Blame for log.jsonl

{"event": "start", "run": "pilot_C_s1", "condition": "C", "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": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 1.3}
{"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": 3.1}
{"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": 7.3206, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 3.4}
{"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": 8.011944}, "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": 8.04201}, "vi": {"chunks": 14105.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.765121, "cps": 110.195312, "mean_p": 0.507011, "boundary_entropy": 0.688289, "bpb": 7.873429}, "tr": {"chunks": 5611.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.309949, "cps": 43.835938, "mean_p": 0.48594, "boundary_entropy": 0.686368, "bpb": 7.996299}, "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": 8.050098}, "ru": {"chunks": 17480.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.922427, "cps": 136.5625, "mean_p": 0.502276, "boundary_entropy": 0.690066, "bpb": 8.084058}, "el": {"chunks": 31424.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.53116, "cps": 245.5, "mean_p": 0.517879, "boundary_entropy": 0.68971, "bpb": 7.792261}, "hi": {"chunks": 1660.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.098663, "cps": 12.96875, "mean_p": 0.476276, "boundary_entropy": 0.689496, "bpb": 8.025216}, "ta": {"chunks": 39578.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.117037, "cps": 309.203125, "mean_p": 0.514531, "boundary_entropy": 0.690545, "bpb": 8.080304}, "th": {"chunks": 44407.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.760771, "cps": 346.929688, "mean_p": 0.52071, "boundary_entropy": 0.69023, "bpb": 8.182809}, "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": 8.116865}, "my": {"chunks": 7051.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.372124, "cps": 55.085938, "mean_p": 0.49115, "boundary_entropy": 0.691021, "bpb": 8.097577}}, "bpb": {"en": 8.01194, "de": 8.04201, "vi": 7.87343, "tr": 7.9963, "fi": 8.0501, "ru": 8.08406, "el": 7.79226, "hi": 8.02522, "ta": 8.0803, "th": 8.18281, "zh": 8.11687, "my": 8.09758}, "gini_cpc": 0.50776, "gini_cps": 0.501327, "bpic": 2.089889, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.651971123407348, "vi": 5.020888176263182, "tr": 2.0339502114971566, "fi": 1.6381995611983071, "ru": 6.053169211666805, "el": 10.047805541206058, "hi": 0.647446094679728, "ta": 13.892454245878406, "th": 18.116778300234753, "zh": 14.369988990412386, "my": 2.4419565885008456}, "t": 4.8}
{"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2631.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.151181, "cps": 20.554688, "mean_p": 0.477867, "boundary_entropy": 0.685416, "bpb": 7.961926}, "de": {"chunks": 2014.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.098086, "cps": 15.734375, "mean_p": 0.469405, "boundary_entropy": 0.685239, "bpb": 7.99997}, "vi": {"chunks": 13976.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.758123, "cps": 109.1875, "mean_p": 0.506724, "boundary_entropy": 0.688298, "bpb": 7.840344}, "tr": {"chunks": 5558.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.307021, "cps": 43.421875, "mean_p": 0.485757, "boundary_entropy": 0.686358, "bpb": 7.961989}, "fi": {"chunks": 4632.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.246239, "cps": 36.1875, "mean_p": 0.484106, "boundary_entropy": 0.686566, "bpb": 8.009839}, "ru": {"chunks": 17252.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.910396, "cps": 134.78125, "mean_p": 0.502047, "boundary_entropy": 0.690066, "bpb": 8.040021}, "el": {"chunks": 31366.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.528334, "cps": 245.046875, "mean_p": 0.517784, "boundary_entropy": 0.689716, "bpb": 7.769483}, "hi": {"chunks": 1669.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.099198, "cps": 13.039062, "mean_p": 0.476368, "boundary_entropy": 0.689506, "bpb": 7.983978}, "ta": {"chunks": 39401.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.107569, "cps": 307.820312, "mean_p": 0.514419, "boundary_entropy": 0.69055, "bpb": 8.042722}, "th": {"chunks": 44381.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.759154, "cps": 346.726562, "mean_p": 0.520516, "boundary_entropy": 0.690244, "bpb": 8.123064}, "zh": {"chunks": 12928.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.18896, "cps": 101.0, "mean_p": 0.52826, "boundary_entropy": 0.684757, "bpb": 8.106759}, "my": {"chunks": 7073.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.373285, "cps": 55.257812, "mean_p": 0.491161, "boundary_entropy": 0.691021, "bpb": 8.084054}}, "bpb": {"en": 7.96193, "de": 7.99997, "vi": 7.84034, "tr": 7.96199, "fi": 8.00984, "ru": 8.04002, "el": 7.76948, "hi": 7.98398, "ta": 8.04272, "th": 8.12306, "zh": 8.10676, "my": 8.08405}, "gini_cpc": 0.50878, "gini_cps": 0.501997, "bpic": 2.098529, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6487992380483157, "vi": 5.014677660330188, "tr": 2.0308192113152597, "fi": 1.6287706670088653, "ru": 6.0218995556331105, "el": 10.109311181920043, "hi": 0.6561521192909098, "ta": 13.940714945520163, "th": 18.250690087473583, "zh": 14.479086841705922, "my": 2.4691277141135712}, "t": 6.3}
{"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2557.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.146929, "cps": 19.976562, "mean_p": 0.477478, "boundary_entropy": 0.685387, "bpb": 7.79171}, "de": {"chunks": 1968.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.095846, "cps": 15.375, "mean_p": 0.468985, "boundary_entropy": 0.685193, "bpb": 7.852599}, "vi": {"chunks": 13540.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.734472, "cps": 105.78125, "mean_p": 0.505671, "boundary_entropy": 0.688326, "bpb": 7.729999}, "tr": {"chunks": 5388.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.29763, "cps": 42.09375, "mean_p": 0.485237, "boundary_entropy": 0.686332, "bpb": 7.842935}, "fi": {"chunks": 4455.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.23683, "cps": 34.804688, "mean_p": 0.483396, "boundary_entropy": 0.686515, "bpb": 7.870379}, "ru": {"chunks": 16097.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.849446, "cps": 125.757812, "mean_p": 0.500927, "boundary_entropy": 0.690065, "bpb": 7.865187}, "el": {"chunks": 31135.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.517078, "cps": 243.242188, "mean_p": 0.517302, "boundary_entropy": 0.689748, "bpb": 7.685975}, "hi": {"chunks": 1730.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.102823, "cps": 13.515625, "mean_p": 0.476709, "boundary_entropy": 0.689543, "bpb": 7.829456}, "ta": {"chunks": 38745.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.072479, "cps": 302.695312, "mean_p": 0.513991, "boundary_entropy": 0.69057, "bpb": 7.91292}, "th": {"chunks": 44277.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.752689, "cps": 345.914062, "mean_p": 0.519897, "boundary_entropy": 0.69029, "bpb": 7.955002}, "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": 8.079666}, "my": {"chunks": 7122.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.375871, "cps": 55.640625, "mean_p": 0.49119, "boundary_entropy": 0.691022, "bpb": 8.001211}}, "bpb": {"en": 7.79171, "de": 7.8526, "vi": 7.73, "tr": 7.84293, "fi": 7.87038, "ru": 7.86519, "el": 7.68598, "hi": 7.82946, "ta": 7.91292, "th": 7.955, "zh": 8.07967, "my": 8.00121}, "gini_cpc": 0.512209, "gini_cps": 0.503659, "bpic": 2.132863, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6523280884338518, "vi": 4.998836296476146, "tr": 2.0256780762909914, "fi": 1.6118670486425006, "ru": 5.7813481332737595, "el": 10.32527000097242, "hi": 0.6998168939850459, "ta": 14.105340938555994, "th": 18.73486268745588, "zh": 14.892352036950685, "my": 2.558185219924295}, "t": 8.0}
{"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": 0.476347, "boundary_entropy": 0.685317, "bpb": 7.297454}, "de": {"chunks": 1784.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.086885, "cps": 13.9375, "mean_p": 0.467596, "boundary_entropy": 0.685051, "bpb": 7.408973}, "vi": {"chunks": 12231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.663466, "cps": 95.554688, "mean_p": 0.502574, "boundary_entropy": 0.688389, "bpb": 7.434503}, "tr": {"chunks": 4854.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.268132, "cps": 37.921875, "mean_p": 0.483425, "boundary_entropy": 0.686243, "bpb": 7.494894}, "fi": {"chunks": 3816.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.20286, "cps": 29.8125, "mean_p": 0.480873, "boundary_entropy": 0.68633, "bpb": 7.453883}, "ru": {"chunks": 12099.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.63847, "cps": 94.523438, "mean_p": 0.496808, "boundary_entropy": 0.69002, "bpb": 7.244503}, "el": {"chunks": 30500.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.486138, "cps": 238.28125, "mean_p": 0.516119, "boundary_entropy": 0.689821, "bpb": 7.512205}, "hi": {"chunks": 1892.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.112452, "cps": 14.78125, "mean_p": 0.477696, "boundary_entropy": 0.689646, "bpb": 7.380599}, "ta": {"chunks": 34880.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.86574, "cps": 272.5, "mean_p": 0.511349, "boundary_entropy": 0.690681, "bpb": 7.492788}, "th": {"chunks": 43413.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.698974, "cps": 339.164062, "mean_p": 0.517056, "boundary_entropy": 0.690483, "bpb": 7.519766}, "zh": {"chunks": 12916.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.186929, "cps": 100.90625, "mean_p": 0.528222, "boundary_entropy": 0.684759, "bpb": 7.97214}, "my": {"chunks": 7554.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.39867, "cps": 59.015625, "mean_p": 0.491562, "boundary_entropy": 0.691038, "bpb": 7.746589}}, "bpb": {"en": 7.29745, "de": 7.40897, "vi": 7.4345, "tr": 7.49489, "fi": 7.45388, "ru": 7.2445, "el": 7.5122, "hi": 7.3806, "ta": 7.49279, "th": 7.51977, "zh": 7.97214, "my": 7.74659}, "gini_cpc": 0.522773, "gini_cps": 0.507407, "bpic": 2.281165, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6574136649592702, "vi": 5.020131672975555, "tr": 2.0288292836402326, "fi": 1.5349448172091722, "ru": 4.83099453940576, "el": 11.244891733346892, "hi": 0.8508682989857227, "ta": 14.117158505529263, "th": 20.421846954359314, "zh": 16.547442394617118, "my": 3.0165455571770794}, "t": 9.8}
{"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2293.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.131759, "cps": 17.914062, "mean_p": 0.478674, "boundary_entropy": 0.685749, "bpb": 6.63594}, "de": {"chunks": 1720.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.083768, "cps": 13.4375, "mean_p": 0.470347, "boundary_entropy": 0.685666, "bpb": 6.756497}, "vi": {"chunks": 9740.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.528343, "cps": 76.09375, "mean_p": 0.496629, "boundary_entropy": 0.688418, "bpb": 7.048651}, "tr": {"chunks": 4403.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.243219, "cps": 34.398438, "mean_p": 0.482546, "boundary_entropy": 0.686344, "bpb": 7.014899}, "fi": {"chunks": 2968.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15778, "cps": 23.1875, "mean_p": 0.478142, "boundary_entropy": 0.686222, "bpb": 6.85417}, "ru": {"chunks": 6002.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.316728, "cps": 46.890625, "mean_p": 0.487676, "boundary_entropy": 0.689678, "bpb": 6.435858}, "el": {"chunks": 23101.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.125615, "cps": 180.476562, "mean_p": 0.50639, "boundary_entropy": 0.69025, "bpb": 6.833865}, "hi": {"chunks": 3520.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.209212, "cps": 27.5, "mean_p": 0.483431, "boundary_entropy": 0.690145, "bpb": 6.500074}, "ta": {"chunks": 21624.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.156673, "cps": 168.9375, "mean_p": 0.498122, "boundary_entropy": 0.690847, "bpb": 6.58499}, "th": {"chunks": 24801.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.541871, "cps": 193.757812, "mean_p": 0.50306, "boundary_entropy": 0.690961, "bpb": 6.461709}, "zh": {"chunks": 13086.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.215713, "cps": 102.234375, "mean_p": 0.52884, "boundary_entropy": 0.684759, "bpb": 7.681181}, "my": {"chunks": 8521.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.449704, "cps": 66.570312, "mean_p": 0.492476, "boundary_entropy": 0.691083, "bpb": 6.949068}}, "bpb": {"en": 6.63594, "de": 6.7565, "vi": 7.04865, "tr": 7.0149, "fi": 6.85417, "ru": 6.43586, "el": 6.83387, "hi": 6.50007, "ta": 6.58499, "th": 6.46171, "zh": 7.68118, "my": 6.94907}, "gini_cpc": 0.497373, "gini_cps": 0.436662, "bpic": 3.151455, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.635764252917065, "vi": 4.0099215889304025, "tr": 1.8459426301197108, "fi": 1.1974905052608262, "ru": 2.403847110685613, "el": 8.542992014809249, "hi": 1.5878433555449138, "ta": 8.77870850584596, "th": 11.702218222778274, "zh": 16.81641973709647, "my": 3.41308618321532}, "t": 11.9}
