary_entropy": 0.67047, "router_noise": 0.0, "grad_norm": 0.8299, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 174.1} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4114.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.236396, "cps": 32.140625, "mean_p": 0.437665, "boundary_entropy": 0.660761, "bpb": 2.184325}, "de": {"chunks": 4611.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.224565, "cps": 36.023438, "mean_p": 0.442835, "boundary_entropy": 0.668134, "bpb": 2.430491}, "vi": {"chunks": 4711.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.255547, "cps": 36.804688, "mean_p": 0.443722, "boundary_entropy": 0.672323, "bpb": 2.024409}, "tr": {"chunks": 5048.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.278849, "cps": 39.4375, "mean_p": 0.456171, "boundary_entropy": 0.671137, "bpb": 2.396543}, "fi": {"chunks": 4219.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.224284, "cps": 32.960938, "mean_p": 0.450673, "boundary_entropy": 0.672926, "bpb": 2.644233}, "ru": {"chunks": 7352.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.387968, "cps": 57.4375, "mean_p": 0.440009, "boundary_entropy": 0.667385, "bpb": 1.568506}, "el": {"chunks": 8452.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.411831, "cps": 66.03125, "mean_p": 0.450268, "boundary_entropy": 0.676091, "bpb": 1.641384}, "hi": {"chunks": 9138.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.54312, "cps": 71.390625, "mean_p": 0.447599, "boundary_entropy": 0.669945, "bpb": 1.273369}, "ta": {"chunks": 10987.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.587697, "cps": 85.835938, "mean_p": 0.441597, "boundary_entropy": 0.665596, "bpb": 1.047992}, "th": {"chunks": 16678.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.036867, "cps": 130.296875, "mean_p": 0.432823, "boundary_entropy": 0.665218, "bpb": 1.200652}, "zh": {"chunks": 2899.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.490857, "cps": 22.648438, "mean_p": 0.450821, "boundary_entropy": 0.675583, "bpb": 3.349314}, "my": {"chunks": 7557.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.398828, "cps": 59.039062, "mean_p": 0.461682, "boundary_entropy": 0.684357, "bpb": 1.178607}}, "bpb": {"en": 2.18433, "de": 2.43049, "vi": 2.02441, "tr": 2.39654, "fi": 2.64423, "ru": 1.56851, "el": 1.64138, "hi": 1.27337, "ta": 1.04799, "th": 1.20065, "zh": 3.34931, "my": 1.17861}, "gini_cpc": 0.261648, "gini_cps": 0.270023, "bpic": 4.474745, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9499539389987035, "vi": 1.081010208251172, "tr": 1.1795833333221442, "fi": 0.9487624213296728, "ru": 1.6411796260404332, "el": 1.7421216425286725, "hi": 2.2975020821880996, "ta": 2.4860707727597124, "th": 4.386142498627475, "zh": 2.076417141932407, "my": 1.6871196316629016}, "t": 177.2} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.19415, "de": 2.43679, "vi": 2.02759, "tr": 2.39964, "fi": 2.65238, "ru": 1.54842, "el": 1.63444, "hi": 1.27072, "ta": 1.05556, "th": 1.21977, "zh": 3.35327, "my": 1.18895}, "bytes_per_chunk": {"en": 4.208, "de": 4.345, "vi": 5.24, "tr": 3.821, "fi": 4.565, "ru": 4.688, "el": 4.333, "hi": 4.993, "ta": 4.38, "th": 2.777, "zh": 5.329, "my": 7.666}, "bpb_mean": 1.91514, "t": 178.5} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "global", "lm_loss": 1.41361, "bpb": 2.03941, "chunk_loss": 0.034122, "F": 0.23676, "G": 0.43981, "chunks": 242.44, "bytes_per_chunk": 4.224, "boundary_entropy": 0.66962, "router_noise": 0.0, "grad_norm": 0.8355, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 178.6} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "global", "lm_loss": 1.33061, "bpb": 1.91967, "chunk_loss": 0.033766, "F": 0.22955, "G": 0.44317, "chunks": 235.06, "bytes_per_chunk": 4.356, "boundary_entropy": 0.67166, "router_noise": 0.0, "grad_norm": 0.7472, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 180.7} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "global", "lm_loss": 1.41493, "bpb": 2.04132, "chunk_loss": 0.032549, "F": 0.21057, "G": 0.43671, "chunks": 215.62, "bytes_per_chunk": 4.749, "boundary_entropy": 0.66882, "router_noise": 0.0, "grad_norm": 0.7705, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 182.9} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "global", "lm_loss": 1.31365, "bpb": 1.8952, "chunk_loss": 0.032373, "F": 0.20685, "G": 0.43512, "chunks": 211.81, "bytes_per_chunk": 4.834, "boundary_entropy": 0.66835, "router_noise": 0.0, "grad_norm": 0.706, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 185.0} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3917.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.225076, "cps": 30.601562, "mean_p": 0.435626, "boundary_entropy": 0.659712, "bpb": 2.170835}, "de": {"chunks": 4573.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.222715, "cps": 35.726562, "mean_p": 0.44127, "boundary_entropy": 0.667616, "bpb": 2.423268}, "vi": {"chunks": 3867.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.209764, "cps": 30.210938, "mean_p": 0.438883, "boundary_entropy": 0.671159, "bpb": 2.016807}, "tr": {"chunks": 5252.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.290118, "cps": 41.03125, "mean_p": 0.456058, "boundary_entropy": 0.671234, "bpb": 2.373505}, "fi": {"chunks": 4386.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.233161, "cps": 34.265625, "mean_p": 0.4489, "boundary_entropy": 0.671897, "bpb": 2.613099}, "ru": {"chunks": 5475.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.288918, "cps": 42.773438, "mean_p": 0.433138, "boundary_entropy": 0.665904, "bpb": 1.543773}, "el": {"chunks": 6681.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325537, "cps": 52.195312, "mean_p": 0.440418, "boundary_entropy": 0.673399, "bpb": 1.61254}, "hi": {"chunks": 9017.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.535929, "cps": 70.445312, "mean_p": 0.446527, "boundary_entropy": 0.669848, "bpb": 1.264665}, "ta": {"chunks": 6986.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.373683, "cps": 54.578125, "mean_p": 0.428959, "boundary_entropy": 0.660974, "bpb": 1.0372}, "th": {"chunks": 4881.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.30345, "cps": 38.132812, "mean_p": 0.417476, "boundary_entropy": 0.661591, "bpb": 1.186308}, "zh": {"chunks": 2648.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.448358, "cps": 20.6875, "mean_p": 0.447178, "boundary_entropy": 0.674601, "bpb": 3.350275}, "my": {"chunks": 6654.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.351172, "cps": 51.984375, "mean_p": 0.451696, "boundary_entropy": 0.681885, "bpb": 1.167839}}, "bpb": {"en": 2.17084, "de": 2.42327, "vi": 2.01681, "tr": 2.37351, "fi": 2.6131, "ru": 1.54377, "el": 1.61254, "hi": 1.26466, "ta": 1.0372, "th": 1.18631, "zh": 3.35027, "my": 1.16784}, "gini_cpc": 0.161334, "gini_cps": 0.170839, "bpic": 5.965168, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9895080754517088, "vi": 0.9319692405036734, "tr": 1.2889756543172222, "fi": 1.0359225634646034, "ru": 1.2836465498136465, "el": 1.4463425910773833, "hi": 2.3810995087136844, "ta": 1.6602506557743408, "th": 1.3482123188902078, "zh": 1.9920263863242194, "my": 1.5602348315902024}, "t": 188.0} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.1784, "de": 2.42752, "vi": 2.01858, "tr": 2.38757, "fi": 2.62456, "ru": 1.52042, "el": 1.60599, "hi": 1.26255, "ta": 1.04467, "th": 1.20706, "zh": 3.34884, "my": 1.17753}, "bytes_per_chunk": {"en": 4.421, "de": 4.411, "vi": 6.312, "tr": 3.683, "fi": 4.312, "ru": 6.171, "el": 5.309, "hi": 5.067, "ta": 6.93, "th": 9.629, "zh": 5.839, "my": 8.926}, "bpb_mean": 1.90031, "t": 189.3} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "global", "lm_loss": 1.52433, "bpb": 2.19914, "chunk_loss": 0.031544, "F": 0.19241, "G": 0.43924, "chunks": 197.03, "bytes_per_chunk": 5.197, "boundary_entropy": 0.67215, "router_noise": 0.0, "grad_norm": 0.7584, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 189.4} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "global", "lm_loss": 1.48586, "bpb": 2.14364, "chunk_loss": 0.034048, "F": 0.23453, "G": 0.44146, "chunks": 240.16, "bytes_per_chunk": 4.264, "boundary_entropy": 0.67086, "router_noise": 0.0, "grad_norm": 0.8535, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 191.5} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "global", "lm_loss": 1.41935, "bpb": 2.04769, "chunk_loss": 0.032383, "F": 0.20667, "G": 0.44309, "chunks": 211.62, "bytes_per_chunk": 4.839, "boundary_entropy": 0.67356, "router_noise": 0.0, "grad_norm": 0.8046, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 193.6} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "global", "lm_loss": 1.33017, "bpb": 1.91903, "chunk_loss": 0.032697, "F": 0.21167, "G": 0.43989, "chunks": 216.75, "bytes_per_chunk": 4.724, "boundary_entropy": 0.67113, "router_noise": 0.0, "grad_norm": 0.8042, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 195.8} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4066.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.233638, "cps": 31.765625, "mean_p": 0.436435, "boundary_entropy": 0.659921, "bpb": 2.145876}, "de": {"chunks": 4632.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.225588, "cps": 36.1875, "mean_p": 0.442931, "boundary_entropy": 0.667863, "bpb": 2.388712}, "vi": {"chunks": 4447.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.241226, "cps": 34.742188, "mean_p": 0.441591, "boundary_entropy": 0.671284, "bpb": 2.000163}, "tr": {"chunks": 5154.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.284704, "cps": 40.265625, "mean_p": 0.455716, "boundary_entropy": 0.670699, "bpb": 2.352019}, "fi": {"chunks": 4360.