| {"event": "start", "run": "pilot_B_s1", "condition": "B", "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": "parity_a", "p3_objective": "none", "freeze_in_p3": true, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 1.3} |
| {"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 1.3} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2674.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.153652, "cps": 20.890625, "mean_p": 0.478126, "boundary_entropy": 0.685438, "bpb": 8.041379}, "de": {"chunks": 2056.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.100131, "cps": 16.0625, "mean_p": 0.469682, "boundary_entropy": 0.685273, "bpb": 8.067143}, "vi": {"chunks": 14185.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.76946, "cps": 110.820312, "mean_p": 0.507219, "boundary_entropy": 0.688283, "bpb": 7.894113}, "tr": {"chunks": 5645.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.311827, "cps": 44.101562, "mean_p": 0.486063, "boundary_entropy": 0.686375, "bpb": 8.01701}, "fi": {"chunks": 4732.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.251555, "cps": 36.96875, "mean_p": 0.48451, "boundary_entropy": 0.686595, "bpb": 8.075294}, "ru": {"chunks": 17548.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.926016, "cps": 137.09375, "mean_p": 0.502342, "boundary_entropy": 0.690066, "bpb": 8.095633}, "el": {"chunks": 31473.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.533548, "cps": 245.882812, "mean_p": 0.517965, "boundary_entropy": 0.689704, "bpb": 7.804942}, "hi": {"chunks": 1659.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.098603, "cps": 12.960938, "mean_p": 0.476241, "boundary_entropy": 0.689492, "bpb": 8.047677}, "ta": {"chunks": 39693.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.123188, "cps": 310.101562, "mean_p": 0.514597, "boundary_entropy": 0.690542, "bpb": 8.096414}, "th": {"chunks": 44424.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.761828, "cps": 347.0625, "mean_p": 0.52078, "boundary_entropy": 0.690224, "bpb": 8.201461}, "zh": {"chunks": 12933.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.189807, "cps": 101.039062, "mean_p": 0.528309, "boundary_entropy": 0.684752, "bpb": 8.121995}, "my": {"chunks": 7048.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.371965, "cps": 55.0625, "mean_p": 0.491145, "boundary_entropy": 0.691021, "bpb": 8.112229}}, "bpb": {"en": 8.04138, "de": 8.06714, "vi": 7.89411, "tr": 8.01701, "fi": 8.07529, "ru": 8.09563, "el": 7.80494, "hi": 8.04768, "ta": 8.09641, "th": 8.20146, "zh": 8.12199, "my": 8.11223}, "gini_cpc": 0.507116, "gini_cps": 0.500863, "bpic": 2.084973, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6516785410034256, "vi": 5.0078223656304, "tr": 2.029439514461233, "fi": 1.6371767549398106, "ru": 6.026721581613623, "el": 9.980677296913099, "hi": 0.6417324941848254, "ta": 13.818190381639045, "th": 17.9746032262692, "zh": 14.25176170627287, "my": 2.4208352616119475}, "t": 3.2} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.57556, "bpb": 8.04383, "chunk_loss": 0.043969, "F": 0.34537, "G": 0.49096, "chunks": 353.66, "bytes_per_chunk": 2.895, "boundary_entropy": 0.6904, "router_noise": 0.0, "grad_norm": 7.3222, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 3.6} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2652.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.152388, "cps": 20.71875, "mean_p": 0.478021, "boundary_entropy": 0.685429, "bpb": 8.012078}, "de": {"chunks": 2040.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.099352, "cps": 15.9375, "mean_p": 0.469566, "boundary_entropy": 0.685259, "bpb": 8.041798}, "vi": {"chunks": 14104.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.765066, "cps": 110.1875, "mean_p": 0.507011, "boundary_entropy": 0.688289, "bpb": 7.873437}, "tr": {"chunks": 5609.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.309838, "cps": 43.820312, "mean_p": 0.48594, "boundary_entropy": 0.686368, "bpb": 7.996385}, "fi": {"chunks": 4696.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.249641, "cps": 36.6875, "mean_p": 0.484353, "boundary_entropy": 0.686583, "bpb": 8.050017}, "ru": {"chunks": 17478.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.922322, "cps": 136.546875, "mean_p": 0.502276, "boundary_entropy": 0.690066, "bpb": 8.084027}, "el": {"chunks": 31425.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.531209, "cps": 245.507812, "mean_p": 0.517879, "boundary_entropy": 0.68971, "bpb": 7.792315}, "hi": {"chunks": 1661.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.098722, "cps": 12.976562, "mean_p": 0.476278, "boundary_entropy": 0.689497, "bpb": 8.025307}, "ta": {"chunks": 39575.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.116876, "cps": 309.179688, "mean_p": 0.514531, "boundary_entropy": 0.690545, "bpb": 8.080302}, "th": {"chunks": 44406.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.760709, "cps": 346.921875, "mean_p": 0.520711, "boundary_entropy": 0.69023, "bpb": 8.182805}, "zh": {"chunks": 12933.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.189807, "cps": 101.039062, "mean_p": 0.528283, "boundary_entropy": 0.684755, "bpb": 8.116868}, "my": {"chunks": 7054.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.372282, "cps": 55.109375, "mean_p": 0.49115, "boundary_entropy": 0.691021, "bpb": 8.097738}}, "bpb": {"en": 8.01208, "de": 8.0418, "vi": 7.87344, "tr": 7.99638, "fi": 8.05002, "ru": 8.08403, "el": 7.79231, "hi": 8.02531, "ta": 8.0803, "th": 8.1828, "zh": 8.11687, "my": 8.09774}, "gini_cpc": 0.507752, "gini_cps": 0.501315, "bpic": 2.089935, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.651971123407348, "vi": 5.0205322111319335, "tr": 2.0332252247883713, "fi": 1.6381995611983071, "ru": 6.052476629377141, "el": 10.048125290618648, "hi": 0.6478361224476075, "ta": 13.891401202199152, "th": 18.11637032900724, "zh": 14.369988990412386, "my": 2.44299557159055}, "t": 4.9} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2633.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.151296, "cps": 20.570312, "mean_p": 0.477866, "boundary_entropy": 0.685416, "bpb": 7.961917}, "de": {"chunks": 2015.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.098135, "cps": 15.742188, "mean_p": 0.469403, "boundary_entropy": 0.685239, "bpb": 8.000014}, "vi": {"chunks": 13973.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.75796, "cps": 109.164062, "mean_p": 0.506724, "boundary_entropy": 0.688298, "bpb": 7.840276}, "tr": {"chunks": 5554.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.3068, "cps": 43.390625, "mean_p": 0.485757, "boundary_entropy": 0.686358, "bpb": 7.962004}, "fi": {"chunks": 4630.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.246133, "cps": 36.171875, "mean_p": 0.484105, "boundary_entropy": 0.686566, "bpb": 8.009966}, "ru": {"chunks": 17258.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.910712, "cps": 134.828125, "mean_p": 0.502049, "boundary_entropy": 0.690066, "bpb": 8.040027}, "el": {"chunks": 31368.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.528432, "cps": 245.0625, "mean_p": 0.517785, "boundary_entropy": 0.689716, "bpb": 7.769565}, "hi": {"chunks": 1671.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.099316, "cps": 13.054688, "mean_p": 0.476368, "boundary_entropy": 0.689506, "bpb": 7.98396}, "ta": {"chunks": 39397.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.107355, "cps": 307.789062, "mean_p": 0.514418, "boundary_entropy": 0.69055, "bpb": 8.042672}, "th": {"chunks": 44381.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.759154, "cps": 346.726562, "mean_p": 0.520515, "boundary_entropy": 0.690244, "bpb": 8.123019}, "zh": {"chunks": 12927.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.188791, "cps": 100.992188, "mean_p": 0.52826, "boundary_entropy": 0.684757, "bpb": 8.106766}, "my": {"chunks": 7072.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.373232, "cps": 55.25, "mean_p": 0.491162, "boundary_entropy": 0.691021, "bpb": 8.083851}}, "bpb": {"en": 7.96192, "de": 8.00001, "vi": 7.84028, "tr": 7.962, "fi": 8.00997, "ru": 8.04003, "el": 7.76956, "hi": 7.98396, "ta": 8.04267, "th": 8.12302, "zh": 8.10677, "my": 8.08385}, "gini_cpc": 0.508769, "gini_cps": 0.501992, "bpic": 2.098552, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6486283166599119, "vi": 5.00979296078409, "tr": 2.0278161851375045, "fi": 1.6268307347300228, "ru": 6.0194181221296175, "el": 10.102276365558682, "hi": 0.6564393973168309, "ta": 13.928711530274855, "th": 18.23682704904785, "zh": 14.466969546570148, "my": 2.4669033630009527}, "t": 6.5} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2558.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.146986, "cps": 19.984375, "mean_p": 0.477477, "boundary_entropy": 0.685387, "bpb": 7.791563}, "de": {"chunks": 1969.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.095894, "cps": 15.382812, "mean_p": 0.468984, "boundary_entropy": 0.685192, "bpb": 7.852521}, "vi": {"chunks": 13537.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.73431, "cps": 105.757812, "mean_p": 0.50567, "boundary_entropy": 0.688326, "bpb": 7.729936}, "tr": {"chunks": 5388.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.29763, "cps": 42.09375, "mean_p": 0.485236, "boundary_entropy": 0.686332, "bpb": 7.842911}, "fi": {"chunks": 4452.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.23667, "cps": 34.78125, "mean_p": 0.483395, "boundary_entropy": 0.686515, "bpb": 7.870383}, "ru": {"chunks": 16097.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.849446, "cps": 125.757812, "mean_p": 0.500929, "boundary_entropy": 0.690065, "bpb": 7.865283}, "el": {"chunks": 31134.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.51703, "cps": 243.234375, "mean_p": 0.517303, "boundary_entropy": 0.689748, "bpb": 7.686066}, "hi": {"chunks": 1732.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.102942, "cps": 13.53125, "mean_p": 0.476707, "boundary_entropy": 0.689543, "bpb": 7.829999}, "ta": {"chunks": 38746.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.072533, "cps": 302.703125, "mean_p": 0.513987, "boundary_entropy": 0.690571, "bpb": 7.912712}, "th": {"chunks": 44274.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.752502, "cps": 345.890625, "mean_p": 0.519894, "boundary_entropy": 0.690291, "bpb": 7.955047}, "zh": {"chunks": 12923.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.188114, "cps": 100.960938, "mean_p": 0.528242, "boundary_entropy": 0.684757, "bpb": 8.079683}, "my": {"chunks": 7120.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.375765, "cps": 55.625, "mean_p": 0.49119, "boundary_entropy": 0.691022, "bpb": 8.000772}}, "bpb": {"en": 7.79156, "de": 7.85252, "vi": 7.72994, "tr": 7.84291, "fi": 7.87038, "ru": 7.86528, "el": 7.68607, "hi": 7.83, "ta": 7.91271, "th": 7.95505, "zh": 8.07968, "my": 8.00077}, "gini_cpc": 0.512203, "gini_cps": 0.50365, "bpic": 2.132946, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6524044114877986, "vi": 4.995774961495422, "tr": 2.024886177121214, "fi": 1.610151912761772, "ru": 5.779088028452308, "el": 10.320902039501288, "hi": 0.7003520348712253, "ta": 14.100190645230702, "th": 18.72626977042165, "zh": 14.886530163597694, "my": 2.5564670387974298}, "t": 8.1} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2300.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.132161, "cps": 17.96875, "mean_p": 0.476349, "boundary_entropy": 0.685317, "bpb": 7.297383}, "de": {"chunks": 1784.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.086885, "cps": 13.9375, "mean_p": 0.467598, "boundary_entropy": 0.685052, "bpb": 7.408897}, "vi": {"chunks": 12230.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.663412, "cps": 95.546875, "mean_p": 0.502575, "boundary_entropy": 0.688389, "bpb": 7.434477}, "tr": {"chunks": 4853.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.268077, "cps": 37.914062, "mean_p": 0.483427, "boundary_entropy": 0.686243, "bpb": 7.494908}, "fi": {"chunks": 3815.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.202807, "cps": 29.804688, "mean_p": 0.480875, "boundary_entropy": 0.68633, "bpb": 7.453954}, "ru": {"chunks": 12105.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.638786, "cps": 94.570312, "mean_p": 0.496809, "boundary_entropy": 0.69002, "bpb": 7.244612}, "el": {"chunks": 30498.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.48604, "cps": 238.265625, "mean_p": 0.516116, "boundary_entropy": 0.689821, "bpb": 7.511993}, "hi": {"chunks": 1891.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.112392, "cps": 14.773438, "mean_p": 0.477698, "boundary_entropy": 0.689646, "bpb": 7.379035}, "ta": {"chunks": 34872.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.865312, "cps": 272.4375, "mean_p": 0.511346, "boundary_entropy": 0.690681, "bpb": 7.492708}, "th": {"chunks": 43411.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.69885, "cps": 339.148438, "mean_p": 0.517046, "boundary_entropy": 0.690483, "bpb": 7.519822}, "zh": {"chunks": 12918.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.187267, "cps": 100.921875, "mean_p": 0.528222, "boundary_entropy": 0.684759, "bpb": 7.972126}, "my": {"chunks": 7556.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.398776, "cps": 59.03125, "mean_p": 0.491562, "boundary_entropy": 0.691038, "bpb": 7.746845}}, "bpb": {"en": 7.29738, "de": 7.4089, "vi": 7.43448, "tr": 7.49491, "fi": 7.45395, "ru": 7.24461, "el": 7.51199, "hi": 7.37903, "ta": 7.49271, "th": 7.51982, "zh": 7.97213, "my": 7.74685}, "gini_cpc": 0.522773, "gini_cps": 0.507385, "bpic": 2.281247, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6574136649592702, "vi": 5.019721229702481, "tr": 2.0284113130420374, "fi": 1.5345425780013082, "ru": 4.833390271882529, "el": 11.244154363397165, "hi": 0.8504185800116286, "ta": 14.113920625138087, "th": 20.420906137232908, "zh": 16.550004711494577, "my": 3.0173442189608175}, "t": 9.8} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2297.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.131989, "cps": 17.945312, "mean_p": 0.478691, "boundary_entropy": 0.685751, "bpb": 6.635969}, "de": {"chunks": 1724.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.083962, "cps": 13.46875, "mean_p": 0.470362, "boundary_entropy": 0.685668, "bpb": 6.75635}, "vi": {"chunks": 9748.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.528777, "cps": 76.15625, "mean_p": 0.496644, "boundary_entropy": 0.688419, "bpb": 7.048647}, "tr": {"chunks": 4407.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.24344, "cps": 34.429688, "mean_p": 0.482564, "boundary_entropy": 0.686346, "bpb": 7.015074}, "fi": {"chunks": 2975.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158152, "cps": 23.242188, "mean_p": 0.478165, "boundary_entropy": 0.686224, "bpb": 6.854253}, "ru": {"chunks": 6018.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.317573, "cps": 47.015625, "mean_p": 0.487701, "boundary_entropy": 0.689679, "bpb": 6.435902}, "el": {"chunks": 23085.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.124836, "cps": 180.351562, "mean_p": 0.506375, "boundary_entropy": 0.69025, "bpb": 6.833856}, "hi": {"chunks": 3520.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.209212, "cps": 27.5, "mean_p": 0.483426, "boundary_entropy": 0.690144, "bpb": 6.500127}, "ta": {"chunks": 21612.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.156031, "cps": 168.84375, "mean_p": 0.498104, "boundary_entropy": 0.690847, "bpb": 6.585306}, "th": {"chunks": 24782.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.54069, "cps": 193.609375, "mean_p": 0.50304, "boundary_entropy": 0.690961, "bpb": 6.462071}, "zh": {"chunks": 13087.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.215882, "cps": 102.242188, "mean_p": 0.528833, "boundary_entropy": 0.68476, "bpb": 7.68123}, "my": {"chunks": 8526.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.449968, "cps": 66.609375, "mean_p": 0.492476, "boundary_entropy": 0.691083, "bpb": 6.94941}}, "bpb": {"en": 6.63597, "de": 6.75635, "vi": 7.04865, "tr": 7.01507, "fi": 6.85425, "ru": 6.4359, "el": 6.83386, "hi": 6.50013, "ta": 6.58531, "th": 6.46207, "zh": 7.68123, "my": 6.94941}, "gini_cpc": 0.497121, "gini_cps": 0.436272, "bpic": 3.151403, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6361330787028944, "vi": 4.006226538808004, "tr": 1.8444021682054395, "fi": 1.198224544975919, "ru": 2.4060580107355816, "el": 8.522208568951946, "hi": 1.585078282222241, "ta": 8.758558082958306, "th": 11.672890522689851, "zh": 16.788418424715406, "my": 3.409141890887776}, "t": 11.8} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3543.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203586, "cps": 27.679688, "mean_p": 0.488237, "boundary_entropy": 0.686849, "bpb": 6.226829}, "de": {"chunks": 3111.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151512, "cps": 24.304688, "mean_p": 0.483352, "boundary_entropy": 0.687367, "bpb": 6.333018}, "vi": {"chunks": 6111.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.331489, "cps": 47.742188, "mean_p": 0.489049, "boundary_entropy": 0.688304, "bpb": 6.743617}, "tr": {"chunks": 4961.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.274043, "cps": 38.757812, "mean_p": 0.48712, "boundary_entropy": 0.687061, "bpb": 6.642139}, "fi": {"chunks": 3340.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177556, "cps": 26.09375, "mean_p": 0.483102, "boundary_entropy": 0.686944, "bpb": 6.410081}, "ru": {"chunks": 2363.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.124697, "cps": 18.460938, "mean_p": 0.481959, "boundary_entropy": 0.689457, "bpb": 5.796101}, "el": {"chunks": 1338.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.065195, "cps": 10.453125, "mean_p": 0.47371, "boundary_entropy": 0.688749, "bpb": 6.232684}, "hi": {"chunks": 4972.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.295513, "cps": 38.84375, "mean_p": 0.488157, "boundary_entropy": 0.690517, "bpb": 5.768992}, "ta": {"chunks": 4398.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.23525, "cps": 34.359375, "mean_p": 0.482205, "boundary_entropy": 0.690299, "bpb": 6.128617}, "th": {"chunks": 4784.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.29742, "cps": 37.375, "mean_p": 0.489181, "boundary_entropy": 0.690753, "bpb": 5.552717}, "zh": {"chunks": 4833.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.81832, "cps": 37.757812, "mean_p": 0.491118, "boundary_entropy": 0.686034, "bpb": 7.308257}, "my": {"chunks": 4123.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.217596, "cps": 32.210938, "mean_p": 0.489019, "boundary_entropy": 0.690966, "bpb": 5.72524}}, "bpb": {"en": 6.22683, "de": 6.33302, "vi": 6.74362, "tr": 6.64214, "fi": 6.41008, "ru": 5.7961, "el": 6.23268, "hi": 5.76899, "ta": 6.12862, "th": 5.55272, "zh": 7.30826, "my": 5.72524}, "gini_cpc": 0.312892, "gini_cps": 0.174748, "bpic": 8.015978, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.7442186887937731, "vi": 1.6282538346163042, "tr": 1.3460823924889673, "fi": 0.872142702458805, "ru": 0.6125019492894309, "el": 0.32023458697317475, "hi": 1.4515400226216733, "ta": 1.1555339863430047, "th": 1.460908693943611, "zh": 4.019539680796492, "my": 1.0688159510310047}, "t": 14.2} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.18188, "bpb": 6.03318, "chunk_loss": 0.02876, "F": 0.14755, "G": 0.48498, "chunks": 151.09, "bytes_per_chunk": 6.777, "boundary_entropy": 0.6915, "router_noise": 0.0, "grad_norm": 1.6425, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.235, "alpha": 0.03, "t": 15.3} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5811.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.333908, "cps": 45.398438, "mean_p": 0.49668, "boundary_entropy": 0.687369, "bpb": 5.700957}, "de": {"chunks": 5758.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.280427, "cps": 44.984375, "mean_p": 0.493273, "boundary_entropy": 0.688122, "bpb": 5.815575}, "vi": {"chunks": 4999.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.271169, "cps": 39.054688, "mean_p": 0.485756, "boundary_entropy": 0.688071, "bpb": 6.193973}, "tr": {"chunks": 4523.