{"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3539.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203356, "cps": 27.648438, "mean_p": 0.488237, "boundary_entropy": 0.686849, "bpb": 6.226757}, "de": {"chunks": 3112.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151561, "cps": 24.3125, "mean_p": 0.483349, "boundary_entropy": 0.687367, "bpb": 6.332953}, "vi": {"chunks": 6121.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.332031, "cps": 47.820312, "mean_p": 0.489068, "boundary_entropy": 0.688305, "bpb": 6.743709}, "tr": {"chunks": 4961.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.274043, "cps": 38.757812, "mean_p": 0.487125, "boundary_entropy": 0.687061, "bpb": 6.642056}, "fi": {"chunks": 3338.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177449, "cps": 26.078125, "mean_p": 0.483104, "boundary_entropy": 0.686944, "bpb": 6.409974}, "ru": {"chunks": 2384.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.125805, "cps": 18.625, "mean_p": 0.482019, "boundary_entropy": 0.689461, "bpb": 5.796463}, "el": {"chunks": 1344.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.065488, "cps": 10.5, "mean_p": 0.473752, "boundary_entropy": 0.688754, "bpb": 6.225464}, "hi": {"chunks": 4974.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.295632, "cps": 38.859375, "mean_p": 0.488172, "boundary_entropy": 0.690518, "bpb": 5.769182}, "ta": {"chunks": 4405.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.235624, "cps": 34.414062, "mean_p": 0.482215, "boundary_entropy": 0.6903, "bpb": 6.128234}, "th": {"chunks": 4758.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.295804, "cps": 37.171875, "mean_p": 0.489154, "boundary_entropy": 0.690751, "bpb": 5.553382}, "zh": {"chunks": 4842.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.819844, "cps": 37.828125, "mean_p": 0.491142, "boundary_entropy": 0.686035, "bpb": 7.308059}, "my": {"chunks": 4180.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.220604, "cps": 32.65625, "mean_p": 0.489087, "boundary_entropy": 0.69097, "bpb": 5.723713}}, "bpb": {"en": 6.22676, "de": 6.33295, "vi": 6.74371, "tr": 6.64206, "fi": 6.40997, "ru": 5.79646, "el": 6.22546, "hi": 5.76918, "ta": 6.12823, "th": 5.55338, "zh": 7.30806, "my": 5.72371}, "gini_cpc": 0.312292, "gini_cps": 0.174114, "bpic": 8.00244, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.7452993435594143, "vi": 1.632761664686714, "tr": 1.347603819324219, "fi": 0.8726056214906391, "ru": 0.6186436995678012, "el": 0.32203418883839474, "hi": 1.4537651900054873, "ta": 1.1586813076859306, "th": 1.4546112095401629, "zh": 4.031576453445507, "my": 1.084816952367011}, "t": 14.3}
{"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.18179, "bpb": 6.03305, "chunk_loss": 0.028848, "F": 0.14883, "G": 0.48507, "chunks": 152.41, "bytes_per_chunk": 6.719, "boundary_entropy": 0.69151, "router_noise": 0.0, "grad_norm": 1.6412, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 15.3}
{"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5782.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.332242, "cps": 45.171875, "mean_p": 0.496587, "boundary_entropy": 0.687366, "bpb": 5.700707}, "de": {"chunks": 5723.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.278722, "cps": 44.710938, "mean_p": 0.493197, "boundary_entropy": 0.688118, "bpb": 5.815287}, "vi": {"chunks": 4982.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.270247, "cps": 38.921875, "mean_p": 0.48571, "boundary_entropy": 0.688067, "bpb": 6.193767}, "tr": {"chunks": 4493.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.248191, "cps": 35.101562, "mean_p": 0.487258, "boundary_entropy": 0.687242, "bpb": 6.161315}, "fi": {"chunks": 4482.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.238265, "cps": 35.015625, "mean_p": 0.489507, "boundary_entropy": 0.687551, "bpb": 5.889532}, "ru": {"chunks": 9880.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.521372, "cps": 77.1875, "mean_p": 0.494674, "boundary_entropy": 0.690222, "bpb": 5.366178}, "el": {"chunks": 2398.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.116845, "cps": 18.734375, "mean_p": 0.484165, "boundary_entropy": 0.689908, "bpb": 5.639318}, "hi": {"chunks": 8923.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.530342, "cps": 69.710938, "mean_p": 0.491209, "boundary_entropy": 0.690636, "bpb": 5.210893}, "ta": {"chunks": 9734.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.520674, "cps": 76.046875, "mean_p": 0.486957, "boundary_entropy": 0.690595, "bpb": 5.234915}, "th": {"chunks": 2675.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.166304, "cps": 20.898438, "mean_p": 0.48666, "boundary_entropy": 0.690635, "bpb": 5.048885}, "zh": {"chunks": 2599.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.440061, "cps": 20.304688, "mean_p": 0.484989, "boundary_entropy": 0.686178, "bpb": 6.980692}, "my": {"chunks": 25781.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.360619, "cps": 201.414062, "mean_p": 0.499834, "boundary_entropy": 0.691347, "bpb": 4.951099}}, "bpb": {"en": 5.70071, "de": 5.81529, "vi": 6.19377, "tr": 6.16131, "fi": 5.88953, "ru": 5.36618, "el": 5.63932, "hi": 5.21089, "ta": 5.23491, "th": 5.04889, "zh": 6.98069, "my": 4.9511}, "gini_cpc": 0.346097, "gini_cps": 0.380178, "bpic": 4.388476, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8389138638945569, "vi": 0.8134045812612807, "tr": 0.7470194335335412, "fi": 0.7171433929759998, "ru": 1.5692558746140555, "el": 0.351685422062259, "hi": 1.5962534647928728, "ta": 1.5671548290759314, "th": 0.5005514847994381, "zh": 1.3245210652347401, "my": 4.095268828004361}, "t": 17.1}
{"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.3387, "bpb": 4.81673, "chunk_loss": 0.032543, "F": 0.20331, "G": 0.48392, "chunks": 208.19, "bytes_per_chunk": 4.919, "boundary_entropy": 0.69112, "router_noise": 0.0, "grad_norm": 1.5938, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.079, "alpha": 0.03, "t": 19.2}
{"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2715.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.156008, "cps": 21.210938, "mean_p": 0.47475, "boundary_entropy": 0.685083, "bpb": 4.482317}, "de": {"chunks": 3245.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.158038, "cps": 25.351562, "mean_p": 0.473739, "boundary_entropy": 0.685689, "bpb": 4.67298}, "vi": {"chunks": 5543.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.300678, "cps": 43.304688, "mean_p": 0.477864, "boundary_entropy": 0.685176, "bpb": 4.932839}, "tr": {"chunks": 3377.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.186544, "cps": 26.382812, "mean_p": 0.470891, "boundary_entropy": 0.684429, "bpb": 4.975747}, "fi": {"chunks": 2485.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.132104, "cps": 19.414062, "mean_p": 0.474055, "boundary_entropy": 0.685658, "bpb": 4.753759}, "ru": {"chunks": 5156.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.272084, "cps": 40.28125, "mean_p": 0.478192, "boundary_entropy": 0.687743, "bpb": 3.650538}, "el": {"chunks": 4685.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.22828, "cps": 36.601562, "mean_p": 0.48079, "boundary_entropy": 0.689261, "bpb": 4.448169}, "hi": {"chunks": 3569.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.212125, "cps": 27.882812, "mean_p": 0.481384, "boundary_entropy": 0.689923, "bpb": 3.448518}, "ta": {"chunks": 6197.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.331479, "cps": 48.414062, "mean_p": 0.480998, "boundary_entropy": 0.689831, "bpb": 3.594822}, "th": {"chunks": 10834.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.673547, "cps": 84.640625, "mean_p": 0.480509, "boundary_entropy": 0.689149, "bpb": 3.340747}, "zh": {"chunks": 2517.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.426177, "cps": 19.664062, "mean_p": 0.483815, "boundary_entropy": 0.686003, "bpb": 6.267792}, "my": {"chunks": 12782.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.674583, "cps": 99.859375, "mean_p": 0.489728, "boundary_entropy": 0.690784, "bpb": 3.994131}}, "bpb": {"en": 4.48232, "de": 4.67298, "vi": 4.93284, "tr": 4.97575, "fi": 4.75376, "ru": 3.65054, "el": 4.44817, "hi": 3.44852, "ta": 3.59482, "th": 3.34075, "zh": 6.26779, "my": 3.99413}, "gini_cpc": 0.300673, "gini_cps": 0.304304, "bpic": 6.081626, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0130166423918592, "vi": 1.92732974000033, "tr": 1.1957345978795981, "fi": 0.846776498175659, "ru": 1.7440461814312165, "el": 1.4632651737183158, "hi": 1.3597083398323662, "ta": 2.124762108815694, "th": 4.317397664685364, "zh": 2.7317695460932137, "my": 4.3240402054124285}, "t": 21.6}
{"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.91709, "bpb": 4.20848, "chunk_loss": 0.03018, "F": 0.16965, "G": 0.46173, "chunks": 173.72, "bytes_per_chunk": 5.895, "boundary_entropy": 0.68535, "router_noise": 0.0, "grad_norm": 1.6716, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.073, "alpha": 0.03, "t": 22.9}
{"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3576.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.205482, "cps": 27.9375, "mean_p": 0.372777, "boundary_entropy": 0.544916, "bpb": 3.877723}, "de": {"chunks": 3981.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.193883, "cps": 31.101562, "mean_p": 0.350828, "boundary_entropy": 0.54017, "bpb": 4.089538}, "vi": {"chunks": 7358.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.399132, "cps": 57.484375, "mean_p": 0.431657, "boundary_entropy": 0.569742, "bpb": 3.651378}, "tr": {"chunks": 4926.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.27211, "cps": 38.484375, "mean_p": 0.383854, "boundary_entropy": 0.573822, "bpb": 4.133699}, "fi": {"chunks": 3848.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.204561, "cps": 30.0625, "mean_p": 0.361295, "boundary_entropy": 0.567272, "bpb": 4.131899}, "ru": {"chunks": 5967.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.314881, "cps": 46.617188, "mean_p": 0.43055, "boundary_entropy": 0.665582, "bpb": 2.824642}, "el": {"chunks": 8366.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.40764, "cps": 65.359375, "mean_p": 0.452858, "boundary_entropy": 0.678113, "bpb": 2.981847}, "hi": {"chunks": 7764.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.461456, "cps": 60.65625, "mean_p": 0.481461, "boundary_entropy": 0.688168, "bpb": 2.451144}, "ta": {"chunks": 13974.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.747473, "cps": 109.171875, "mean_p": 0.443281, "boundary_entropy": 0.675847, "bpb": 2.381503}, "th": {"chunks": 3680.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.228785, "cps": 28.75, "mean_p": 0.455727, "boundary_entropy": 0.681391, "bpb": 2.334986}, "zh": {"chunks": 2914.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.493397, "cps": 22.765625, "mean_p": 0.457425, "boundary_entropy": 0.677562, "bpb": 5.260118}, "my": {"chunks": 10359.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.546707, "cps": 80.929688, "mean_p": 0.376521, "boundary_entropy": 0.640169, "bpb": 2.465449}}, "bpb": {"en": 3.87772, "de": 4.08954, "vi": 3.65138, "tr": 4.1337, "fi": 4.1319, "ru": 2.82464, "el": 2.98185, "hi": 2.45114, "ta": 2.3815, "th": 2.33499, "zh": 5.26012, "my": 2.46545}, "gini_cpc": 0.239049, "gini_cps": 0.267513, "bpic": 5.002816, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9435531753271947, "vi": 1.9424204942815588, "tr": 1.32425147747027, "fi": 0.9955195428635037, "ru": 1.5324045527793029, "el": 1.983826216256961, "hi": 2.2457275347288013, "ta": 3.6376581148101548, "th": 1.1134055030636323, "zh": 2.401168928424512, "my": 2.6606090689035202}, "t": 26.2}