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.231779, "cps": 34.0625, "mean_p": 0.448109, "boundary_entropy": 0.67128, "bpb": 2.56888}, "ru": {"chunks": 6392.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.337309, "cps": 49.9375, "mean_p": 0.436233, "boundary_entropy": 0.666616, "bpb": 1.527195}, "el": {"chunks": 6796.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.331141, "cps": 53.09375, "mean_p": 0.442371, "boundary_entropy": 0.673351, "bpb": 1.607513}, "hi": {"chunks": 10081.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.599168, "cps": 78.757812, "mean_p": 0.449282, "boundary_entropy": 0.670803, "bpb": 1.25938}, "ta": {"chunks": 10064.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.538326, "cps": 78.625, "mean_p": 0.436988, "boundary_entropy": 0.664041, "bpb": 1.034644}, "th": {"chunks": 12549.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.780168, "cps": 98.039062, "mean_p": 0.423784, "boundary_entropy": 0.663177, "bpb": 1.170295}, "zh": {"chunks": 2719.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.460379, "cps": 21.242188, "mean_p": 0.448517, "boundary_entropy": 0.674878, "bpb": 3.29791}, "my": {"chunks": 10782.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.569031, "cps": 84.234375, "mean_p": 0.460263, "boundary_entropy": 0.683692, "bpb": 1.152608}}, "bpb": {"en": 2.14588, "de": 2.38871, "vi": 2.00016, "tr": 2.35202, "fi": 2.56888, "ru": 1.52719, "el": 1.60751, "hi": 1.25938, "ta": 1.03464, "th": 1.17029, "zh": 3.29791, "my": 1.15261}, "gini_cpc": 0.237514, "gini_cps": 0.248283, "bpic": 4.67786, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9655458230963323, "vi": 1.0324778262012835, "tr": 1.218570355417444, "fi": 0.9920449524005366, "ru": 1.4437244048399303, "el": 1.417324423291095, "hi": 2.5645152604609383, "ta": 2.304103080135244, "th": 3.3392182088483757, "zh": 1.9704821761624027, "my": 2.435525591514551}, "t": 198.9} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.1569, "de": 2.39683, "vi": 2.00249, "tr": 2.35887, "fi": 2.58104, "ru": 1.5046, "el": 1.59977, "hi": 1.25585, "ta": 1.04172, "th": 1.18958, "zh": 3.29954, "my": 1.1611}, "bytes_per_chunk": {"en": 4.304, "de": 4.436, "vi": 5.365, "tr": 3.749, "fi": 4.3, "ru": 5.337, "el": 5.336, "hi": 4.488, "ta": 4.878, "th": 3.669, "zh": 5.815, "my": 5.17}, "bpb_mean": 1.87902, "t": 200.2} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "global", "lm_loss": 1.40464, "bpb": 2.02647, "chunk_loss": 0.032865, "F": 0.21552, "G": 0.43839, "chunks": 220.69, "bytes_per_chunk": 4.64, "boundary_entropy": 0.67024, "router_noise": 0.0, "grad_norm": 0.7888, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 200.3} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "global", "lm_loss": 1.38541, "bpb": 1.99872, "chunk_loss": 0.033607, "F": 0.22717, "G": 0.44193, "chunks": 232.62, "bytes_per_chunk": 4.402, "boundary_entropy": 0.6699, "router_noise": 0.0, "grad_norm": 0.7671, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 202.4} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "global", "lm_loss": 1.28775, "bpb": 1.85784, "chunk_loss": 0.031911, "F": 0.19839, "G": 0.43791, "chunks": 203.16, "bytes_per_chunk": 5.04, "boundary_entropy": 0.66983, "router_noise": 0.0, "grad_norm": 0.817, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 204.5} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "global", "lm_loss": 1.36667, "bpb": 1.97168, "chunk_loss": 0.032275, "F": 0.20526, "G": 0.43878, "chunks": 210.19, "bytes_per_chunk": 4.872, "boundary_entropy": 0.67107, "router_noise": 0.0, "grad_norm": 0.733, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 206.7} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4009.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.230363, "cps": 31.320312, "mean_p": 0.436433, "boundary_entropy": 0.660228, "bpb": 2.131115}, "de": {"chunks": 4779.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.232747, "cps": 37.335938, "mean_p": 0.444358, "boundary_entropy": 0.668424, "bpb": 2.372207}, "vi": {"chunks": 4937.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.267806, "cps": 38.570312, "mean_p": 0.445061, "boundary_entropy": 0.672217, "bpb": 1.992261}, "tr": {"chunks": 4832.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.266917, "cps": 37.75, "mean_p": 0.454536, "boundary_entropy": 0.670729, "bpb": 2.336608}, "fi": {"chunks": 4427.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.235341, "cps": 34.585938, "mean_p": 0.450326, "boundary_entropy": 0.672212, "bpb": 2.561587}, "ru": {"chunks": 6362.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.335726, "cps": 49.703125, "mean_p": 0.436707, "boundary_entropy": 0.666885, "bpb": 1.515152}, "el": {"chunks": 6443.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.31394, "cps": 50.335938, "mean_p": 0.4457, "boundary_entropy": 0.67495, "bpb": 1.575157}, "hi": {"chunks": 9380.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.557504, "cps": 73.28125, "mean_p": 0.444676, "boundary_entropy": 0.668857, "bpb": 1.255657}, "ta": {"chunks": 8453.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.452153, "cps": 66.039062, "mean_p": 0.436994, "boundary_entropy": 0.664342, "bpb": 1.018015}, "th": {"chunks": 7190.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.447, "cps": 56.171875, "mean_p": 0.421792, "boundary_entropy": 0.662409, "bpb": 1.16059}, "zh": {"chunks": 2651.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.448866, "cps": 20.710938, "mean_p": 0.445871, "boundary_entropy": 0.674044, "bpb": 3.278901}, "my": {"chunks": 5428.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.286468, "cps": 42.40625, "mean_p": 0.450623, "boundary_entropy": 0.681857, "bpb": 1.175871}}, "bpb": {"en": 2.13111, "de": 2.37221, "vi": 1.99226, "tr": 2.33661, "fi": 2.56159, "ru": 1.51515, "el": 1.57516, "hi": 1.25566, "ta": 1.01802, "th": 1.16059, "zh": 3.2789, "my": 1.17587}, "gini_cpc": 0.167686, "gini_cps": 0.177523, "bpic": 5.570844, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0103519576933975, "vi": 1.162540386652408, "tr": 1.158682474458431, "fi": 1.0216113340963906, "ru": 1.4573790238454243, "el": 1.3628101207307677, "hi": 2.4201140302076642, "ta": 1.9627883130834733, "th": 1.9404206559487345, "zh": 1.9485176723520063, "my": 1.2435536508972607}, "t": 209.8} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.14023, "de": 2.37584, "vi": 1.994, "tr": 2.34275, "fi": 2.57213, "ru": 1.4939, "el": 1.5672, "hi": 1.25312, "ta": 1.02528, "th": 1.17808, "zh": 3.28021, "my": 1.18332}, "bytes_per_chunk": {"en": 4.435, "de": 4.339, "vi": 4.972, "tr": 4.088, "fi": 4.36, "ru": 5.357, "el": 5.658, "hi": 4.924, "ta": 5.726, "th": 6.299, "zh": 5.927, "my": 10.836}, "bpb_mean": 1.86717, "t": 211.0} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "global", "lm_loss": 1.31474, "bpb": 1.89677, "chunk_loss": 0.031789, "F": 0.19736, "G": 0.43723, "chunks": 202.09, "bytes_per_chunk": 5.067, "boundary_entropy": 0.6697, "router_noise": 0.0, "grad_norm": 0.8021, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 211.1} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "global", "lm_loss": 1.21882, "bpb": 1.75838, "chunk_loss": 0.03283, "F": 0.21448, "G": 0.43758, "chunks": 219.62, "bytes_per_chunk": 4.662, "boundary_entropy": 0.66783, "router_noise": 0.0, "grad_norm": 0.7401, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 213.2} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "global", "lm_loss": 1.32901, "bpb": 1.91735, "chunk_loss": 0.033534, "F": 0.22534, "G": 0.44234, "chunks": 230.75, "bytes_per_chunk": 4.438, "boundary_entropy": 0.67052, "router_noise": 0.0, "grad_norm": 0.7853, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 215.4} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "global", "lm_loss": 1.37678, "bpb": 1.98627, "chunk_loss": 0.032549, "F": 0.20978, "G": 0.44168, "chunks": 214.81, "bytes_per_chunk": 4.767, "boundary_entropy": 0.67221, "router_noise": 0.0, "grad_norm": 0.8073, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 217.5} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4058.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.233178, "cps": 31.703125, "mean_p": 0.435708, "boundary_entropy": 0.658941, "bpb": 2.11484}, "de": {"chunks": 4661.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.227, "cps": 36.414062, "mean_p": 0.441723, "boundary_entropy": 0.667726, "bpb": 2.345734}, "vi": {"chunks": 4355.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.236235, "cps": 34.023438, "mean_p": 0.441185, "boundary_entropy": 0.671073, "bpb": 1.976499}, "tr": {"chunks": 5037.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.278241, "cps": 39.351562, "mean_p": 0.455275, "boundary_entropy": 0.67057, "bpb": 2.308164}, "fi": {"chunks": 4700.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.249854, "cps": 36.71875, "mean_p": 0.451178, "boundary_entropy": 0.672086, "bpb": 2.522392}, "ru": {"chunks": 6878.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.362955, "cps": 53.734375, "mean_p": 0.438523, "boundary_entropy": 0.667354, "bpb": 1.501617}, "el": {"chunks": 7440.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.36252, "cps": 58.125, "mean_p": 0.44555, "boundary_entropy": 0.674456, "bpb": 1.56276}, "hi": {"chunks": 10858.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.645349, "cps": 84.828125, "mean_p": 0.448117, "boundary_entropy": 0.669775, "bpb": 1.238915}, "ta": {"chunks": 9089.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.486173, "cps": 71.007812, "mean_p": 0.438236, "boundary_entropy": 0.664669, "bpb": 1.00678}, "th": {"chunks": 9089.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.565061, "cps": 71.007812, "mean_p": 0.425256, "boundary_entropy": 0.664745, "bpb": 1.147342}, "zh": {"chunks": 2589.