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.249848, "cps": 35.335938, "mean_p": 0.487362, "boundary_entropy": 0.68725, "bpb": 6.161437}, "fi": {"chunks": 4506.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.239541, "cps": 35.203125, "mean_p": 0.489602, "boundary_entropy": 0.687557, "bpb": 5.889799}, "ru": {"chunks": 10041.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.529868, "cps": 78.445312, "mean_p": 0.494861, "boundary_entropy": 0.690227, "bpb": 5.366184}, "el": {"chunks": 2472.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.12045, "cps": 19.3125, "mean_p": 0.484382, "boundary_entropy": 0.689923, "bpb": 5.639089}, "hi": {"chunks": 8847.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.525825, "cps": 69.117188, "mean_p": 0.49112, "boundary_entropy": 0.690632, "bpb": 5.21088}, "ta": {"chunks": 9979.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.533779, "cps": 77.960938, "mean_p": 0.487149, "boundary_entropy": 0.690607, "bpb": 5.234383}, "th": {"chunks": 2652.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.164874, "cps": 20.71875, "mean_p": 0.486587, "boundary_entropy": 0.69063, "bpb": 5.049163}, "zh": {"chunks": 2592.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.438876, "cps": 20.25, "mean_p": 0.48492, "boundary_entropy": 0.686171, "bpb": 6.981327}, "my": {"chunks": 23598.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.245408, "cps": 184.359375, "mean_p": 0.498896, "boundary_entropy": 0.69134, "bpb": 4.9514}}, "bpb": {"en": 5.70096, "de": 5.81557, "vi": 6.19397, "tr": 6.16144, "fi": 5.8898, "ru": 5.36618, "el": 5.63909, "hi": 5.21088, "ta": 5.23438, "th": 5.04916, "zh": 6.98133, "my": 4.9514}, "gini_cpc": 0.33211, "gini_cps": 0.366298, "bpic": 4.474119, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8398321540446331, "vi": 0.8121069731250978, "tr": 0.748254403022959, "fi": 0.7173854246190545, "ru": 1.5868687127361492, "el": 0.3607288406618149, "hi": 1.574759362719114, "ta": 1.5985815559749659, "th": 0.49377113614126295, "zh": 1.3143614091075098, "my": 3.729795885138023}, "t": 17.1} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.33762, "bpb": 4.81517, "chunk_loss": 0.032752, "F": 0.20636, "G": 0.48421, "chunks": 211.31, "bytes_per_chunk": 4.846, "boundary_entropy": 0.69114, "router_noise": 0.0, "grad_norm": 1.5993, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.076, "alpha": 0.03, "t": 19.1} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2403.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.13808, "cps": 18.773438, "mean_p": 0.473514, "boundary_entropy": 0.684965, "bpb": 4.478942}, "de": {"chunks": 2819.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.137291, "cps": 22.023438, "mean_p": 0.471835, "boundary_entropy": 0.685519, "bpb": 4.661277}, "vi": {"chunks": 5016.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.272091, "cps": 39.1875, "mean_p": 0.475184, "boundary_entropy": 0.684849, "bpb": 4.897379}, "tr": {"chunks": 3066.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169364, "cps": 23.953125, "mean_p": 0.470316, "boundary_entropy": 0.684478, "bpb": 4.965014}, "fi": {"chunks": 2239.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.119026, "cps": 17.492188, "mean_p": 0.472517, "boundary_entropy": 0.685494, "bpb": 4.720231}, "ru": {"chunks": 5428.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.286438, "cps": 42.40625, "mean_p": 0.479972, "boundary_entropy": 0.687867, "bpb": 3.661614}, "el": {"chunks": 6262.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.305121, "cps": 48.921875, "mean_p": 0.484808, "boundary_entropy": 0.689565, "bpb": 4.431959}, "hi": {"chunks": 2775.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.164933, "cps": 21.679688, "mean_p": 0.47941, "boundary_entropy": 0.689738, "bpb": 3.449549}, "ta": {"chunks": 6964.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.372506, "cps": 54.40625, "mean_p": 0.482293, "boundary_entropy": 0.68992, "bpb": 3.588186}, "th": {"chunks": 8227.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.51147, "cps": 64.273438, "mean_p": 0.479318, "boundary_entropy": 0.68912, "bpb": 3.322672}, "zh": {"chunks": 2123.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.359465, "cps": 16.585938, "mean_p": 0.481728, "boundary_entropy": 0.685845, "bpb": 6.316049}, "my": {"chunks": 12627.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.666403, "cps": 98.648438, "mean_p": 0.487172, "boundary_entropy": 0.690466, "bpb": 3.949849}}, "bpb": {"en": 4.47894, "de": 4.66128, "vi": 4.89738, "tr": 4.96501, "fi": 4.72023, "ru": 3.66161, "el": 4.43196, "hi": 3.44955, "ta": 3.58819, "th": 3.32267, "zh": 6.31605, "my": 3.94985}, "gini_cpc": 0.295625, "gini_cps": 0.317735, "bpic": 6.401792, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9942898776325045, "vi": 1.9705376416176281, "tr": 1.2265688990237098, "fi": 0.8620105071733005, "ru": 2.0744404586615017, "el": 2.2097470740631855, "hi": 1.1944782988574831, "ta": 2.697762057925053, "th": 3.7041689029527514, "zh": 2.6033160800945994, "my": 4.826220430604873}, "t": 21.5} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.91637, "bpb": 4.20744, "chunk_loss": 0.029548, "F": 0.16, "G": 0.46086, "chunks": 163.84, "bytes_per_chunk": 6.25, "boundary_entropy": 0.68531, "router_noise": 0.0, "grad_norm": 1.6821, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.07, "alpha": 0.03, "t": 22.6} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3485.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200253, "cps": 27.226562, "mean_p": 0.367531, "boundary_entropy": 0.538634, "bpb": 3.879564}, "de": {"chunks": 3920.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.190912, "cps": 30.625, "mean_p": 0.34767, "boundary_entropy": 0.536819, "bpb": 4.087636}, "vi": {"chunks": 7316.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.396854, "cps": 57.15625, "mean_p": 0.427289, "boundary_entropy": 0.566078, "bpb": 3.604652}, "tr": {"chunks": 4945.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.273159, "cps": 38.632812, "mean_p": 0.37742, "boundary_entropy": 0.5665, "bpb": 4.128808}, "fi": {"chunks": 3691.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.196215, "cps": 28.835938, "mean_p": 0.348897, "boundary_entropy": 0.55542, "bpb": 4.14556}, "ru": {"chunks": 7854.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.414459, "cps": 61.359375, "mean_p": 0.434928, "boundary_entropy": 0.664023, "bpb": 2.816809}, "el": {"chunks": 8268.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.402865, "cps": 64.59375, "mean_p": 0.448533, "boundary_entropy": 0.676065, "bpb": 2.976759}, "hi": {"chunks": 12593.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.74847, "cps": 98.382812, "mean_p": 0.483812, "boundary_entropy": 0.687613, "bpb": 2.432862}, "ta": {"chunks": 15629.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.835999, "cps": 122.101562, "mean_p": 0.457597, "boundary_entropy": 0.675096, "bpb": 2.448266}, "th": {"chunks": 4872.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.302891, "cps": 38.0625, "mean_p": 0.467285, "boundary_entropy": 0.685227, "bpb": 2.320388}, "zh": {"chunks": 2187.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.370301, "cps": 17.085938, "mean_p": 0.44507, "boundary_entropy": 0.674858, "bpb": 5.268797}, "my": {"chunks": 13436.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.709099, "cps": 104.96875, "mean_p": 0.381384, "boundary_entropy": 0.63787, "bpb": 2.456563}}, "bpb": {"en": 3.87956, "de": 4.08764, "vi": 3.60465, "tr": 4.12881, "fi": 4.14556, "ru": 2.81681, "el": 2.97676, "hi": 2.43286, "ta": 2.44827, "th": 2.32039, "zh": 5.2688, "my": 2.45656}, "gini_cpc": 0.275401, "gini_cps": 0.314444, "bpic": 4.351456, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9533557661072728, "vi": 1.9817638069933565, "tr": 1.3640713338808672, "fi": 0.9798362431334822, "ru": 2.069679129931141, "el": 2.0117821968437832, "hi": 3.7376227815665097, "ta": 4.174717182837232, "th": 1.5125423804403526, "zh": 1.8491693149636026, "my": 3.5410165524372523}, "t": 25.8} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.8834, "de": 4.0879, "vi": 3.60864, "tr": 4.13905, "fi": 4.14674, "ru": 2.80202, "el": 2.96998, "hi": 2.42892, "ta": 2.46585, "th": 2.35258, "zh": 5.27759, "my": 2.47297}, "bytes_per_chunk": {"en": 5.096, "de": 5.439, "vi": 3.251, "tr": 4.025, "fi": 5.263, "ru": 4.317, "el": 4.739, "hi": 3.872, "ta": 3.297, "th": 9.123, "zh": 7.928, "my": 4.034}, "bpb_mean": 3.3863, "t": 27.2} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.53521, "bpb": 3.65753, "chunk_loss": 0.032324, "F": 0.21033, "G": 0.39309, "chunks": 215.38, "bytes_per_chunk": 4.754, "boundary_entropy": 0.58806, "router_noise": 0.0, "grad_norm": 1.4269, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 27.2} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.18169, "bpb": 3.14751, "chunk_loss": 0.030813, "F": 0.18866, "G": 0.35209, "chunks": 193.19, "bytes_per_chunk": 5.301, "boundary_entropy": 0.52347, "router_noise": 0.0, "grad_norm": 0.8862, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 29.4} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3015.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.173246, "cps": 23.554688, "mean_p": 0.284727, "boundary_entropy": 0.392967, "bpb": 3.555239}, "de": {"chunks": 3561.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.173428, "cps": 27.820312, "mean_p": 0.274969, "boundary_entropy": 0.408041, "bpb": 3.82544}, "vi": {"chunks": 7175.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.389205, "cps": 56.054688, "mean_p": 0.351164, "boundary_entropy": 0.443111, "bpb": 3.167883}, "tr": {"chunks": 4284.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.236646, "cps": 33.46875, "mean_p": 0.305201, "boundary_entropy": 0.454378, "bpb": 3.878894}, "fi": {"chunks": 3681.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.195683, "cps": 28.757812, "mean_p": 0.28466, "boundary_entropy": 0.444877, "bpb": 3.888859}, "ru": {"chunks": 10489.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.553509, "cps": 81.945312, "mean_p": 0.316496, "boundary_entropy": 0.500547, "bpb": 2.672706}, "el": {"chunks": 9421.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.459046, "cps": 73.601562, "mean_p": 0.386945, "boundary_entropy": 0.616431, "bpb": 2.731627}, "hi": {"chunks": 5437.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.32315, "cps": 42.476562, "mean_p": 0.454402, "boundary_entropy": 0.673323, "bpb": 2.162595}, "ta": {"chunks": 13884.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.742658, "cps": 108.46875, "mean_p": 0.476724, "boundary_entropy": 0.674417, "bpb": 1.949048}, "th": {"chunks": 15256.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.948461, "cps": 119.1875, "mean_p": 0.461841, "boundary_entropy": 0.67911, "bpb": 2.088107}, "zh": {"chunks": 4530.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.767017, "cps": 35.390625, "mean_p": 0.453434, "boundary_entropy": 0.661581, "bpb": 4.986644}, "my": {"chunks": 6448.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.3403, "cps": 50.375, "mean_p": 0.337194, "boundary_entropy": 0.610311, "bpb": 2.270899}}, "bpb": {"en": 3.55524, "de": 3.82544, "vi": 3.16788, "tr": 3.87889, "fi": 3.88886, "ru": 2.67271, "el": 2.73163, "hi": 2.1626, "ta": 1.94905, "th": 2.08811, "zh": 4.98664, "my": 2.2709}, "gini_cpc": 0.31023, "gini_cps": 0.29617, "bpic": 4.402117, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0010513836118196, "vi": 2.246547301463931, "tr": 1.3659528683709554, "fi": 1.1295117099632876, "ru": 3.1949324091941347, "el": 2.6496771611416197, "hi": 1.865267244756899, "ta": 4.286728113910256, "th": 5.47465074379424, "zh": 4.4273266245811245, "my": 1.9642576646007874}, "t": 30.7} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.17759, "bpb": 3.1416, "chunk_loss": 0.034032, "F": 0.24261, "G": 0.32749, "chunks": 248.44, "bytes_per_chunk": 4.122, "boundary_entropy": 0.45157, "router_noise": 0.0, "grad_norm": 0.8537, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.072, "alpha": 0.03, "t": 33.0} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.44153, "bpb": 3.52238, "chunk_loss": 0.03358, "F": 0.23276, "G": 0.34812, "chunks": 238.34, "bytes_per_chunk": 4.296, "boundary_entropy": 0.43088, "router_noise": 0.0, "grad_norm": 0.9308, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.053, "alpha": 0.03, "t": 35.7} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3191.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183359, "cps": 24.929688, "mean_p": 0.301227, "boundary_entropy": 0.378825, "bpb": 3.188475}, "de": {"chunks": 3656.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178055, "cps": 28.5625, "mean_p": 0.291257, "boundary_entropy": 0.397045, "bpb": 3.444483}, "vi": {"chunks": 4626.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.250936, "cps": 36.140625, "mean_p": 0.298321, "boundary_entropy": 0.383547, "bpb": 2.756291}, "tr": {"chunks": 3964.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.218969, "cps": 30.96875, "mean_p": 0.291957, "boundary_entropy": 0.400217, "bpb": 3.459199}, "fi": {"chunks": 3631.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.193025, "cps": 28.367188, "mean_p": 0.281296, "boundary_entropy": 0.4036, "bpb": 3.623437}, "ru": {"chunks": 8316.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.438839, "cps": 64.96875, "mean_p": 0.277634, "boundary_entropy": 0.267203, "bpb": 2.553491}, "el": {"chunks": 8303.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.40457, "cps": 64.867188, "mean_p": 0.279817, "boundary_entropy": 0.328976, "bpb": 2.567341}, "hi": {"chunks": 4809.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.285825, "cps": 37.570312, "mean_p": 0.317423, "boundary_entropy": 0.536976, "bpb": 1.945983}, "ta": {"chunks": 8711.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.465953, "cps": 68.054688, "mean_p": 0.306656, "boundary_entropy": 0.492786, "bpb": 1.737444}, "th": {"chunks": 5349.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.332546, "cps": 41.789062, "mean_p": 0.367361, "boundary_entropy": 0.614014, "bpb": 1.93672}, "zh": {"chunks": 3637.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.615814, "cps": 28.414062, "mean_p": 0.298984, "boundary_entropy": 0.480241, "bpb": 4.162665}, "my": {"chunks": 3522.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.185877, "cps": 27.515625, "mean_p": 0.361249, "boundary_entropy": 0.606629, "bpb": 2.002539}}, "bpb": {"en": 3.18848, "de": 3.44448, "vi": 2.75629, "tr": 3.4592, "fi": 3.62344, "ru": 2.55349, "el": 2.56734, "hi": 1.94598, "ta": 1.73744, "th": 1.93672, "zh": 4.16267, "my": 2.00254}, "gini_cpc": 0.234493, "gini_cps": 0.201533, "bpic": 6.218602, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9710712489283747, "vi": 1.3685472694638783, "tr": 1.1942091939412887, "fi": 1.0527171095690842, "ru": 2.3933299872911036, "el": 2.206436883885028, "hi": 1.5588237720967637, "ta": 2.541205930795098, "th": 1.8136306582655388, "zh": 3.358514088313273, "my": 1.0137323167259722}, "t": 38.8} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.1984, "de": 3.44307, "vi": 2.76182, "tr": 3.47125, "fi": 3.61595, "ru": 2.5395, "el": 2.56225, "hi": 1.9438, "ta": 1.75198, "th": 1.96624, "zh": 4.17144, "my": 2.0161}, "bytes_per_chunk": {"en": 5.475, "de": 5.725, "vi": 5.319, "tr": 4.989, "fi": 5.482, "ru": 4.06, "el": 4.394, "hi": 7.904, "ta": 5.554, "th": 6.993, "zh": 3.618, "my": 15.329}, "bpb_mean": 2.78682, "t": 40.1} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.00947, "bpb": 2.89905, "chunk_loss": 0.030649, "F": 0.18927, "G": 0.30399, "chunks": 193.81, "bytes_per_chunk": 5.283, "boundary_entropy": 0.40297, "router_noise": 0.0, "grad_norm": 1.1301, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 40.2} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.07845, "bpb": 2.99857, "chunk_loss": 0.029904, "F": 0.16235, "G": 0.28422, "chunks": 166.25, "bytes_per_chunk": 6.159, "boundary_entropy": 0.39823, "router_noise": 0.0, "grad_norm": 1.0395, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 42.3} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.05666, "bpb": 2.96713, "chunk_loss": 0.030013, "F": 0.17099, "G": 0.27758, "chunks": 175.09, "bytes_per_chunk": 5.848, "boundary_entropy": 0.38092, "router_noise": 0.0, "grad_norm": 0.9061, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 44.3} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.94234, "bpb": 2.80221, "chunk_loss": 0.030009, "F": 0.168, "G": 0.29229, "chunks": 172.03, "bytes_per_chunk": 5.952, "boundary_entropy": 0.4168, "router_noise": 0.0, "grad_norm": 0.7726, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 46.3} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2881.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165546, "cps": 22.507812, "mean_p": 0.275815, "boundary_entropy": 0.351453, "bpb": 2.909661}, "de": {"chunks": 3312.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161301, "cps": 25.875, "mean_p": 0.286557, "boundary_entropy": 0.396892, "bpb": 3.17768}, "vi": {"chunks": 4381.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237646, "cps": 34.226562, "mean_p": 0.295135, "boundary_entropy": 0.380142, "bpb": 2.512123}, "tr": {"chunks": 3678.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203171, "cps": 28.734375, "mean_p": 0.298379, "boundary_entropy": 0.39117, "bpb": 3.154069}, "fi": {"chunks": 3692.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.196268, "cps": 28.84375, "mean_p": 0.340333, "boundary_entropy": 0.458616, "bpb": 3.375687}, "ru": {"chunks": 6847.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.361319, "cps": 53.492188, "mean_p": 0.235267, "boundary_entropy": 0.248614, "bpb": 2.135893}, "el": {"chunks": 7495.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.3652, "cps": 58.554688, "mean_p": 0.235438, "boundary_entropy": 0.244871, "bpb": 2.380012}, "hi": {"chunks": 4914.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.292065, "cps": 38.390625, "mean_p": 0.346735, "boundary_entropy": 0.514961, "bpb": 1.756015}, "ta": {"chunks": 11283.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.60353, "cps": 88.148438, "mean_p": 0.257573, "boundary_entropy": 0.346192, "bpb": 1.433973}, "th": {"chunks": 7613.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.473298, "cps": 59.476562, "mean_p": 0.363317, "boundary_entropy": 0.540329, "bpb": 1.758108}, "zh": {"chunks": 2720.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.460549, "cps": 21.25, "mean_p": 0.331034, "boundary_entropy": 0.557691, "bpb": 3.870194}, "my": {"chunks": 12228.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.645345, "cps": 95.53125, "mean_p": 0.401947, "boundary_entropy": 0.591263, "bpb": 1.73675}}, "bpb": {"en": 2.90966, "de": 3.17768, "vi": 2.51212, "tr": 3.15407, "fi": 3.37569, "ru": 2.13589, "el": 2.38001, "hi": 1.75601, "ta": 1.43397, "th": 1.75811, "zh": 3.87019, "my": 1.73675}, "gini_cpc": 0.258976, "gini_cps": 0.280807, "bpic": 5.402018, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9743585105667051, "vi": 1.4355257037423437, "tr": 1.2272754176930756, "fi": 1.185579471537061, "ru": 2.182589067303844, "el": 2.206031218045879, "hi": 1.7642533110046712, "ta": 3.6456920445756937, "th": 2.859009684428772, "zh": 2.78199486035592, "my": 3.8982789778811147}, "t": 49.2} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.91833, "de": 3.17669, "vi": 2.51861, "tr": 3.16377, "fi": 3.37054, "ru": 2.12025, "el": 2.37326, "hi": 1.75193, "ta": 1.44632, "th": 1.78566, "zh": 3.88231, "my": 1.75023}, "bytes_per_chunk": {"en": 5.685, "de": 6.263, "vi": 5.503, "tr": 5.319, "fi": 5.804, "ru": 5.121, "el": 5.095, "hi": 7.676, "ta": 4.19, "th": 5.622, "zh": 6.18, "my": 4.972}, "bpb_mean": 2.52149, "t": 50.4} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.85736, "bpb": 2.67961, "chunk_loss": 0.030378, "F": 0.18198, "G": 0.29982, "chunks": 186.34, "bytes_per_chunk": 5.495, "boundary_entropy": 0.4046, "router_noise": 0.0, "grad_norm": 0.6343, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 50.5} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.883, "bpb": 2.7166, "chunk_loss": 0.030193, "F": 0.17682, "G": 0.27644, "chunks": 181.06, "bytes_per_chunk": 5.656, "boundary_entropy": 0.35351, "router_noise": 0.0, "grad_norm": 0.8137, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 52.4} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.8343, "bpb": 2.64633, "chunk_loss": 0.030053, "F": 0.1731, "G": 0.27824, "chunks": 177.25, "bytes_per_chunk": 5.777, "boundary_entropy": 0.35557, "router_noise": 0.0, "grad_norm": 0.7153, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 54.4} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.9116, "bpb": 2.75785, "chunk_loss": 0.030105, "F": 0.17282, "G": 0.28612, "chunks": 176.97, "bytes_per_chunk": 5.786, "boundary_entropy": 0.37407, "router_noise": 0.0, "grad_norm": 0.8308, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 56.4} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.270053, "boundary_entropy": 0.326158, "bpb": 2.721808}, "de": {"chunks": 3321.