{"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.88109, "de": 4.0892, "vi": 3.65456, "tr": 4.14384, "fi": 4.13267, "ru": 2.80958, "el": 2.97551, "hi": 2.44697, "ta": 2.39878, "th": 2.36825, "zh": 5.27032, "my": 2.48183}, "bytes_per_chunk": {"en": 5.023, "de": 5.313, "vi": 3.245, "tr": 4.081, "fi": 5.085, "ru": 5.708, "el": 4.682, "hi": 5.761, "ta": 3.714, "th": 11.907, "zh": 5.453, "my": 5.099}, "bpb_mean": 3.38772, "t": 27.6}
{"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.53667, "bpb": 3.65964, "chunk_loss": 0.031904, "F": 0.20346, "G": 0.39577, "chunks": 208.34, "bytes_per_chunk": 4.915, "boundary_entropy": 0.59204, "router_noise": 0.0, "grad_norm": 1.2222, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.095, "alpha": 0.03, "t": 27.6}
{"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.18804, "bpb": 3.15667, "chunk_loss": 0.029834, "F": 0.16544, "G": 0.34261, "chunks": 169.41, "bytes_per_chunk": 6.045, "boundary_entropy": 0.51687, "router_noise": 0.0, "grad_norm": 1.1053, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 29.9}
{"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3024.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.173763, "cps": 23.625, "mean_p": 0.28449, "boundary_entropy": 0.382427, "bpb": 3.57941}, "de": {"chunks": 3583.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.1745, "cps": 27.992188, "mean_p": 0.269079, "boundary_entropy": 0.389816, "bpb": 3.846663}, "vi": {"chunks": 7144.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.387524, "cps": 55.8125, "mean_p": 0.3568, "boundary_entropy": 0.444701, "bpb": 3.195352}, "tr": {"chunks": 4259.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.235265, "cps": 33.273438, "mean_p": 0.292809, "boundary_entropy": 0.428917, "bpb": 3.898855}, "fi": {"chunks": 3688.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.196055, "cps": 28.8125, "mean_p": 0.274431, "boundary_entropy": 0.420231, "bpb": 3.952397}, "ru": {"chunks": 8898.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.469551, "cps": 69.515625, "mean_p": 0.322619, "boundary_entropy": 0.528001, "bpb": 2.656169}, "el": {"chunks": 5273.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.256931, "cps": 41.195312, "mean_p": 0.372671, "boundary_entropy": 0.631817, "bpb": 2.858066}, "hi": {"chunks": 6281.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.373314, "cps": 49.070312, "mean_p": 0.445198, "boundary_entropy": 0.673449, "bpb": 2.161146}, "ta": {"chunks": 9246.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.494571, "cps": 72.234375, "mean_p": 0.442997, "boundary_entropy": 0.66192, "bpb": 1.945902}, "th": {"chunks": 12392.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.770407, "cps": 96.8125, "mean_p": 0.443041, "boundary_entropy": 0.673296, "bpb": 2.092576}, "zh": {"chunks": 3480.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.589231, "cps": 27.1875, "mean_p": 0.436887, "boundary_entropy": 0.659992, "bpb": 4.987597}, "my": {"chunks": 12245.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.646242, "cps": 95.664062, "mean_p": 0.355792, "boundary_entropy": 0.618284, "bpb": 2.198795}}, "bpb": {"en": 3.57941, "de": 3.84666, "vi": 3.19535, "tr": 3.89886, "fi": 3.9524, "ru": 2.65617, "el": 2.85807, "hi": 2.16115, "ta": 1.9459, "th": 2.09258, "zh": 4.9876, "my": 2.19879}, "gini_cpc": 0.271108, "gini_cps": 0.269649, "bpic": 4.826645, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0042381930287985, "vi": 2.230183699915189, "tr": 1.353940009504695, "fi": 1.1282916193075143, "ru": 2.702249968589019, "el": 1.4786291357973131, "hi": 2.1484045024646807, "ta": 2.8462349877099813, "th": 4.4336629315491605, "zh": 3.39100269298373, "my": 3.7190990651942997}, "t": 31.2}
{"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.19993, "bpb": 3.17382, "chunk_loss": 0.031291, "F": 0.2103, "G": 0.31186, "chunks": 215.34, "bytes_per_chunk": 4.755, "boundary_entropy": 0.45763, "router_noise": 0.0, "grad_norm": 0.8413, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.067, "alpha": 0.03, "t": 33.3}
{"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.39524, "bpb": 3.45559, "chunk_loss": 0.031609, "F": 0.20819, "G": 0.31283, "chunks": 213.19, "bytes_per_chunk": 4.803, "boundary_entropy": 0.44076, "router_noise": 0.0, "grad_norm": 0.8464, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 35.5}
{"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3029.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.17405, "cps": 23.664062, "mean_p": 0.26677, "boundary_entropy": 0.324397, "bpb": 3.182764}, "de": {"chunks": 3512.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.171042, "cps": 27.4375, "mean_p": 0.255561, "boundary_entropy": 0.3396, "bpb": 3.444012}, "vi": {"chunks": 4275.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.231896, "cps": 33.398438, "mean_p": 0.281036, "boundary_entropy": 0.345107, "bpb": 2.788893}, "tr": {"chunks": 3992.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.220516, "cps": 31.1875, "mean_p": 0.269896, "boundary_entropy": 0.361855, "bpb": 3.430189}, "fi": {"chunks": 3471.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18452, "cps": 27.117188, "mean_p": 0.25516, "boundary_entropy": 0.357563, "bpb": 3.636505}, "ru": {"chunks": 6969.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.367757, "cps": 54.445312, "mean_p": 0.256465, "boundary_entropy": 0.247528, "bpb": 2.541964}, "el": {"chunks": 8743.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.42601, "cps": 68.304688, "mean_p": 0.294931, "boundary_entropy": 0.320449, "bpb": 2.555983}, "hi": {"chunks": 4995.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.29688, "cps": 39.023438, "mean_p": 0.371724, "boundary_entropy": 0.576156, "bpb": 1.956}, "ta": {"chunks": 10145.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.542658, "cps": 79.257812, "mean_p": 0.332057, "boundary_entropy": 0.509488, "bpb": 1.692436}, "th": {"chunks": 5750.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.357476, "cps": 44.921875, "mean_p": 0.368479, "boundary_entropy": 0.603697, "bpb": 1.93809}, "zh": {"chunks": 4419.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.748222, "cps": 34.523438, "mean_p": 0.353283, "boundary_entropy": 0.525192, "bpb": 4.143967}, "my": {"chunks": 4479.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.236384, "cps": 34.992188, "mean_p": 0.287497, "boundary_entropy": 0.527286, "bpb": 2.010932}}, "bpb": {"en": 3.18276, "de": 3.44401, "vi": 2.78889, "tr": 3.43019, "fi": 3.63651, "ru": 2.54196, "el": 2.55598, "hi": 1.956, "ta": 1.69244, "th": 1.93809, "zh": 4.14397, "my": 2.01093}, "gini_cpc": 0.263858, "gini_cps": 0.210274, "bpic": 6.017357, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9827135559685005, "vi": 1.3323484590644115, "tr": 1.2669656138628036, "fi": 1.060150633258825, "ru": 2.11293480523802, "el": 2.4476227438228197, "hi": 1.7057102785995897, "ta": 3.1178228007393267, "th": 2.053863733366802, "zh": 4.298880826577982, "my": 1.3581337235936088}, "t": 38.6}
{"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.19259, "de": 3.44221, "vi": 2.79364, "tr": 3.44104, "fi": 3.63199, "ru": 2.52918, "el": 2.55012, "hi": 1.95301, "ta": 1.70759, "th": 1.96795, "zh": 4.1538, "my": 2.02651}, "bytes_per_chunk": {"en": 5.648, "de": 5.968, "vi": 5.8, "tr": 4.883, "fi": 5.575, "ru": 4.994, "el": 4.043, "hi": 7.086, "ta": 4.644, "th": 7.069, "zh": 3.93, "my": 12.131}, "bpb_mean": 2.78247, "t": 39.9}
{"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.00512, "bpb": 2.89278, "chunk_loss": 0.030673, "F": 0.18643, "G": 0.29049, "chunks": 190.91, "bytes_per_chunk": 5.364, "boundary_entropy": 0.37329, "router_noise": 0.0, "grad_norm": 1.1395, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 39.9}
{"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.06131, "bpb": 2.97384, "chunk_loss": 0.030278, "F": 0.17746, "G": 0.27218, "chunks": 181.72, "bytes_per_chunk": 5.635, "boundary_entropy": 0.36822, "router_noise": 0.0, "grad_norm": 0.9951, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 42.2}
{"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.05651, "bpb": 2.96692, "chunk_loss": 0.030242, "F": 0.17599, "G": 0.26587, "chunks": 180.22, "bytes_per_chunk": 5.682, "boundary_entropy": 0.34953, "router_noise": 0.0, "grad_norm": 0.8775, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 44.4}
{"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.95388, "bpb": 2.81885, "chunk_loss": 0.030242, "F": 0.17554, "G": 0.2772, "chunks": 179.75, "bytes_per_chunk": 5.697, "boundary_entropy": 0.37558, "router_noise": 0.0, "grad_norm": 0.7948, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 46.5}
{"event": "probe", "step": 600, "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": 0.257274, "boundary_entropy": 0.315934, "bpb": 2.920643}, "de": {"chunks": 3315.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161447, "cps": 25.898438, "mean_p": 0.269026, "boundary_entropy": 0.359544, "bpb": 3.17561}, "vi": {"chunks": 4205.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228099, "cps": 32.851562, "mean_p": 0.2616, "boundary_entropy": 0.316396, "bpb": 2.51629}, "tr": {"chunks": 3897.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215268, "cps": 30.445312, "mean_p": 0.298201, "boundary_entropy": 0.372193, "bpb": 3.152927}, "fi": {"chunks": 3669.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.195045, "cps": 28.664062, "mean_p": 0.325706, "boundary_entropy": 0.438942, "bpb": 3.379113}, "ru": {"chunks": 7019.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.370396, "cps": 54.835938, "mean_p": 0.239687, "boundary_entropy": 0.225227, "bpb": 2.155404}, "el": {"chunks": 7750.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.377625, "cps": 60.546875, "mean_p": 0.241728, "boundary_entropy": 0.220728, "bpb": 2.411864}, "hi": {"chunks": 4575.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.271917, "cps": 35.742188, "mean_p": 0.336588, "boundary_entropy": 0.516138, "bpb": 1.720448}, "ta": {"chunks": 8153.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.436106, "cps": 63.695312, "mean_p": 0.210278, "boundary_entropy": 0.309443, "bpb": 1.454351}, "th": {"chunks": 4281.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.266149, "cps": 33.445312, "mean_p": 0.293301, "boundary_entropy": 0.521403, "bpb": 1.810941}, "zh": {"chunks": 3555.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.60193, "cps": 27.773438, "mean_p": 0.321625, "boundary_entropy": 0.390929, "bpb": 3.857597}, "my": {"chunks": 4057.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.214112, "cps": 31.695312, "mean_p": 0.222301, "boundary_entropy": 0.416158, "bpb": 1.827451}}, "bpb": {"en": 2.92064, "de": 3.17561, "vi": 2.51629, "tr": 3.15293, "fi": 3.37911, "ru": 2.1554, "el": 2.41186, "hi": 1.72045, "ta": 1.45435, "th": 1.81094, "zh": 3.8576, "my": 1.82745}, "gini_cpc": 0.226885, "gini_cps": 0.185698, "bpic": 6.689576, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9708602475203754, "vi": 1.3716662458676885, "tr": 1.29451011357455, "fi": 1.1729011760201329, "ru": 2.227366182056224, "el": 2.270839556650938, "hi": 1.6351651364559145, "ta": 2.6225124959960873, "th": 1.6004781741091672, "zh": 3.6196931493653977, "my": 1.2875592553963935}, "t": 49.5}
{"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.92956, "de": 3.17322, "vi": 2.52329, "tr": 3.16332, "fi": 3.37519, "ru": 2.14078, "el": 2.40472, "hi": 1.71534, "ta": 1.46621, "th": 1.83869, "zh": 3.86899, "my": 1.84262}, "bytes_per_chunk": {"en": 5.696, "de": 6.275, "vi": 5.82, "tr": 5.047, "fi": 5.675, "ru": 5.007, "el": 4.859, "hi": 8.284, "ta": 5.931, "th": 9.746, "zh": 4.917, "my": 14.228}, "bpb_mean": 2.53683, "t": 50.8}
{"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.86706, "bpb": 2.6936, "chunk_loss": 0.030141, "F": 0.17148, "G": 0.27484, "chunks": 175.59, "bytes_per_chunk": 5.832, "boundary_entropy": 0.3482, "router_noise": 0.0, "grad_norm": 0.6507, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 50.8}
{"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.89122, "bpb": 2.72845, "chunk_loss": 0.03024, "F": 0.18079, "G": 0.25599, "chunks": 185.12, "bytes_per_chunk": 5.531, "boundary_entropy": 0.29801, "router_noise": 0.0, "grad_norm": 0.7961, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 52.9}