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.438368, "cps": 20.226562, "mean_p": 0.449581, "boundary_entropy": 0.675462, "bpb": 3.266543}, "my": {"chunks": 6851.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.361569, "cps": 53.523438, "mean_p": 0.458221, "boundary_entropy": 0.684083, "bpb": 1.143087}}, "bpb": {"en": 2.11484, "de": 2.34573, "vi": 1.9765, "tr": 2.30816, "fi": 2.52239, "ru": 1.50162, "el": 1.56276, "hi": 1.23891, "ta": 1.00678, "th": 1.14734, "zh": 3.26654, "my": 1.14309}, "gini_cpc": 0.197459, "gini_cps": 0.211206, "bpic": 5.076133, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9735063154416324, "vi": 1.013111149519919, "tr": 1.1932555882584917, "fi": 1.0715144990168213, "ru": 1.5565570208233908, "el": 1.5546912986372357, "hi": 2.76762243165043, "ta": 2.0849839272248043, "th": 2.4232996281919625, "zh": 1.8799689884417232, "my": 1.55061031602873}, "t": 220.6} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.1244, "de": 2.35361, "vi": 1.97967, "tr": 2.31359, "fi": 2.53134, "ru": 1.47985, "el": 1.55618, "hi": 1.23481, "ta": 1.01536, "th": 1.16495, "zh": 3.26655, "my": 1.15126}, "bytes_per_chunk": {"en": 4.233, "de": 4.298, "vi": 5.545, "tr": 3.83, "fi": 4.013, "ru": 4.913, "el": 4.696, "hi": 3.975, "ta": 5.29, "th": 4.977, "zh": 6.098, "my": 8.167}, "bpb_mean": 1.84763, "t": 221.9} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "global", "lm_loss": 1.324, "bpb": 1.91012, "chunk_loss": 0.03347, "F": 0.22513, "G": 0.43907, "chunks": 230.53, "bytes_per_chunk": 4.442, "boundary_entropy": 0.66978, "router_noise": 0.0, "grad_norm": 0.817, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 221.9} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "global", "lm_loss": 1.31152, "bpb": 1.89213, "chunk_loss": 0.033834, "F": 0.23175, "G": 0.43915, "chunks": 237.31, "bytes_per_chunk": 4.315, "boundary_entropy": 0.66883, "router_noise": 0.0, "grad_norm": 0.7454, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 224.0} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "global", "lm_loss": 1.53576, "bpb": 2.21563, "chunk_loss": 0.033266, "F": 0.22125, "G": 0.44125, "chunks": 226.56, "bytes_per_chunk": 4.52, "boundary_entropy": 0.67122, "router_noise": 0.0, "grad_norm": 0.8324, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 226.2} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "global", "lm_loss": 1.43245, "bpb": 2.06658, "chunk_loss": 0.03377, "F": 0.23093, "G": 0.43876, "chunks": 236.47, "bytes_per_chunk": 4.33, "boundary_entropy": 0.66733, "router_noise": 0.0, "grad_norm": 0.7562, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 228.3} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4013.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.230592, "cps": 31.351562, "mean_p": 0.433803, "boundary_entropy": 0.657903, "bpb": 2.102862}, "de": {"chunks": 4820.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.234744, "cps": 37.65625, "mean_p": 0.441883, "boundary_entropy": 0.666983, "bpb": 2.32951}, "vi": {"chunks": 4598.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.249417, "cps": 35.921875, "mean_p": 0.443062, "boundary_entropy": 0.671499, "bpb": 1.965737}, "tr": {"chunks": 5040.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.278407, "cps": 39.375, "mean_p": 0.454118, "boundary_entropy": 0.670063, "bpb": 2.283317}, "fi": {"chunks": 4749.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.252459, "cps": 37.101562, "mean_p": 0.450742, "boundary_entropy": 0.671434, "bpb": 2.494209}, "ru": {"chunks": 6764.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.356939, "cps": 52.84375, "mean_p": 0.438771, "boundary_entropy": 0.667616, "bpb": 1.47904}, "el": {"chunks": 6112.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.297812, "cps": 47.75, "mean_p": 0.442143, "boundary_entropy": 0.673974, "bpb": 1.541305}, "hi": {"chunks": 9615.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.571471, "cps": 75.117188, "mean_p": 0.447277, "boundary_entropy": 0.670519, "bpb": 1.223031}, "ta": {"chunks": 8590.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.459481, "cps": 67.109375, "mean_p": 0.438215, "boundary_entropy": 0.665191, "bpb": 0.998599}, "th": {"chunks": 11585.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.720236, "cps": 90.507812, "mean_p": 0.423918, "boundary_entropy": 0.66262, "bpb": 1.143231}, "zh": {"chunks": 2700.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.457162, "cps": 21.09375, "mean_p": 0.450862, "boundary_entropy": 0.67575, "bpb": 3.246234}, "my": {"chunks": 14863.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.78441, "cps": 116.117188, "mean_p": 0.46228, "boundary_entropy": 0.68285, "bpb": 1.135521}}, "bpb": {"en": 2.10286, "de": 2.32951, "vi": 1.96574, "tr": 2.28332, "fi": 2.49421, "ru": 1.47904, "el": 1.54131, "hi": 1.22303, "ta": 0.9986, "th": 1.14323, "zh": 3.24623, "my": 1.13552}, "gini_cpc": 0.244327, "gini_cps": 0.262804, "bpic": 4.598989, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0180042510018548, "vi": 1.081635133334969, "tr": 1.2073548901770512, "fi": 1.0948263632535602, "ru": 1.5479229706619713, "el": 1.2915090711801178, "hi": 2.478273175330251, "ta": 1.9926116025564682, "th": 3.123416738311089, "zh": 1.9825551718175665, "my": 3.4017160741207286}, "t": 231.4} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.11117, "de": 2.33575, "vi": 1.96895, "tr": 2.29347, "fi": 2.50438, "ru": 1.45744, "el": 1.53418, "hi": 1.21953, "ta": 1.00673, "th": 1.1605, "zh": 3.2465, "my": 1.14312}, "bytes_per_chunk": {"en": 4.359, "de": 4.259, "vi": 5.175, "tr": 3.84, "fi": 4.023, "ru": 4.944, "el": 5.999, "hi": 4.552, "ta": 5.653, "th": 3.97, "zh": 5.803, "my": 3.656}, "bpb_mean": 1.83181, "t": 232.7} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "global", "lm_loss": 1.38226, "bpb": 1.99418, "chunk_loss": 0.032858, "F": 0.21463, "G": 0.44063, "chunks": 219.78, "bytes_per_chunk": 4.659, "boundary_entropy": 0.66959, "router_noise": 0.0, "grad_norm": 0.7699, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 232.8} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "global", "lm_loss": 1.54108, "bpb": 2.22331, "chunk_loss": 0.032564, "F": 0.20999, "G": 0.44032, "chunks": 215.03, "bytes_per_chunk": 4.762, "boundary_entropy": 0.67083, "router_noise": 0.0, "grad_norm": 0.7778, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 234.9} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "global", "lm_loss": 1.47414, "bpb": 2.12674, "chunk_loss": 0.032986, "F": 0.21725, "G": 0.43936, "chunks": 222.47, "bytes_per_chunk": 4.603, "boundary_entropy": 0.66975, "router_noise": 0.0, "grad_norm": 0.7418, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 237.0} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "global", "lm_loss": 1.4306, "bpb": 2.06391, "chunk_loss": 0.03285, "F": 0.21448, "G": 0.43828, "chunks": 219.62, "bytes_per_chunk": 4.662, "boundary_entropy": 0.6692, "router_noise": 0.0, "grad_norm": 0.8028, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 239.1} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4169.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.239556, "cps": 32.570312, "mean_p": 0.436431, "boundary_entropy": 0.659176, "bpb": 2.093123}, "de": {"chunks": 4851.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.236254, "cps": 37.898438, "mean_p": 0.443275, "boundary_entropy": 0.667582, "bpb": 2.313747}, "vi": {"chunks": 4774.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.258964, "cps": 37.296875, "mean_p": 0.443906, "boundary_entropy": 0.67162, "bpb": 1.959057}, "tr": {"chunks": 4986.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.275424, "cps": 38.953125, "mean_p": 0.453835, "boundary_entropy": 0.669993, "bpb": 2.26576}, "fi": {"chunks": 4687.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.249163, "cps": 36.617188, "mean_p": 0.450067, "boundary_entropy": 0.671183, "bpb": 2.478751}, "ru": {"chunks": 6643.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.350554, "cps": 51.898438, "mean_p": 0.438731, "boundary_entropy": 0.667706, "bpb": 1.471186}, "el": {"chunks": 7488.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.364859, "cps": 58.5, "mean_p": 0.445512, "boundary_entropy": 0.674606, "bpb": 1.524936}, "hi": {"chunks": 7957.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.472927, "cps": 62.164062, "mean_p": 0.443862, "boundary_entropy": 0.670618, "bpb": 1.218313}, "ta": {"chunks": 8465.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.452795, "cps": 66.132812, "mean_p": 0.435922, "boundary_entropy": 0.664854, "bpb": 0.990216}, "th": {"chunks": 9491.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.590053, "cps": 74.148438, "mean_p": 0.424213, "boundary_entropy": 0.663693, "bpb": 1.127264}, "zh": {"chunks": 2738.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.463596, "cps": 21.390625, "mean_p": 0.449562, "boundary_entropy": 0.675128, "bpb": 3.207492}, "my": {"chunks": 6971.