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.16174, "cps": 25.945312, "mean_p": 0.286811, "boundary_entropy": 0.380933, "bpb": 3.028152}, "vi": {"chunks": 4502.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.244209, "cps": 35.171875, "mean_p": 0.284206, "boundary_entropy": 0.347088, "bpb": 2.376336}, "tr": {"chunks": 3865.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213501, "cps": 30.195312, "mean_p": 0.289217, "boundary_entropy": 0.355518, "bpb": 2.984668}, "fi": {"chunks": 3291.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.174951, "cps": 25.710938, "mean_p": 0.316541, "boundary_entropy": 0.421318, "bpb": 3.22565}, "ru": {"chunks": 6724.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.354828, "cps": 52.53125, "mean_p": 0.24595, "boundary_entropy": 0.249817, "bpb": 1.99193}, "el": {"chunks": 6910.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.336695, "cps": 53.984375, "mean_p": 0.22542, "boundary_entropy": 0.289523, "bpb": 2.145686}, "hi": {"chunks": 6995.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.41575, "cps": 54.648438, "mean_p": 0.366063, "boundary_entropy": 0.543075, "bpb": 1.634097}, "ta": {"chunks": 9020.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.482482, "cps": 70.46875, "mean_p": 0.24329, "boundary_entropy": 0.345045, "bpb": 1.345011}, "th": {"chunks": 8403.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.522412, "cps": 65.648438, "mean_p": 0.388614, "boundary_entropy": 0.563385, "bpb": 1.587703}, "zh": {"chunks": 3097.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.524382, "cps": 24.195312, "mean_p": 0.324883, "boundary_entropy": 0.534157, "bpb": 3.707646}, "my": {"chunks": 10368.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.547182, "cps": 81.0, "mean_p": 0.395744, "boundary_entropy": 0.582224, "bpb": 1.584386}}, "bpb": {"en": 2.72181, "de": 3.02815, "vi": 2.37634, "tr": 2.98467, "fi": 3.22565, "ru": 1.99193, "el": 2.14569, "hi": 1.6341, "ta": 1.34501, "th": 1.5877, "zh": 3.70765, "my": 1.58439}, "gini_cpc": 0.236756, "gini_cps": 0.241478, "bpic": 5.531022, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9736267489057571, "vi": 1.4700712263501752, "tr": 1.28521260209532, "fi": 1.0531543535403904, "ru": 2.135966903017188, "el": 2.02681089779, "hi": 2.5026993132861475, "ta": 2.9044044705701464, "th": 3.1447731788932174, "zh": 3.156630811105802, "my": 3.2938789968386963}, "t": 59.2} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.73255, "de": 3.02658, "vi": 2.3819, "tr": 2.99482, "fi": 3.22176, "ru": 1.97502, "el": 2.13317, "hi": 1.62968, "ta": 1.35619, "th": 1.6171, "zh": 3.72071, "my": 1.59373}, "bytes_per_chunk": {"en": 5.739, "de": 6.263, "vi": 5.438, "tr": 5.144, "fi": 6.068, "ru": 5.218, "el": 5.546, "hi": 6.17, "ta": 5.567, "th": 5.39, "zh": 5.627, "my": 5.563}, "bpb_mean": 2.36527, "t": 60.5} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.64328, "bpb": 2.37076, "chunk_loss": 0.030268, "F": 0.1778, "G": 0.27683, "chunks": 182.06, "bytes_per_chunk": 5.624, "boundary_entropy": 0.33801, "router_noise": 0.0, "grad_norm": 0.7396, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 60.5} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.71079, "bpb": 2.46815, "chunk_loss": 0.03004, "F": 0.1727, "G": 0.29556, "chunks": 176.84, "bytes_per_chunk": 5.79, "boundary_entropy": 0.38987, "router_noise": 0.0, "grad_norm": 0.7178, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 62.5} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.6264, "bpb": 2.34639, "chunk_loss": 0.030294, "F": 0.18124, "G": 0.26554, "chunks": 185.59, "bytes_per_chunk": 5.517, "boundary_entropy": 0.3193, "router_noise": 0.0, "grad_norm": 0.6772, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 64.5} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.71602, "bpb": 2.4757, "chunk_loss": 0.030091, "F": 0.17358, "G": 0.27255, "chunks": 177.75, "bytes_per_chunk": 5.761, "boundary_entropy": 0.34732, "router_noise": 0.0, "grad_norm": 0.679, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 66.5} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2904.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166868, "cps": 22.6875, "mean_p": 0.255689, "boundary_entropy": 0.29333, "bpb": 2.588163}, "de": {"chunks": 3329.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162129, "cps": 26.007812, "mean_p": 0.283204, "boundary_entropy": 0.368345, "bpb": 2.91214}, "vi": {"chunks": 4363.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.236669, "cps": 34.085938, "mean_p": 0.263216, "boundary_entropy": 0.302411, "bpb": 2.284005}, "tr": {"chunks": 3865.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213501, "cps": 30.195312, "mean_p": 0.272469, "boundary_entropy": 0.328347, "bpb": 2.901413}, "fi": {"chunks": 3031.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161129, "cps": 23.679688, "mean_p": 0.26591, "boundary_entropy": 0.351865, "bpb": 3.183524}, "ru": {"chunks": 6760.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.356728, "cps": 52.8125, "mean_p": 0.251068, "boundary_entropy": 0.266064, "bpb": 1.875513}, "el": {"chunks": 7440.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.36252, "cps": 58.125, "mean_p": 0.246988, "boundary_entropy": 0.307869, "bpb": 2.033438}, "hi": {"chunks": 6089.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.361902, "cps": 47.570312, "mean_p": 0.372814, "boundary_entropy": 0.538745, "bpb": 1.5696}, "ta": {"chunks": 8906.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.476384, "cps": 69.578125, "mean_p": 0.253168, "boundary_entropy": 0.366624, "bpb": 1.277539}, "th": {"chunks": 7540.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.46876, "cps": 58.90625, "mean_p": 0.396641, "boundary_entropy": 0.583815, "bpb": 1.486955}, "zh": {"chunks": 3617.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.612428, "cps": 28.257812, "mean_p": 0.342657, "boundary_entropy": 0.542461, "bpb": 3.646391}, "my": {"chunks": 7959.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.420044, "cps": 62.179688, "mean_p": 0.341182, "boundary_entropy": 0.532234, "bpb": 1.437301}}, "bpb": {"en": 2.58816, "de": 2.91214, "vi": 2.284, "tr": 2.90141, "fi": 3.18352, "ru": 1.87551, "el": 2.03344, "hi": 1.5696, "ta": 1.27754, "th": 1.48696, "zh": 3.64639, "my": 1.4373}, "gini_cpc": 0.235915, "gini_cps": 0.2145, "bpic": 5.832272, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9716031097692768, "vi": 1.4183048212728662, "tr": 1.2794592398958577, "fi": 0.9656095694843575, "ru": 2.137789746832683, "el": 2.1724990667596082, "hi": 2.1687945304729044, "ta": 2.854859433401338, "th": 2.8091684927952487, "zh": 3.6701395205324183, "my": 2.5172284804693836}, "t": 69.3} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.60129, "de": 2.91534, "vi": 2.28955, "tr": 2.91044, "fi": 3.18231, "ru": 1.85755, "el": 2.02296, "hi": 1.5645, "ta": 1.28642, "th": 1.50977, "zh": 3.65665, "my": 1.44794}, "bytes_per_chunk": {"en": 5.724, "de": 6.223, "vi": 5.557, "tr": 5.174, "fi": 6.285, "ru": 5.156, "el": 5.176, "hi": 6.775, "ta": 5.789, "th": 6.13, "zh": 5.301, "my": 6.599}, "bpb_mean": 2.27039, "t": 70.6} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.7374, "bpb": 2.50654, "chunk_loss": 0.030225, "F": 0.1781, "G": 0.28661, "chunks": 182.38, "bytes_per_chunk": 5.615, "boundary_entropy": 0.36699, "router_noise": 0.0, "grad_norm": 0.6856, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 70.6} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.51598, "bpb": 2.1871, "chunk_loss": 0.030067, "F": 0.17459, "G": 0.27648, "chunks": 178.78, "bytes_per_chunk": 5.728, "boundary_entropy": 0.3396, "router_noise": 0.0, "grad_norm": 0.7322, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 72.6} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.77031, "bpb": 2.55402, "chunk_loss": 0.030121, "F": 0.17261, "G": 0.2831, "chunks": 176.75, "bytes_per_chunk": 5.793, "boundary_entropy": 0.35867, "router_noise": 0.0, "grad_norm": 0.7602, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 74.5} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.66811, "bpb": 2.40658, "chunk_loss": 0.030251, "F": 0.17612, "G": 0.26526, "chunks": 180.34, "bytes_per_chunk": 5.678, "boundary_entropy": 0.33279, "router_noise": 0.0, "grad_norm": 0.7345, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 76.5} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2895.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166351, "cps": 22.617188, "mean_p": 0.252275, "boundary_entropy": 0.284002, "bpb": 2.500307}, "de": {"chunks": 3310.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161204, "cps": 25.859375, "mean_p": 0.275231, "boundary_entropy": 0.351194, "bpb": 2.802935}, "vi": {"chunks": 4425.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.240033, "cps": 34.570312, "mean_p": 0.257293, "boundary_entropy": 0.283297, "bpb": 2.220374}, "tr": {"chunks": 3917.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216373, "cps": 30.601562, "mean_p": 0.26277, "boundary_entropy": 0.306538, "bpb": 2.796633}, "fi": {"chunks": 3278.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.17426, "cps": 25.609375, "mean_p": 0.306614, "boundary_entropy": 0.394116, "bpb": 3.065875}, "ru": {"chunks": 6520.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.344063, "cps": 50.9375, "mean_p": 0.240977, "boundary_entropy": 0.277528, "bpb": 1.806957}, "el": {"chunks": 6632.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.32315, "cps": 51.8125, "mean_p": 0.236367, "boundary_entropy": 0.302797, "bpb": 1.913993}, "hi": {"chunks": 7612.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.452422, "cps": 59.46875, "mean_p": 0.39521, "boundary_entropy": 0.546286, "bpb": 1.647152}, "ta": {"chunks": 9298.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.497352, "cps": 72.640625, "mean_p": 0.257395, "boundary_entropy": 0.364035, "bpb": 1.178719}, "th": {"chunks": 7144.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.444141, "cps": 55.8125, "mean_p": 0.375606, "boundary_entropy": 0.578505, "bpb": 1.426822}, "zh": {"chunks": 2916.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.493735, "cps": 22.78125, "mean_p": 0.336671, "boundary_entropy": 0.544219, "bpb": 3.574339}, "my": {"chunks": 9245.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.487914, "cps": 72.226562, "mean_p": 0.396572, "boundary_entropy": 0.58086, "bpb": 1.368769}}, "bpb": {"en": 2.50031, "de": 2.80293, "vi": 2.22037, "tr": 2.79663, "fi": 3.06588, "ru": 1.80696, "el": 1.91399, "hi": 1.64715, "ta": 1.17872, "th": 1.42682, "zh": 3.57434, "my": 1.36877}, "gini_cpc": 0.221885, "gini_cps": 0.232099, "bpic": 5.711707, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9690610514755851, "vi": 1.4429314030109992, "tr": 1.3007042888434317, "fi": 1.0475448303178359, "ru": 2.0683019125869824, "el": 1.9425813929474802, "hi": 2.7196890883754588, "ta": 2.9897826993723906, "th": 2.6699057629400804, "zh": 2.968039176809743, "my": 2.9330474677902214}, "t": 79.4} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.51173, "de": 2.80266, "vi": 2.22509, "tr": 2.80516, "fi": 3.06915, "ru": 1.78967, "el": 1.9005, "hi": 1.64274, "ta": 1.18835, "th": 1.44916, "zh": 3.58455, "my": 1.37753}, "bytes_per_chunk": {"en": 5.729, "de": 6.263, "vi": 5.505, "tr": 5.033, "fi": 6.023, "ru": 5.342, "el": 5.784, "hi": 5.498, "ta": 5.471, "th": 6.28, "zh": 6.016, "my": 5.901}, "bpb_mean": 2.19552, "t": 80.6} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.62063, "bpb": 2.33807, "chunk_loss": 0.030308, "F": 0.18033, "G": 0.27135, "chunks": 184.66, "bytes_per_chunk": 5.545, "boundary_entropy": 0.3294, "router_noise": 0.0, "grad_norm": 0.7498, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 80.7} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.63914, "bpb": 2.36478, "chunk_loss": 0.030307, "F": 0.17722, "G": 0.27433, "chunks": 181.47, "bytes_per_chunk": 5.643, "boundary_entropy": 0.33287, "router_noise": 0.0, "grad_norm": 0.7605, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 82.6} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.57151, "bpb": 2.26721, "chunk_loss": 0.030036, "F": 0.17358, "G": 0.27263, "chunks": 177.75, "bytes_per_chunk": 5.761, "boundary_entropy": 0.33729, "router_noise": 0.0, "grad_norm": 0.7441, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 84.6} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.71433, "bpb": 2.47326, "chunk_loss": 0.030048, "F": 0.1694, "G": 0.26954, "chunks": 173.47, "bytes_per_chunk": 5.903, "boundary_entropy": 0.33253, "router_noise": 0.0, "grad_norm": 0.8034, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 86.5} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2890.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166063, "cps": 22.578125, "mean_p": 0.238185, "boundary_entropy": 0.256071, "bpb": 2.438134}, "de": {"chunks": 3299.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160668, "cps": 25.773438, "mean_p": 0.26785, "boundary_entropy": 0.336365, "bpb": 2.737711}, "vi": {"chunks": 4346.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.235747, "cps": 33.953125, "mean_p": 0.25059, "boundary_entropy": 0.279455, "bpb": 2.184822}, "tr": {"chunks": 3973.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.219466, "cps": 31.039062, "mean_p": 0.254024, "boundary_entropy": 0.285709, "bpb": 2.713117}, "fi": {"chunks": 3030.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161076, "cps": 23.671875, "mean_p": 0.261694, "boundary_entropy": 0.338129, "bpb": 2.986336}, "ru": {"chunks": 6547.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.345488, "cps": 51.148438, "mean_p": 0.244266, "boundary_entropy": 0.251133, "bpb": 1.724482}, "el": {"chunks": 6838.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.333187, "cps": 53.421875, "mean_p": 0.241279, "boundary_entropy": 0.311905, "bpb": 1.85853}, "hi": {"chunks": 7019.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.417177, "cps": 54.835938, "mean_p": 0.381573, "boundary_entropy": 0.532865, "bpb": 1.447522}, "ta": {"chunks": 8318.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.444932, "cps": 64.984375, "mean_p": 0.25555, "boundary_entropy": 0.372638, "bpb": 1.143627}, "th": {"chunks": 8098.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.50345, "cps": 63.265625, "mean_p": 0.391048, "boundary_entropy": 0.588382, "bpb": 1.360183}, "zh": {"chunks": 2966.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.502201, "cps": 23.171875, "mean_p": 0.320724, "boundary_entropy": 0.523287, "bpb": 3.560106}, "my": {"chunks": 8738.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.461157, "cps": 68.265625, "mean_p": 0.38146, "boundary_entropy": 0.582471, "bpb": 1.418417}}, "bpb": {"en": 2.43813, "de": 2.73771, "vi": 2.18482, "tr": 2.71312, "fi": 2.98634, "ru": 1.72448, "el": 1.85853, "hi": 1.44752, "ta": 1.14363, "th": 1.36018, "zh": 3.56011, "my": 1.41842}, "gini_cpc": 0.223389, "gini_cps": 0.223739, "bpic": 5.809406, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9675116114038386, "vi": 1.4196224460204796, "tr": 1.3215825360724207, "fi": 0.9699671417316564, "ru": 2.08046016196328, "el": 2.00638618466209, "hi": 2.512155087224996, "ta": 2.6792900050066124, "th": 3.0316773886134754, "zh": 3.0241545457847687, "my": 2.7769940021607122}, "t": 89.4} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.45005, "de": 2.74078, "vi": 2.19106, "tr": 2.71995, "fi": 2.99162, "ru": 1.70275, "el": 1.84657, "hi": 1.4441, "ta": 1.15255, "th": 1.38142, "zh": 3.56577, "my": 1.42841}, "bytes_per_chunk": {"en": 5.736, "de": 6.263, "vi": 5.552, "tr": 4.897, "fi": 6.296, "ru": 5.317, "el": 5.628, "hi": 5.955, "ta": 6.201, "th": 5.563, "zh": 6.402, "my": 7.286}, "bpb_mean": 2.13459, "t": 90.6} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.56349, "bpb": 2.25564, "chunk_loss": 0.030244, "F": 0.1763, "G": 0.26159, "chunks": 180.53, "bytes_per_chunk": 5.672, "boundary_entropy": 0.30112, "router_noise": 0.0, "grad_norm": 0.7646, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 90.6} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.72723, "bpb": 2.49187, "chunk_loss": 0.030147, "F": 0.17267, "G": 0.25635, "chunks": 176.81, "bytes_per_chunk": 5.791, "boundary_entropy": 0.3006, "router_noise": 0.0, "grad_norm": 0.8144, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 92.6} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.53868, "bpb": 2.21985, "chunk_loss": 0.030173, "F": 0.17581, "G": 0.28177, "chunks": 180.03, "bytes_per_chunk": 5.688, "boundary_entropy": 0.35915, "router_noise": 0.0, "grad_norm": 0.7143, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 94.6} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.60674, "bpb": 2.31804, "chunk_loss": 0.029939, "F": 0.16794, "G": 0.25971, "chunks": 171.97, "bytes_per_chunk": 5.955, "boundary_entropy": 0.32713, "router_noise": 0.0, "grad_norm": 0.7267, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 96.5} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2894.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166293, "cps": 22.609375, "mean_p": 0.229878, "boundary_entropy": 0.235655, "bpb": 2.378939}, "de": {"chunks": 3299.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160668, "cps": 25.773438, "mean_p": 0.255734, "boundary_entropy": 0.314626, "bpb": 2.666278}, "vi": {"chunks": 4296.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.233035, "cps": 33.5625, "mean_p": 0.238972, "boundary_entropy": 0.242859, "bpb": 2.13937}, "tr": {"chunks": 3914.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216207, "cps": 30.578125, "mean_p": 0.24219, "boundary_entropy": 0.257971, "bpb": 2.638326}, "fi": {"chunks": 3045.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161873, "cps": 23.789062, "mean_p": 0.253183, "boundary_entropy": 0.318782, "bpb": 2.906831}, "ru": {"chunks": 6493.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.342639, "cps": 50.726562, "mean_p": 0.243267, "boundary_entropy": 0.245427, "bpb": 1.658714}, "el": {"chunks": 6617.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.322419, "cps": 51.695312, "mean_p": 0.233742, "boundary_entropy": 0.308332, "bpb": 1.800185}, "hi": {"chunks": 6799.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.404101, "cps": 53.117188, "mean_p": 0.367064, "boundary_entropy": 0.524846, "bpb": 1.414479}, "ta": {"chunks": 10000.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.534902, "cps": 78.125, "mean_p": 0.267378, "boundary_entropy": 0.367025, "bpb": 1.14737}, "th": {"chunks": 5211.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.323966, "cps": 40.710938, "mean_p": 0.344537, "boundary_entropy": 0.567248, "bpb": 1.319864}, "zh": {"chunks": 2921.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.494582, "cps": 22.820312, "mean_p": 0.312974, "boundary_entropy": 0.508089, "bpb": 3.505199}, "my": {"chunks": 8524.