{"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.83047, "bpb": 2.64082, "chunk_loss": 0.030327, "F": 0.18362, "G": 0.26325, "chunks": 188.03, "bytes_per_chunk": 5.446, "boundary_entropy": 0.30247, "router_noise": 0.0, "grad_norm": 0.7395, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 55.1}
{"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.90956, "bpb": 2.75491, "chunk_loss": 0.030398, "F": 0.18246, "G": 0.2644, "chunks": 186.84, "bytes_per_chunk": 5.481, "boundary_entropy": 0.30737, "router_noise": 0.0, "grad_norm": 0.8473, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 57.3}
{"event": "probe", "step": 800, "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": 0.248639, "boundary_entropy": 0.285962, "bpb": 2.722468}, "de": {"chunks": 3295.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160473, "cps": 25.742188, "mean_p": 0.263136, "boundary_entropy": 0.34059, "bpb": 3.035858}, "vi": {"chunks": 4223.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229075, "cps": 32.992188, "mean_p": 0.253042, "boundary_entropy": 0.290001, "bpb": 2.386674}, "tr": {"chunks": 3803.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210076, "cps": 29.710938, "mean_p": 0.280054, "boundary_entropy": 0.335873, "bpb": 2.991263}, "fi": {"chunks": 3286.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.174685, "cps": 25.671875, "mean_p": 0.289789, "boundary_entropy": 0.388734, "bpb": 3.234428}, "ru": {"chunks": 6909.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.364591, "cps": 53.976562, "mean_p": 0.227778, "boundary_entropy": 0.213351, "bpb": 1.981479}, "el": {"chunks": 7033.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.342689, "cps": 54.945312, "mean_p": 0.223816, "boundary_entropy": 0.264042, "bpb": 2.125414}, "hi": {"chunks": 5770.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.342942, "cps": 45.078125, "mean_p": 0.340917, "boundary_entropy": 0.523948, "bpb": 1.622136}, "ta": {"chunks": 7360.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.393688, "cps": 57.5, "mean_p": 0.21781, "boundary_entropy": 0.345423, "bpb": 1.361223}, "th": {"chunks": 6881.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.42779, "cps": 53.757812, "mean_p": 0.339066, "boundary_entropy": 0.545545, "bpb": 1.595136}, "zh": {"chunks": 3066.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.519133, "cps": 23.953125, "mean_p": 0.255207, "boundary_entropy": 0.275847, "bpb": 3.680655}, "my": {"chunks": 8144.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.429808, "cps": 63.625, "mean_p": 0.259086, "boundary_entropy": 0.456724, "bpb": 1.632714}}, "bpb": {"en": 2.72247, "de": 3.03586, "vi": 2.38667, "tr": 2.99126, "fi": 3.23443, "ru": 1.98148, "el": 2.12541, "hi": 1.62214, "ta": 1.36122, "th": 1.59514, "zh": 3.68066, "my": 1.63271}, "gini_cpc": 0.208754, "gini_cps": 0.201671, "bpic": 6.125696, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9693572742503906, "vi": 1.3837537199050258, "tr": 1.2689854305293002, "fi": 1.0552042642120212, "ru": 2.2023525435960654, "el": 2.0700490402290415, "hi": 2.071579488094618, "ta": 2.378116941245866, "th": 2.5841121290627056, "zh": 3.135880971268842, "my": 2.5963022567765615}, "t": 60.5}
{"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.73395, "de": 3.03358, "vi": 2.39278, "tr": 3.00283, "fi": 3.22862, "ru": 1.96668, "el": 2.11244, "hi": 1.61903, "ta": 1.37148, "th": 1.62608, "zh": 3.69173, "my": 1.64254}, "bytes_per_chunk": {"en": 5.688, "de": 6.249, "vi": 5.806, "tr": 5.184, "fi": 6.031, "ru": 5.047, "el": 5.366, "hi": 6.935, "ta": 6.73, "th": 6.35, "zh": 5.233, "my": 6.436}, "bpb_mean": 2.36848, "t": 61.8}
{"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.64196, "bpb": 2.36885, "chunk_loss": 0.030114, "F": 0.17435, "G": 0.25445, "chunks": 178.53, "bytes_per_chunk": 5.736, "boundary_entropy": 0.30099, "router_noise": 0.0, "grad_norm": 0.7741, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 61.8}
{"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.70559, "bpb": 2.46065, "chunk_loss": 0.030147, "F": 0.17804, "G": 0.2656, "chunks": 182.31, "bytes_per_chunk": 5.617, "boundary_entropy": 0.31215, "router_noise": 0.0, "grad_norm": 0.6774, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.05, "alpha": 0.03, "t": 64.3}
{"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.6164, "bpb": 2.33198, "chunk_loss": 0.030375, "F": 0.18692, "G": 0.25011, "chunks": 191.41, "bytes_per_chunk": 5.35, "boundary_entropy": 0.26158, "router_noise": 0.0, "grad_norm": 0.6701, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.049, "alpha": 0.03, "t": 66.8}
{"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.7161, "bpb": 2.4758, "chunk_loss": 0.030332, "F": 0.18414, "G": 0.24499, "chunks": 188.56, "bytes_per_chunk": 5.431, "boundary_entropy": 0.25771, "router_noise": 0.0, "grad_norm": 0.7154, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.049, "alpha": 0.03, "t": 69.2}
{"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2903.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.16681, "cps": 22.679688, "mean_p": 0.23241, "boundary_entropy": 0.241227, "bpb": 2.596756}, "de": {"chunks": 3296.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160522, "cps": 25.75, "mean_p": 0.241997, "boundary_entropy": 0.291212, "bpb": 2.922417}, "vi": {"chunks": 4166.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225983, "cps": 32.546875, "mean_p": 0.23285, "boundary_entropy": 0.238963, "bpb": 2.278332}, "tr": {"chunks": 3892.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.214992, "cps": 30.40625, "mean_p": 0.255829, "boundary_entropy": 0.28947, "bpb": 2.881058}, "fi": {"chunks": 2943.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.156451, "cps": 22.992188, "mean_p": 0.239709, "boundary_entropy": 0.311523, "bpb": 3.180772}, "ru": {"chunks": 6889.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.363536, "cps": 53.820312, "mean_p": 0.230568, "boundary_entropy": 0.227803, "bpb": 1.867747}, "el": {"chunks": 7696.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.374994, "cps": 60.125, "mean_p": 0.241296, "boundary_entropy": 0.245467, "bpb": 1.985278}, "hi": {"chunks": 6408.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.380862, "cps": 50.0625, "mean_p": 0.336869, "boundary_entropy": 0.511047, "bpb": 1.550283}, "ta": {"chunks": 8505.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.454934, "cps": 66.445312, "mean_p": 0.235777, "boundary_entropy": 0.35841, "bpb": 1.271345}, "th": {"chunks": 7020.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.436431, "cps": 54.84375, "mean_p": 0.326922, "boundary_entropy": 0.536078, "bpb": 1.524263}, "zh": {"chunks": 3107.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.526075, "cps": 24.273438, "mean_p": 0.253619, "boundary_entropy": 0.257805, "bpb": 3.613505}, "my": {"chunks": 8419.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.444321, "cps": 65.773438, "mean_p": 0.314923, "boundary_entropy": 0.484419, "bpb": 1.431178}}, "bpb": {"en": 2.59676, "de": 2.92242, "vi": 2.27833, "tr": 2.88106, "fi": 3.18077, "ru": 1.86775, "el": 1.98528, "hi": 1.55028, "ta": 1.27135, "th": 1.52426, "zh": 3.6135, "my": 1.43118}, "gini_cpc": 0.218503, "gini_cps": 0.220815, "bpic": 5.882242, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9623030897608696, "vi": 1.3547314343597305, "tr": 1.2888410633157086, "fi": 0.9378976831246615, "ru": 2.1793353068475247, "el": 2.2480258364001813, "hi": 2.283202937577, "ta": 2.727256169135285, "th": 2.61633367601993, "zh": 3.153732800205866, "my": 2.663631963789288}, "t": 72.7}
{"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.60848, "de": 2.92403, "vi": 2.28274, "tr": 2.88984, "fi": 3.17877, "ru": 1.85127, "el": 1.97442, "hi": 1.5458, "ta": 1.2801, "th": 1.54825, "zh": 3.62122, "my": 1.4405}, "bytes_per_chunk": {"en": 5.673, "de": 6.239, "vi": 5.817, "tr": 5.035, "fi": 6.455, "ru": 5.083, "el": 4.813, "hi": 6.617, "ta": 5.897, "th": 6.159, "zh": 5.196, "my": 6.295}, "bpb_mean": 2.26212, "t": 74.0}
{"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.73295, "bpb": 2.50012, "chunk_loss": 0.030222, "F": 0.18121, "G": 0.25283, "chunks": 185.56, "bytes_per_chunk": 5.518, "boundary_entropy": 0.27899, "router_noise": 0.0, "grad_norm": 0.6951, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.096, "alpha": 0.03, "t": 74.0}
{"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.51605, "bpb": 2.1872, "chunk_loss": 0.030108, "F": 0.17642, "G": 0.262, "chunks": 180.66, "bytes_per_chunk": 5.668, "boundary_entropy": 0.30746, "router_noise": 0.0, "grad_norm": 0.7106, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 76.1}
{"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.77144, "bpb": 2.55565, "chunk_loss": 0.030417, "F": 0.18372, "G": 0.25779, "chunks": 188.12, "bytes_per_chunk": 5.443, "boundary_entropy": 0.27918, "router_noise": 0.0, "grad_norm": 0.7116, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 78.2}
{"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.65861, "bpb": 2.39287, "chunk_loss": 0.030288, "F": 0.18359, "G": 0.24137, "chunks": 188.0, "bytes_per_chunk": 5.447, "boundary_entropy": 0.24286, "router_noise": 0.0, "grad_norm": 0.7138, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 80.2}
{"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2900.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166638, "cps": 22.65625, "mean_p": 0.226857, "boundary_entropy": 0.227937, "bpb": 2.506918}, "de": {"chunks": 3324.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161886, "cps": 25.96875, "mean_p": 0.246003, "boundary_entropy": 0.296157, "bpb": 2.810333}, "vi": {"chunks": 4236.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22978, "cps": 33.09375, "mean_p": 0.243178, "boundary_entropy": 0.254807, "bpb": 2.230745}, "tr": {"chunks": 3940.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.217643, "cps": 30.78125, "mean_p": 0.254412, "boundary_entropy": 0.267979, "bpb": 2.797438}, "fi": {"chunks": 2974.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158099, "cps": 23.234375, "mean_p": 0.252384, "boundary_entropy": 0.326154, "bpb": 3.076182}, "ru": {"chunks": 6966.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.367599, "cps": 54.421875, "mean_p": 0.24427, "boundary_entropy": 0.215658, "bpb": 1.785088}, "el": {"chunks": 7383.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.359743, "cps": 57.679688, "mean_p": 0.228492, "boundary_entropy": 0.22004, "bpb": 1.867273}, "hi": {"chunks": 8738.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.519346, "cps": 68.265625, "mean_p": 0.354656, "boundary_entropy": 0.508621, "bpb": 1.569697}, "ta": {"chunks": 8404.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449532, "cps": 65.65625, "mean_p": 0.244206, "boundary_entropy": 0.379475, "bpb": 1.216972}, "th": {"chunks": 9985.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.620765, "cps": 78.007812, "mean_p": 0.402325, "boundary_entropy": 0.589268, "bpb": 1.433304}, "zh": {"chunks": 3076.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.520826, "cps": 24.03125, "mean_p": 0.242322, "boundary_entropy": 0.232553, "bpb": 3.547964}, "my": {"chunks": 7178.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.378826, "cps": 56.078125, "mean_p": 0.325171, "boundary_entropy": 0.53632, "bpb": 1.377376}}, "bpb": {"en": 2.50692, "de": 2.81033, "vi": 2.23074, "tr": 2.79744, "fi": 3.07618, "ru": 1.78509, "el": 1.86727, "hi": 1.5697, "ta": 1.21697, "th": 1.4333, "zh": 3.54796, "my": 1.37738}, "gini_cpc": 0.249019, "gini_cps": 0.240997, "bpic": 5.553673, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9714819374027007, "vi": 1.3789195589349343, "tr": 1.3060860554642304, "fi": 0.9487574584936546, "ru": 2.2059739423164406, "el": 2.1588285136776735, "hi": 3.116614520674284, "ta": 2.6976567955658437, "th": 3.725230295949321, "zh": 3.125496455971134, "my": 2.27334945731694}, "t": 83.2}
{"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.51902, "de": 2.81128, "vi": 2.23535, "tr": 2.80485, "fi": 3.07969, "ru": 1.76937, "el": 1.85369, "hi": 1.56622, "ta": 1.22634, "th": 1.45353, "zh": 3.55291, "my": 1.38586}, "bytes_per_chunk": {"en": 5.653, "de": 6.221, "vi": 5.775, "tr": 4.926, "fi": 6.407, "ru": 5.041, "el": 5.076, "hi": 4.759, "ta": 5.957, "th": 4.198, "zh": 5.272, "my": 7.309}, "bpb_mean": 2.18818, "t": 84.5}