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.367902, "cps": 54.460938, "mean_p": 0.455951, "boundary_entropy": 0.683363, "bpb": 1.124715}}, "bpb": {"en": 2.09312, "de": 2.31375, "vi": 1.95906, "tr": 2.26576, "fi": 2.47875, "ru": 1.47119, "el": 1.52494, "hi": 1.21831, "ta": 0.99022, "th": 1.12726, "zh": 3.20749, "my": 1.12471}, "gini_cpc": 0.168787, "gini_cps": 0.179168, "bpic": 5.241478, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.986213839180561, "vi": 1.0810144463933395, "tr": 1.1497249282764797, "fi": 1.0401004760602364, "ru": 1.4633468033626764, "el": 1.523060712663677, "hi": 1.9741789198346102, "ta": 1.890138890501608, "th": 2.4631061762391613, "zh": 1.9352283885940909, "my": 1.5357620505158025}, "t": 242.2} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.10219, "de": 2.32003, "vi": 1.96032, "tr": 2.27324, "fi": 2.48891, "ru": 1.44875, "el": 1.51872, "hi": 1.21512, "ta": 0.99821, "th": 1.14493, "zh": 3.20602, "my": 1.13278}, "bytes_per_chunk": {"en": 4.186, "de": 4.141, "vi": 4.976, "tr": 3.857, "fi": 4.011, "ru": 5.088, "el": 4.734, "hi": 5.635, "ta": 5.815, "th": 4.677, "zh": 5.598, "my": 7.988}, "bpb_mean": 1.81744, "t": 243.5} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "global", "lm_loss": 1.3127, "bpb": 1.89382, "chunk_loss": 0.033209, "F": 0.22095, "G": 0.44029, "chunks": 226.25, "bytes_per_chunk": 4.526, "boundary_entropy": 0.66996, "router_noise": 0.0, "grad_norm": 0.7595, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 243.6} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "global", "lm_loss": 1.47064, "bpb": 2.12168, "chunk_loss": 0.032388, "F": 0.20758, "G": 0.43753, "chunks": 212.56, "bytes_per_chunk": 4.817, "boundary_entropy": 0.67, "router_noise": 0.0, "grad_norm": 0.7741, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 245.7} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "global", "lm_loss": 1.32871, "bpb": 1.91692, "chunk_loss": 0.032769, "F": 0.21301, "G": 0.43936, "chunks": 218.12, "bytes_per_chunk": 4.695, "boundary_entropy": 0.66931, "router_noise": 0.0, "grad_norm": 0.7691, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 247.8} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "global", "lm_loss": 1.48111, "bpb": 2.1368, "chunk_loss": 0.033603, "F": 0.22702, "G": 0.44208, "chunks": 232.47, "bytes_per_chunk": 4.405, "boundary_entropy": 0.67094, "router_noise": 0.0, "grad_norm": 0.7397, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 249.9} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4085.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.23473, "cps": 31.914062, "mean_p": 0.435775, "boundary_entropy": 0.658757, "bpb": 2.079091}, "de": {"chunks": 4881.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.237715, "cps": 38.132812, "mean_p": 0.442829, "boundary_entropy": 0.667025, "bpb": 2.297773}, "vi": {"chunks": 4495.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.24383, "cps": 35.117188, "mean_p": 0.442488, "boundary_entropy": 0.671521, "bpb": 1.953501}, "tr": {"chunks": 5170.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.285588, "cps": 40.390625, "mean_p": 0.456607, "boundary_entropy": 0.670594, "bpb": 2.251139}, "fi": {"chunks": 4705.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.25012, "cps": 36.757812, "mean_p": 0.450821, "boundary_entropy": 0.671495, "bpb": 2.462099}, "ru": {"chunks": 7012.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.370026, "cps": 54.78125, "mean_p": 0.439094, "boundary_entropy": 0.66771, "bpb": 1.456721}, "el": {"chunks": 7790.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.379574, "cps": 60.859375, "mean_p": 0.445854, "boundary_entropy": 0.674268, "bpb": 1.514031}, "hi": {"chunks": 7782.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.462526, "cps": 60.796875, "mean_p": 0.442901, "boundary_entropy": 0.66917, "bpb": 1.20949}, "ta": {"chunks": 8549.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.457288, "cps": 66.789062, "mean_p": 0.435801, "boundary_entropy": 0.664375, "bpb": 0.983199}, "th": {"chunks": 10755.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.668635, "cps": 84.023438, "mean_p": 0.426011, "boundary_entropy": 0.664533, "bpb": 1.122216}, "zh": {"chunks": 2612.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.442262, "cps": 20.40625, "mean_p": 0.446526, "boundary_entropy": 0.674302, "bpb": 3.193816}, "my": {"chunks": 11595.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.611938, "cps": 90.585938, "mean_p": 0.46167, "boundary_entropy": 0.684065, "bpb": 1.117981}}, "bpb": {"en": 2.07909, "de": 2.29777, "vi": 1.9535, "tr": 2.25114, "fi": 2.4621, "ru": 1.45672, "el": 1.51403, "hi": 1.20949, "ta": 0.9832, "th": 1.12222, "zh": 3.19382, "my": 1.11798}, "gini_cpc": 0.200737, "gini_cps": 0.223665, "bpic": 4.831627, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0127178382758006, "vi": 1.0387681220763414, "tr": 1.2166679028496734, "fi": 1.065564648200905, "ru": 1.576393924616413, "el": 1.6170694478911352, "hi": 1.970462675573484, "ta": 1.9481478203055358, "th": 2.848534008246429, "zh": 1.8841340113843938, "my": 2.606990425900532}, "t": 253.0} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.08829, "de": 2.30337, "vi": 1.95546, "tr": 2.2582, "fi": 2.471, "ru": 1.43297, "el": 1.50681, "hi": 1.20484, "ta": 0.99235, "th": 1.14016, "zh": 3.19506, "my": 1.12537}, "bytes_per_chunk": {"en": 4.235, "de": 4.145, "vi": 5.361, "tr": 3.745, "fi": 4.065, "ru": 4.831, "el": 4.662, "hi": 5.842, "ta": 5.777, "th": 4.142, "zh": 6.028, "my": 4.747}, "bpb_mean": 1.80616, "t": 254.3} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 254.3} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.29735, "bpb": 1.87168, "chunk_loss": 0.032761, "F": 0.21338, "G": 0.43883, "chunks": 218.5, "bytes_per_chunk": 4.686, "boundary_entropy": 0.67032, "router_noise": 0.0, "grad_norm": 0.7334, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 254.3} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.31127, "bpb": 1.89177, "chunk_loss": 0.034146, "F": 0.23679, "G": 0.43882, "chunks": 242.47, "bytes_per_chunk": 4.223, "boundary_entropy": 0.66886, "router_noise": 0.0, "grad_norm": 0.7625, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 256.5} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.35679, "bpb": 1.95743, "chunk_loss": 0.033385, "F": 0.22455, "G": 0.43627, "chunks": 229.94, "bytes_per_chunk": 4.453, "boundary_entropy": 0.66735, "router_noise": 0.0, "grad_norm": 0.7509, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 258.6} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.48579, "bpb": 2.14355, "chunk_loss": 0.03279, "F": 0.21338, "G": 0.44218, "chunks": 218.5, "bytes_per_chunk": 4.686, "boundary_entropy": 0.67227, "router_noise": 0.0, "grad_norm": 0.8083, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 260.8} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3997.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.229673, "cps": 31.226562, "mean_p": 0.434035, "boundary_entropy": 0.658353, "bpb": 2.065376}, "de": {"chunks": 4741.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.230897, "cps": 37.039062, "mean_p": 0.442088, "boundary_entropy": 0.667299, "bpb": 2.287168}, "vi": {"chunks": 3996.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.216762, "cps": 31.21875, "mean_p": 0.439304, "boundary_entropy": 0.67083, "bpb": 1.94631}, "tr": {"chunks": 4987.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.275479, "cps": 38.960938, "mean_p": 0.455466, "boundary_entropy": 0.670496, "bpb": 2.238298}, "fi": {"chunks": 4474.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.23784, "cps": 34.953125, "mean_p": 0.44961, "boundary_entropy": 0.671421, "bpb": 2.445548}, "ru": {"chunks": 5798.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305963, "cps": 45.296875, "mean_p": 0.43421, "boundary_entropy": 0.666915, "bpb": 1.447379}, "el": {"chunks": 6229.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.303513, "cps": 48.664062, "mean_p": 0.441462, "boundary_entropy": 0.673606, "bpb": 1.506159}, "hi": {"chunks": 8560.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.508767, "cps": 66.875, "mean_p": 0.444538, "boundary_entropy": 0.669804, "bpb": 1.198777}, "ta": {"chunks": 8384.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.448462, "cps": 65.5, "mean_p": 0.436393, "boundary_entropy": 0.665295, "bpb": 0.974756}, "th": {"chunks": 7074.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.439789, "cps": 55.265625, "mean_p": 0.420162, "boundary_entropy": 0.662123, "bpb": 1.117313}, "zh": {"chunks": 2731.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.462411, "cps": 21.335938, "mean_p": 0.450331, "boundary_entropy": 0.675515, "bpb": 3.182054}, "my": {"chunks": 8362.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.441313, "cps": 65.328125, "mean_p": 0.457607, "boundary_entropy": 0.683306, "bpb": 1.114587}}, "bpb": {"en": 2.06538, "de": 2.28717, "vi": 1.94631, "tr": 2.2383, "fi": 2.44555, "ru": 1.44738, "el": 1.50616, "hi": 1.19878, "ta": 0.97476, "th": 1.11731, "zh": 3.18205, "my": 1.11459}, "gini_cpc": 0.169525, "gini_cps": 0.183223, "bpic": 5.535329, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0053274017754585, "vi": 0.9437833460530711, "tr": 1.1994407201302208, "fi": 1.0355571420471685, "ru": 1.3321679122138437, "el": 1.3215008830076729, "hi": 2.215178174120932, "ta": 1.9526111821095027, "th": 1.9148464861235295, "zh": 2.013345134146919, "my": 1.9214839406945963}, "t": 263.8} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.07484, "de": 2.29261, "vi": 1.947, "tr": 2.24558, "fi": 2.45504, "ru": 1.42401, "el": 1.49815, "hi": 1.19448, "ta": 0.98342, "th": 1.13583, "zh": 3.1819, "my": 1.12232}, "bytes_per_chunk": {"en": 4.361, "de": 4.308, "vi": 6.19, "tr": 3.848, "fi": 4.184, "ru": 5.787, "el": 5.781, "hi": 5.34, "ta": 5.83, "th": 6.273, "zh": 5.818, "my": 6.55}, "bpb_mean": 1.79626, "t": 265.1} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.45359, "bpb": 2.09709, "chunk_loss": 0.032975, "F": 0.21686, "G": 0.43828, "chunks": 222.06, "bytes_per_chunk": 4.611, "boundary_entropy": 0.66934, "router_noise": 0.0, "grad_norm": 0.8055, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 265.1} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.3815, "bpb": 1.99309, "chunk_loss": 0.033484, "F": 0.22549, "G": 0.44041, "chunks": 230.91, "bytes_per_chunk": 4.435, "boundary_entropy": 0.67002, "router_noise": 0.0, "grad_norm": 0.8063, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 267.2} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.31539, "bpb": 1.89771, "chunk_loss": 0.032515, "F": 0.2099, "G": 0.43487, "chunks": 214.94, "bytes_per_chunk": 4.764, "boundary_entropy": 0.66721, "router_noise": 0.0, "grad_norm": 0.7751, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 269.4} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.36912, "bpb": 1.97523, "chunk_loss": 0.03263, "F": 0.21069, "G": 0.44032, "chunks": 215.75, "bytes_per_chunk": 4.746, "boundary_entropy": 0.6709, "router_noise": 0.0, "grad_norm": 0.7988, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 271.5} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4064.