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.449863, "cps": 66.59375, "mean_p": 0.37135, "boundary_entropy": 0.573211, "bpb": 1.27938}}, "bpb": {"en": 2.37894, "de": 2.66628, "vi": 2.13937, "tr": 2.63833, "fi": 2.90683, "ru": 1.65871, "el": 1.80019, "hi": 1.41448, "ta": 1.14737, "th": 1.31986, "zh": 3.5052, "my": 1.27938}, "gini_cpc": 0.22669, "gini_cps": 0.234475, "bpic": 5.99536, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.966174345873218, "vi": 1.4013503429839689, "tr": 1.300157193874978, "fi": 0.9734216628458175, "ru": 2.0604485852815304, "el": 1.9388574640463556, "hi": 2.4300519700030083, "ta": 3.2166227106538545, "th": 1.9481643927313408, "zh": 2.974155749450443, "my": 2.705239115848868}, "t": 99.4} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.38957, "de": 2.67096, "vi": 2.14348, "tr": 2.64392, "fi": 2.91598, "ru": 1.63854, "el": 1.78789, "hi": 1.41172, "ta": 1.15542, "th": 1.33876, "zh": 3.51206, "my": 1.28696}, "bytes_per_chunk": {"en": 5.733, "de": 6.272, "vi": 5.606, "tr": 5.047, "fi": 6.3, "ru": 5.357, "el": 5.818, "hi": 6.252, "ta": 5.108, "th": 8.304, "zh": 6.533, "my": 6.33}, "bpb_mean": 2.0746, "t": 100.6} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.5759, "bpb": 2.27354, "chunk_loss": 0.030205, "F": 0.17465, "G": 0.25853, "chunks": 178.84, "bytes_per_chunk": 5.726, "boundary_entropy": 0.32813, "router_noise": 0.0, "grad_norm": 0.7967, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 100.6} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.50234, "bpb": 2.16742, "chunk_loss": 0.030594, "F": 0.18515, "G": 0.2799, "chunks": 189.59, "bytes_per_chunk": 5.401, "boundary_entropy": 0.33812, "router_noise": 0.0, "grad_norm": 0.7174, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 102.6} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.42955, "bpb": 2.06241, "chunk_loss": 0.03002, "F": 0.17026, "G": 0.25769, "chunks": 174.34, "bytes_per_chunk": 5.873, "boundary_entropy": 0.3194, "router_noise": 0.0, "grad_norm": 0.6709, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 104.5} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.62079, "bpb": 2.3383, "chunk_loss": 0.030264, "F": 0.17719, "G": 0.26373, "chunks": 181.44, "bytes_per_chunk": 5.644, "boundary_entropy": 0.31233, "router_noise": 0.0, "grad_norm": 0.8243, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 106.5} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.228722, "boundary_entropy": 0.232444, "bpb": 2.340004}, "de": {"chunks": 3297.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160571, "cps": 25.757812, "mean_p": 0.25518, "boundary_entropy": 0.312279, "bpb": 2.605572}, "vi": {"chunks": 4359.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.236452, "cps": 34.054688, "mean_p": 0.247591, "boundary_entropy": 0.253685, "bpb": 2.124105}, "tr": {"chunks": 3778.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.208695, "cps": 29.515625, "mean_p": 0.232362, "boundary_entropy": 0.255945, "bpb": 2.627624}, "fi": {"chunks": 3002.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159587, "cps": 23.453125, "mean_p": 0.246339, "boundary_entropy": 0.308869, "bpb": 2.844734}, "ru": {"chunks": 6402.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.337836, "cps": 50.015625, "mean_p": 0.237169, "boundary_entropy": 0.237915, "bpb": 1.604147}, "el": {"chunks": 6583.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.320762, "cps": 51.429688, "mean_p": 0.242102, "boundary_entropy": 0.304014, "bpb": 1.748685}, "hi": {"chunks": 6275.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.372957, "cps": 49.023438, "mean_p": 0.362829, "boundary_entropy": 0.520326, "bpb": 1.392183}, "ta": {"chunks": 7898.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.422466, "cps": 61.703125, "mean_p": 0.268164, "boundary_entropy": 0.410453, "bpb": 1.092737}, "th": {"chunks": 6818.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.423873, "cps": 53.265625, "mean_p": 0.366957, "boundary_entropy": 0.581638, "bpb": 1.313362}, "zh": {"chunks": 2973.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503386, "cps": 23.226562, "mean_p": 0.317251, "boundary_entropy": 0.502697, "bpb": 3.486846}, "my": {"chunks": 8068.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.425797, "cps": 63.03125, "mean_p": 0.369756, "boundary_entropy": 0.571747, "bpb": 1.27052}}, "bpb": {"en": 2.34, "de": 2.60557, "vi": 2.1241, "tr": 2.62762, "fi": 2.84473, "ru": 1.60415, "el": 1.74869, "hi": 1.39218, "ta": 1.09274, "th": 1.31336, "zh": 3.48685, "my": 1.27052}, "gini_cpc": 0.211083, "gini_cps": 0.206153, "bpic": 6.156059, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9672597549450447, "vi": 1.4243617593852949, "tr": 1.257152549656864, "fi": 0.961336391352732, "ru": 2.0350872520734185, "el": 1.9322334149933547, "hi": 2.246649045901938, "ta": 2.5448854472947575, "th": 2.553362700942004, "zh": 3.032341048176362, "my": 2.5649511117599877}, "t": 109.4} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.3522, "de": 2.60962, "vi": 2.12969, "tr": 2.6354, "fi": 2.84767, "ru": 1.58131, "el": 1.73821, "hi": 1.39031, "ta": 1.09878, "th": 1.33203, "zh": 3.49045, "my": 1.27656}, "bytes_per_chunk": {"en": 5.736, "de": 6.262, "vi": 5.538, "tr": 5.227, "fi": 6.363, "ru": 5.437, "el": 5.839, "hi": 6.554, "ta": 6.437, "th": 6.713, "zh": 5.959, "my": 6.49}, "bpb_mean": 2.04019, "t": 110.6} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.5852, "bpb": 2.28696, "chunk_loss": 0.030066, "F": 0.16998, "G": 0.23724, "chunks": 174.06, "bytes_per_chunk": 5.883, "boundary_entropy": 0.26152, "router_noise": 0.0, "grad_norm": 1.0371, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 110.6} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.55101, "bpb": 2.23764, "chunk_loss": 0.030035, "F": 0.17117, "G": 0.25505, "chunks": 175.28, "bytes_per_chunk": 5.842, "boundary_entropy": 0.28576, "router_noise": 0.0, "grad_norm": 0.7069, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 112.6} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.59394, "bpb": 2.29956, "chunk_loss": 0.029981, "F": 0.16922, "G": 0.25314, "chunks": 173.28, "bytes_per_chunk": 5.909, "boundary_entropy": 0.3114, "router_noise": 0.0, "grad_norm": 0.6726, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 114.5} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.60035, "bpb": 2.30882, "chunk_loss": 0.030179, "F": 0.17609, "G": 0.26045, "chunks": 180.31, "bytes_per_chunk": 5.679, "boundary_entropy": 0.29936, "router_noise": 0.0, "grad_norm": 0.6857, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 116.5} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2892.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166178, "cps": 22.59375, "mean_p": 0.223216, "boundary_entropy": 0.217673, "bpb": 2.277787}, "de": {"chunks": 3303.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160863, "cps": 25.804688, "mean_p": 0.252772, "boundary_entropy": 0.306686, "bpb": 2.561253}, "vi": {"chunks": 4321.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234391, "cps": 33.757812, "mean_p": 0.237627, "boundary_entropy": 0.231592, "bpb": 2.087209}, "tr": {"chunks": 3774.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.208474, "cps": 29.484375, "mean_p": 0.225441, "boundary_entropy": 0.243747, "bpb": 2.558354}, "fi": {"chunks": 3038.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161501, "cps": 23.734375, "mean_p": 0.248101, "boundary_entropy": 0.306012, "bpb": 2.780492}, "ru": {"chunks": 6495.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.342744, "cps": 50.742188, "mean_p": 0.237463, "boundary_entropy": 0.234407, "bpb": 1.568598}, "el": {"chunks": 6606.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.321883, "cps": 51.609375, "mean_p": 0.237744, "boundary_entropy": 0.292747, "bpb": 1.717283}, "hi": {"chunks": 7000.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.416048, "cps": 54.6875, "mean_p": 0.371284, "boundary_entropy": 0.520438, "bpb": 1.364198}, "ta": {"chunks": 8719.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.466381, "cps": 68.117188, "mean_p": 0.316718, "boundary_entropy": 0.459902, "bpb": 1.07425}, "th": {"chunks": 7282.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.45272, "cps": 56.890625, "mean_p": 0.379762, "boundary_entropy": 0.587694, "bpb": 1.259582}, "zh": {"chunks": 2858.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.483915, "cps": 22.328125, "mean_p": 0.302689, "boundary_entropy": 0.485287, "bpb": 3.458801}, "my": {"chunks": 8159.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.4306, "cps": 63.742188, "mean_p": 0.386853, "boundary_entropy": 0.580967, "bpb": 1.239214}}, "bpb": {"en": 2.27779, "de": 2.56125, "vi": 2.08721, "tr": 2.55835, "fi": 2.78049, "ru": 1.5686, "el": 1.71728, "hi": 1.3642, "ta": 1.07425, "th": 1.25958, "zh": 3.4588, "my": 1.23921}, "gini_cpc": 0.216253, "gini_cps": 0.220772, "bpic": 5.954986, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9680148017976526, "vi": 1.4104800763476473, "tr": 1.2545188058282448, "fi": 0.9718555466905896, "ru": 2.062508621727849, "el": 1.9369729498914587, "hi": 2.5036222278396725, "ta": 2.8065128803827326, "th": 2.7243032025146494, "zh": 2.912021743132028, "my": 2.591190773700812}, "t": 119.3} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.28966, "de": 2.56837, "vi": 2.08997, "tr": 2.56655, "fi": 2.78706, "ru": 1.54651, "el": 1.70582, "hi": 1.36173, "ta": 1.08272, "th": 1.27672, "zh": 3.4621, "my": 1.24672}, "bytes_per_chunk": {"en": 5.734, "de": 6.266, "vi": 5.601, "tr": 5.226, "fi": 6.275, "ru": 5.372, "el": 5.796, "hi": 6.088, "ta": 5.83, "th": 6.267, "zh": 6.108, "my": 6.669}, "bpb_mean": 1.99866, "t": 120.6} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.51595, "bpb": 2.18706, "chunk_loss": 0.029998, "F": 0.17068, "G": 0.26616, "chunks": 174.78, "bytes_per_chunk": 5.859, "boundary_entropy": 0.3182, "router_noise": 0.0, "grad_norm": 0.6824, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 120.6} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.39505, "bpb": 2.01264, "chunk_loss": 0.03024, "F": 0.17969, "G": 0.25164, "chunks": 184.0, "bytes_per_chunk": 5.565, "boundary_entropy": 0.27686, "router_noise": 0.0, "grad_norm": 0.7128, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 122.6} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.56502, "bpb": 2.25784, "chunk_loss": 0.030201, "F": 0.17233, "G": 0.25307, "chunks": 176.47, "bytes_per_chunk": 5.803, "boundary_entropy": 0.29404, "router_noise": 0.0, "grad_norm": 0.7085, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 124.5} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.48136, "bpb": 2.13715, "chunk_loss": 0.030194, "F": 0.17728, "G": 0.25274, "chunks": 181.53, "bytes_per_chunk": 5.641, "boundary_entropy": 0.28092, "router_noise": 0.0, "grad_norm": 0.7019, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 126.5} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2893.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166236, "cps": 22.601562, "mean_p": 0.222983, "boundary_entropy": 0.215947, "bpb": 2.248048}, "de": {"chunks": 3307.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161058, "cps": 25.835938, "mean_p": 0.256854, "boundary_entropy": 0.315515, "bpb": 2.50879}, "vi": {"chunks": 4419.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.239707, "cps": 34.523438, "mean_p": 0.241198, "boundary_entropy": 0.233284, "bpb": 2.051314}, "tr": {"chunks": 3735.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.206319, "cps": 29.179688, "mean_p": 0.230566, "boundary_entropy": 0.252927, "bpb": 2.514941}, "fi": {"chunks": 3059.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162618, "cps": 23.898438, "mean_p": 0.252265, "boundary_entropy": 0.316245, "bpb": 2.735276}, "ru": {"chunks": 6425.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.33905, "cps": 50.195312, "mean_p": 0.237822, "boundary_entropy": 0.232007, "bpb": 1.533159}, "el": {"chunks": 6825.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332554, "cps": 53.320312, "mean_p": 0.249803, "boundary_entropy": 0.304253, "bpb": 1.672468}, "hi": {"chunks": 7058.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.419495, "cps": 55.140625, "mean_p": 0.358231, "boundary_entropy": 0.504131, "bpb": 1.331459}, "ta": {"chunks": 9640.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.515646, "cps": 75.3125, "mean_p": 0.335209, "boundary_entropy": 0.46228, "bpb": 1.06645}, "th": {"chunks": 8460.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.525956, "cps": 66.09375, "mean_p": 0.392317, "boundary_entropy": 0.587189, "bpb": 1.233922}, "zh": {"chunks": 3049.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.516255, "cps": 23.820312, "mean_p": 0.295127, "boundary_entropy": 0.458483, "bpb": 3.411967}, "my": {"chunks": 7749.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.408961, "cps": 60.539062, "mean_p": 0.332501, "boundary_entropy": 0.537017, "bpb": 1.211823}}, "bpb": {"en": 2.24805, "de": 2.50879, "vi": 2.05131, "tr": 2.51494, "fi": 2.73528, "ru": 1.53316, "el": 1.67247, "hi": 1.33146, "ta": 1.06645, "th": 1.23392, "zh": 3.41197, "my": 1.21182}, "gini_cpc": 0.234164, "gini_cps": 0.232387, "bpic": 5.760834, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9688520762670342, "vi": 1.4419710662109966, "tr": 1.241125621775721, "fi": 0.9782351865776073, "ru": 2.039574644282852, "el": 2.0004951237210546, "hi": 2.52349394886155, "ta": 3.101896130019182, "th": 3.16391628856422, "zh": 3.105558169122451, "my": 2.460129514867076}, "t": 129.3} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.25913, "de": 2.5162, "vi": 2.05486, "tr": 2.52261, "fi": 2.7417, "ru": 1.51285, "el": 1.66514, "hi": 1.32718, "ta": 1.07328, "th": 1.25309, "zh": 3.41821, "my": 1.21791}, "bytes_per_chunk": {"en": 5.737, "de": 6.269, "vi": 5.478, "tr": 5.224, "fi": 6.268, "ru": 5.405, "el": 5.649, "hi": 5.847, "ta": 5.215, "th": 5.422, "zh": 5.836, "my": 6.653}, "bpb_mean": 1.96351, "t": 130.6} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.48476, "bpb": 2.14206, "chunk_loss": 0.030143, "F": 0.17297, "G": 0.25805, "chunks": 177.12, "bytes_per_chunk": 5.781, "boundary_entropy": 0.31156, "router_noise": 0.0, "grad_norm": 0.6632, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 130.6} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.71696, "bpb": 2.47705, "chunk_loss": 0.030263, "F": 0.18024, "G": 0.24393, "chunks": 184.56, "bytes_per_chunk": 5.548, "boundary_entropy": 0.26045, "router_noise": 0.0, "grad_norm": 0.7395, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.038, "alpha": 0.03, "t": 132.5} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.49145, "bpb": 2.15171, "chunk_loss": 0.030093, "F": 0.17072, "G": 0.25217, "chunks": 174.81, "bytes_per_chunk": 5.858, "boundary_entropy": 0.29301, "router_noise": 0.0, "grad_norm": 0.668, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 134.5} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.49409, "bpb": 2.15552, "chunk_loss": 0.030157, "F": 0.17499, "G": 0.25636, "chunks": 179.19, "bytes_per_chunk": 5.715, "boundary_entropy": 0.29541, "router_noise": 0.0, "grad_norm": 0.7065, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 136.4} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2890.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166063, "cps": 22.578125, "mean_p": 0.2183, "boundary_entropy": 0.201127, "bpb": 2.220133}, "de": {"chunks": 3309.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161155, "cps": 25.851562, "mean_p": 0.248736, "boundary_entropy": 0.293654, "bpb": 2.480704}, "vi": {"chunks": 4338.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.235313, "cps": 33.890625, "mean_p": 0.23378, "boundary_entropy": 0.216755, "bpb": 2.041782}, "tr": {"chunks": 3625.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200243, "cps": 28.320312, "mean_p": 0.219697, "boundary_entropy": 0.23087, "bpb": 2.477259}, "fi": {"chunks": 3017.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160385, "cps": 23.570312, "mean_p": 0.240444, "boundary_entropy": 0.294091, "bpb": 2.706622}, "ru": {"chunks": 6434.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.339525, "cps": 50.265625, "mean_p": 0.234826, "boundary_entropy": 0.21381, "bpb": 1.511648}, "el": {"chunks": 6694.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326171, "cps": 52.296875, "mean_p": 0.237969, "boundary_entropy": 0.280057, "bpb": 1.658995}, "hi": {"chunks": 6811.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.404814, "cps": 53.210938, "mean_p": 0.352474, "boundary_entropy": 0.501753, "bpb": 1.318303}, "ta": {"chunks": 8129.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.434822, "cps": 63.507812, "mean_p": 0.304155, "boundary_entropy": 0.451645, "bpb": 1.044}, "th": {"chunks": 8731.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.542804, "cps": 68.210938, "mean_p": 0.403885, "boundary_entropy": 0.600009, "bpb": 1.220699}, "zh": {"chunks": 3025.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.512191, "cps": 23.632812, "mean_p": 0.293379, "boundary_entropy": 0.453938, "bpb": 3.403304}, "my": {"chunks": 8023.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.423422, "cps": 62.679688, "mean_p": 0.353634, "boundary_entropy": 0.551084, "bpb": 1.223929}}, "bpb": {"en": 2.22013, "de": 2.4807, "vi": 2.04178, "tr": 2.47726, "fi": 2.70662, "ru": 1.51165, "el": 1.659, "hi": 1.3183, "ta": 1.044, "th": 1.2207, "zh": 3.4033, "my": 1.22393}, "gini_cpc": 0.232174, "gini_cps": 0.223062, "bpic": 5.901962, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9704443534814495, "vi": 1.4170092431746069, "tr": 1.205823481817902, "fi": 0.9658055665361079, "ru": 2.0445518072509152, "el": 1.9641341211067607, "hi": 2.437710257741765, "ta": 2.6184116915963873, "th": 3.2686558755228767, "zh": 3.0843113624406353, "my": 2.5497622887772255}, "t": 139.2} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.22985, "de": 2.48802, "vi": 2.04395, "tr": 2.48255, "fi": 2.71029, "ru": 1.4889, "el": 1.64888, "hi": 1.315, "ta": 1.05111, "th": 1.238, "zh": 3.40707, "my": 1.22993}, "bytes_per_chunk": {"en": 5.737, "de": 6.262, "vi": 5.551, "tr": 5.334, "fi": 6.335, "ru": 5.408, "el": 5.729, "hi": 6.066, "ta": 6.218, "th": 5.379, "zh": 5.933, "my": 6.462}, "bpb_mean": 1.94446, "t": 140.5} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 140.5} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "parity_a", "lm_loss": 1.50529, "bpb": 2.17168, "chunk_loss": 0.035386, "F": 0.17905, "G": 0.26075, "chunks": 183.34, "bytes_per_chunk": 5.585, "boundary_entropy": 0.30281, "router_noise": 0.0, "grad_norm": 0.7361, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 10.0837, "alpha": 0.03, "t": 140.5} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "parity_a", "lm_loss": 1.31279, "bpb": 1.89396, "chunk_loss": 0.032004, "F": 0.15973, "G": 0.21285, "chunks": 163.56, "bytes_per_chunk": 6.261, "boundary_entropy": 0.23745, "router_noise": 0.0, "grad_norm": 0.6257, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.1484, "alpha": 0.03, "t": 142.5} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "parity_a", "lm_loss": 1.41165, "bpb": 2.03659, "chunk_loss": 0.031172, "F": 0.15958, "G": 0.20688, "chunks": 163.41, "bytes_per_chunk": 6.267, "boundary_entropy": 0.22629, "router_noise": 0.0, "grad_norm": 0.6959, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.8824, "alpha": 0.03, "t": 144.4} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "parity_a", "lm_loss": 1.58224, "bpb": 2.28269, "chunk_loss": 0.031071, "F": 0.17175, "G": 0.21007, "chunks": 175.88, "bytes_per_chunk": 5.822, "boundary_entropy": 0.19848, "router_noise": 0.0, "grad_norm": 0.6767, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 7.9249, "alpha": 0.03, "t": 146.4} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2886.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165833, "cps": 22.546875, "mean_p": 0.197434, "boundary_entropy": 0.14518, "bpb": 2.193387}, "de": {"chunks": 3274.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159451, "cps": 25.578125, "mean_p": 0.205954, "boundary_entropy": 0.20142, "bpb": 2.458297}, "vi": {"chunks": 4252.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230648, "cps": 33.21875, "mean_p": 0.2013, "boundary_entropy": 0.153238, "bpb": 2.022737}, "tr": {"chunks": 3350.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185052, "cps": 26.171875, "mean_p": 0.197399, "boundary_entropy": 0.177151, "bpb": 2.471049}, "fi": {"chunks": 3024.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160757, "cps": 23.625, "mean_p": 0.210092, "boundary_entropy": 0.224057, "bpb": 2.65907}, "ru": {"chunks": 6317.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.333351, "cps": 49.351562, "mean_p": 0.196805, "boundary_entropy": 0.157834, "bpb": 1.488541}, "el": {"chunks": 6275.