{"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.61774, "bpb": 2.3339, "chunk_loss": 0.030706, "F": 0.19244, "G": 0.25667, "chunks": 197.06, "bytes_per_chunk": 5.196, "boundary_entropy": 0.27437, "router_noise": 0.0, "grad_norm": 0.7116, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 84.5}
{"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.63373, "bpb": 2.35698, "chunk_loss": 0.030334, "F": 0.18301, "G": 0.25075, "chunks": 187.41, "bytes_per_chunk": 5.464, "boundary_entropy": 0.25979, "router_noise": 0.0, "grad_norm": 0.7146, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 86.5}
{"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.57162, "bpb": 2.26737, "chunk_loss": 0.030236, "F": 0.18213, "G": 0.25518, "chunks": 186.5, "bytes_per_chunk": 5.491, "boundary_entropy": 0.28135, "router_noise": 0.0, "grad_norm": 0.7405, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 88.6}
{"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.71839, "bpb": 2.47912, "chunk_loss": 0.030111, "F": 0.17456, "G": 0.24856, "chunks": 178.75, "bytes_per_chunk": 5.729, "boundary_entropy": 0.2625, "router_noise": 0.0, "grad_norm": 0.7589, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 90.7}
{"event": "probe", "step": 1400, "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": 0.218759, "boundary_entropy": 0.205253, "bpb": 2.44696}, "de": {"chunks": 3303.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160863, "cps": 25.804688, "mean_p": 0.23136, "boundary_entropy": 0.26209, "bpb": 2.756324}, "vi": {"chunks": 4288.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.232601, "cps": 33.5, "mean_p": 0.234429, "boundary_entropy": 0.234392, "bpb": 2.191622}, "tr": {"chunks": 3947.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.21803, "cps": 30.835938, "mean_p": 0.239295, "boundary_entropy": 0.233282, "bpb": 2.727454}, "fi": {"chunks": 3014.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160225, "cps": 23.546875, "mean_p": 0.243176, "boundary_entropy": 0.303893, "bpb": 2.981567}, "ru": {"chunks": 6943.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.366385, "cps": 54.242188, "mean_p": 0.232014, "boundary_entropy": 0.197735, "bpb": 1.715086}, "el": {"chunks": 7392.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.360181, "cps": 57.75, "mean_p": 0.232299, "boundary_entropy": 0.220733, "bpb": 1.81187}, "hi": {"chunks": 5736.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.340921, "cps": 44.8125, "mean_p": 0.306573, "boundary_entropy": 0.461926, "bpb": 1.479479}, "ta": {"chunks": 8022.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.429099, "cps": 62.671875, "mean_p": 0.248127, "boundary_entropy": 0.368771, "bpb": 1.155612}, "th": {"chunks": 6337.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.39397, "cps": 49.507812, "mean_p": 0.334051, "boundary_entropy": 0.542427, "bpb": 1.374511}, "zh": {"chunks": 3137.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.531155, "cps": 24.507812, "mean_p": 0.241589, "boundary_entropy": 0.229129, "bpb": 3.507108}, "my": {"chunks": 8137.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.429438, "cps": 63.570312, "mean_p": 0.333573, "boundary_entropy": 0.529345, "bpb": 1.373031}}, "bpb": {"en": 2.44696, "de": 2.75632, "vi": 2.19162, "tr": 2.72745, "fi": 2.98157, "ru": 1.71509, "el": 1.81187, "hi": 1.47948, "ta": 1.15561, "th": 1.37451, "zh": 3.50711, "my": 1.37303}, "gini_cpc": 0.211829, "gini_cps": 0.210099, "bpic": 6.07633, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9640147406331996, "vi": 1.393924151642918, "tr": 1.3066043000954592, "fi": 0.9601937454390805, "ru": 2.195661865718834, "el": 2.1584829437482558, "hi": 2.043062149251532, "ta": 2.5714891505440756, "th": 2.360968267751126, "zh": 3.183087552090184, "my": 2.573525335542075}, "t": 93.9}
{"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.45989, "de": 2.75796, "vi": 2.19682, "tr": 2.73264, "fi": 2.98772, "ru": 1.69354, "el": 1.80024, "hi": 1.47661, "ta": 1.16447, "th": 1.39582, "zh": 3.50929, "my": 1.38078}, "bytes_per_chunk": {"en": 5.673, "de": 6.244, "vi": 5.745, "tr": 4.924, "fi": 6.346, "ru": 5.095, "el": 5.102, "hi": 7.462, "ta": 6.261, "th": 7.138, "zh": 5.14, "my": 6.647}, "bpb_mean": 2.12965, "t": 95.2}
{"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.57414, "bpb": 2.271, "chunk_loss": 0.030111, "F": 0.17722, "G": 0.23789, "chunks": 181.47, "bytes_per_chunk": 5.643, "boundary_entropy": 0.23314, "router_noise": 0.0, "grad_norm": 0.8199, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 95.2}
{"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.72337, "bpb": 2.4863, "chunk_loss": 0.030182, "F": 0.17709, "G": 0.22697, "chunks": 181.34, "bytes_per_chunk": 5.647, "boundary_entropy": 0.21345, "router_noise": 0.0, "grad_norm": 0.8052, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 97.5}
{"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.5302, "bpb": 2.20762, "chunk_loss": 0.03054, "F": 0.19055, "G": 0.25721, "chunks": 195.12, "bytes_per_chunk": 5.248, "boundary_entropy": 0.27042, "router_noise": 0.0, "grad_norm": 0.7057, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 99.7}
{"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.60602, "bpb": 2.317, "chunk_loss": 0.030111, "F": 0.17896, "G": 0.24266, "chunks": 183.25, "bytes_per_chunk": 5.588, "boundary_entropy": 0.24858, "router_noise": 0.0, "grad_norm": 0.7132, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 101.8}
{"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2907.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.16704, "cps": 22.710938, "mean_p": 0.213351, "boundary_entropy": 0.192835, "bpb": 2.397359}, "de": {"chunks": 3295.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160473, "cps": 25.742188, "mean_p": 0.225914, "boundary_entropy": 0.250925, "bpb": 2.688162}, "vi": {"chunks": 4203.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22799, "cps": 32.835938, "mean_p": 0.231145, "boundary_entropy": 0.221584, "bpb": 2.145786}, "tr": {"chunks": 3920.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216539, "cps": 30.625, "mean_p": 0.233768, "boundary_entropy": 0.224931, "bpb": 2.648825}, "fi": {"chunks": 2972.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157993, "cps": 23.21875, "mean_p": 0.231509, "boundary_entropy": 0.289701, "bpb": 2.924094}, "ru": {"chunks": 6848.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.361372, "cps": 53.5, "mean_p": 0.231071, "boundary_entropy": 0.193073, "bpb": 1.641419}, "el": {"chunks": 7159.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.348828, "cps": 55.929688, "mean_p": 0.225627, "boundary_entropy": 0.210103, "bpb": 1.759144}, "hi": {"chunks": 7501.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.445825, "cps": 58.601562, "mean_p": 0.341017, "boundary_entropy": 0.497113, "bpb": 1.420601}, "ta": {"chunks": 7593.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.406151, "cps": 59.320312, "mean_p": 0.238691, "boundary_entropy": 0.375829, "bpb": 1.209277}, "th": {"chunks": 7065.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.439229, "cps": 55.195312, "mean_p": 0.357809, "boundary_entropy": 0.554259, "bpb": 1.315166}, "zh": {"chunks": 3089.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.523027, "cps": 24.132812, "mean_p": 0.23551, "boundary_entropy": 0.213878, "bpb": 3.463418}, "my": {"chunks": 9619.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.507653, "cps": 75.148438, "mean_p": 0.399681, "boundary_entropy": 0.583531, "bpb": 1.281304}}, "bpb": {"en": 2.39736, "de": 2.68816, "vi": 2.14579, "tr": 2.64883, "fi": 2.92409, "ru": 1.64142, "el": 1.75914, "hi": 1.4206, "ta": 1.20928, "th": 1.31517, "zh": 3.46342, "my": 1.2813}, "gini_cpc": 0.226141, "gini_cps": 0.223795, "bpic": 5.799837, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.960687412148392, "vi": 1.364882723249036, "tr": 1.2963271115683805, "fi": 0.9458363705797812, "ru": 2.1633840615818607, "el": 2.088289023259514, "hi": 2.6689668581908403, "ta": 2.4314593809539975, "th": 2.6294819221800596, "zh": 3.1311476985326006, "my": 3.0391045244025636}, "t": 104.9}
{"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.40843, "de": 2.68963, "vi": 2.14825, "tr": 2.65388, "fi": 2.93053, "ru": 1.62363, "el": 1.7476, "hi": 1.41923, "ta": 1.21503, "th": 1.33498, "zh": 3.46777, "my": 1.28912}, "bytes_per_chunk": {"en": 5.67, "de": 6.265, "vi": 5.807, "tr": 4.995, "fi": 6.438, "ru": 5.181, "el": 5.273, "hi": 5.983, "ta": 6.651, "th": 6.456, "zh": 5.168, "my": 5.661}, "bpb_mean": 2.07734, "t": 106.3}
{"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.57827, "bpb": 2.27696, "chunk_loss": 0.030208, "F": 0.17566, "G": 0.23026, "chunks": 179.88, "bytes_per_chunk": 5.693, "boundary_entropy": 0.23394, "router_noise": 0.0, "grad_norm": 0.8564, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 106.3}
{"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.50153, "bpb": 2.16624, "chunk_loss": 0.030294, "F": 0.18271, "G": 0.25488, "chunks": 187.09, "bytes_per_chunk": 5.473, "boundary_entropy": 0.27375, "router_noise": 0.0, "grad_norm": 0.724, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 108.4}
{"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.42088, "bpb": 2.0499, "chunk_loss": 0.030077, "F": 0.17493, "G": 0.23763, "chunks": 179.12, "bytes_per_chunk": 5.717, "boundary_entropy": 0.24574, "router_noise": 0.0, "grad_norm": 0.6628, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 110.5}
{"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.62176, "bpb": 2.33971, "chunk_loss": 0.030281, "F": 0.18097, "G": 0.23945, "chunks": 185.31, "bytes_per_chunk": 5.526, "boundary_entropy": 0.23592, "router_noise": 0.0, "grad_norm": 0.7966, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 112.7}
{"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2912.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167327, "cps": 22.75, "mean_p": 0.219666, "boundary_entropy": 0.205154, "bpb": 2.344513}, "de": {"chunks": 3323.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161837, "cps": 25.960938, "mean_p": 0.243951, "boundary_entropy": 0.279759, "bpb": 2.626822}, "vi": {"chunks": 4301.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.233306, "cps": 33.601562, "mean_p": 0.238075, "boundary_entropy": 0.228051, "bpb": 2.129502}, "tr": {"chunks": 3697.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.20422, "cps": 28.882812, "mean_p": 0.221617, "boundary_entropy": 0.226104, "bpb": 2.635944}, "fi": {"chunks": 3017.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160385, "cps": 23.570312, "mean_p": 0.234272, "boundary_entropy": 0.286685, "bpb": 2.872972}, "ru": {"chunks": 6910.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.364644, "cps": 53.984375, "mean_p": 0.231025, "boundary_entropy": 0.192107, "bpb": 1.587275}, "el": {"chunks": 7132.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.347513, "cps": 55.71875, "mean_p": 0.223331, "boundary_entropy": 0.206936, "bpb": 1.697754}, "hi": {"chunks": 7669.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.45581, "cps": 59.914062, "mean_p": 0.341778, "boundary_entropy": 0.485894, "bpb": 1.390629}, "ta": {"chunks": 9292.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.497031, "cps": 72.59375, "mean_p": 0.282236, "boundary_entropy": 0.395225, "bpb": 1.113422}, "th": {"chunks": 8496.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.528194, "cps": 66.375, "mean_p": 0.36803, "boundary_entropy": 0.557958, "bpb": 1.307751}, "zh": {"chunks": 2998.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.507619, "cps": 23.421875, "mean_p": 0.228576, "boundary_entropy": 0.208737, "bpb": 3.425839}, "my": {"chunks": 9015.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.475776, "cps": 70.429688, "mean_p": 0.397233, "boundary_entropy": 0.583337, "bpb": 1.272134}}, "bpb": {"en": 2.34451, "de": 2.62682, "vi": 2.1295, "tr": 2.63594, "fi": 2.87297, "ru": 1.58727, "el": 1.69775, "hi": 1.39063, "ta": 1.11342, "th": 1.30775, "zh": 3.42584, "my": 1.27213}, "gini_cpc": 0.2337, "gini_cps": 0.239146, "bpic": 5.581295, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9671875192333114, "vi": 1.3943090625810317, "tr": 1.2204827587619804, "fi": 0.9585089585471676, "ru": 2.1792225419699034, "el": 2.076840952498751, "hi": 2.7240584025929495, "ta": 2.970410566967133, "th": 3.156648232091191, "zh": 3.0336881519259604, "my": 2.8433813109389026}, "t": 115.9}