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.233523, "cps": 31.75, "mean_p": 0.434724, "boundary_entropy": 0.65847, "bpb": 2.060952}, "de": {"chunks": 4869.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.23713, "cps": 38.039062, "mean_p": 0.442966, "boundary_entropy": 0.667334, "bpb": 2.271174}, "vi": {"chunks": 4403.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.238839, "cps": 34.398438, "mean_p": 0.442493, "boundary_entropy": 0.671875, "bpb": 1.934881}, "tr": {"chunks": 5066.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.279843, "cps": 39.578125, "mean_p": 0.45464, "boundary_entropy": 0.669967, "bpb": 2.224028}, "fi": {"chunks": 4756.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.252831, "cps": 37.15625, "mean_p": 0.451056, "boundary_entropy": 0.671552, "bpb": 2.429466}, "ru": {"chunks": 6763.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.356887, "cps": 52.835938, "mean_p": 0.438184, "boundary_entropy": 0.667299, "bpb": 1.436715}, "el": {"chunks": 7394.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.360279, "cps": 57.765625, "mean_p": 0.446165, "boundary_entropy": 0.674919, "bpb": 1.48955}, "hi": {"chunks": 9144.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.543477, "cps": 71.4375, "mean_p": 0.447015, "boundary_entropy": 0.670664, "bpb": 1.193225}, "ta": {"chunks": 8656.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.463012, "cps": 67.625, "mean_p": 0.439414, "boundary_entropy": 0.666509, "bpb": 0.96933}, "th": {"chunks": 6052.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.376251, "cps": 47.28125, "mean_p": 0.420968, "boundary_entropy": 0.663211, "bpb": 1.105827}, "zh": {"chunks": 2834.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.479851, "cps": 22.140625, "mean_p": 0.450662, "boundary_entropy": 0.675345, "bpb": 3.175681}, "my": {"chunks": 11532.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.608613, "cps": 90.09375, "mean_p": 0.461671, "boundary_entropy": 0.683811, "bpb": 1.10866}}, "bpb": {"en": 2.06095, "de": 2.27117, "vi": 1.93488, "tr": 2.22403, "fi": 2.42947, "ru": 1.43671, "el": 1.48955, "hi": 1.19322, "ta": 0.96933, "th": 1.10583, "zh": 3.17568, "my": 1.10866}, "gini_cpc": 0.186466, "gini_cps": 0.210754, "bpic": 5.080971, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0154482332262527, "vi": 1.022765251500804, "tr": 1.1983537933200266, "fi": 1.082680671009633, "ru": 1.528271780794881, "el": 1.5427978394511095, "hi": 2.3272956909361073, "ta": 1.982723706073248, "th": 1.6111956290278762, "zh": 2.0548343836909484, "my": 2.6062236327686716}, "t": 274.6} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.07054, "de": 2.2768, "vi": 1.9365, "tr": 2.23131, "fi": 2.43762, "ru": 1.41386, "el": 1.48224, "hi": 1.18826, "ta": 0.97849, "th": 1.1246, "zh": 3.17448, "my": 1.11642}, "bytes_per_chunk": {"en": 4.331, "de": 4.211, "vi": 5.465, "tr": 3.763, "fi": 3.955, "ru": 4.97, "el": 4.982, "hi": 5.007, "ta": 5.729, "th": 7.375, "zh": 5.525, "my": 4.679}, "bpb_mean": 1.78593, "t": 275.9} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.41642, "bpb": 2.04346, "chunk_loss": 0.03328, "F": 0.22122, "G": 0.4435, "chunks": 226.53, "bytes_per_chunk": 4.52, "boundary_entropy": 0.67206, "router_noise": 0.0, "grad_norm": 0.7848, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 275.9} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.51301, "bpb": 2.18282, "chunk_loss": 0.032266, "F": 0.20398, "G": 0.44125, "chunks": 208.88, "bytes_per_chunk": 4.902, "boundary_entropy": 0.67246, "router_noise": 0.0, "grad_norm": 0.7965, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 278.0} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.45882, "bpb": 2.10463, "chunk_loss": 0.03302, "F": 0.21701, "G": 0.4426, "chunks": 222.22, "bytes_per_chunk": 4.608, "boundary_entropy": 0.67238, "router_noise": 0.0, "grad_norm": 0.8256, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 280.1} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.46527, "bpb": 2.11394, "chunk_loss": 0.034118, "F": 0.23639, "G": 0.43923, "chunks": 242.06, "bytes_per_chunk": 4.23, "boundary_entropy": 0.66818, "router_noise": 0.0, "grad_norm": 0.7768, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 282.2} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4152.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.23858, "cps": 32.4375, "mean_p": 0.436071, "boundary_entropy": 0.658858, "bpb": 2.049368}, "de": {"chunks": 4833.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.235377, "cps": 37.757812, "mean_p": 0.442691, "boundary_entropy": 0.66731, "bpb": 2.262656}, "vi": {"chunks": 4720.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.256035, "cps": 36.875, "mean_p": 0.443522, "boundary_entropy": 0.67175, "bpb": 1.929906}, "tr": {"chunks": 5029.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.277799, "cps": 39.289062, "mean_p": 0.454828, "boundary_entropy": 0.670102, "bpb": 2.213485}, "fi": {"chunks": 4651.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.247249, "cps": 36.335938, "mean_p": 0.45024, "boundary_entropy": 0.671262, "bpb": 2.410573}, "ru": {"chunks": 6631.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.349921, "cps": 51.804688, "mean_p": 0.438178, "boundary_entropy": 0.667895, "bpb": 1.428883}, "el": {"chunks": 6708.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326853, "cps": 52.40625, "mean_p": 0.444393, "boundary_entropy": 0.674466, "bpb": 1.479427}, "hi": {"chunks": 9945.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.591085, "cps": 77.695312, "mean_p": 0.446879, "boundary_entropy": 0.669911, "bpb": 1.183791}, "ta": {"chunks": 8013.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.428617, "cps": 62.601562, "mean_p": 0.435018, "boundary_entropy": 0.665083, "bpb": 0.963584}, "th": {"chunks": 7436.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.462294, "cps": 58.09375, "mean_p": 0.425309, "boundary_entropy": 0.664909, "bpb": 1.102422}, "zh": {"chunks": 2689.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.4553, "cps": 21.007812, "mean_p": 0.44906, "boundary_entropy": 0.674995, "bpb": 3.156962}, "my": {"chunks": 10303.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.543751, "cps": 80.492188, "mean_p": 0.4624, "boundary_entropy": 0.684327, "bpb": 1.105949}}, "bpb": {"en": 2.04937, "de": 2.26266, "vi": 1.92991, "tr": 2.21348, "fi": 2.41057, "ru": 1.42888, "el": 1.47943, "hi": 1.18379, "ta": 0.96358, "th": 1.10242, "zh": 3.15696, "my": 1.10595}, "gini_cpc": 0.181977, "gini_cps": 0.199964, "bpic": 5.109586, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9865774015549198, "vi": 1.0731628546478722, "tr": 1.1643883714224865, "fi": 1.036337542939302, "ru": 1.4666841170100813, "el": 1.3699950063075996, "hi": 2.4775161257075125, "ta": 1.796538168995141, "th": 1.9376935392001668, "zh": 1.9083768294807473, "my": 2.2791195114465994}, "t": 285.3} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.0593, "de": 2.2685, "vi": 1.93069, "tr": 2.22149, "fi": 2.41938, "ru": 1.4054, "el": 1.47112, "hi": 1.17913, "ta": 0.97281, "th": 1.12042, "zh": 3.156, "my": 1.11212}, "bytes_per_chunk": {"en": 4.246, "de": 4.205, "vi": 5.131, "tr": 3.82, "fi": 4.052, "ru": 5.043, "el": 5.406, "hi": 4.564, "ta": 6.194, "th": 6.09, "zh": 5.863, "my": 5.225}, "bpb_mean": 1.77636, "t": 286.6} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.35354, "bpb": 1.95274, "chunk_loss": 0.033094, "F": 0.21851, "G": 0.44074, "chunks": 223.75, "bytes_per_chunk": 4.577, "boundary_entropy": 0.66964, "router_noise": 0.0, "grad_norm": 0.7645, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 286.6} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.28892, "bpb": 1.85951, "chunk_loss": 0.033611, "F": 0.22833, "G": 0.43791, "chunks": 233.81, "bytes_per_chunk": 4.38, "boundary_entropy": 0.66774, "router_noise": 0.0, "grad_norm": 0.7235, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 288.8} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.39046, "bpb": 2.00601, "chunk_loss": 0.033375, "F": 0.22238, "G": 0.44415, "chunks": 227.72, "bytes_per_chunk": 4.497, "boundary_entropy": 0.67195, "router_noise": 0.0, "grad_norm": 0.8336, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 290.9} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.26466, "bpb": 1.82452, "chunk_loss": 0.032574, "F": 0.21063, "G": 0.43664, "chunks": 215.69, "bytes_per_chunk": 4.748, "boundary_entropy": 0.66807, "router_noise": 0.0, "grad_norm": 0.7852, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 293.1} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3991.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.229328, "cps": 31.179688, "mean_p": 0.434388, "boundary_entropy": 0.658312, "bpb": 2.04571}, "de": {"chunks": 4751.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.231384, "cps": 37.117188, "mean_p": 0.442055, "boundary_entropy": 0.667286, "bpb": 2.255464}, "vi": {"chunks": 4371.