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.305755, "cps": 49.023438, "mean_p": 0.195279, "boundary_entropy": 0.207182, "bpb": 1.625981}, "hi": {"chunks": 4655.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.276672, "cps": 36.367188, "mean_p": 0.140728, "boundary_entropy": 0.197097, "bpb": 1.319277}, "ta": {"chunks": 3485.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.186413, "cps": 27.226562, "mean_p": 0.112104, "boundary_entropy": 0.208314, "bpb": 1.097738}, "th": {"chunks": 2936.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.18253, "cps": 22.9375, "mean_p": 0.159349, "boundary_entropy": 0.342861, "bpb": 1.240463}, "zh": {"chunks": 2773.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.469523, "cps": 21.664062, "mean_p": 0.25545, "boundary_entropy": 0.398941, "bpb": 3.374023}, "my": {"chunks": 3670.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.193688, "cps": 28.671875, "mean_p": 0.124114, "boundary_entropy": 0.248858, "bpb": 1.240065}}, "bpb": {"en": 2.19339, "de": 2.4583, "vi": 2.02274, "tr": 2.47105, "fi": 2.65907, "ru": 1.48854, "el": 1.62598, "hi": 1.31928, "ta": 1.09774, "th": 1.24046, "zh": 3.37402, "my": 1.24006}, "gini_cpc": 0.191787, "gini_cps": 0.158114, "bpic": 8.183487, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9615105666827286, "vi": 1.3908423539017518, "tr": 1.1158917044681835, "fi": 0.9693881283908395, "ru": 2.0101545812099904, "el": 1.843744247998012, "hi": 1.6683701303314975, "ta": 1.1241004104069094, "th": 1.1006843193352362, "zh": 2.831289122095082, "my": 1.1679667561554832}, "t": 149.2} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.20576, "de": 2.46603, "vi": 2.02649, "tr": 2.47731, "fi": 2.66416, "ru": 1.46553, "el": 1.61709, "hi": 1.3164, "ta": 1.10396, "th": 1.26401, "zh": 3.37838, "my": 1.2472}, "bytes_per_chunk": {"en": 5.741, "de": 6.302, "vi": 5.648, "tr": 6.061, "fi": 6.31, "ru": 5.544, "el": 6.07, "hi": 8.593, "ta": 13.711, "th": 15.067, "zh": 6.487, "my": 13.364}, "bpb_mean": 1.93603, "t": 150.5} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "parity_a", "lm_loss": 1.67974, "bpb": 2.42335, "chunk_loss": 0.030734, "F": 0.16083, "G": 0.20786, "chunks": 164.69, "bytes_per_chunk": 6.218, "boundary_entropy": 0.2293, "router_noise": 0.0, "grad_norm": 0.7102, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 7.6058, "alpha": 0.03, "t": 150.5} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "parity_a", "lm_loss": 1.43461, "bpb": 2.0697, "chunk_loss": 0.030986, "F": 0.15985, "G": 0.19256, "chunks": 163.69, "bytes_per_chunk": 6.256, "boundary_entropy": 0.17784, "router_noise": 0.0, "grad_norm": 0.6715, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.0748, "alpha": 0.03, "t": 152.4} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "parity_a", "lm_loss": 1.37351, "bpb": 1.98155, "chunk_loss": 0.031398, "F": 0.15933, "G": 0.1884, "chunks": 163.16, "bytes_per_chunk": 6.276, "boundary_entropy": 0.17291, "router_noise": 0.0, "grad_norm": 0.6903, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.9458, "alpha": 0.03, "t": 154.4} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "parity_a", "lm_loss": 1.34001, "bpb": 1.93323, "chunk_loss": 0.031014, "F": 0.15863, "G": 0.18834, "chunks": 162.44, "bytes_per_chunk": 6.304, "boundary_entropy": 0.16697, "router_noise": 0.0, "grad_norm": 0.7328, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.1192, "alpha": 0.03, "t": 156.3} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.191864, "boundary_entropy": 0.125757, "bpb": 2.170659}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.1915, "boundary_entropy": 0.160664, "bpb": 2.425538}, "vi": {"chunks": 4284.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.232384, "cps": 33.46875, "mean_p": 0.194463, "boundary_entropy": 0.138928, "bpb": 2.005612}, "tr": {"chunks": 3249.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179473, "cps": 25.382812, "mean_p": 0.194658, "boundary_entropy": 0.165402, "bpb": 2.427586}, "fi": {"chunks": 3000.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159481, "cps": 23.4375, "mean_p": 0.185691, "boundary_entropy": 0.169721, "bpb": 2.632797}, "ru": {"chunks": 6256.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.330132, "cps": 48.875, "mean_p": 0.1887, "boundary_entropy": 0.146601, "bpb": 1.459377}, "el": {"chunks": 6300.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.306973, "cps": 49.21875, "mean_p": 0.1788, "boundary_entropy": 0.163187, "bpb": 1.590393}, "hi": {"chunks": 4384.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.260565, "cps": 34.25, "mean_p": 0.113159, "boundary_entropy": 0.1725, "bpb": 1.313438}, "ta": {"chunks": 3389.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.181278, "cps": 26.476562, "mean_p": 0.098827, "boundary_entropy": 0.189692, "bpb": 1.066739}, "th": {"chunks": 2815.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.175008, "cps": 21.992188, "mean_p": 0.129661, "boundary_entropy": 0.283769, "bpb": 1.221082}, "zh": {"chunks": 2910.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.492719, "cps": 22.734375, "mean_p": 0.255532, "boundary_entropy": 0.383657, "bpb": 3.347347}, "my": {"chunks": 3138.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.165611, "cps": 24.515625, "mean_p": 0.098286, "boundary_entropy": 0.200888, "bpb": 1.217426}}, "bpb": {"en": 2.17066, "de": 2.42554, "vi": 2.00561, "tr": 2.42759, "fi": 2.6328, "ru": 1.45938, "el": 1.59039, "hi": 1.31344, "ta": 1.06674, "th": 1.22108, "zh": 3.34735, "my": 1.21743}, "gini_cpc": 0.202772, "gini_cps": 0.156449, "bpic": 8.362517, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9619789919517142, "vi": 1.3998545026856168, "tr": 1.0811245722168215, "fi": 0.9606959274011311, "ru": 1.9886763275494073, "el": 1.8491676309369791, "hi": 1.5696110625074255, "ta": 1.092000098870845, "th": 1.0542264598345175, "zh": 2.968083568850728, "my": 0.9976222841723899}, "t": 159.2} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.18345, "de": 2.43651, "vi": 2.00694, "tr": 2.43267, "fi": 2.64285, "ru": 1.43837, "el": 1.5795, "hi": 1.31016, "ta": 1.07276, "th": 1.23982, "zh": 3.35058, "my": 1.22559}, "bytes_per_chunk": {"en": 5.736, "de": 6.301, "vi": 5.617, "tr": 6.224, "fi": 6.344, "ru": 5.565, "el": 6.048, "hi": 9.087, "ta": 14.12, "th": 15.686, "zh": 6.204, "my": 15.477}, "bpb_mean": 1.90993, "t": 160.4} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "parity_a", "lm_loss": 1.46663, "bpb": 2.1159, "chunk_loss": 0.031206, "F": 0.16504, "G": 0.19176, "chunks": 169.0, "bytes_per_chunk": 6.059, "boundary_entropy": 0.16368, "router_noise": 0.0, "grad_norm": 0.7153, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 8.8073, "alpha": 0.03, "t": 160.5} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "parity_a", "lm_loss": 1.41307, "bpb": 2.03863, "chunk_loss": 0.030898, "F": 0.15649, "G": 0.18207, "chunks": 160.25, "bytes_per_chunk": 6.39, "boundary_entropy": 0.15745, "router_noise": 0.0, "grad_norm": 0.7419, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.5124, "alpha": 0.03, "t": 162.4} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "parity_a", "lm_loss": 1.47806, "bpb": 2.13239, "chunk_loss": 0.030938, "F": 0.16278, "G": 0.18903, "chunks": 166.69, "bytes_per_chunk": 6.143, "boundary_entropy": 0.16434, "router_noise": 0.0, "grad_norm": 0.6852, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.031, "alpha": 0.03, "t": 164.3} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "parity_a", "lm_loss": 1.37066, "bpb": 1.97744, "chunk_loss": 0.031117, "F": 0.15775, "G": 0.1847, "chunks": 161.53, "bytes_per_chunk": 6.339, "boundary_entropy": 0.16826, "router_noise": 0.0, "grad_norm": 0.648, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.8764, "alpha": 0.03, "t": 166.3} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.188422, "boundary_entropy": 0.113392, "bpb": 2.150368}, "de": {"chunks": 3285.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159986, "cps": 25.664062, "mean_p": 0.187168, "boundary_entropy": 0.145937, "bpb": 2.392711}, "vi": {"chunks": 4250.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23054, "cps": 33.203125, "mean_p": 0.191007, "boundary_entropy": 0.123097, "bpb": 1.991195}, "tr": {"chunks": 3263.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180246, "cps": 25.492188, "mean_p": 0.193019, "boundary_entropy": 0.157862, "bpb": 2.39801}, "fi": {"chunks": 3000.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159481, "cps": 23.4375, "mean_p": 0.182418, "boundary_entropy": 0.157943, "bpb": 2.584868}, "ru": {"chunks": 6274.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.331082, "cps": 49.015625, "mean_p": 0.184779, "boundary_entropy": 0.134746, "bpb": 1.445658}, "el": {"chunks": 6110.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.297715, "cps": 47.734375, "mean_p": 0.173133, "boundary_entropy": 0.156306, "bpb": 1.568282}, "hi": {"chunks": 4116.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.244636, "cps": 32.15625, "mean_p": 0.09839, "boundary_entropy": 0.156191, "bpb": 1.305174}, "ta": {"chunks": 3268.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.174806, "cps": 25.53125, "mean_p": 0.088803, "boundary_entropy": 0.173249, "bpb": 1.058839}, "th": {"chunks": 2668.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.165869, "cps": 20.84375, "mean_p": 0.130821, "boundary_entropy": 0.283623, "bpb": 1.204756}, "zh": {"chunks": 3128.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.529631, "cps": 24.4375, "mean_p": 0.264936, "boundary_entropy": 0.378311, "bpb": 3.334507}, "my": {"chunks": 2757.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.145503, "cps": 21.539062, "mean_p": 0.093275, "boundary_entropy": 0.19937, "bpb": 1.199242}}, "bpb": {"en": 2.15037, "de": 2.39271, "vi": 1.9912, "tr": 2.39801, "fi": 2.58487, "ru": 1.44566, "el": 1.56828, "hi": 1.30517, "ta": 1.05884, "th": 1.20476, "zh": 3.33451, "my": 1.19924}, "gini_cpc": 0.219194, "gini_cps": 0.158168, "bpic": 8.526951, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9637392462828244, "vi": 1.388744546315096, "tr": 1.0857831576311139, "fi": 0.9606959274011311, "ru": 1.994398222353737, "el": 1.7933990833372924, "hi": 1.4736585614234863, "ta": 1.053011603160201, "th": 0.9991744919497308, "zh": 3.1904348465172085, "my": 0.8764960603770805}, "t": 169.1} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.1618, "de": 2.40228, "vi": 1.99424, "tr": 2.40252, "fi": 2.59404, "ru": 1.42362, "el": 1.55877, "hi": 1.30271, "ta": 1.06576, "th": 1.2238, "zh": 3.33793, "my": 1.20763}, "bytes_per_chunk": {"en": 5.739, "de": 6.287, "vi": 5.647, "tr": 6.135, "fi": 6.334, "ru": 5.556, "el": 6.199, "hi": 9.692, "ta": 14.772, "th": 16.417, "zh": 5.726, "my": 18.805}, "bpb_mean": 1.88959, "t": 170.4} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "parity_a", "lm_loss": 1.48545, "bpb": 2.14305, "chunk_loss": 0.030491, "F": 0.15375, "G": 0.18796, "chunks": 157.44, "bytes_per_chunk": 6.504, "boundary_entropy": 0.17766, "router_noise": 0.0, "grad_norm": 0.7394, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 8.834, "alpha": 0.03, "t": 170.4} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "parity_a", "lm_loss": 1.39005, "bpb": 2.00542, "chunk_loss": 0.030433, "F": 0.15417, "G": 0.17983, "chunks": 157.88, "bytes_per_chunk": 6.486, "boundary_entropy": 0.14713, "router_noise": 0.0, "grad_norm": 0.6981, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.4055, "alpha": 0.03, "t": 172.3} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "parity_a", "lm_loss": 1.42293, "bpb": 2.05285, "chunk_loss": 0.030666, "F": 0.16159, "G": 0.18357, "chunks": 165.47, "bytes_per_chunk": 6.188, "boundary_entropy": 0.13267, "router_noise": 0.0, "grad_norm": 0.666, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.0087, "alpha": 0.03, "t": 174.3} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "parity_a", "lm_loss": 1.36763, "bpb": 1.97307, "chunk_loss": 0.030866, "F": 0.14874, "G": 0.17404, "chunks": 152.31, "bytes_per_chunk": 6.723, "boundary_entropy": 0.15499, "router_noise": 0.0, "grad_norm": 0.6851, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.8608, "alpha": 0.03, "t": 176.2} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.184874, "boundary_entropy": 0.100236, "bpb": 2.145423}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.182289, "boundary_entropy": 0.129614, "bpb": 2.378673}, "vi": {"chunks": 4227.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229292, "cps": 33.023438, "mean_p": 0.184318, "boundary_entropy": 0.120298, "bpb": 1.980519}, "tr": {"chunks": 3031.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167431, "cps": 23.679688, "mean_p": 0.179719, "boundary_entropy": 0.14253, "bpb": 2.383394}, "fi": {"chunks": 2998.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159375, "cps": 23.421875, "mean_p": 0.18043, "boundary_entropy": 0.147214, "bpb": 2.566303}, "ru": {"chunks": 6254.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.330026, "cps": 48.859375, "mean_p": 0.178994, "boundary_entropy": 0.130929, "bpb": 1.422961}, "el": {"chunks": 6080.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.296253, "cps": 47.5, "mean_p": 0.170129, "boundary_entropy": 0.150731, "bpb": 1.539136}, "hi": {"chunks": 3378.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.200773, "cps": 26.390625, "mean_p": 0.085659, "boundary_entropy": 0.151425, "bpb": 1.293282}, "ta": {"chunks": 3149.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.168441, "cps": 24.601562, "mean_p": 0.083267, "boundary_entropy": 0.169571, "bpb": 1.051845}, "th": {"chunks": 2767.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.172024, "cps": 21.617188, "mean_p": 0.129112, "boundary_entropy": 0.270698, "bpb": 1.187854}, "zh": {"chunks": 3057.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.517609, "cps": 23.882812, "mean_p": 0.253925, "boundary_entropy": 0.350772, "bpb": 3.307937}, "my": {"chunks": 2886.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.152312, "cps": 22.546875, "mean_p": 0.079721, "boundary_entropy": 0.16218, "bpb": 1.193186}}, "bpb": {"en": 2.14542, "de": 2.37867, "vi": 1.98052, "tr": 2.38339, "fi": 2.5663, "ru": 1.42296, "el": 1.53914, "hi": 1.29328, "ta": 1.05184, "th": 1.18785, "zh": 3.30794, "my": 1.19319}, "gini_cpc": 0.212379, "gini_cps": 0.150027, "bpic": 8.723287, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9619789919517142, "vi": 1.3812289875938615, "tr": 1.008583742194271, "fi": 0.9600554634495304, "ru": 1.9880405614600372, "el": 1.7845935231899737, "hi": 1.2094311517221905, "ta": 1.0146675453951877, "th": 1.0362503070558116, "zh": 3.1180176872772076, "my": 0.9175073015046262}, "t": 179.1} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.1555, "de": 2.38854, "vi": 1.9833, "tr": 2.38768, "fi": 2.57403, "ru": 1.40118, "el": 1.52913, "hi": 1.29172, "ta": 1.05819, "th": 1.20634, "zh": 3.30967, "my": 1.20202}, "bytes_per_chunk": {"en": 5.741, "de": 6.3, "vi": 5.681, "tr": 6.368, "fi": 6.337, "ru": 5.573, "el": 6.288, "hi": 10.735, "ta": 14.915, "th": 16.024, "zh": 5.755, "my": 16.552}, "bpb_mean": 1.87394, "t": 180.3} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "parity_a", "lm_loss": 1.32138, "bpb": 1.90636, "chunk_loss": 0.030943, "F": 0.1539, "G": 0.17358, "chunks": 157.59, "bytes_per_chunk": 6.498, "boundary_entropy": 0.13873, "router_noise": 0.0, "grad_norm": 0.7439, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.495, "alpha": 0.03, "t": 180.4} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "parity_a", "lm_loss": 1.55698, "bpb": 2.24624, "chunk_loss": 0.030512, "F": 0.16183, "G": 0.18846, "chunks": 165.72, "bytes_per_chunk": 6.179, "boundary_entropy": 0.15606, "router_noise": 0.0, "grad_norm": 0.6945, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 7.9529, "alpha": 0.03, "t": 182.4} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "parity_a", "lm_loss": 1.50965, "bpb": 2.17797, "chunk_loss": 0.030522, "F": 0.16257, "G": 0.18818, "chunks": 166.47, "bytes_per_chunk": 6.151, "boundary_entropy": 0.14125, "router_noise": 0.0, "grad_norm": 0.7152, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 7.6934, "alpha": 0.03, "t": 184.3} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "parity_a", "lm_loss": 1.33558, "bpb": 1.92684, "chunk_loss": 0.031385, "F": 0.16095, "G": 0.18234, "chunks": 164.81, "bytes_per_chunk": 6.213, "boundary_entropy": 0.17054, "router_noise": 0.0, "grad_norm": 0.6269, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.8711, "alpha": 0.03, "t": 186.2} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2890.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166063, "cps": 22.578125, "mean_p": 0.185481, "boundary_entropy": 0.099323, "bpb": 2.118712}, "de": {"chunks": 3281.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159792, "cps": 25.632812, "mean_p": 0.181217, "boundary_entropy": 0.12434, "bpb": 2.347964}, "vi": {"chunks": 4237.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229835, "cps": 33.101562, "mean_p": 0.184648, "boundary_entropy": 0.107084, "bpb": 1.956353}, "tr": {"chunks": 3080.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170138, "cps": 24.0625, "mean_p": 0.185418, "boundary_entropy": 0.142973, "bpb": 2.358416}, "fi": {"chunks": 2998.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159375, "cps": 23.421875, "mean_p": 0.17766, "boundary_entropy": 0.139099, "bpb": 2.54296}, "ru": {"chunks": 6218.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328127, "cps": 48.578125, "mean_p": 0.180382, "boundary_entropy": 0.127516, "bpb": 1.408499}, "el": {"chunks": 6035.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.29406, "cps": 47.148438, "mean_p": 0.171895, "boundary_entropy": 0.152878, "bpb": 1.517538}, "hi": {"chunks": 1900.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.112927, "cps": 14.84375, "mean_p": 0.078927, "boundary_entropy": 0.155081, "bpb": 1.293759}, "ta": {"chunks": 1963.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.105001, "cps": 15.335938, "mean_p": 0.076826, "boundary_entropy": 0.175803, "bpb": 1.049482}, "th": {"chunks": 2834.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.176189, "cps": 22.140625, "mean_p": 0.126468, "boundary_entropy": 0.264701, "bpb": 1.177251}, "zh": {"chunks": 3092.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.523535, "cps": 24.15625, "mean_p": 0.253808, "boundary_entropy": 0.346244, "bpb": 3.284}, "my": {"chunks": 2936.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.15495, "cps": 22.9375, "mean_p": 0.081712, "boundary_entropy": 0.1622, "bpb": 1.180044}}, "bpb": {"en": 2.11871, "de": 2.34796, "vi": 1.95635, "tr": 2.35842, "fi": 2.54296, "ru": 1.4085, "el": 1.51754, "hi": 1.29376, "ta": 1.04948, "th": 1.17725, "zh": 3.284, "my": 1.18004}, "gini_cpc": 0.255815, "gini_cps": 0.193887, "bpic": 9.255764, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.962232675664139, "vi": 1.3840175572454607, "tr": 1.02453415834459, "fi": 0.9597232643272297, "ru": 1.9759128283317053, "el": 1.770772246919525, "hi": 0.6800248847025918, "ta": 0.6322969800225992, "th": 1.0609747739356123, "zh": 3.1526250355922136, "my": 0.933080154536948}, "t": 189.1} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.12999, "de": 2.35953, "vi": 1.9596, "tr": 2.36543, "fi": 2.55023, "ru": 1.38698, "el": 1.50762, "hi": 1.29106, "ta": 1.055, "th": 1.19542, "zh": 3.28393, "my": 1.18822}, "bytes_per_chunk": {"en": 5.738, "de": 6.299, "vi": 5.661, "tr": 6.298, "fi": 6.342, "ru": 5.6, "el": 6.306, "hi": 18.16, "ta": 21.45, "th": 15.412, "zh": 5.669, "my": 16.682}, "bpb_mean": 1.85608, "t": 190.3} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "parity_a", "lm_loss": 1.37122, "bpb": 1.97825, "chunk_loss": 0.030588, "F": 0.15134, "G": 0.17698, "chunks": 154.97, "bytes_per_chunk": 6.608, "boundary_entropy": 0.14919, "router_noise": 0.0, "grad_norm": 0.6766, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 8.9964, "alpha": 0.03, "t": 190.4} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "parity_a", "lm_loss": 1.37064, "bpb": 1.97742, "chunk_loss": 0.030752, "F": 0.15002, "G": 0.17705, "chunks": 153.62, "bytes_per_chunk": 6.666, "boundary_entropy": 0.1534, "router_noise": 0.0, "grad_norm": 0.7382, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.522, "alpha": 0.03, "t": 192.3} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "parity_a", "lm_loss": 1.43764, "bpb": 2.07408, "chunk_loss": 0.031097, "F": 0.16162, "G": 0.18294, "chunks": 165.5, "bytes_per_chunk": 6.187, "boundary_entropy": 0.15835, "router_noise": 0.0, "grad_norm": 0.6968, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.4057, "alpha": 0.03, "t": 194.3} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "parity_a", "lm_loss": 1.4001, "bpb": 2.01991, "chunk_loss": 0.030551, "F": 0.15466, "G": 0.17851, "chunks": 158.38, "bytes_per_chunk": 6.466, "boundary_entropy": 0.14171, "router_noise": 0.0, "grad_norm": 0.6886, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.9203, "alpha": 0.03, "t": 196.2} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.183532, "boundary_entropy": 0.093306, "bpb": 2.104236}, "de": {"chunks": 3270.