{"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.35591, "de": 2.62888, "vi": 2.13288, "tr": 2.64233, "fi": 2.87182, "ru": 1.56635, "el": 1.687, "hi": 1.38962, "ta": 1.12131, "th": 1.32526, "zh": 3.43033, "my": 1.27933}, "bytes_per_chunk": {"en": 5.66, "de": 6.219, "vi": 5.737, "tr": 5.182, "fi": 6.41, "ru": 5.143, "el": 5.284, "hi": 5.593, "ta": 5.441, "th": 5.271, "zh": 5.233, "my": 5.701}, "bpb_mean": 2.03592, "t": 117.2}
{"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.58499, "bpb": 2.28665, "chunk_loss": 0.030104, "F": 0.17365, "G": 0.22456, "chunks": 177.81, "bytes_per_chunk": 5.759, "boundary_entropy": 0.21459, "router_noise": 0.0, "grad_norm": 1.0087, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 117.3}
{"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.55297, "bpb": 2.24046, "chunk_loss": 0.030152, "F": 0.17618, "G": 0.23635, "chunks": 180.41, "bytes_per_chunk": 5.676, "boundary_entropy": 0.23108, "router_noise": 0.0, "grad_norm": 0.718, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 119.6}
{"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.58662, "bpb": 2.28901, "chunk_loss": 0.030209, "F": 0.18176, "G": 0.23123, "chunks": 186.12, "bytes_per_chunk": 5.502, "boundary_entropy": 0.21493, "router_noise": 0.0, "grad_norm": 0.6796, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 121.9}
{"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.59706, "bpb": 2.30407, "chunk_loss": 0.03018, "F": 0.18152, "G": 0.23927, "chunks": 185.88, "bytes_per_chunk": 5.509, "boundary_entropy": 0.2249, "router_noise": 0.0, "grad_norm": 0.6799, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.048, "alpha": 0.03, "t": 124.3}
{"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2906.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166983, "cps": 22.703125, "mean_p": 0.20792, "boundary_entropy": 0.176197, "bpb": 2.29701}, "de": {"chunks": 3299.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160668, "cps": 25.773438, "mean_p": 0.229124, "boundary_entropy": 0.252522, "bpb": 2.576236}, "vi": {"chunks": 4209.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228316, "cps": 32.882812, "mean_p": 0.228142, "boundary_entropy": 0.213883, "bpb": 2.087831}, "tr": {"chunks": 3642.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201182, "cps": 28.453125, "mean_p": 0.213252, "boundary_entropy": 0.213036, "bpb": 2.566184}, "fi": {"chunks": 2972.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157993, "cps": 23.21875, "mean_p": 0.229029, "boundary_entropy": 0.280296, "bpb": 2.807451}, "ru": {"chunks": 6782.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.357889, "cps": 52.984375, "mean_p": 0.2316, "boundary_entropy": 0.198822, "bpb": 1.562809}, "el": {"chunks": 7241.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.352824, "cps": 56.570312, "mean_p": 0.227072, "boundary_entropy": 0.197261, "bpb": 1.682745}, "hi": {"chunks": 6973.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.414443, "cps": 54.476562, "mean_p": 0.327072, "boundary_entropy": 0.478261, "bpb": 1.374091}, "ta": {"chunks": 8494.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.454346, "cps": 66.359375, "mean_p": 0.239979, "boundary_entropy": 0.354297, "bpb": 1.083228}, "th": {"chunks": 7592.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.471993, "cps": 59.3125, "mean_p": 0.363572, "boundary_entropy": 0.559265, "bpb": 1.261246}, "zh": {"chunks": 3064.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.518794, "cps": 23.9375, "mean_p": 0.232934, "boundary_entropy": 0.206878, "bpb": 3.389033}, "my": {"chunks": 10090.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.53251, "cps": 78.828125, "mean_p": 0.400202, "boundary_entropy": 0.587733, "bpb": 1.233757}}, "bpb": {"en": 2.29701, "de": 2.57624, "vi": 2.08783, "tr": 2.56618, "fi": 2.80745, "ru": 1.56281, "el": 1.68274, "hi": 1.37409, "ta": 1.08323, "th": 1.26125, "zh": 3.38903, "my": 1.23376}, "gini_cpc": 0.234514, "gini_cps": 0.238728, "bpic": 5.705593, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9621846376314842, "vi": 1.3673015119247682, "tr": 1.2048081599597986, "fi": 0.9461618476515568, "ru": 2.1432709688080522, "el": 2.1129353671282605, "hi": 2.4819504248524082, "ta": 2.720917020737315, "th": 2.8265953775762407, "zh": 3.106875343911694, "my": 3.18901307900886}, "t": 127.8}
{"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.30611, "de": 2.58001, "vi": 2.09058, "tr": 2.57305, "fi": 2.81213, "ru": 1.54201, "el": 1.67152, "hi": 1.37121, "ta": 1.09259, "th": 1.28117, "zh": 3.39301, "my": 1.24143}, "bytes_per_chunk": {"en": 5.659, "de": 6.252, "vi": 5.821, "tr": 5.223, "fi": 6.439, "ru": 5.221, "el": 5.22, "hi": 6.337, "ta": 6.016, "th": 5.98, "zh": 5.217, "my": 5.764}, "bpb_mean": 1.99624, "t": 129.1}
{"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.51336, "bpb": 2.18331, "chunk_loss": 0.030098, "F": 0.17456, "G": 0.24425, "chunks": 178.75, "bytes_per_chunk": 5.729, "boundary_entropy": 0.25695, "router_noise": 0.0, "grad_norm": 0.6679, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.097, "alpha": 0.03, "t": 129.2}
{"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.38765, "bpb": 2.00195, "chunk_loss": 0.030206, "F": 0.18314, "G": 0.23106, "chunks": 187.53, "bytes_per_chunk": 5.46, "boundary_entropy": 0.21191, "router_noise": 0.0, "grad_norm": 0.6816, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 131.3}
{"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.57001, "bpb": 2.26505, "chunk_loss": 0.030174, "F": 0.17459, "G": 0.23857, "chunks": 178.78, "bytes_per_chunk": 5.728, "boundary_entropy": 0.2377, "router_noise": 0.0, "grad_norm": 0.7289, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 133.4}
{"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.47502, "bpb": 2.12801, "chunk_loss": 0.030161, "F": 0.17825, "G": 0.23359, "chunks": 182.53, "bytes_per_chunk": 5.61, "boundary_entropy": 0.22494, "router_noise": 0.0, "grad_norm": 0.6663, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 135.6}
{"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2898.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166523, "cps": 22.640625, "mean_p": 0.20511, "boundary_entropy": 0.167612, "bpb": 2.258969}, "de": {"chunks": 3326.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161983, "cps": 25.984375, "mean_p": 0.230396, "boundary_entropy": 0.251387, "bpb": 2.532457}, "vi": {"chunks": 4187.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227122, "cps": 32.710938, "mean_p": 0.228726, "boundary_entropy": 0.214828, "bpb": 2.062475}, "tr": {"chunks": 3862.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213335, "cps": 30.171875, "mean_p": 0.221529, "boundary_entropy": 0.208919, "bpb": 2.520424}, "fi": {"chunks": 3021.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160598, "cps": 23.601562, "mean_p": 0.236378, "boundary_entropy": 0.288733, "bpb": 2.762348}, "ru": {"chunks": 6962.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.367388, "cps": 54.390625, "mean_p": 0.232066, "boundary_entropy": 0.182165, "bpb": 1.525683}, "el": {"chunks": 7535.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.367149, "cps": 58.867188, "mean_p": 0.231827, "boundary_entropy": 0.196232, "bpb": 1.624865}, "hi": {"chunks": 7285.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.432987, "cps": 56.914062, "mean_p": 0.326058, "boundary_entropy": 0.475346, "bpb": 1.347822}, "ta": {"chunks": 9334.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.499278, "cps": 72.921875, "mean_p": 0.282133, "boundary_entropy": 0.384249, "bpb": 1.056411}, "th": {"chunks": 7439.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.462481, "cps": 58.117188, "mean_p": 0.359833, "boundary_entropy": 0.557031, "bpb": 1.245275}, "zh": {"chunks": 3022.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.511683, "cps": 23.609375, "mean_p": 0.22414, "boundary_entropy": 0.182791, "bpb": 3.358283}, "my": {"chunks": 7817.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.41255, "cps": 61.070312, "mean_p": 0.348086, "boundary_entropy": 0.547577, "bpb": 1.221706}}, "bpb": {"en": 2.25897, "de": 2.53246, "vi": 2.06247, "tr": 2.52042, "fi": 2.76235, "ru": 1.52568, "el": 1.62487, "hi": 1.34782, "ta": 1.05641, "th": 1.24528, "zh": 3.35828, "my": 1.22171}, "gini_cpc": 0.222275, "gini_cps": 0.223878, "bpic": 5.754873, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9727373165592448, "vi": 1.3639095113995174, "tr": 1.281113055738448, "fi": 0.964416386474017, "ru": 2.2062287702737398, "el": 2.2047946711103, "hi": 2.6001608938684977, "ta": 2.998251544765958, "th": 2.7772772235490675, "zh": 3.0727466681249864, "my": 2.477436175983}, "t": 138.9}
{"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.27017, "de": 2.53726, "vi": 2.06636, "tr": 2.52535, "fi": 2.76727, "ru": 1.50553, "el": 1.61589, "hi": 1.34487, "ta": 1.06421, "th": 1.26502, "zh": 3.36266, "my": 1.2293}, "bytes_per_chunk": {"en": 5.683, "de": 6.24, "vi": 5.823, "tr": 5.029, "fi": 6.365, "ru": 5.107, "el": 4.981, "hi": 6.04, "ta": 5.549, "th": 6.005, "zh": 5.256, "my": 6.616}, "bpb_mean": 1.96283, "t": 140.2}
{"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.47748, "bpb": 2.13155, "chunk_loss": 0.030153, "F": 0.17877, "G": 0.23617, "chunks": 183.06, "bytes_per_chunk": 5.594, "boundary_entropy": 0.22947, "router_noise": 0.0, "grad_norm": 0.6771, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.094, "alpha": 0.03, "t": 140.3}
{"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.71729, "bpb": 2.47752, "chunk_loss": 0.03025, "F": 0.18286, "G": 0.22467, "chunks": 187.25, "bytes_per_chunk": 5.469, "boundary_entropy": 0.19138, "router_noise": 0.0, "grad_norm": 0.7597, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 142.3}
{"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.48193, "bpb": 2.13797, "chunk_loss": 0.030078, "F": 0.17261, "G": 0.22769, "chunks": 176.75, "bytes_per_chunk": 5.793, "boundary_entropy": 0.22021, "router_noise": 0.0, "grad_norm": 0.6355, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 144.4}
{"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.48451, "bpb": 2.1417, "chunk_loss": 0.03008, "F": 0.17715, "G": 0.23468, "chunks": 181.41, "bytes_per_chunk": 5.645, "boundary_entropy": 0.22659, "router_noise": 0.0, "grad_norm": 0.6734, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 146.5}
{"event": "probe", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2908.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167098, "cps": 22.71875, "mean_p": 0.206123, "boundary_entropy": 0.166133, "bpb": 2.23476}, "de": {"chunks": 3317.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161545, "cps": 25.914062, "mean_p": 0.225429, "boundary_entropy": 0.240454, "bpb": 2.503216}, "vi": {"chunks": 4356.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23629, "cps": 34.03125, "mean_p": 0.23158, "boundary_entropy": 0.21202, "bpb": 2.052536}, "tr": {"chunks": 3830.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.211567, "cps": 29.921875, "mean_p": 0.215054, "boundary_entropy": 0.193422, "bpb": 2.491197}, "fi": {"chunks": 3017.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160385, "cps": 23.570312, "mean_p": 0.239746, "boundary_entropy": 0.293632, "bpb": 2.716313}, "ru": {"chunks": 6811.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.35942, "cps": 53.210938, "mean_p": 0.223318, "boundary_entropy": 0.163322, "bpb": 1.50081}, "el": {"chunks": 7170.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.349364, "cps": 56.015625, "mean_p": 0.216983, "boundary_entropy": 0.171435, "bpb": 1.606404}, "hi": {"chunks": 6633.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.394235, "cps": 51.820312, "mean_p": 0.320359, "boundary_entropy": 0.469093, "bpb": 1.325936}, "ta": {"chunks": 8895.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.475796, "cps": 69.492188, "mean_p": 0.260684, "boundary_entropy": 0.358774, "bpb": 1.042731}, "th": {"chunks": 7506.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.466646, "cps": 58.640625, "mean_p": 0.375357, "boundary_entropy": 0.567009, "bpb": 1.225283}, "zh": {"chunks": 2981.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.504741, "cps": 23.289062, "mean_p": 0.220427, "boundary_entropy": 0.184766, "bpb": 3.329918}, "my": {"chunks": 7136.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.37661, "cps": 55.75, "mean_p": 0.358977, "boundary_entropy": 0.565549, "bpb": 1.223274}}, "bpb": {"en": 2.23476, "de": 2.50322, "vi": 2.05254, "tr": 2.4912, "fi": 2.71631, "ru": 1.50081, "el": 1.6064, "hi": 1.32594, "ta": 1.04273, "th": 1.22528, "zh": 3.32992, "my": 1.22327}, "gini_cpc": 0.21692, "gini_cps": 0.21423, "bpic": 5.944563, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9667691476082619, "vi": 1.4140815214167477, "tr": 1.266128954437017, "fi": 0.9598274027817579, "ru": 2.1509552939212364, "el": 2.090778514607793, "hi": 2.3593080113066174, "ta": 2.8474112784939445, "th": 2.7926545231894355, "zh": 3.0206350080653763, "my": 2.253830145139133}, "t": 149.6}