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237103, "cps": 34.148438, "mean_p": 0.441849, "boundary_entropy": 0.671372, "bpb": 1.921536}, "tr": {"chunks": 4804.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.26537, "cps": 37.53125, "mean_p": 0.452307, "boundary_entropy": 0.66944, "bpb": 2.212791}, "fi": {"chunks": 4273.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.227154, "cps": 33.382812, "mean_p": 0.447542, "boundary_entropy": 0.670781, "bpb": 2.401466}, "ru": {"chunks": 6833.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.36058, "cps": 53.382812, "mean_p": 0.438144, "boundary_entropy": 0.667413, "bpb": 1.42143}, "el": {"chunks": 7867.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.383326, "cps": 61.460938, "mean_p": 0.44656, "boundary_entropy": 0.674389, "bpb": 1.468712}, "hi": {"chunks": 11307.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.672036, "cps": 88.335938, "mean_p": 0.451312, "boundary_entropy": 0.671186, "bpb": 1.176522}, "ta": {"chunks": 8383.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.448409, "cps": 65.492188, "mean_p": 0.437373, "boundary_entropy": 0.665925, "bpb": 0.958254}, "th": {"chunks": 11478.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.713584, "cps": 89.671875, "mean_p": 0.425843, "boundary_entropy": 0.663943, "bpb": 1.100425}, "zh": {"chunks": 2739.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.463766, "cps": 21.398438, "mean_p": 0.450232, "boundary_entropy": 0.675311, "bpb": 3.147428}, "my": {"chunks": 10030.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.529343, "cps": 78.359375, "mean_p": 0.460809, "boundary_entropy": 0.684008, "bpb": 1.098801}}, "bpb": {"en": 2.04571, "de": 2.25546, "vi": 1.92154, "tr": 2.21279, "fi": 2.40147, "ru": 1.42143, "el": 1.46871, "hi": 1.17652, "ta": 0.95825, "th": 1.10043, "zh": 3.14743, "my": 1.0988}, "gini_cpc": 0.22973, "gini_cps": 0.240956, "bpic": 4.748178, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0089624779505897, "vi": 1.0339036224507583, "tr": 1.1571638019494037, "fi": 0.9905203485259302, "ru": 1.5723332511343135, "el": 1.671516613300845, "hi": 2.930452671557066, "ta": 1.9553134563465846, "th": 3.1116271168724596, "zh": 2.0222785809215336, "my": 2.3082346140119894}, "t": 296.1} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.0552, "de": 2.26226, "vi": 1.92219, "tr": 2.21873, "fi": 2.41185, "ru": 1.39872, "el": 1.46146, "hi": 1.1724, "ta": 0.96659, "th": 1.11772, "zh": 3.1467, "my": 1.10668}, "bytes_per_chunk": {"en": 4.386, "de": 4.259, "vi": 5.572, "tr": 3.981, "fi": 4.389, "ru": 4.863, "el": 4.64, "hi": 3.839, "ta": 5.846, "th": 3.865, "zh": 5.84, "my": 5.368}, "bpb_mean": 1.77004, "t": 297.4} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.37241, "bpb": 1.97996, "chunk_loss": 0.032599, "F": 0.21048, "G": 0.44076, "chunks": 215.53, "bytes_per_chunk": 4.751, "boundary_entropy": 0.67167, "router_noise": 0.0, "grad_norm": 0.7703, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 297.5} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.38579, "bpb": 1.99927, "chunk_loss": 0.03251, "F": 0.20972, "G": 0.43767, "chunks": 214.75, "bytes_per_chunk": 4.768, "boundary_entropy": 0.66925, "router_noise": 0.0, "grad_norm": 0.9128, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 299.6} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.49132, "bpb": 2.15152, "chunk_loss": 0.032637, "F": 0.21121, "G": 0.4403, "chunks": 216.28, "bytes_per_chunk": 4.735, "boundary_entropy": 0.6714, "router_noise": 0.0, "grad_norm": 0.7813, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 301.7} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.343, "bpb": 1.93755, "chunk_loss": 0.032974, "F": 0.21741, "G": 0.43892, "chunks": 222.62, "bytes_per_chunk": 4.6, "boundary_entropy": 0.66911, "router_noise": 0.0, "grad_norm": 0.7456, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 303.8} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3951.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.22703, "cps": 30.867188, "mean_p": 0.433235, "boundary_entropy": 0.657593, "bpb": 2.039593}, "de": {"chunks": 4739.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.230799, "cps": 37.023438, "mean_p": 0.441269, "boundary_entropy": 0.666692, "bpb": 2.24973}, "vi": {"chunks": 4654.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.252455, "cps": 36.359375, "mean_p": 0.442179, "boundary_entropy": 0.670914, "bpb": 1.9189}, "tr": {"chunks": 4904.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.270894, "cps": 38.3125, "mean_p": 0.452684, "boundary_entropy": 0.669194, "bpb": 2.195603}, "fi": {"chunks": 4435.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.235766, "cps": 34.648438, "mean_p": 0.44766, "boundary_entropy": 0.670295, "bpb": 2.389766}, "ru": {"chunks": 6079.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.320792, "cps": 47.492188, "mean_p": 0.434681, "boundary_entropy": 0.666701, "bpb": 1.416736}, "el": {"chunks": 6580.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.320616, "cps": 51.40625, "mean_p": 0.442447, "boundary_entropy": 0.67394, "bpb": 1.465603}, "hi": {"chunks": 8030.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.477266, "cps": 62.734375, "mean_p": 0.443754, "boundary_entropy": 0.669892, "bpb": 1.173641}, "ta": {"chunks": 8274.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.442578, "cps": 64.640625, "mean_p": 0.436072, "boundary_entropy": 0.665794, "bpb": 0.952286}, "th": {"chunks": 8104.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.503823, "cps": 63.3125, "mean_p": 0.421884, "boundary_entropy": 0.663303, "bpb": 1.094289}, "zh": {"chunks": 2637.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.446495, "cps": 20.601562, "mean_p": 0.449463, "boundary_entropy": 0.675221, "bpb": 3.144915}, "my": {"chunks": 9201.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.485592, "cps": 71.882812, "mean_p": 0.459355, "boundary_entropy": 0.683889, "bpb": 1.095005}}, "bpb": {"en": 2.03959, "de": 2.24973, "vi": 1.9189, "tr": 2.1956, "fi": 2.38977, "ru": 1.41674, "el": 1.4656, "hi": 1.17364, "ta": 0.95229, "th": 1.09429, "zh": 3.14491, "my": 1.09501}, "gini_cpc": 0.168359, "gini_cps": 0.186684, "bpic": 5.360968, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0166030118895593, "vi": 1.1119885810289143, "tr": 1.1932103192014678, "fi": 1.0384816602928697, "ru": 1.4129930300535116, "el": 1.4122192886836855, "hi": 2.102217092518568, "ta": 1.9494277214681668, "th": 2.219194961089636, "zh": 1.966680345920561, "my": 2.1388914535549395}, "t": 306.9} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.05073, "de": 2.25671, "vi": 1.91936, "tr": 2.20352, "fi": 2.39799, "ru": 1.39463, "el": 1.45899, "hi": 1.16982, "ta": 0.96166, "th": 1.11192, "zh": 3.14345, "my": 1.10285}, "bytes_per_chunk": {"en": 4.409, "de": 4.279, "vi": 5.223, "tr": 3.912, "fi": 4.205, "ru": 5.545, "el": 5.527, "hi": 5.633, "ta": 6.006, "th": 5.553, "zh": 5.981, "my": 5.754}, "bpb_mean": 1.7643, "t": 308.1} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.25901, "bpb": 1.81636, "chunk_loss": 0.032033, "F": 0.20099, "G": 0.43786, "chunks": 205.81, "bytes_per_chunk": 4.975, "boundary_entropy": 0.66983, "router_noise": 0.0, "grad_norm": 0.7883, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 308.2} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.30563, "bpb": 1.88363, "chunk_loss": 0.032528, "F": 0.20935, "G": 0.43901, "chunks": 214.38, "bytes_per_chunk": 4.777, "boundary_entropy": 0.67034, "router_noise": 0.0, "grad_norm": 0.7483, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 310.3} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.23772, "bpb": 1.78565, "chunk_loss": 0.033095, "F": 0.21896, "G": 0.43786, "chunks": 224.22, "bytes_per_chunk": 4.567, "boundary_entropy": 0.66887, "router_noise": 0.0, "grad_norm": 0.7853, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 312.4} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.46187, "bpb": 2.10904, "chunk_loss": 0.032235, "F": 0.20441, "G": 0.44208, "chunks": 209.31, "bytes_per_chunk": 4.892, "boundary_entropy": 0.67334, "router_noise": 0.0, "grad_norm": 0.8466, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 314.5} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3961.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.227604, "cps": 30.945312, "mean_p": 0.433929, "boundary_entropy": 0.65799, "bpb": 2.034833}, "de": {"chunks": 4743.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.230994, "cps": 37.054688, "mean_p": 0.441414, "boundary_entropy": 0.666493, "bpb": 2.240733}, "vi": {"chunks": 4373.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237212, "cps": 34.164062, "mean_p": 0.443358, "boundary_entropy": 0.672009, "bpb": 1.914145}, "tr": {"chunks": 4891.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.270176, "cps": 38.210938, "mean_p": 0.453227, "boundary_entropy": 0.669429, "bpb": 2.186953}, "fi": {"chunks": 4477.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.237999, "cps": 34.976562, "mean_p": 0.448572, "boundary_entropy": 0.670583, "bpb": 2.378534}, "ru": {"chunks": 6422.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.338892, "cps": 50.171875, "mean_p": 0.436953, "boundary_entropy": 0.667707, "bpb": 1.408304}, "el": {"chunks": 6829.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332749, "cps": 53.351562, "mean_p": 0.444006, "boundary_entropy": 0.674288, "bpb": 1.455789}, "hi": {"chunks": 8939.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.531293, "cps": 69.835938, "mean_p": 0.446014, "boundary_entropy": 0.670149, "bpb": 1.165514}, "ta": {"chunks": 8055.