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159256, "cps": 25.546875, "mean_p": 0.177847, "boundary_entropy": 0.116142, "bpb": 2.332598}, "vi": {"chunks": 4226.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229238, "cps": 33.015625, "mean_p": 0.180051, "boundary_entropy": 0.105138, "bpb": 1.946851}, "tr": {"chunks": 3024.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167044, "cps": 23.625, "mean_p": 0.178412, "boundary_entropy": 0.13475, "bpb": 2.338203}, "fi": {"chunks": 2978.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158312, "cps": 23.265625, "mean_p": 0.174871, "boundary_entropy": 0.132963, "bpb": 2.515246}, "ru": {"chunks": 6206.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327493, "cps": 48.484375, "mean_p": 0.17224, "boundary_entropy": 0.128837, "bpb": 1.406383}, "el": {"chunks": 5935.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.289188, "cps": 46.367188, "mean_p": 0.166747, "boundary_entropy": 0.153509, "bpb": 1.498213}, "hi": {"chunks": 2657.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.15792, "cps": 20.757812, "mean_p": 0.082184, "boundary_entropy": 0.149036, "bpb": 1.27176}, "ta": {"chunks": 2519.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.134742, "cps": 19.679688, "mean_p": 0.077371, "boundary_entropy": 0.172459, "bpb": 1.026496}, "th": {"chunks": 2489.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.15474, "cps": 19.445312, "mean_p": 0.115694, "boundary_entropy": 0.246634, "bpb": 1.173721}, "zh": {"chunks": 3089.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.523027, "cps": 24.132812, "mean_p": 0.249483, "boundary_entropy": 0.329919, "bpb": 3.255683}, "my": {"chunks": 2484.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.131096, "cps": 19.40625, "mean_p": 0.080321, "boundary_entropy": 0.169154, "bpb": 1.176485}}, "bpb": {"en": 2.10424, "de": 2.3326, "vi": 1.94685, "tr": 2.3382, "fi": 2.51525, "ru": 1.40638, "el": 1.49821, "hi": 1.27176, "ta": 1.0265, "th": 1.17372, "zh": 3.25568, "my": 1.17649}, "gini_cpc": 0.236615, "gini_cps": 0.174836, "bpic": 9.189278, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9600032024955444, "vi": 1.3818588588952525, "tr": 1.0069515429748193, "fi": 0.9543114756993237, "ru": 1.974148841179594, "el": 1.7432401279455303, "hi": 0.9519492977434841, "ta": 0.8122318852770576, "th": 0.9327841710328498, "zh": 3.1528390577188325, "my": 0.7902519055853663}, "t": 199.1} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.11588, "de": 2.34226, "vi": 1.94975, "tr": 2.3459, "fi": 2.52476, "ru": 1.38533, "el": 1.48718, "hi": 1.26826, "ta": 1.0333, "th": 1.19205, "zh": 3.25595, "my": 1.18455}, "bytes_per_chunk": {"en": 5.741, "de": 6.316, "vi": 5.686, "tr": 6.408, "fi": 6.373, "ru": 5.607, "el": 6.482, "hi": 11.509, "ta": 16.944, "th": 17.357, "zh": 5.614, "my": 20.49}, "bpb_mean": 1.84043, "t": 200.4} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "parity_a", "lm_loss": 1.46997, "bpb": 2.12072, "chunk_loss": 0.031985, "F": 0.16553, "G": 0.17997, "chunks": 169.5, "bytes_per_chunk": 6.041, "boundary_entropy": 0.14093, "router_noise": 0.0, "grad_norm": 1.3831, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.551, "alpha": 0.03, "t": 200.4} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "parity_a", "lm_loss": 1.32537, "bpb": 1.9121, "chunk_loss": 0.031096, "F": 0.16757, "G": 0.1826, "chunks": 171.59, "bytes_per_chunk": 5.968, "boundary_entropy": 0.12501, "router_noise": 0.0, "grad_norm": 0.6826, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.7039, "alpha": 0.03, "t": 202.3} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "parity_a", "lm_loss": 1.58747, "bpb": 2.29024, "chunk_loss": 0.030552, "F": 0.16541, "G": 0.19027, "chunks": 169.38, "bytes_per_chunk": 6.046, "boundary_entropy": 0.15954, "router_noise": 0.0, "grad_norm": 0.7556, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 7.5805, "alpha": 0.03, "t": 204.3} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "parity_a", "lm_loss": 1.30908, "bpb": 1.88861, "chunk_loss": 0.03078, "F": 0.14798, "G": 0.16633, "chunks": 151.53, "bytes_per_chunk": 6.758, "boundary_entropy": 0.13153, "router_noise": 0.0, "grad_norm": 0.6865, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.2592, "alpha": 0.03, "t": 206.2} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2885.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165776, "cps": 22.539062, "mean_p": 0.183223, "boundary_entropy": 0.090891, "bpb": 2.090483}, "de": {"chunks": 3276.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159548, "cps": 25.59375, "mean_p": 0.177268, "boundary_entropy": 0.113064, "bpb": 2.313311}, "vi": {"chunks": 4228.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229346, "cps": 33.03125, "mean_p": 0.181662, "boundary_entropy": 0.098767, "bpb": 1.93905}, "tr": {"chunks": 3045.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168204, "cps": 23.789062, "mean_p": 0.179974, "boundary_entropy": 0.131312, "bpb": 2.315649}, "fi": {"chunks": 2992.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159056, "cps": 23.375, "mean_p": 0.17609, "boundary_entropy": 0.133981, "bpb": 2.508004}, "ru": {"chunks": 6206.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327493, "cps": 48.484375, "mean_p": 0.175238, "boundary_entropy": 0.123584, "bpb": 1.381202}, "el": {"chunks": 5965.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.29065, "cps": 46.601562, "mean_p": 0.168633, "boundary_entropy": 0.133611, "bpb": 1.483459}, "hi": {"chunks": 2631.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.156374, "cps": 20.554688, "mean_p": 0.080965, "boundary_entropy": 0.144892, "bpb": 1.256846}, "ta": {"chunks": 2635.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.140947, "cps": 20.585938, "mean_p": 0.079954, "boundary_entropy": 0.17627, "bpb": 1.022339}, "th": {"chunks": 2912.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.181038, "cps": 22.75, "mean_p": 0.120893, "boundary_entropy": 0.249876, "bpb": 1.14737}, "zh": {"chunks": 3103.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.525398, "cps": 24.242188, "mean_p": 0.250035, "boundary_entropy": 0.330647, "bpb": 3.250629}, "my": {"chunks": 2375.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.125343, "cps": 18.554688, "mean_p": 0.076769, "boundary_entropy": 0.163069, "bpb": 1.163575}}, "bpb": {"en": 2.09048, "de": 2.31331, "vi": 1.93905, "tr": 2.31565, "fi": 2.508, "ru": 1.3812, "el": 1.48346, "hi": 1.25685, "ta": 1.02234, "th": 1.14737, "zh": 3.25063, "my": 1.16358}, "gini_cpc": 0.235746, "gini_cps": 0.169375, "bpic": 9.082929, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9624314108725177, "vi": 1.3834712529243456, "tr": 1.0146471696572406, "fi": 0.9594625070492788, "ru": 1.9755174019013821, "el": 1.7532663821990815, "hi": 0.9432874966844439, "ta": 0.8502241815738645, "th": 1.0920653008362742, "zh": 3.1693239727422275, "my": 0.756098785730828}, "t": 209.1} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.10266, "de": 2.32441, "vi": 1.94, "tr": 2.32135, "fi": 2.51954, "ru": 1.36042, "el": 1.47187, "hi": 1.25376, "ta": 1.02893, "th": 1.16637, "zh": 3.25018, "my": 1.17172}, "bytes_per_chunk": {"en": 5.74, "de": 6.308, "vi": 5.669, "tr": 6.36, "fi": 6.333, "ru": 5.605, "el": 6.412, "hi": 11.646, "ta": 16.327, "th": 15.126, "zh": 5.55, "my": 20.37}, "bpb_mean": 1.82593, "t": 210.4} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "parity_a", "lm_loss": 1.37117, "bpb": 1.97817, "chunk_loss": 0.030629, "F": 0.16098, "G": 0.17739, "chunks": 164.84, "bytes_per_chunk": 6.212, "boundary_entropy": 0.12364, "router_noise": 0.0, "grad_norm": 0.709, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 8.2652, "alpha": 0.03, "t": 210.4} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "parity_a", "lm_loss": 1.4894, "bpb": 2.14875, "chunk_loss": 0.030449, "F": 0.15372, "G": 0.17973, "chunks": 157.41, "bytes_per_chunk": 6.505, "boundary_entropy": 0.16396, "router_noise": 0.0, "grad_norm": 0.6959, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.011, "alpha": 0.03, "t": 212.4} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "parity_a", "lm_loss": 1.49497, "bpb": 2.15678, "chunk_loss": 0.030661, "F": 0.15829, "G": 0.18039, "chunks": 162.09, "bytes_per_chunk": 6.317, "boundary_entropy": 0.15614, "router_noise": 0.0, "grad_norm": 0.7086, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.7756, "alpha": 0.03, "t": 214.3} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "parity_a", "lm_loss": 1.42614, "bpb": 2.05748, "chunk_loss": 0.030597, "F": 0.15555, "G": 0.17515, "chunks": 159.28, "bytes_per_chunk": 6.429, "boundary_entropy": 0.1409, "router_noise": 0.0, "grad_norm": 0.7033, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.8445, "alpha": 0.03, "t": 216.3} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.182117, "boundary_entropy": 0.086479, "bpb": 2.073829}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.177436, "boundary_entropy": 0.111565, "bpb": 2.300359}, "vi": {"chunks": 4239.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229943, "cps": 33.117188, "mean_p": 0.180406, "boundary_entropy": 0.094803, "bpb": 1.921636}, "tr": {"chunks": 3062.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169143, "cps": 23.921875, "mean_p": 0.17911, "boundary_entropy": 0.125931, "bpb": 2.293111}, "fi": {"chunks": 2981.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158471, "cps": 23.289062, "mean_p": 0.170214, "boundary_entropy": 0.117847, "bpb": 2.472318}, "ru": {"chunks": 6177.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.325963, "cps": 48.257812, "mean_p": 0.175058, "boundary_entropy": 0.121667, "bpb": 1.370673}, "el": {"chunks": 5979.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.291332, "cps": 46.710938, "mean_p": 0.166072, "boundary_entropy": 0.135413, "bpb": 1.459553}, "hi": {"chunks": 3493.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.207608, "cps": 27.289062, "mean_p": 0.086032, "boundary_entropy": 0.145906, "bpb": 1.247561}, "ta": {"chunks": 3161.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.169083, "cps": 24.695312, "mean_p": 0.081023, "boundary_entropy": 0.164118, "bpb": 1.026544}, "th": {"chunks": 2863.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177992, "cps": 22.367188, "mean_p": 0.111302, "boundary_entropy": 0.227014, "bpb": 1.146254}, "zh": {"chunks": 3189.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.539959, "cps": 24.914062, "mean_p": 0.251384, "boundary_entropy": 0.317095, "bpb": 3.216873}, "my": {"chunks": 2452.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.129407, "cps": 19.15625, "mean_p": 0.072869, "boundary_entropy": 0.148753, "bpb": 1.157567}}, "bpb": {"en": 2.07383, "de": 2.30036, "vi": 1.92164, "tr": 2.29311, "fi": 2.47232, "ru": 1.37067, "el": 1.45955, "hi": 1.24756, "ta": 1.02654, "th": 1.14625, "zh": 3.21687, "my": 1.15757}, "gini_cpc": 0.226012, "gini_cps": 0.155601, "bpic": 8.769531, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.962312087170534, "vi": 1.3856297712934642, "tr": 1.0192519857515847, "fi": 0.9549420640160005, "ru": 1.9642434713015005, "el": 1.7555558063832981, "hi": 1.2510378344782735, "ta": 1.0188868511067366, "th": 1.0725738739056794, "zh": 3.253778670547722, "my": 0.7798014189000717}, "t": 219.2} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.08747, "de": 2.3101, "vi": 1.92365, "tr": 2.29902, "fi": 2.48221, "ru": 1.34994, "el": 1.45005, "hi": 1.24442, "ta": 1.03281, "th": 1.16384, "zh": 3.21565, "my": 1.16569}, "bytes_per_chunk": {"en": 5.74, "de": 6.304, "vi": 5.669, "tr": 6.346, "fi": 6.378, "ru": 5.631, "el": 6.407, "hi": 10.477, "ta": 15.359, "th": 15.446, "zh": 5.323, "my": 20.085}, "bpb_mean": 1.8104, "t": 220.4} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "parity_a", "lm_loss": 1.40449, "bpb": 2.02625, "chunk_loss": 0.030796, "F": 0.16122, "G": 0.18287, "chunks": 165.09, "bytes_per_chunk": 6.203, "boundary_entropy": 0.15189, "router_noise": 0.0, "grad_norm": 0.7335, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 8.8695, "alpha": 0.03, "t": 220.5} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "parity_a", "lm_loss": 1.30533, "bpb": 1.88319, "chunk_loss": 0.030924, "F": 0.1561, "G": 0.17025, "chunks": 159.84, "bytes_per_chunk": 6.406, "boundary_entropy": 0.12665, "router_noise": 0.0, "grad_norm": 0.7253, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.3813, "alpha": 0.03, "t": 222.4} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "parity_a", "lm_loss": 1.44647, "bpb": 2.08682, "chunk_loss": 0.030626, "F": 0.16055, "G": 0.17842, "chunks": 164.41, "bytes_per_chunk": 6.228, "boundary_entropy": 0.13287, "router_noise": 0.0, "grad_norm": 0.685, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.3701, "alpha": 0.03, "t": 224.3} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "parity_a", "lm_loss": 1.24078, "bpb": 1.79007, "chunk_loss": 0.031068, "F": 0.15472, "G": 0.16872, "chunks": 158.44, "bytes_per_chunk": 6.463, "boundary_entropy": 0.13233, "router_noise": 0.0, "grad_norm": 0.6742, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.0005, "alpha": 0.03, "t": 226.3} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2886.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165833, "cps": 22.546875, "mean_p": 0.181859, "boundary_entropy": 0.084568, "bpb": 2.062646}, "de": {"chunks": 3277.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159597, "cps": 25.601562, "mean_p": 0.174633, "boundary_entropy": 0.103494, "bpb": 2.287781}, "vi": {"chunks": 4218.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228804, "cps": 32.953125, "mean_p": 0.178918, "boundary_entropy": 0.089568, "bpb": 1.915456}, "tr": {"chunks": 3015.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.166547, "cps": 23.554688, "mean_p": 0.173582, "boundary_entropy": 0.119174, "bpb": 2.287238}, "fi": {"chunks": 2986.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158737, "cps": 23.328125, "mean_p": 0.169964, "boundary_entropy": 0.12041, "bpb": 2.455647}, "ru": {"chunks": 6180.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.326121, "cps": 48.28125, "mean_p": 0.175547, "boundary_entropy": 0.119369, "bpb": 1.357453}, "el": {"chunks": 5955.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.290162, "cps": 46.523438, "mean_p": 0.164946, "boundary_entropy": 0.13369, "bpb": 1.446281}, "hi": {"chunks": 3703.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.220089, "cps": 28.929688, "mean_p": 0.086861, "boundary_entropy": 0.14817, "bpb": 1.235929}, "ta": {"chunks": 3318.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.177481, "cps": 25.921875, "mean_p": 0.085061, "boundary_entropy": 0.169926, "bpb": 1.00233}, "th": {"chunks": 2915.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.181225, "cps": 22.773438, "mean_p": 0.107066, "boundary_entropy": 0.215247, "bpb": 1.136315}, "zh": {"chunks": 2964.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.501863, "cps": 23.15625, "mean_p": 0.237051, "boundary_entropy": 0.305099, "bpb": 3.205727}, "my": {"chunks": 2575.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.135898, "cps": 20.117188, "mean_p": 0.076976, "boundary_entropy": 0.157365, "bpb": 1.147464}}, "bpb": {"en": 2.06265, "de": 2.28778, "vi": 1.91546, "tr": 2.28724, "fi": 2.45565, "ru": 1.35745, "el": 1.44628, "hi": 1.23593, "ta": 1.00233, "th": 1.13632, "zh": 3.20573, "my": 1.14746}, "gini_cpc": 0.212768, "gini_cps": 0.1541, "bpic": 8.723882, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9623916087413871, "vi": 1.379720848720035, "tr": 1.0043025340213652, "fi": 0.9572066638144997, "ru": 1.9665593338416556, "el": 1.749720636944727, "hi": 1.3271696224742286, "ta": 1.070233905804914, "th": 1.0928115772691465, "zh": 3.026303987699179, "my": 0.8194862117439697}, "t": 229.2} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.0745, "de": 2.29731, "vi": 1.91901, "tr": 2.2931, "fi": 2.46694, "ru": 1.33576, "el": 1.43611, "hi": 1.23302, "ta": 1.00931, "th": 1.15464, "zh": 3.20539, "my": 1.15682}, "bytes_per_chunk": {"en": 5.741, "de": 6.308, "vi": 5.707, "tr": 6.383, "fi": 6.364, "ru": 5.638, "el": 6.415, "hi": 10.087, "ta": 14.474, "th": 15.254, "zh": 5.81, "my": 19.415}, "bpb_mean": 1.79849, "t": 230.4} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584, "t": 230.4} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "none", "lm_loss": 1.25533, "bpb": 1.81106, "chunk_loss": 0.0, "F": 0.15329, "G": 0.16849, "chunks": 156.97, "bytes_per_chunk": 6.524, "boundary_entropy": 0.119, "router_noise": 0.0, "grad_norm": 0.6967, "lr": 9.12e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 230.5} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "none", "lm_loss": 1.27541, "bpb": 1.84002, "chunk_loss": 0.0, "F": 0.159, "G": 0.1719, "chunks": 162.81, "bytes_per_chunk": 6.289, "boundary_entropy": 0.11325, "router_noise": 0.0, "grad_norm": 0.6944, "lr": 8.81e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 232.4} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "none", "lm_loss": 1.36267, "bpb": 1.96591, "chunk_loss": 0.0, "F": 0.15326, "G": 0.17435, "chunks": 156.94, "bytes_per_chunk": 6.525, "boundary_entropy": 0.13993, "router_noise": 0.0, "grad_norm": 0.7352, "lr": 8.5e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 234.3} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "none", "lm_loss": 1.3273, "bpb": 1.91489, "chunk_loss": 0.0, "F": 0.15298, "G": 0.17064, "chunks": 156.66, "bytes_per_chunk": 6.537, "boundary_entropy": 0.12617, "router_noise": 0.0, "grad_norm": 0.7039, "lr": 8.21e-05, "router_trainable": false, "sec_per_step": 0.038, "alpha": 0.0, "t": 236.2} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.182806, "boundary_entropy": 0.087103, "bpb": 2.05349}, "de": {"chunks": 3284.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159938, "cps": 25.65625, "mean_p": 0.176612, "boundary_entropy": 0.10724, "bpb": 2.270588}, "vi": {"chunks": 4234.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229672, "cps": 33.078125, "mean_p": 0.182236, "boundary_entropy": 0.092506, "bpb": 1.91058}, "tr": {"chunks": 3052.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168591, "cps": 23.84375, "mean_p": 0.17907, "boundary_entropy": 0.126634, "bpb": 2.267177}, "fi": {"chunks": 2991.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159003, "cps": 23.367188, "mean_p": 0.172752, "boundary_entropy": 0.12219, "bpb": 2.437002}, "ru": {"chunks": 6209.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327652, "cps": 48.507812, "mean_p": 0.187681, "boundary_entropy": 0.125521, "bpb": 1.347296}, "el": {"chunks": 5968.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.290796, "cps": 46.625, "mean_p": 0.172022, "boundary_entropy": 0.132858, "bpb": 1.436928}, "hi": {"chunks": 4270.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.253789, "cps": 33.359375, "mean_p": 0.097202, "boundary_entropy": 0.149027, "bpb": 1.230298}, "ta": {"chunks": 3259.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.174325, "cps": 25.460938, "mean_p": 0.085502, "boundary_entropy": 0.162351, "bpb": 1.000356}, "th": {"chunks": 3177.