{"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.24572, "de": 2.50837, "vi": 2.05416, "tr": 2.49586, "fi": 2.71737, "ru": 1.47711, "el": 1.59642, "hi": 1.32453, "ta": 1.05036, "th": 1.24181, "zh": 3.33191, "my": 1.23132}, "bytes_per_chunk": {"en": 5.686, "de": 6.266, "vi": 5.652, "tr": 5.055, "fi": 6.332, "ru": 5.194, "el": 5.221, "hi": 6.605, "ta": 5.781, "th": 6.2, "zh": 5.259, "my": 7.784}, "bpb_mean": 1.93958, "t": 150.8}
{"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 150.8}
{"event": "train", "step": 2400, "phase": 2, "objective": "parity_a", "lm_loss": 1.50272, "bpb": 2.16796, "chunk_loss": 0.034967, "F": 0.18619, "G": 0.2415, "chunks": 190.66, "bytes_per_chunk": 5.371, "boundary_entropy": 0.22907, "router_noise": 0.0, "grad_norm": 0.7161, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.088, "n_target_mean": 10.0837, "alpha": 0.03, "t": 150.9}
{"event": "train", "step": 2450, "phase": 2, "objective": "parity_a", "lm_loss": 1.30911, "bpb": 1.88864, "chunk_loss": 0.031903, "F": 0.16162, "G": 0.19911, "chunks": 165.5, "bytes_per_chunk": 6.187, "boundary_entropy": 0.18338, "router_noise": 0.0, "grad_norm": 0.6003, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.046, "n_target_mean": 10.1484, "alpha": 0.03, "t": 153.2}
{"event": "train", "step": 2500, "phase": 2, "objective": "parity_a", "lm_loss": 1.4021, "bpb": 2.02281, "chunk_loss": 0.031311, "F": 0.16489, "G": 0.19324, "chunks": 168.84, "bytes_per_chunk": 6.065, "boundary_entropy": 0.15172, "router_noise": 0.0, "grad_norm": 0.6697, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 8.8824, "alpha": 0.03, "t": 155.3}
{"event": "train", "step": 2550, "phase": 2, "objective": "parity_a", "lm_loss": 1.57993, "bpb": 2.27935, "chunk_loss": 0.031084, "F": 0.1734, "G": 0.19889, "chunks": 177.56, "bytes_per_chunk": 5.767, "boundary_entropy": 0.1412, "router_noise": 0.0, "grad_norm": 0.6771, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 7.9249, "alpha": 0.03, "t": 157.4}
{"event": "probe", "step": 2600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2895.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166351, "cps": 22.617188, "mean_p": 0.189531, "boundary_entropy": 0.115779, "bpb": 2.205942}, "de": {"chunks": 3289.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160181, "cps": 25.695312, "mean_p": 0.192695, "boundary_entropy": 0.160496, "bpb": 2.474082}, "vi": {"chunks": 4174.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226417, "cps": 32.609375, "mean_p": 0.194995, "boundary_entropy": 0.128365, "bpb": 2.03126}, "tr": {"chunks": 3238.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178865, "cps": 25.296875, "mean_p": 0.193341, "boundary_entropy": 0.161012, "bpb": 2.479499}, "fi": {"chunks": 2918.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155122, "cps": 22.796875, "mean_p": 0.199427, "boundary_entropy": 0.20663, "bpb": 2.681845}, "ru": {"chunks": 6457.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.340739, "cps": 50.445312, "mean_p": 0.19691, "boundary_entropy": 0.140813, "bpb": 1.486659}, "el": {"chunks": 6814.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332018, "cps": 53.234375, "mean_p": 0.194662, "boundary_entropy": 0.151308, "bpb": 1.576999}, "hi": {"chunks": 4133.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.245646, "cps": 32.289062, "mean_p": 0.124741, "boundary_entropy": 0.169076, "bpb": 1.321755}, "ta": {"chunks": 3210.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.171704, "cps": 25.078125, "mean_p": 0.096415, "boundary_entropy": 0.172715, "bpb": 1.127988}, "th": {"chunks": 2715.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.168791, "cps": 21.210938, "mean_p": 0.115884, "boundary_entropy": 0.243716, "bpb": 1.229611}, "zh": {"chunks": 2972.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503217, "cps": 23.21875, "mean_p": 0.212672, "boundary_entropy": 0.160647, "bpb": 3.302322}, "my": {"chunks": 2612.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.137851, "cps": 20.40625, "mean_p": 0.123875, "boundary_entropy": 0.284304, "bpb": 1.21021}}, "bpb": {"en": 2.20594, "de": 2.47408, "vi": 2.03126, "tr": 2.4795, "fi": 2.68184, "ru": 1.48666, "el": 1.577, "hi": 1.32176, "ta": 1.12799, "th": 1.22961, "zh": 3.30232, "my": 1.21021}, "gini_cpc": 0.223335, "gini_cps": 0.175936, "bpic": 8.448302, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9629129300009666, "vi": 1.3610837686255166, "tr": 1.075231168566513, "fi": 0.9325002485867738, "ru": 2.048316786744501, "el": 1.9958910753233008, "hi": 1.4766782714471587, "ta": 1.0321792283270996, "th": 1.0146688334801677, "zh": 3.0250385574343475, "my": 0.8286771212404606}, "t": 160.6}
{"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.21694, "de": 2.48005, "vi": 2.03408, "tr": 2.48372, "fi": 2.68395, "ru": 1.46389, "el": 1.5682, "hi": 1.31873, "ta": 1.13498, "th": 1.25064, "zh": 3.30278, "my": 1.21818}, "bytes_per_chunk": {"en": 5.676, "de": 6.284, "vi": 5.826, "tr": 6.023, "fi": 6.49, "ru": 5.463, "el": 5.489, "hi": 9.646, "ta": 16.228, "th": 15.44, "zh": 5.255, "my": 19.531}, "bpb_mean": 1.92968, "t": 161.9}
{"event": "train", "step": 2600, "phase": 2, "objective": "parity_a", "lm_loss": 1.6684, "bpb": 2.407, "chunk_loss": 0.031009, "F": 0.17148, "G": 0.19586, "chunks": 175.59, "bytes_per_chunk": 5.832, "boundary_entropy": 0.1432, "router_noise": 0.0, "grad_norm": 0.69, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.092, "n_target_mean": 7.6058, "alpha": 0.03, "t": 162.0}
{"event": "train", "step": 2650, "phase": 2, "objective": "parity_a", "lm_loss": 1.43095, "bpb": 2.06442, "chunk_loss": 0.03085, "F": 0.15811, "G": 0.18088, "chunks": 161.91, "bytes_per_chunk": 6.325, "boundary_entropy": 0.13463, "router_noise": 0.0, "grad_norm": 0.6769, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.045, "n_target_mean": 9.0748, "alpha": 0.03, "t": 164.2}
{"event": "train", "step": 2700, "phase": 2, "objective": "parity_a", "lm_loss": 1.37075, "bpb": 1.97757, "chunk_loss": 0.031426, "F": 0.15903, "G": 0.17935, "chunks": 162.84, "bytes_per_chunk": 6.288, "boundary_entropy": 0.12882, "router_noise": 0.0, "grad_norm": 0.6564, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.047, "n_target_mean": 9.9458, "alpha": 0.03, "t": 166.5}
{"event": "train", "step": 2750, "phase": 2, "objective": "parity_a", "lm_loss": 1.33633, "bpb": 1.92792, "chunk_loss": 0.031083, "F": 0.1593, "G": 0.18137, "chunks": 163.12, "bytes_per_chunk": 6.277, "boundary_entropy": 0.13383, "router_noise": 0.0, "grad_norm": 0.6797, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.048, "n_target_mean": 9.1192, "alpha": 0.03, "t": 168.9}
{"event": "probe", "step": 2800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2910.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167213, "cps": 22.734375, "mean_p": 0.186643, "boundary_entropy": 0.105336, "bpb": 2.181576}, "de": {"chunks": 3290.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.16023, "cps": 25.703125, "mean_p": 0.184287, "boundary_entropy": 0.13556, "bpb": 2.442413}, "vi": {"chunks": 4164.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225875, "cps": 32.53125, "mean_p": 0.18584, "boundary_entropy": 0.111922, "bpb": 2.012788}, "tr": {"chunks": 3449.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190521, "cps": 26.945312, "mean_p": 0.192803, "boundary_entropy": 0.146416, "bpb": 2.431394}, "fi": {"chunks": 2884.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.153315, "cps": 22.53125, "mean_p": 0.18413, "boundary_entropy": 0.1735, "bpb": 2.655718}, "ru": {"chunks": 6467.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.341266, "cps": 50.523438, "mean_p": 0.197382, "boundary_entropy": 0.122607, "bpb": 1.460276}, "el": {"chunks": 6821.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332359, "cps": 53.289062, "mean_p": 0.187611, "boundary_entropy": 0.116164, "bpb": 1.539324}, "hi": {"chunks": 4269.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.25373, "cps": 33.351562, "mean_p": 0.111541, "boundary_entropy": 0.148893, "bpb": 1.311139}, "ta": {"chunks": 3640.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.194704, "cps": 28.4375, "mean_p": 0.094406, "boundary_entropy": 0.162297, "bpb": 1.071362}, "th": {"chunks": 2900.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.180292, "cps": 22.65625, "mean_p": 0.10943, "boundary_entropy": 0.215344, "bpb": 1.210141}, "zh": {"chunks": 2992.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506603, "cps": 23.375, "mean_p": 0.216828, "boundary_entropy": 0.157518, "bpb": 3.268649}, "my": {"chunks": 3817.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.201446, "cps": 29.820312, "mean_p": 0.112694, "boundary_entropy": 0.219192, "bpb": 1.186712}}, "bpb": {"en": 2.18158, "de": 2.44241, "vi": 2.01279, "tr": 2.43139, "fi": 2.65572, "ru": 1.46028, "el": 1.53932, "hi": 1.31114, "ta": 1.07136, "th": 1.21014, "zh": 3.26865, "my": 1.18671}, "gini_cpc": 0.20497, "gini_cps": 0.162026, "bpic": 8.062118, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9582407198670885, "vi": 1.3508238195498814, "tr": 1.139393595980845, "fi": 0.9168842314789494, "ru": 2.0409143432255257, "el": 1.98764278046607, "hi": 1.5174074539299338, "ta": 1.1644129997821788, "th": 1.0782216895423475, "zh": 3.0296975642453416, "my": 1.2047305267373105}, "t": 172.2}
{"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.19425, "de": 2.45065, "vi": 2.01491, "tr": 2.43647, "fi": 2.66341, "ru": 1.4364, "el": 1.52892, "hi": 1.30874, "ta": 1.07771, "th": 1.22709, "zh": 3.27037, "my": 1.19514}, "bytes_per_chunk": {"en": 5.668, "de": 6.289, "vi": 5.835, "tr": 5.509, "fi": 6.607, "ru": 5.455, "el": 5.484, "hi": 9.328, "ta": 13.793, "th": 14.616, "zh": 5.247, "my": 14.506}, "bpb_mean": 1.90034, "t": 173.5}
{"event": "train", "step": 2800, "phase": 2, "objective": "parity_a", "lm_loss": 1.45895, "bpb": 2.10482, "chunk_loss": 0.03142, "F": 0.17142, "G": 0.18977, "chunks": 175.53, "bytes_per_chunk": 5.834, "boundary_entropy": 0.12426, "router_noise": 0.0, "grad_norm": 0.683, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.093, "n_target_mean": 8.8073, "alpha": 0.03, "t": 173.6}
{"event": "train", "step": 2850, "phase": 2, "objective": "parity_a", "lm_loss": 1.4031, "bpb": 2.02425, "chunk_loss": 0.030962, "F": 0.15924, "G": 0.17748, "chunks": 163.06, "bytes_per_chunk": 6.28, "boundary_entropy": 0.12371, "router_noise": 0.0, "grad_norm": 0.7184, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.5124, "alpha": 0.03, "t": 175.7}
{"event": "train", "step": 2900, "phase": 2, "objective": "parity_a", "lm_loss": 1.47768, "bpb": 2.13184, "chunk_loss": 0.030961, "F": 0.16486, "G": 0.18539, "chunks": 168.81, "bytes_per_chunk": 6.066, "boundary_entropy": 0.1238, "router_noise": 0.0, "grad_norm": 0.699, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.031, "alpha": 0.03, "t": 177.8}
{"event": "train", "step": 2950, "phase": 2, "objective": "parity_a", "lm_loss": 1.36489, "bpb": 1.96912, "chunk_loss": 0.031161, "F": 0.15945, "G": 0.17716, "chunks": 163.28, "bytes_per_chunk": 6.271, "boundary_entropy": 0.12479, "router_noise": 0.0, "grad_norm": 0.6504, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.8764, "alpha": 0.03, "t": 179.9}