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.430864, "cps": 62.929688, "mean_p": 0.436003, "boundary_entropy": 0.665575, "bpb": 0.94978}, "th": {"chunks": 10344.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.643084, "cps": 80.8125, "mean_p": 0.42283, "boundary_entropy": 0.662923, "bpb": 1.088038}, "zh": {"chunks": 2744.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.464612, "cps": 21.4375, "mean_p": 0.450768, "boundary_entropy": 0.675558, "bpb": 3.130673}, "my": {"chunks": 9174.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.484167, "cps": 71.671875, "mean_p": 0.458634, "boundary_entropy": 0.68374, "bpb": 1.094291}}, "bpb": {"en": 2.03483, "de": 2.24073, "vi": 1.91414, "tr": 2.18695, "fi": 2.37853, "ru": 1.4083, "el": 1.45579, "hi": 1.16551, "ta": 0.94978, "th": 1.08804, "zh": 3.13067, "my": 1.09429}, "gini_cpc": 0.199845, "gini_cps": 0.209198, "bpic": 5.120357, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0148923882396788, "vi": 1.0422109053918505, "tr": 1.1870428301638374, "fi": 1.0456696155760652, "ru": 1.488950859268368, "el": 1.461960185481888, "hi": 2.3342810380406585, "ta": 1.8930380922472307, "th": 2.8254441325145847, "zh": 2.041314601995258, "my": 2.127230922648451}, "t": 317.6} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.04504, "de": 2.24723, "vi": 1.91388, "tr": 2.19439, "fi": 2.38816, "ru": 1.38545, "el": 1.44873, "hi": 1.16042, "ta": 0.95855, "th": 1.10529, "zh": 3.13043, "my": 1.10211}, "bytes_per_chunk": {"en": 4.374, "de": 4.242, "vi": 5.61, "tr": 3.916, "fi": 4.162, "ru": 5.221, "el": 5.34, "hi": 4.949, "ta": 6.125, "th": 4.412, "zh": 5.776, "my": 5.762}, "bpb_mean": 1.75664, "t": 318.9} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.35647, "bpb": 1.95697, "chunk_loss": 0.032635, "F": 0.21213, "G": 0.43667, "chunks": 217.22, "bytes_per_chunk": 4.714, "boundary_entropy": 0.66946, "router_noise": 0.0, "grad_norm": 0.8261, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 318.9} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.21103, "bpb": 1.74715, "chunk_loss": 0.032409, "F": 0.20758, "G": 0.43925, "chunks": 212.56, "bytes_per_chunk": 4.817, "boundary_entropy": 0.67022, "router_noise": 0.0, "grad_norm": 0.7772, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 321.0} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.32316, "bpb": 1.90892, "chunk_loss": 0.033294, "F": 0.22192, "G": 0.44098, "chunks": 227.25, "bytes_per_chunk": 4.506, "boundary_entropy": 0.67022, "router_noise": 0.0, "grad_norm": 0.8264, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 323.1} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.39504, "bpb": 2.01262, "chunk_loss": 0.032704, "F": 0.21243, "G": 0.43985, "chunks": 217.53, "bytes_per_chunk": 4.707, "boundary_entropy": 0.67002, "router_noise": 0.0, "grad_norm": 0.864, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 325.2} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3957.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.227375, "cps": 30.914062, "mean_p": 0.433263, "boundary_entropy": 0.657875, "bpb": 2.030313}, "de": {"chunks": 4658.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.226854, "cps": 36.390625, "mean_p": 0.441055, "boundary_entropy": 0.666592, "bpb": 2.235148}, "vi": {"chunks": 4385.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237863, "cps": 34.257812, "mean_p": 0.442543, "boundary_entropy": 0.671644, "bpb": 1.91102}, "tr": {"chunks": 4890.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.270121, "cps": 38.203125, "mean_p": 0.452786, "boundary_entropy": 0.669378, "bpb": 2.183486}, "fi": {"chunks": 4420.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.234969, "cps": 34.53125, "mean_p": 0.447895, "boundary_entropy": 0.670443, "bpb": 2.377023}, "ru": {"chunks": 6268.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.330765, "cps": 48.96875, "mean_p": 0.436141, "boundary_entropy": 0.667228, "bpb": 1.402874}, "el": {"chunks": 7318.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.356576, "cps": 57.171875, "mean_p": 0.44527, "boundary_entropy": 0.674411, "bpb": 1.448991}, "hi": {"chunks": 8059.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.47899, "cps": 62.960938, "mean_p": 0.445268, "boundary_entropy": 0.670309, "bpb": 1.16401}, "ta": {"chunks": 8195.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.438353, "cps": 64.023438, "mean_p": 0.436394, "boundary_entropy": 0.665918, "bpb": 0.947005}, "th": {"chunks": 5773.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.358906, "cps": 45.101562, "mean_p": 0.420584, "boundary_entropy": 0.662707, "bpb": 1.087003}, "zh": {"chunks": 2752.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.465967, "cps": 21.5, "mean_p": 0.450478, "boundary_entropy": 0.675401, "bpb": 3.124528}, "my": {"chunks": 8610.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.454402, "cps": 67.265625, "mean_p": 0.458194, "boundary_entropy": 0.683578, "bpb": 1.088647}}, "bpb": {"en": 2.03031, "de": 2.23515, "vi": 1.91102, "tr": 2.18349, "fi": 2.37702, "ru": 1.40287, "el": 1.44899, "hi": 1.16401, "ta": 0.94701, "th": 1.087, "zh": 3.12453, "my": 1.08865}, "gini_cpc": 0.158278, "gini_cps": 0.178611, "bpic": 5.539164, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9977118879658466, "vi": 1.046127275864033, "tr": 1.1879998276227217, "fi": 1.0333999967515337, "ru": 1.4547147535211973, "el": 1.5682295390252998, "hi": 2.1066100551207803, "ta": 1.9278869267465586, "th": 1.578478104968794, "zh": 2.0493354697437, "my": 1.9984709759106083}, "t": 328.3} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.04018, "de": 2.2409, "vi": 1.91153, "tr": 2.1895, "fi": 2.38571, "ru": 1.38047, "el": 1.44193, "hi": 1.15928, "ta": 0.95577, "th": 1.10498, "zh": 3.12383, "my": 1.09601}, "bytes_per_chunk": {"en": 4.386, "de": 4.339, "vi": 5.607, "tr": 3.938, "fi": 4.239, "ru": 5.307, "el": 4.972, "hi": 5.623, "ta": 6.06, "th": 7.82, "zh": 5.639, "my": 6.162}, "bpb_mean": 1.75251, "t": 329.5} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.34361, "bpb": 1.93842, "chunk_loss": 0.031966, "F": 0.19968, "G": 0.43669, "chunks": 204.47, "bytes_per_chunk": 5.008, "boundary_entropy": 0.66893, "router_noise": 0.0, "grad_norm": 0.7782, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 329.6} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.31981, "bpb": 1.90408, "chunk_loss": 0.032684, "F": 0.21231, "G": 0.43824, "chunks": 217.41, "bytes_per_chunk": 4.71, "boundary_entropy": 0.66969, "router_noise": 0.0, "grad_norm": 0.8412, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 331.7} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.2959, "bpb": 1.86959, "chunk_loss": 0.032546, "F": 0.20993, "G": 0.43551, "chunks": 214.97, "bytes_per_chunk": 4.763, "boundary_entropy": 0.66742, "router_noise": 0.0, "grad_norm": 0.7966, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 333.8} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.20743, "bpb": 1.74195, "chunk_loss": 0.032563, "F": 0.2106, "G": 0.43735, "chunks": 215.66, "bytes_per_chunk": 4.748, "boundary_entropy": 0.66833, "router_noise": 0.0, "grad_norm": 0.7787, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 335.9} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4013.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.230592, "cps": 31.351562, "mean_p": 0.434595, "boundary_entropy": 0.658248, "bpb": 2.025719}, "de": {"chunks": 4729.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.230312, "cps": 36.945312, "mean_p": 0.44221, "boundary_entropy": 0.666788, "bpb": 2.230589}, "vi": {"chunks": 4676.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.253648, "cps": 36.53125, "mean_p": 0.443942, "boundary_entropy": 0.672019, "bpb": 1.905695}, "tr": {"chunks": 4922.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.271889, "cps": 38.453125, "mean_p": 0.453433, "boundary_entropy": 0.669418, "bpb": 2.178028}, "fi": {"chunks": 4585.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.24374, "cps": 35.820312, "mean_p": 0.449228, "boundary_entropy": 0.670644, "bpb": 2.365266}, "ru": {"chunks": 6375.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.336412, "cps": 49.804688, "mean_p": 0.436076, "boundary_entropy": 0.666933, "bpb": 1.400753}, "el": {"chunks": 7634.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.371973, "cps": 59.640625, "mean_p": 0.44466, "boundary_entropy": 0.673933, "bpb": 1.441807}, "hi": {"chunks": 8991.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.534383, "cps": 70.242188, "mean_p": 0.446988, "boundary_entropy": 0.670746, "bpb": 1.158128}, "ta": {"chunks": 8310.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.444504, "cps": 64.921875, "mean_p": 0.436597, "boundary_entropy": 0.666171, "bpb": 0.942998}, "th": {"chunks": 12119.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.753435, "cps": 94.679688, "mean_p": 0.425341, "boundary_entropy": 0.663483, "bpb": 1.081961}, "zh": {"chunks": 2813.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.476295, "cps": 21.976562, "mean_p": 0.450953, "boundary_entropy": 0.675445, "bpb": 3.119445}, "my": {"chunks": 9797.