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.197513, "cps": 24.820312, "mean_p": 0.116079, "boundary_entropy": 0.222434, "bpb": 1.124087}, "zh": {"chunks": 3230.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.546901, "cps": 25.234375, "mean_p": 0.253625, "boundary_entropy": 0.318231, "bpb": 3.196164}, "my": {"chunks": 2649.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.139804, "cps": 20.695312, "mean_p": 0.078904, "boundary_entropy": 0.154855, "bpb": 1.138411}}, "bpb": {"en": 2.05349, "de": 2.27059, "vi": 1.91058, "tr": 2.26718, "fi": 2.437, "ru": 1.3473, "el": 1.43693, "hi": 1.2303, "ta": 1.00036, "th": 1.12409, "zh": 3.19616, "my": 1.13841}, "gini_cpc": 0.223814, "gini_cps": 0.151315, "bpic": 8.488664, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9637794737017485, "vi": 1.383995388454005, "tr": 1.0159232725388103, "fi": 0.9581454926104855, "ru": 1.9744192509812233, "el": 1.7523259830231683, "hi": 1.5293248076788513, "ta": 1.050475244465946, "th": 1.1902085914768925, "zh": 3.2956115101502483, "my": 0.8424526748231198}, "t": 239.0} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.06713, "de": 2.28092, "vi": 1.91212, "tr": 2.27311, "fi": 2.44645, "ru": 1.32619, "el": 1.427, "hi": 1.22545, "ta": 1.00643, "th": 1.14252, "zh": 3.19265, "my": 1.14749}, "bytes_per_chunk": {"en": 5.739, "de": 6.299, "vi": 5.704, "tr": 6.356, "fi": 6.343, "ru": 5.614, "el": 6.385, "hi": 9.373, "ta": 15.011, "th": 14.239, "zh": 5.357, "my": 18.922}, "bpb_mean": 1.78729, "t": 240.3} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "none", "lm_loss": 1.32281, "bpb": 1.90842, "chunk_loss": 0.0, "F": 0.15927, "G": 0.18021, "chunks": 163.09, "bytes_per_chunk": 6.279, "boundary_entropy": 0.14197, "router_noise": 0.0, "grad_norm": 0.6843, "lr": 7.92e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 240.3} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "none", "lm_loss": 1.16307, "bpb": 1.67796, "chunk_loss": 0.0, "F": 0.14862, "G": 0.16633, "chunks": 152.19, "bytes_per_chunk": 6.729, "boundary_entropy": 0.12814, "router_noise": 0.0, "grad_norm": 0.7862, "lr": 7.63e-05, "router_trainable": false, "sec_per_step": 0.041, "alpha": 0.0, "t": 242.3} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "none", "lm_loss": 1.29185, "bpb": 1.86375, "chunk_loss": 0.0, "F": 0.15567, "G": 0.17313, "chunks": 159.41, "bytes_per_chunk": 6.424, "boundary_entropy": 0.11987, "router_noise": 0.0, "grad_norm": 0.7365, "lr": 7.36e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 244.3} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "none", "lm_loss": 1.29209, "bpb": 1.86409, "chunk_loss": 0.0, "F": 0.16696, "G": 0.18381, "chunks": 170.97, "bytes_per_chunk": 5.989, "boundary_entropy": 0.11849, "router_noise": 0.0, "grad_norm": 0.686, "lr": 7.09e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 246.3} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.183249, "boundary_entropy": 0.087736, "bpb": 2.042294}, "de": {"chunks": 3278.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159645, "cps": 25.609375, "mean_p": 0.176669, "boundary_entropy": 0.106304, "bpb": 2.268052}, "vi": {"chunks": 4224.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229129, "cps": 33.0, "mean_p": 0.182918, "boundary_entropy": 0.088057, "bpb": 1.898462}, "tr": {"chunks": 3072.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169696, "cps": 24.0, "mean_p": 0.180756, "boundary_entropy": 0.127784, "bpb": 2.259394}, "fi": {"chunks": 2998.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159375, "cps": 23.421875, "mean_p": 0.172756, "boundary_entropy": 0.122714, "bpb": 2.42853}, "ru": {"chunks": 6215.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327968, "cps": 48.554688, "mean_p": 0.186106, "boundary_entropy": 0.121206, "bpb": 1.337607}, "el": {"chunks": 6090.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.29674, "cps": 47.578125, "mean_p": 0.173904, "boundary_entropy": 0.131335, "bpb": 1.419704}, "hi": {"chunks": 4141.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.246122, "cps": 32.351562, "mean_p": 0.100541, "boundary_entropy": 0.140224, "bpb": 1.221738}, "ta": {"chunks": 3233.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.172934, "cps": 25.257812, "mean_p": 0.08818, "boundary_entropy": 0.162383, "bpb": 0.994547}, "th": {"chunks": 3095.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.192415, "cps": 24.179688, "mean_p": 0.113926, "boundary_entropy": 0.221912, "bpb": 1.118149}, "zh": {"chunks": 3341.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.565696, "cps": 26.101562, "mean_p": 0.257355, "boundary_entropy": 0.313011, "bpb": 3.185725}, "my": {"chunks": 2588.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.136584, "cps": 20.21875, "mean_p": 0.080927, "boundary_entropy": 0.163073, "bpb": 1.129222}}, "bpb": {"en": 2.04229, "de": 2.26805, "vi": 1.89846, "tr": 2.25939, "fi": 2.42853, "ru": 1.33761, "el": 1.4197, "hi": 1.22174, "ta": 0.99455, "th": 1.11815, "zh": 3.18573, "my": 1.12922}, "gini_cpc": 0.230191, "gini_cps": 0.153933, "bpic": 8.497686, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9620186098642911, "vi": 1.3807266227750867, "tr": 1.0225806989643595, "fi": 0.9603878926266252, "ru": 1.9763272096711713, "el": 1.7881476602900628, "hi": 1.4831227233250874, "ta": 1.042094650309421, "th": 1.1594886970793146, "zh": 3.408866271025381, "my": 0.8230530473545619}, "t": 249.2} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.05384, "de": 2.27707, "vi": 1.90008, "tr": 2.26485, "fi": 2.43933, "ru": 1.3169, "el": 1.41078, "hi": 1.21681, "ta": 1.00163, "th": 1.13685, "zh": 3.18195, "my": 1.13868}, "bytes_per_chunk": {"en": 5.739, "de": 6.299, "vi": 5.707, "tr": 6.331, "fi": 6.353, "ru": 5.606, "el": 6.225, "hi": 9.675, "ta": 15.163, "th": 14.437, "zh": 4.998, "my": 19.142}, "bpb_mean": 1.77823, "t": 250.4} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "none", "lm_loss": 1.45319, "bpb": 2.09651, "chunk_loss": 0.0, "F": 0.16446, "G": 0.19171, "chunks": 168.41, "bytes_per_chunk": 6.081, "boundary_entropy": 0.16508, "router_noise": 0.0, "grad_norm": 0.6687, "lr": 6.82e-05, "router_trainable": false, "sec_per_step": 0.084, "alpha": 0.0, "t": 250.5} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "none", "lm_loss": 1.34528, "bpb": 1.94083, "chunk_loss": 0.0, "F": 0.15076, "G": 0.17118, "chunks": 154.38, "bytes_per_chunk": 6.633, "boundary_entropy": 0.13851, "router_noise": 0.0, "grad_norm": 0.7203, "lr": 6.57e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 252.5} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "none", "lm_loss": 1.40214, "bpb": 2.02285, "chunk_loss": 0.0, "F": 0.1633, "G": 0.18473, "chunks": 167.22, "bytes_per_chunk": 6.124, "boundary_entropy": 0.13687, "router_noise": 0.0, "grad_norm": 0.7707, "lr": 6.32e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 254.4} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "none", "lm_loss": 1.37085, "bpb": 1.97772, "chunk_loss": 0.0, "F": 0.16269, "G": 0.18111, "chunks": 166.59, "bytes_per_chunk": 6.147, "boundary_entropy": 0.12672, "router_noise": 0.0, "grad_norm": 0.6852, "lr": 6.08e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 256.4} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2893.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166236, "cps": 22.601562, "mean_p": 0.183784, "boundary_entropy": 0.089928, "bpb": 2.033994}, "de": {"chunks": 3283.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159889, "cps": 25.648438, "mean_p": 0.177566, "boundary_entropy": 0.109466, "bpb": 2.254465}, "vi": {"chunks": 4209.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228316, "cps": 32.882812, "mean_p": 0.182899, "boundary_entropy": 0.087244, "bpb": 1.892081}, "tr": {"chunks": 3081.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170193, "cps": 24.070312, "mean_p": 0.180839, "boundary_entropy": 0.128668, "bpb": 2.244077}, "fi": {"chunks": 2995.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159215, "cps": 23.398438, "mean_p": 0.17369, "boundary_entropy": 0.123183, "bpb": 2.410055}, "ru": {"chunks": 6223.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328391, "cps": 48.617188, "mean_p": 0.187143, "boundary_entropy": 0.119581, "bpb": 1.329951}, "el": {"chunks": 6110.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.297715, "cps": 47.734375, "mean_p": 0.175316, "boundary_entropy": 0.131735, "bpb": 1.408539}, "hi": {"chunks": 4501.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.267519, "cps": 35.164062, "mean_p": 0.105798, "boundary_entropy": 0.149975, "bpb": 1.202263}, "ta": {"chunks": 3379.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.180744, "cps": 26.398438, "mean_p": 0.091678, "boundary_entropy": 0.167909, "bpb": 0.985456}, "th": {"chunks": 3172.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.197202, "cps": 24.78125, "mean_p": 0.114197, "boundary_entropy": 0.215893, "bpb": 1.110811}, "zh": {"chunks": 3205.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.542668, "cps": 25.039062, "mean_p": 0.251566, "boundary_entropy": 0.309939, "bpb": 3.165816}, "my": {"chunks": 2996.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.158117, "cps": 23.40625, "mean_p": 0.084371, "boundary_entropy": 0.159375, "bpb": 1.121825}}, "bpb": {"en": 2.03399, "de": 2.25446, "vi": 1.89208, "tr": 2.24408, "fi": 2.41006, "ru": 1.32995, "el": 1.40854, "hi": 1.20226, "ta": 0.98546, "th": 1.11081, "zh": 3.16582, "my": 1.12182}, "gini_cpc": 0.216332, "gini_cps": 0.147565, "bpic": 8.334549, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9618207941894991, "vi": 1.3734456251826397, "tr": 1.0238040269587674, "fi": 0.9577686772801354, "ru": 1.975451052353648, "el": 1.7909194440931346, "hi": 1.6092726358495093, "ta": 1.0872725127940681, "th": 1.1862816155231333, "zh": 3.2644519291693848, "my": 0.9511611855131964}, "t": 259.3} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.04634, "de": 2.26399, "vi": 1.89303, "tr": 2.25044, "fi": 2.4217, "ru": 1.30874, "el": 1.39978, "hi": 1.1971, "ta": 0.99292, "th": 1.12992, "zh": 3.16385, "my": 1.13099}, "bytes_per_chunk": {"en": 5.737, "de": 6.304, "vi": 5.706, "tr": 6.349, "fi": 6.351, "ru": 5.597, "el": 6.201, "hi": 8.901, "ta": 14.357, "th": 14.312, "zh": 5.301, "my": 16.531}, "bpb_mean": 1.76657, "t": 260.6} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "none", "lm_loss": 1.25092, "bpb": 1.80469, "chunk_loss": 0.0, "F": 0.16138, "G": 0.17894, "chunks": 165.25, "bytes_per_chunk": 6.197, "boundary_entropy": 0.1224, "router_noise": 0.0, "grad_norm": 0.7148, "lr": 5.85e-05, "router_trainable": false, "sec_per_step": 0.084, "alpha": 0.0, "t": 260.6} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "none", "lm_loss": 1.33841, "bpb": 1.93092, "chunk_loss": 0.0, "F": 0.16171, "G": 0.18191, "chunks": 165.59, "bytes_per_chunk": 6.184, "boundary_entropy": 0.13955, "router_noise": 0.0, "grad_norm": 0.7406, "lr": 5.62e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 262.5} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "none", "lm_loss": 1.37105, "bpb": 1.978, "chunk_loss": 0.0, "F": 0.17029, "G": 0.19264, "chunks": 174.38, "bytes_per_chunk": 5.872, "boundary_entropy": 0.14552, "router_noise": 0.0, "grad_norm": 0.7126, "lr": 5.41e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 264.5} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "none", "lm_loss": 1.26882, "bpb": 1.83053, "chunk_loss": 0.0, "F": 0.14484, "G": 0.16996, "chunks": 148.31, "bytes_per_chunk": 6.904, "boundary_entropy": 0.15478, "router_noise": 0.0, "grad_norm": 0.7399, "lr": 5.2e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 266.5} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.183083, "boundary_entropy": 0.087655, "bpb": 2.025631}, "de": {"chunks": 3283.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159889, "cps": 25.648438, "mean_p": 0.177439, "boundary_entropy": 0.108538, "bpb": 2.244985}, "vi": {"chunks": 4229.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229401, "cps": 33.039062, "mean_p": 0.182782, "boundary_entropy": 0.086682, "bpb": 1.883209}, "tr": {"chunks": 3078.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170027, "cps": 24.046875, "mean_p": 0.180757, "boundary_entropy": 0.127979, "bpb": 2.233454}, "fi": {"chunks": 2993.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159109, "cps": 23.382812, "mean_p": 0.172422, "boundary_entropy": 0.121672, "bpb": 2.397696}, "ru": {"chunks": 6219.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328179, "cps": 48.585938, "mean_p": 0.188804, "boundary_entropy": 0.119579, "bpb": 1.323646}, "el": {"chunks": 6067.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.29562, "cps": 47.398438, "mean_p": 0.175936, "boundary_entropy": 0.133703, "bpb": 1.404755}, "hi": {"chunks": 4497.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.267281, "cps": 35.132812, "mean_p": 0.106023, "boundary_entropy": 0.15078, "bpb": 1.198689}, "ta": {"chunks": 3514.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.187965, "cps": 27.453125, "mean_p": 0.093341, "boundary_entropy": 0.171147, "bpb": 0.98679}, "th": {"chunks": 3018.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.187628, "cps": 23.578125, "mean_p": 0.116043, "boundary_entropy": 0.22671, "bpb": 1.109015}, "zh": {"chunks": 3248.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.549949, "cps": 25.375, "mean_p": 0.253951, "boundary_entropy": 0.310903, "bpb": 3.159752}, "my": {"chunks": 2758.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.145556, "cps": 21.546875, "mean_p": 0.082495, "boundary_entropy": 0.163213, "bpb": 1.122537}}, "bpb": {"en": 2.02563, "de": 2.24498, "vi": 1.88321, "tr": 2.23345, "fi": 2.3977, "ru": 1.32365, "el": 1.40476, "hi": 1.19869, "ta": 0.98679, "th": 1.10902, "zh": 3.15975, "my": 1.12254}, "gini_cpc": 0.222656, "gini_cps": 0.153991, "bpic": 8.381145, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9638197289886459, "vi": 1.382839828269764, "tr": 1.024932820527941, "fi": 0.9591182829980108, "ru": 1.9782841835797447, "el": 1.7820114332342938, "hi": 1.6111840391239922, "ta": 1.133061867750528, "th": 1.1310335991069267, "zh": 3.315125043532136, "my": 0.8774213991966346}, "t": 269.4} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.03862, "de": 2.25523, "vi": 1.88548, "tr": 2.2392, "fi": 2.40793, "ru": 1.30262, "el": 1.39526, "hi": 1.19516, "ta": 0.99284, "th": 1.12651, "zh": 3.15747, "my": 1.13194}, "bytes_per_chunk": {"en": 5.739, "de": 6.3, "vi": 5.699, "tr": 6.333, "fi": 6.355, "ru": 5.608, "el": 6.286, "hi": 8.947, "ta": 13.724, "th": 14.765, "zh": 5.236, "my": 18.255}, "bpb_mean": 1.76069, "t": 270.6} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "none", "lm_loss": 1.45838, "bpb": 2.104, "chunk_loss": 0.0, "F": 0.1698, "G": 0.19006, "chunks": 173.88, "bytes_per_chunk": 5.889, "boundary_entropy": 0.12282, "router_noise": 0.0, "grad_norm": 0.7513, "lr": 5e-05, "router_trainable": false, "sec_per_step": 0.084, "alpha": 0.0, "t": 270.7} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "none", "lm_loss": 1.27151, "bpb": 1.8344, "chunk_loss": 0.0, "F": 0.15958, "G": 0.18165, "chunks": 163.41, "bytes_per_chunk": 6.267, "boundary_entropy": 0.1548, "router_noise": 0.0, "grad_norm": 0.692, "lr": 4.81e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 272.6} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "none", "lm_loss": 1.44451, "bpb": 2.08398, "chunk_loss": 0.0, "F": 0.1701, "G": 0.19245, "chunks": 174.19, "bytes_per_chunk": 5.879, "boundary_entropy": 0.14905, "router_noise": 0.0, "grad_norm": 0.731, "lr": 4.63e-05, "router_trainable": false, "sec_per_step": 0.04, "alpha": 0.0, "t": 274.6} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "none", "lm_loss": 1.2338, "bpb": 1.78, "chunk_loss": 0.0, "F": 0.15607, "G": 0.17873, "chunks": 159.81, "bytes_per_chunk": 6.408, "boundary_entropy": 0.1556, "router_noise": 0.0, "grad_norm": 0.6441, "lr": 4.45e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 276.6} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.182905, "boundary_entropy": 0.087062, "bpb": 2.018638}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.176528, "boundary_entropy": 0.105515, "bpb": 2.235382}, "vi": {"chunks": 4224.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229129, "cps": 33.0, "mean_p": 0.183212, "boundary_entropy": 0.089118, "bpb": 1.877158}, "tr": {"chunks": 3071.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16964, "cps": 23.992188, "mean_p": 0.180703, "boundary_entropy": 0.126725, "bpb": 2.224345}, "fi": {"chunks": 2985.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158684, "cps": 23.320312, "mean_p": 0.172457, "boundary_entropy": 0.120652, "bpb": 2.393266}, "ru": {"chunks": 6228.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328654, "cps": 48.65625, "mean_p": 0.189382, "boundary_entropy": 0.120373, "bpb": 1.316046}, "el": {"chunks": 6050.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.294791, "cps": 47.265625, "mean_p": 0.176687, "boundary_entropy": 0.135356, "bpb": 1.392095}, "hi": {"chunks": 4426.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.263061, "cps": 34.578125, "mean_p": 0.103964, "boundary_entropy": 0.155094, "bpb": 1.193374}, "ta": {"chunks": 3297.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.176357, "cps": 25.757812, "mean_p": 0.090014, "boundary_entropy": 0.173252, "bpb": 0.976529}, "th": {"chunks": 2951.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.183463, "cps": 23.054688, "mean_p": 0.115028, "boundary_entropy": 0.224897, "bpb": 1.101823}, "zh": {"chunks": 3341.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.565696, "cps": 26.101562, "mean_p": 0.261033, "boundary_entropy": 0.318782, "bpb": 3.147793}, "my": {"chunks": 3012.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.158961, "cps": 23.53125, "mean_p": 0.086182, "boundary_entropy": 0.165839, "bpb": 1.114346}}, "bpb": {"en": 2.01864, "de": 2.23538, "vi": 1.87716, "tr": 2.22435, "fi": 2.39327, "ru": 1.31605, "el": 1.3921, "hi": 1.19337, "ta": 0.97653, "th": 1.10182, "zh": 3.14779, "my": 1.11435}, "gini_cpc": 0.22251, "gini_cps": 0.148615, "bpic": 8.388106, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9619789919517142, "vi": 1.3802486973258743, "tr": 1.021893986235106, "fi": 0.9558924477641255, "ru": 1.9797756022982271, "el": 1.7757879630426547, "hi": 1.5846483947668488, "ta": 1.062355953371843, "th": 1.105158892707517, "zh": 3.4076863242372104, "my": 0.957564792838508}, "t": 279.5} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.03117, "de": 2.24662, "vi": 1.88031, "tr": 2.23089, "fi": 2.4029, "ru": 1.29502, "el": 1.38121, "hi": 1.18902, "ta": 0.9841, "th": 1.1199, "zh": 3.1444, "my": 1.12342}, "bytes_per_chunk": {"en": 5.739, "de": 6.303, "vi": 5.702, "tr": 6.328, "fi": 6.362, "ru": 5.596, "el": 6.311, "hi": 9.048, "ta": 14.82, "th": 15.06, "zh": 5.083, "my": 16.6}, "bpb_mean": 1.75241, "t": 280.8} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "none", "lm_loss": 1.34781, "bpb": 1.94448, "chunk_loss": 0.0, "F": 0.15723, "G": 0.17951, "chunks": 161.0, "bytes_per_chunk": 6.36, "boundary_entropy": 0.13037, "router_noise": 0.0, "grad_norm": 0.7633, "lr": 4.29e-05, "router_trainable": false, "sec_per_step": 0.084, "alpha": 0.0, "t": 280.8} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "none", "lm_loss": 1.31237, "bpb": 1.89335, "chunk_loss": 0.0, "F": 0.15677, "G": 0.17687, "chunks": 160.53, "bytes_per_chunk": 6.379, "boundary_entropy": 0.12817, "router_noise": 0.0, "grad_norm": 0.7562, "lr": 4.14e-05, "router_trainable": false, "sec_per_step": 0.04, "alpha": 0.0, "t": 282.8} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "none", "lm_loss": 1.28171, "bpb": 1.84912, "chunk_loss": 0.0, "F": 0.15482, "G": 0.17633, "chunks": 158.53, "bytes_per_chunk": 6.459, "boundary_entropy": 0.13958, "router_noise": 0.0, "grad_norm": 0.7101, "lr": 3.99e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 284.7} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "none", "lm_loss": 1.35634, "bpb": 1.95679, "chunk_loss": 0.0, "F": 0.16165, "G": 0.18173, "chunks": 165.53, "bytes_per_chunk": 6.186, "boundary_entropy": 0.12783, "router_noise": 0.0, "grad_norm": 0.736, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 286.7} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.182728, "boundary_entropy": 0.086595, "bpb": 2.012359}, "de": {"chunks": 3277.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159597, "cps": 25.601562, "mean_p": 0.17661, "boundary_entropy": 0.106455, "bpb": 2.228045}, "vi": {"chunks": 4210.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22837, "cps": 32.890625, "mean_p": 0.18239, "boundary_entropy": 0.089729, "bpb": 1.870715}, "tr": {"chunks": 3074.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169806, "cps": 24.015625, "mean_p": 0.180929, "boundary_entropy": 0.127484, "bpb": 2.217618}, "fi": {"chunks": 2978.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158312, "cps": 23.265625, "mean_p": 0.171647, "boundary_entropy": 0.120461, "bpb": 2.382887}, "ru": {"chunks": 6227.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328602, "cps": 48.648438, "mean_p": 0.190411, "boundary_entropy": 0.124019, "bpb": 1.312229}, "el": {"chunks": 6056.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.295084, "cps": 47.3125, "mean_p": 0.175077, "boundary_entropy": 0.134366, "bpb": 1.384764}, "hi": {"chunks": 4530.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.269242, "cps": 35.390625, "mean_p": 0.106386, "boundary_entropy": 0.153564, "bpb": 1.187286}, "ta": {"chunks": 3244.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.173522, "cps": 25.34375, "mean_p": 0.091757, "boundary_entropy": 0.175958, "bpb": 0.977079}, "th": {"chunks": 2942.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.182903, "cps": 22.984375, "mean_p": 0.113537, "boundary_entropy": 0.223827, "bpb": 1.101279}, "zh": {"chunks": 3384.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.572977, "cps": 26.4375, "mean_p": 0.264418, "boundary_entropy": 0.320032, "bpb": 3.137315}, "my": {"chunks": 3319.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.175164, "cps": 25.929688, "mean_p": 0.093221, "boundary_entropy": 0.168241, "bpb": 1.10886}}, "bpb": {"en": 2.01236, "de": 2.22804, "vi": 1.87072, "tr": 2.21762, "fi": 2.38289, "ru": 1.31223, "el": 1.38476, "hi": 1.18729, "ta": 0.97708, "th": 1.10128, "zh": 3.13732, "my": 1.10886}, "gini_cpc": 0.221574, "gini_cps": 0.147727, "bpic": 8.319914, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.962058255222599, "vi": 1.376627022231191, "tr": 1.0236008740425246, "fi": 0.9543114756993237, "ru": 1.980829009672145, "el": 1.7787804911268965, "hi": 1.6230072708987515, "ta": 1.046002475521546, "th": 1.1025516396860762, "zh": 3.4539356980642704, "my": 1.0558961653131365}, "t": 289.6} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.0248, "de": 2.23811, "vi": 1.87325, "tr": 2.22238, "fi": 2.39155, "ru": 1.29087, "el": 1.37516, "hi": 1.18327, "ta": 0.98451, "th": 1.11944, "zh": 3.13371, "my": 1.11806}, "bytes_per_chunk": {"en": 5.74, "de": 6.305, "vi": 5.711, "tr": 6.334, "fi": 6.374, "ru": 5.598, "el": 6.326, "hi": 8.863, "ta": 15.022, "th": 15.1, "zh": 4.939, "my": 14.894}, "bpb_mean": 1.74626, "t": 290.8} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "none", "lm_loss": 1.22677, "bpb": 1.76986, "chunk_loss": 0.0, "F": 0.15338, "G": 0.17432, "chunks": 157.06, "bytes_per_chunk": 6.52, "boundary_entropy": 0.141, "router_noise": 0.0, "grad_norm": 0.7083, "lr": 3.73e-05, "router_trainable": false, "sec_per_step": 0.084, "alpha": 0.0, "t": 290.9} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "none", "lm_loss": 1.27987, "bpb": 1.84646, "chunk_loss": 0.0, "F": 0.15558, "G": 0.17807, "chunks": 159.31, "bytes_per_chunk": 6.428, "boundary_entropy": 0.15048, "router_noise": 0.0, "grad_norm": 0.7455, "lr": 3.62e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 292.8} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "none", "lm_loss": 1.45763, "bpb": 2.10291, "chunk_loss": 0.0, "F": 0.16895, "G": 0.19029, "chunks": 173.0, "bytes_per_chunk": 5.919, "boundary_entropy": 0.12831, "router_noise": 0.0, "grad_norm": 0.7654, "lr": 3.51e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 294.7} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "none", "lm_loss": 1.30211, "bpb": 1.87854, "chunk_loss": 0.0, "F": 0.15234, "G": 0.1739, "chunks": 156.0, "bytes_per_chunk": 6.564, "boundary_entropy": 0.1254, "router_noise": 0.0, "grad_norm": 0.7295, "lr": 3.41e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 296.7} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.183134, "boundary_entropy": 0.087837, "bpb": 2.008509}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.176327, "boundary_entropy": 0.105979, "bpb": 2.221661}, "vi": {"chunks": 4207.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228207, "cps": 32.867188, "mean_p": 0.183918, "boundary_entropy": 0.088893, "bpb": 1.86684}, "tr": {"chunks": 3093.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170856, "cps": 24.164062, "mean_p": 0.18237, "boundary_entropy": 0.128965, "bpb": 2.212247}, "fi": {"chunks": 2987.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15879, "cps": 23.335938, "mean_p": 0.170867, "boundary_entropy": 0.118352, "bpb": 2.375118}, "ru": {"chunks": 6238.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.329182, "cps": 48.734375, "mean_p": 0.193971, "boundary_entropy": 0.127495, "bpb": 1.305171}, "el": {"chunks": 6153.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.29981, "cps": 48.070312, "mean_p": 0.179361, "boundary_entropy": 0.135326, "bpb": 1.375849}, "hi": {"chunks": 4424.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.262942, "cps": 34.5625, "mean_p": 0.109235, "boundary_entropy": 0.152861, "bpb": 1.182912}, "ta": {"chunks": 3413.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.182562, "cps": 26.664062, "mean_p": 0.094985, "boundary_entropy": 0.173678, "bpb": 0.972408}, "th": {"chunks": 3023.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.187939, "cps": 23.617188, "mean_p": 0.112523, "boundary_entropy": 0.218121, "bpb": 1.095961}, "zh": {"chunks": 3478.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.588893, "cps": 27.171875, "mean_p": 0.267543, "boundary_entropy": 0.31876, "bpb": 3.12618}, "my": {"chunks": 3272.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.172683, "cps": 25.5625, "mean_p": 0.092655, "boundary_entropy": 0.171977, "bpb": 1.108202}}, "bpb": {"en": 2.00851, "de": 2.22166, "vi": 1.86684, "tr": 2.21225, "fi": 2.37512, "ru": 1.30517, "el": 1.37585, "hi": 1.18291, "ta": 0.97241, "th": 1.09596, "zh": 3.12618, "my": 1.1082}, "gini_cpc": 0.224302, "gini_cps": 0.14615, "bpic": 8.26135, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.962312087170534, "vi": 1.3751697211209257, "tr": 1.0295709967111861, "fi": 0.9568641211726915, "ru": 1.983641051315972, "el": 1.8066457395344429, "hi": 1.5844807843492361, "ta": 1.1001141483161316, "th": 1.1325151312668071, "zh": 3.548649174087481, "my": 1.040583296346262}, "t": 299.6} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.02212, "de": 2.23244, "vi": 1.86818, "tr": 2.21743, "fi": 2.38489, "ru": 1.28388, "el": 1.36521, "hi": 1.17855, "ta": 0.97972, "th": 1.11541, "zh": 3.12235, "my": 1.11746}, "bytes_per_chunk": {"en": 5.738, "de": 6.305, "vi": 5.715, "tr": 6.313, "fi": 6.368, "ru": 5.587, "el": 6.171, "hi": 9.097, "ta": 14.226, "th": 14.872, "zh": 4.736, "my": 15.284}, "bpb_mean": 1.74064, "t": 300.8} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "none", "lm_loss": 1.31358, "bpb": 1.89509, "chunk_loss": 0.0, "F": 0.16595, "G": 0.1874, "chunks": 169.94, "bytes_per_chunk": 6.026, "boundary_entropy": 0.13378, "router_noise": 0.0, "grad_norm": 0.719, "lr": 3.33e-05, "router_trainable": false, "sec_per_step": 0.084, "alpha": 0.0, "t": 300.9} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "none", "lm_loss": 1.21745, "bpb": 1.75641, "chunk_loss": 0.0, "F": 0.15817, "G": 0.17658, "chunks": 161.97, "bytes_per_chunk": 6.322, "boundary_entropy": 0.12442, "router_noise": 0.0, "grad_norm": 0.7425, "lr": 3.25e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 302.8} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "none", "lm_loss": 1.28194, "bpb": 1.84945, "chunk_loss": 0.0, "F": 0.15857, "G": 0.17753, "chunks": 162.38, "bytes_per_chunk": 6.306, "boundary_entropy": 0.11496, "router_noise": 0.0, "grad_norm": 0.7624, "lr": 3.18e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 304.7} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "none", "lm_loss": 1.30463, "bpb": 1.88218, "chunk_loss": 0.0, "F": 0.15997, "G": 0.1821, "chunks": 163.81, "bytes_per_chunk": 6.251, "boundary_entropy": 0.14428, "router_noise": 0.0, "grad_norm": 0.7505, "lr": 3.13e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 306.7} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.183411, "boundary_entropy": 0.088851, "bpb": 2.00625}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.177203, "boundary_entropy": 0.107575, "bpb": 2.21629}, "vi": {"chunks": 4210.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22837, "cps": 32.890625, "mean_p": 0.183853, "boundary_entropy": 0.088472, "bpb": 1.863448}, "tr": {"chunks": 3097.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.171077, "cps": 24.195312, "mean_p": 0.182677, "boundary_entropy": 0.129712, "bpb": 2.206129}, "fi": {"chunks": 2985.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158684, "cps": 23.320312, "mean_p": 0.172632, "boundary_entropy": 0.121333, "bpb": 2.371631}, "ru": {"chunks": 6243.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.329446, "cps": 48.773438, "mean_p": 0.194826, "boundary_entropy": 0.124143, "bpb": 1.303021}, "el": {"chunks": 6165.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.300395, "cps": 48.164062, "mean_p": 0.179554, "boundary_entropy": 0.135637, "bpb": 1.375126}, "hi": {"chunks": 4559.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.270966, "cps": 35.617188, "mean_p": 0.109071, "boundary_entropy": 0.154148, "bpb": 1.18032}, "ta": {"chunks": 3565.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.190693, "cps": 27.851562, "mean_p": 0.098452, "boundary_entropy": 0.178352, "bpb": 0.97027}, "th": {"chunks": 3027.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.188188, "cps": 23.648438, "mean_p": 0.115483, "boundary_entropy": 0.224219, "bpb": 1.09271}, "zh": {"chunks": 3336.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.564849, "cps": 26.0625, "mean_p": 0.262884, "boundary_entropy": 0.320754, "bpb": 3.122985}, "my": {"chunks": 3124.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.164872, "cps": 24.40625, "mean_p": 0.090443, "boundary_entropy": 0.172309, "bpb": 1.105498}}, "bpb": {"en": 2.00625, "de": 2.21629, "vi": 1.86345, "tr": 2.20613, "fi": 2.37163, "ru": 1.30302, "el": 1.37513, "hi": 1.18032, "ta": 0.97027, "th": 1.09271, "zh": 3.12298, "my": 1.1055}, "gini_cpc": 0.220165, "gini_cps": 0.149164, "bpic": 8.257262, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.962312087170534, "vi": 1.3761503508246011, "tr": 1.030902481996296, "fi": 0.9562234354537944, "ru": 1.9852310168909286, "el": 1.8101691832000388, "hi": 1.6328318028589892, "ta": 1.149108391077354, "th": 1.1340136627008353, "zh": 3.403764705220195, "my": 0.9935153477340228}, "t": 309.6} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.01965, "de": 2.22746, "vi": 1.86563, "tr": 2.21115, "fi": 2.38162, "ru": 1.28154, "el": 1.36326, "hi": 1.17577, "ta": 0.97777, "th": 1.1104, "zh": 3.11845, "my": 1.1142}, "bytes_per_chunk": {"en": 5.739, "de": 6.304, "vi": 5.707, "tr": 6.295, "fi": 6.36, "ru": 5.582, "el": 6.159, "hi": 8.787, "ta": 13.627, "th": 14.816, "zh": 5.053, "my": 16.043}, "bpb_mean": 1.73724, "t": 310.8} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "none", "lm_loss": 1.34205, "bpb": 1.93616, "chunk_loss": 0.0, "F": 0.15405, "G": 0.17532, "chunks": 157.75, "bytes_per_chunk": 6.491, "boundary_entropy": 0.13248, "router_noise": 0.0, "grad_norm": 0.7368, "lr": 3.08e-05, "router_trainable": false, "sec_per_step": 0.083, "alpha": 0.0, "t": 310.8} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "none", "lm_loss": 1.3039, "bpb": 1.88113, "chunk_loss": 0.0, "F": 0.15701, "G": 0.1796, "chunks": 160.78, "bytes_per_chunk": 6.369, "boundary_entropy": 0.1447, "router_noise": 0.0, "grad_norm": 0.7161, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 312.8} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "none", "lm_loss": 1.2143, "bpb": 1.75187, "chunk_loss": 0.0, "F": 0.15042, "G": 0.17357, "chunks": 154.03, "bytes_per_chunk": 6.648, "boundary_entropy": 0.13769, "router_noise": 0.0, "grad_norm": 0.7212, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 314.7} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "none", "lm_loss": 1.23504, "bpb": 1.78178, "chunk_loss": 0.0, "F": 0.15094, "G": 0.17328, "chunks": 154.56, "bytes_per_chunk": 6.625, "boundary_entropy": 0.13639, "router_noise": 0.0, "grad_norm": 0.7632, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.039, "alpha": 0.0, "t": 316.7} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.183627, "boundary_entropy": 0.089455, "bpb": 2.004428}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.177202, "boundary_entropy": 0.108205, "bpb": 2.213324}, "vi": {"chunks": 4213.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228533, "cps": 32.914062, "mean_p": 0.183982, "boundary_entropy": 0.089736, "bpb": 1.861369}, "tr": {"chunks": 3087.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170524, "cps": 24.117188, "mean_p": 0.182437, "boundary_entropy": 0.12951, "bpb": 2.20494}, "fi": {"chunks": 2989.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158896, "cps": 23.351562, "mean_p": 0.17257, "boundary_entropy": 0.120876, "bpb": 2.367539}, "ru": {"chunks": 6242.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.329393, "cps": 48.765625, "mean_p": 0.194287, "boundary_entropy": 0.125397, "bpb": 1.301157}, "el": {"chunks": 6159.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.300102, "cps": 48.117188, "mean_p": 0.179594, "boundary_entropy": 0.136297, "bpb": 1.370404}, "hi": {"chunks": 4568.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.271501, "cps": 35.6875, "mean_p": 0.111962, "boundary_entropy": 0.152452, "bpb": 1.177103}, "ta": {"chunks": 3390.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.181332, "cps": 26.484375, "mean_p": 0.095591, "boundary_entropy": 0.17573, "bpb": 0.968372}, "th": {"chunks": 2953.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.183587, "cps": 23.070312, "mean_p": 0.116292, "boundary_entropy": 0.229397, "bpb": 1.091369}, "zh": {"chunks": 3475.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.588385, "cps": 27.148438, "mean_p": 0.266842, "boundary_entropy": 0.317486, "bpb": 3.116364}, "my": {"chunks": 3174.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.167511, "cps": 24.796875, "mean_p": 0.091411, "boundary_entropy": 0.170857, "bpb": 1.10566}}, "bpb": {"en": 2.00443, "de": 2.21332, "vi": 1.86137, "tr": 2.20494, "fi": 2.36754, "ru": 1.30116, "el": 1.3704, "hi": 1.1771, "ta": 0.96837, "th": 1.09137, "zh": 3.11636, "my": 1.10566}, "gini_cpc": 0.226215, "gini_cps": 0.14981, "bpic": 8.268113, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.962312087170534, "vi": 1.3771309805282765, "tr": 1.0275737687835214, "fi": 0.9575048068915885, "ru": 1.9849130237759374, "el": 1.808407461367241, "hi": 1.6360552040929726, "ta": 1.092700545793052, "th": 1.1062908311713135, "zh": 3.5455882346043697, "my": 1.0094166817246442}, "t": 319.6} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.01784, "de": 2.22347, "vi": 1.86341, "tr": 2.21028, "fi": 2.37779, "ru": 1.28057, "el": 1.35988, "hi": 1.17251, "ta": 0.97578, "th": 1.10927, "zh": 3.11138, "my": 1.11448}, "bytes_per_chunk": {"en": 5.739, "de": 6.305, "vi": 5.711, "tr": 6.318, "fi": 6.36, "ru": 5.578, "el": 6.178, "hi": 8.759, "ta": 14.372, "th": 15.045, "zh": 4.797, "my": 15.674}, "bpb_mean": 1.73472, "t": 320.8} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.183627, "boundary_entropy": 0.089455, "bpb": 2.004428}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.177202, "boundary_entropy": 0.108205, "bpb": 2.213324}, "vi": {"chunks": 4213.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228533, "cps": 32.914062, "mean_p": 0.183982, "boundary_entropy": 0.089736, "bpb": 1.861369}, "tr": {"chunks": 3087.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170524, "cps": 24.117188, "mean_p": 0.182437, "boundary_entropy": 0.12951, "bpb": 2.20494}, "fi": {"chunks": 2989.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158896, "cps": 23.351562, "mean_p": 0.17257, "boundary_entropy": 0.120876, "bpb": 2.367539}, "ru": {"chunks": 6242.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.329393, "cps": 48.765625, "mean_p": 0.194287, "boundary_entropy": 0.125397, "bpb": 1.301157}, "el": {"chunks": 6159.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.300102, "cps": 48.117188, "mean_p": 0.179594, "boundary_entropy": 0.136297, "bpb": 1.370404}, "hi": {"chunks": 4568.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.271501, "cps": 35.6875, "mean_p": 0.111962, "boundary_entropy": 0.152452, "bpb": 1.177103}, "ta": {"chunks": 3390.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.181332, "cps": 26.484375, "mean_p": 0.095591, "boundary_entropy": 0.17573, "bpb": 0.968372}, "th": {"chunks": 2953.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.183587, "cps": 23.070312, "mean_p": 0.116292, "boundary_entropy": 0.229397, "bpb": 1.091369}, "zh": {"chunks": 3475.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.588385, "cps": 27.148438, "mean_p": 0.266842, "boundary_entropy": 0.317486, "bpb": 3.116364}, "my": {"chunks": 3174.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.167511, "cps": 24.796875, "mean_p": 0.091411, "boundary_entropy": 0.170857, "bpb": 1.10566}}, "bpb": {"en": 2.00443, "de": 2.21332, "vi": 1.86137, "tr": 2.20494, "fi": 2.36754, "ru": 1.30116, "el": 1.3704, "hi": 1.1771, "ta": 0.96837, "th": 1.09137, "zh": 3.11636, "my": 1.10566}, "gini_cpc": 0.226215, "gini_cps": 0.14981, "bpic": 8.268113, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.962312087170534, "vi": 1.3771309805282765, "tr": 1.0275737687835214, "fi": 0.9575048068915885, "ru": 1.9849130237759374, "el": 1.808407461367241, "hi": 1.6360552040929726, "ta": 1.092700545793052, "th": 1.1062908311713135, "zh": 3.5455882346043697, "my": 1.0094166817246442}, "t": 321.9} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_B_s1/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_B_s1/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_B_s1/th0.60"], "t": 323.3} |
| {"event": "final", "run": "pilot_B_s1", "condition": "B", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.16594839970120095, "cps": 22.5625, "bpb": 2.025561883633651, "bytes_per_chunk": 5.738452782277483}, "de": {"cpc": 0.1596941508790727, "cps": 25.6171875, "bpb": 2.2150645576246735, "bytes_per_chunk": 6.295182748186926}, "vi": {"cpc": 0.22853268239761323, "cps": 32.9140625, "bpb": 1.8667046584639686, "bytes_per_chunk": 5.700703497917777}, "tr": {"cpc": 0.17052422250455726, "cps": 24.1171875, "bpb": 2.2021629243559824, "bytes_per_chunk": 6.352157214340235}, "fi": {"cpc": 0.15889639040986656, "cps": 23.3515625, "bpb": 2.392972709489225, "bytes_per_chunk": 6.338794114446688}, "ru": {"cpc": 0.32939313984168866, "cps": 48.765625, "bpb": 1.2800927215140925, "bytes_per_chunk": 5.570302373515225}, "el": {"cpc": 0.30010232422160504, "cps": 48.1171875, "bpb": 1.3561542255135945, "bytes_per_chunk": 6.20202283078015}, "hi": {"cpc": 0.2715007429420505, "cps": 35.6875, "bpb": 1.1812027026424354, "bytes_per_chunk": 8.817787345689394}, "ta": {"cpc": 0.18133190692698584, "cps": 26.484375, "bpb": 0.9769356889124318, "bytes_per_chunk": 14.478694319405706}, "th": {"cpc": 0.18358719303699098, "cps": 23.0703125, "bpb": 1.1062575272807682, "bytes_per_chunk": 14.971102227298687}, "zh": {"cpc": 0.5883846935320014, "cps": 27.1484375, "bpb": 3.1204771658389325, "bytes_per_chunk": 4.7911686222904555}, "my": {"cpc": 0.16751108296390121, "cps": 24.796875, "bpb": 1.123622013599108, "bytes_per_chunk": 15.603345138537543}}, "gini_cpc": 0.2262152618690665, "gini_cps": 0.14981041428786868, "bpic": 8.268112975849366, "seed": 1, "mean_boundary_entropy": 0.15568843483924866, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584}], "t": 325.6} |
|
|