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.517047, "cps": 76.539062, "mean_p": 0.461585, "boundary_entropy": 0.684361, "bpb": 1.088209}}, "bpb": {"en": 2.02572, "de": 2.23059, "vi": 1.90569, "tr": 2.17803, "fi": 2.36527, "ru": 1.40075, "el": 1.44181, "hi": 1.15813, "ta": 0.943, "th": 1.08196, "zh": 3.11945, "my": 1.08821}, "gini_cpc": 0.215617, "gini_cps": 0.225421, "bpic": 4.860202, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.998784668669662, "vi": 1.0999838807034177, "tr": 1.179087454256239, "fi": 1.0570180828632498, "ru": 1.4589013805396316, "el": 1.6131184963005591, "hi": 2.3174367258860413, "ta": 1.9276603512507857, "th": 3.267387781751583, "zh": 2.065528777156598, "my": 2.2422534063217907}, "t": 339.0} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.03544, "de": 2.2376, "vi": 1.90745, "tr": 2.18415, "fi": 2.3738, "ru": 1.37788, "el": 1.43451, "hi": 1.15354, "ta": 0.95287, "th": 1.09924, "zh": 3.11842, "my": 1.09535}, "bytes_per_chunk": {"en": 4.323, "de": 4.278, "vi": 5.18, "tr": 3.896, "fi": 4.109, "ru": 5.222, "el": 4.801, "hi": 5.002, "ta": 6.085, "th": 3.785, "zh": 5.629, "my": 5.371}, "bpb_mean": 1.74752, "t": 340.3} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.29339, "bpb": 1.86597, "chunk_loss": 0.033032, "F": 0.21878, "G": 0.43703, "chunks": 224.03, "bytes_per_chunk": 4.571, "boundary_entropy": 0.66876, "router_noise": 0.0, "grad_norm": 0.8356, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 340.3} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.28687, "bpb": 1.85656, "chunk_loss": 0.032782, "F": 0.21323, "G": 0.44166, "chunks": 218.34, "bytes_per_chunk": 4.69, "boundary_entropy": 0.67103, "router_noise": 0.0, "grad_norm": 0.7797, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 342.4} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.3591, "bpb": 1.96076, "chunk_loss": 0.032922, "F": 0.21564, "G": 0.44207, "chunks": 220.81, "bytes_per_chunk": 4.637, "boundary_entropy": 0.67215, "router_noise": 0.0, "grad_norm": 0.8606, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 344.5} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.40514, "bpb": 2.02719, "chunk_loss": 0.032939, "F": 0.21545, "G": 0.44244, "chunks": 220.62, "bytes_per_chunk": 4.641, "boundary_entropy": 0.67187, "router_noise": 0.0, "grad_norm": 0.8217, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 346.6} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3979.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.228639, "cps": 31.085938, "mean_p": 0.433868, "boundary_entropy": 0.658059, "bpb": 2.022513}, "de": {"chunks": 4752.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.231432, "cps": 37.125, "mean_p": 0.441517, "boundary_entropy": 0.666617, "bpb": 2.228951}, "vi": {"chunks": 4378.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237483, "cps": 34.203125, "mean_p": 0.441779, "boundary_entropy": 0.671324, "bpb": 1.903086}, "tr": {"chunks": 4866.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.268795, "cps": 38.015625, "mean_p": 0.452434, "boundary_entropy": 0.669224, "bpb": 2.17349}, "fi": {"chunks": 4386.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.233161, "cps": 34.265625, "mean_p": 0.447501, "boundary_entropy": 0.670471, "bpb": 2.367439}, "ru": {"chunks": 6309.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.332929, "cps": 49.289062, "mean_p": 0.435936, "boundary_entropy": 0.667, "bpb": 1.398225}, "el": {"chunks": 6979.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.340057, "cps": 54.523438, "mean_p": 0.443426, "boundary_entropy": 0.673794, "bpb": 1.441775}, "hi": {"chunks": 9044.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.537533, "cps": 70.65625, "mean_p": 0.44741, "boundary_entropy": 0.670897, "bpb": 1.15861}, "ta": {"chunks": 8321.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.445092, "cps": 65.007812, "mean_p": 0.437543, "boundary_entropy": 0.666248, "bpb": 0.939992}, "th": {"chunks": 10754.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.668573, "cps": 84.015625, "mean_p": 0.423592, "boundary_entropy": 0.662906, "bpb": 1.080115}, "zh": {"chunks": 2794.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.473078, "cps": 21.828125, "mean_p": 0.451208, "boundary_entropy": 0.675493, "bpb": 3.111745}, "my": {"chunks": 10070.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.531455, "cps": 78.671875, "mean_p": 0.461693, "boundary_entropy": 0.684314, "bpb": 1.084167}}, "bpb": {"en": 2.02251, "de": 2.22895, "vi": 1.90309, "tr": 2.17349, "fi": 2.36744, "ru": 1.39823, "el": 1.44178, "hi": 1.15861, "ta": 0.93999, "th": 1.08012, "zh": 3.11175, "my": 1.08417}, "gini_cpc": 0.209726, "gini_cps": 0.220578, "bpic": 5.008104, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0122183493242607, "vi": 1.0386824563703136, "tr": 1.1756329079876167, "fi": 1.0197810206310258, "ru": 1.4561345082792843, "el": 1.4873135488871023, "hi": 2.3510164171708303, "ta": 1.9467053999262376, "th": 2.9241466408623515, "zh": 2.069107906247045, "my": 2.3244289486237197}, "t": 349.7} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.0335, "de": 2.23654, "vi": 1.90372, "tr": 2.18072, "fi": 2.37655, "ru": 1.37561, "el": 1.43417, "hi": 1.15268, "ta": 0.94915, "th": 1.097, "zh": 3.11017, "my": 1.09135}, "bytes_per_chunk": {"en": 4.43, "de": 4.265, "vi": 5.563, "tr": 3.938, "fi": 4.348, "ru": 5.276, "el": 5.21, "hi": 4.954, "ta": 5.971, "th": 4.236, "zh": 5.565, "my": 5.213}, "bpb_mean": 1.7451, "t": 350.9} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3979.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.228639, "cps": 31.085938, "mean_p": 0.433868, "boundary_entropy": 0.658059, "bpb": 2.022513}, "de": {"chunks": 4752.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.231432, "cps": 37.125, "mean_p": 0.441517, "boundary_entropy": 0.666617, "bpb": 2.228951}, "vi": {"chunks": 4378.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237483, "cps": 34.203125, "mean_p": 0.441779, "boundary_entropy": 0.671324, "bpb": 1.903086}, "tr": {"chunks": 4866.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.268795, "cps": 38.015625, "mean_p": 0.452434, "boundary_entropy": 0.669224, "bpb": 2.17349}, "fi": {"chunks": 4386.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.233161, "cps": 34.265625, "mean_p": 0.447501, "boundary_entropy": 0.670471, "bpb": 2.367439}, "ru": {"chunks": 6309.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.332929, "cps": 49.289062, "mean_p": 0.435936, "boundary_entropy": 0.667, "bpb": 1.398225}, "el": {"chunks": 6979.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.340057, "cps": 54.523438, "mean_p": 0.443426, "boundary_entropy": 0.673794, "bpb": 1.441775}, "hi": {"chunks": 9044.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.537533, "cps": 70.65625, "mean_p": 0.44741, "boundary_entropy": 0.670897, "bpb": 1.15861}, "ta": {"chunks": 8321.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.445092, "cps": 65.007812, "mean_p": 0.437543, "boundary_entropy": 0.666248, "bpb": 0.939992}, "th": {"chunks": 10754.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.668573, "cps": 84.015625, "mean_p": 0.423592, "boundary_entropy": 0.662906, "bpb": 1.080115}, "zh": {"chunks": 2794.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.473078, "cps": 21.828125, "mean_p": 0.451208, "boundary_entropy": 0.675493, "bpb": 3.111745}, "my": {"chunks": 10070.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.531455, "cps": 78.671875, "mean_p": 0.461693, "boundary_entropy": 0.684314, "bpb": 1.084167}}, "bpb": {"en": 2.02251, "de": 2.22895, "vi": 1.90309, "tr": 2.17349, "fi": 2.36744, "ru": 1.39823, "el": 1.44178, "hi": 1.15861, "ta": 0.93999, "th": 1.08012, "zh": 3.11175, "my": 1.08417}, "gini_cpc": 0.209726, "gini_cps": 0.220578, "bpic": 5.008104, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0122183493242607, "vi": 1.0386824563703136, "tr": 1.1756329079876167, "fi": 1.0197810206310258, "ru": 1.4561345082792843, "el": 1.4873135488871023, "hi": 2.3510164171708303, "ta": 1.9467053999262376, "th": 2.9241466408623515, "zh": 2.069107906247045, "my": 2.3244289486237197}, "t": 351.9} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_noema_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_noema_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_noema_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_noema_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_noema_s2/th0.70"], "t": 356.8} |
| {"event": "final", "run": "emactl_noema_s2", "condition": "A", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.22863874044704935, "cps": 31.0859375, "bpb": 2.0440929497706244, "bytes_per_chunk": 4.447226675488375}, "de": {"cpc": 0.23143232844689038, "cps": 37.125, "bpb": 2.229200510378607, "bytes_per_chunk": 4.252719353033265}, "vi": {"cpc": 0.2374830485489558, "cps": 34.203125, "bpb": 1.9064757324673185, "bytes_per_chunk": 5.540809316973674}, "tr": {"cpc": 0.2687952273103905, "cps": 38.015625, "bpb": 2.1740123956039623, "bytes_per_chunk": 3.9388166000540914}, "fi": {"cpc": 0.23316144808888417, "cps": 34.265625, "bpb": 2.3927762408084456, "bytes_per_chunk": 4.315731419210918}, "ru": {"cpc": 0.3329287598944591, "cps": 49.2890625, "bpb": 1.3761500732624246, "bytes_per_chunk": 5.276809887577121}, "el": {"cpc": 0.34005749646737804, "cps": 54.5234375, "bpb": 1.4288364316573776, "bytes_per_chunk": 5.19611496531219}, "hi": {"cpc": 0.5375334323922734, "cps": 70.65625, "bpb": 1.15970682221278, "bytes_per_chunk": 4.935961889698545}, "ta": {"cpc": 0.4450922706606044, "cps": 65.0078125, "bpb": 0.9492078075249067, "bytes_per_chunk": 5.968602361084232}, "th": {"cpc": 0.6685732048492384, "cps": 84.015625, "bpb": 1.0939795622231419, "bytes_per_chunk": 4.236396838992227}, "zh": {"cpc": 0.47307822553335593, "cps": 21.828125, "bpb": 3.1189088345478835, "bytes_per_chunk": 5.605319990591656}, "my": {"cpc": 0.5314545070719865, "cps": 78.671875, "bpb": 1.1001802838363859, "bytes_per_chunk": 5.20814169489505}}, "gini_cpc": 0.2097257925277538, "gini_cps": 0.2205780004871769, "bpic": 5.008103664265581, "seed": 2, "mean_boundary_entropy": 0.6694104671478271, "dechunk": "ste", "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": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 359.2} |
|
|