{"event": "probe", "step": 3000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2893.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166236, "cps": 22.601562, "mean_p": 0.183998, "boundary_entropy": 0.09628, "bpb": 2.160738}, "de": {"chunks": 3280.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159743, "cps": 25.625, "mean_p": 0.180894, "boundary_entropy": 0.12354, "bpb": 2.408427}, "vi": {"chunks": 4159.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225603, "cps": 32.492188, "mean_p": 0.183509, "boundary_entropy": 0.090701, "bpb": 2.002644}, "tr": {"chunks": 3041.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167983, "cps": 23.757812, "mean_p": 0.178503, "boundary_entropy": 0.131465, "bpb": 2.409792}, "fi": {"chunks": 2921.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155281, "cps": 22.820312, "mean_p": 0.181122, "boundary_entropy": 0.156354, "bpb": 2.601984}, "ru": {"chunks": 6353.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.335251, "cps": 49.632812, "mean_p": 0.189533, "boundary_entropy": 0.113669, "bpb": 1.448972}, "el": {"chunks": 6619.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.322516, "cps": 51.710938, "mean_p": 0.181277, "boundary_entropy": 0.120475, "bpb": 1.522925}, "hi": {"chunks": 3813.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.226627, "cps": 29.789062, "mean_p": 0.09029, "boundary_entropy": 0.145636, "bpb": 1.300944}, "ta": {"chunks": 2740.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.146563, "cps": 21.40625, "mean_p": 0.073044, "boundary_entropy": 0.152922, "bpb": 1.061219}, "th": {"chunks": 2659.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.165309, "cps": 20.773438, "mean_p": 0.100954, "boundary_entropy": 0.21224, "bpb": 1.185241}, "zh": {"chunks": 3037.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.514223, "cps": 23.726562, "mean_p": 0.216771, "boundary_entropy": 0.152442, "bpb": 3.241995}, "my": {"chunks": 2667.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.140754, "cps": 20.835938, "mean_p": 0.096153, "boundary_entropy": 0.203786, "bpb": 1.178}}, "bpb": {"en": 2.16074, "de": 2.40843, "vi": 2.00264, "tr": 2.40979, "fi": 2.60198, "ru": 1.44897, "el": 1.52293, "hi": 1.30094, "ta": 1.06122, "th": 1.18524, "zh": 3.24199, "my": 1.178}, "gini_cpc": 0.229084, "gini_cps": 0.174181, "bpic": 8.686365, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9609418839298073, "vi": 1.357130043985412, "tr": 1.01051218629718, "fi": 0.9341042759049334, "ru": 2.016718710525908, "el": 1.9401138789611225, "hi": 1.363287394022257, "ta": 0.8816592734701825, "th": 0.9944271171740261, "zh": 3.093335572195764, "my": 0.8467112422442239}, "t": 183.4}
{"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.17258, "de": 2.41562, "vi": 2.00396, "tr": 2.41196, "fi": 2.61035, "ru": 1.42571, "el": 1.51327, "hi": 1.29752, "ta": 1.06795, "th": 1.2031, "zh": 3.24313, "my": 1.18704}, "bytes_per_chunk": {"en": 5.677, "de": 6.287, "vi": 5.831, "tr": 6.264, "fi": 6.505, "ru": 5.533, "el": 5.634, "hi": 10.368, "ta": 18.065, "th": 15.565, "zh": 5.175, "my": 19.072}, "bpb_mean": 1.87935, "t": 184.7}
{"event": "train", "step": 3000, "phase": 2, "objective": "parity_a", "lm_loss": 1.47603, "bpb": 2.12947, "chunk_loss": 0.030494, "F": 0.15482, "G": 0.17708, "chunks": 158.53, "bytes_per_chunk": 6.459, "boundary_entropy": 0.12906, "router_noise": 0.0, "grad_norm": 0.7024, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.097, "n_target_mean": 8.834, "alpha": 0.03, "t": 184.8}
{"event": "train", "step": 3050, "phase": 2, "objective": "parity_a", "lm_loss": 1.38724, "bpb": 2.00136, "chunk_loss": 0.030464, "F": 0.15573, "G": 0.17349, "chunks": 159.47, "bytes_per_chunk": 6.421, "boundary_entropy": 0.11266, "router_noise": 0.0, "grad_norm": 0.686, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.4055, "alpha": 0.03, "t": 186.9}
{"event": "train", "step": 3100, "phase": 2, "objective": "parity_a", "lm_loss": 1.42416, "bpb": 2.05463, "chunk_loss": 0.030706, "F": 0.16394, "G": 0.17911, "chunks": 167.88, "bytes_per_chunk": 6.1, "boundary_entropy": 0.10374, "router_noise": 0.0, "grad_norm": 0.6927, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 8.0087, "alpha": 0.03, "t": 189.0}
{"event": "train", "step": 3150, "phase": 2, "objective": "parity_a", "lm_loss": 1.37235, "bpb": 1.97988, "chunk_loss": 0.030884, "F": 0.14938, "G": 0.16989, "chunks": 152.97, "bytes_per_chunk": 6.694, "boundary_entropy": 0.12944, "router_noise": 0.0, "grad_norm": 0.6953, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.8608, "alpha": 0.03, "t": 191.1}
{"event": "probe", "step": 3200, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.18086, "boundary_entropy": 0.08467, "bpb": 2.155222}, "de": {"chunks": 3283.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159889, "cps": 25.648438, "mean_p": 0.177155, "boundary_entropy": 0.111745, "bpb": 2.392767}, "vi": {"chunks": 4145.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.224844, "cps": 32.382812, "mean_p": 0.180605, "boundary_entropy": 0.090788, "bpb": 1.995041}, "tr": {"chunks": 2995.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.165442, "cps": 23.398438, "mean_p": 0.172228, "boundary_entropy": 0.120713, "bpb": 2.394178}, "fi": {"chunks": 2877.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152942, "cps": 22.476562, "mean_p": 0.175613, "boundary_entropy": 0.144645, "bpb": 2.585408}, "ru": {"chunks": 6326.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.333826, "cps": 49.421875, "mean_p": 0.18643, "boundary_entropy": 0.1095, "bpb": 1.42351}, "el": {"chunks": 6629.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.323003, "cps": 51.789062, "mean_p": 0.178659, "boundary_entropy": 0.11376, "bpb": 1.496767}, "hi": {"chunks": 3643.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.216523, "cps": 28.460938, "mean_p": 0.085006, "boundary_entropy": 0.149007, "bpb": 1.278282}, "ta": {"chunks": 3576.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.191281, "cps": 27.9375, "mean_p": 0.083619, "boundary_entropy": 0.159111, "bpb": 1.05217}, "th": {"chunks": 2830.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.17594, "cps": 22.109375, "mean_p": 0.098088, "boundary_entropy": 0.192101, "bpb": 1.17293}, "zh": {"chunks": 2996.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.507281, "cps": 23.40625, "mean_p": 0.214954, "boundary_entropy": 0.143731, "bpb": 3.231275}, "my": {"chunks": 2653.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.140015, "cps": 20.726562, "mean_p": 0.091744, "boundary_entropy": 0.190189, "bpb": 1.160358}}, "bpb": {"en": 2.15522, "de": 2.39277, "vi": 1.99504, "tr": 2.39418, "fi": 2.58541, "ru": 1.42351, "el": 1.49677, "hi": 1.27828, "ta": 1.05217, "th": 1.17293, "zh": 3.23128, "my": 1.16036}, "gini_cpc": 0.219833, "gini_cps": 0.166997, "bpic": 8.558135, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9624861838776273, "vi": 1.3534973863219626, "tr": 0.9959150694115093, "fi": 0.9206700319464307, "ru": 2.0095369762770274, "el": 1.9443892100506381, "hi": 1.3034072334955589, "ta": 1.1514579142748163, "th": 1.0591108052204932, "zh": 3.0536861188482347, "my": 0.8428492456224016}, "t": 194.2}
{"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.16483, "de": 2.39884, "vi": 1.99797, "tr": 2.39466, "fi": 2.59215, "ru": 1.40162, "el": 1.48667, "hi": 1.27656, "ta": 1.05883, "th": 1.19126, "zh": 3.22933, "my": 1.17044}, "bytes_per_chunk": {"en": 5.679, "de": 6.287, "vi": 5.834, "tr": 6.364, "fi": 6.589, "ru": 5.543, "el": 5.612, "hi": 10.444, "ta": 13.526, "th": 15.298, "zh": 5.226, "my": 19.136}, "bpb_mean": 1.8636, "t": 195.5}
{"event": "train", "step": 3200, "phase": 2, "objective": "parity_a", "lm_loss": 1.32323, "bpb": 1.90901, "chunk_loss": 0.031053, "F": 0.15845, "G": 0.17035, "chunks": 162.25, "bytes_per_chunk": 6.311, "boundary_entropy": 0.10572, "router_noise": 0.0, "grad_norm": 0.7252, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.088, "n_target_mean": 9.495, "alpha": 0.03, "t": 195.5}
{"event": "train", "step": 3250, "phase": 2, "objective": "parity_a", "lm_loss": 1.54573, "bpb": 2.23002, "chunk_loss": 0.030571, "F": 0.16611, "G": 0.1827, "chunks": 170.09, "bytes_per_chunk": 6.02, "boundary_entropy": 0.10851, "router_noise": 0.0, "grad_norm": 0.6941, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 7.9529, "alpha": 0.03, "t": 197.6}
{"event": "train", "step": 3300, "phase": 2, "objective": "parity_a", "lm_loss": 1.51217, "bpb": 2.18159, "chunk_loss": 0.030519, "F": 0.16187, "G": 0.17995, "chunks": 165.75, "bytes_per_chunk": 6.178, "boundary_entropy": 0.10766, "router_noise": 0.0, "grad_norm": 0.7358, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 7.6934, "alpha": 0.03, "t": 199.7}
{"event": "train", "step": 3350, "phase": 2, "objective": "parity_a", "lm_loss": 1.32564, "bpb": 1.9125, "chunk_loss": 0.031403, "F": 0.16028, "G": 0.17354, "chunks": 164.12, "bytes_per_chunk": 6.239, "boundary_entropy": 0.11372, "router_noise": 0.0, "grad_norm": 0.6515, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.8711, "alpha": 0.03, "t": 201.8}
{"event": "probe", "step": 3400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2894.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166293, "cps": 22.609375, "mean_p": 0.180417, "boundary_entropy": 0.081924, "bpb": 2.133238}, "de": {"chunks": 3280.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159743, "cps": 25.625, "mean_p": 0.175262, "boundary_entropy": 0.105548, "bpb": 2.370796}, "vi": {"chunks": 4151.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22517, "cps": 32.429688, "mean_p": 0.17853, "boundary_entropy": 0.086566, "bpb": 1.963539}, "tr": {"chunks": 3020.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.166823, "cps": 23.59375, "mean_p": 0.17767, "boundary_entropy": 0.127631, "bpb": 2.377331}, "fi": {"chunks": 2870.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15257, "cps": 22.421875, "mean_p": 0.174452, "boundary_entropy": 0.140761, "bpb": 2.565576}, "ru": {"chunks": 6313.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.33314, "cps": 49.320312, "mean_p": 0.185227, "boundary_entropy": 0.102364, "bpb": 1.407119}, "el": {"chunks": 6353.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.309555, "cps": 49.632812, "mean_p": 0.176585, "boundary_entropy": 0.111099, "bpb": 1.484706}, "hi": {"chunks": 2346.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.139435, "cps": 18.328125, "mean_p": 0.074426, "boundary_entropy": 0.148349, "bpb": 1.273858}, "ta": {"chunks": 1855.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.099224, "cps": 14.492188, "mean_p": 0.065356, "boundary_entropy": 0.144142, "bpb": 1.050281}, "th": {"chunks": 2877.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.178862, "cps": 22.476562, "mean_p": 0.099708, "boundary_entropy": 0.190997, "bpb": 1.156893}, "zh": {"chunks": 2980.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.504572, "cps": 23.28125, "mean_p": 0.211686, "boundary_entropy": 0.137036, "bpb": 3.201497}, "my": {"chunks": 2232.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.117796, "cps": 17.4375, "mean_p": 0.079348, "boundary_entropy": 0.17517, "bpb": 1.157191}}, "bpb": {"en": 2.13324, "de": 2.3708, "vi": 1.96354, "tr": 2.37733, "fi": 2.56558, "ru": 1.40712, "el": 1.48471, "hi": 1.27386, "ta": 1.05028, "th": 1.15689, "zh": 3.2015, "my": 1.15719}, "gini_cpc": 0.261394, "gini_cps": 0.205256, "bpic": 9.321634, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9606098376672191, "vi": 1.3540515069195658, "tr": 1.0031872063113012, "fi": 0.9174778891190464, "ru": 2.0033284951304946, "el": 1.8615024133423752, "hi": 0.8384912371859182, "ta": 0.59668351282629, "th": 1.0755841408344018, "zh": 3.0342294191586174, "my": 0.7083638792321296}, "t": 204.7}
{"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.14432, "de": 2.37959, "vi": 1.9663, "tr": 2.37881, "fi": 2.57541, "ru": 1.38422, "el": 1.47289, "hi": 1.2713, "ta": 1.05578, "th": 1.17533, "zh": 3.19853, "my": 1.16659}, "bytes_per_chunk": {"en": 5.674, "de": 6.291, "vi": 5.837, "tr": 6.297, "fi": 6.604, "ru": 5.556, "el": 5.848, "hi": 15.451, "ta": 24.869, "th": 14.534, "zh": 5.223, "my": 21.999}, "bpb_mean": 1.84742, "t": 206.0}
{"event": "train", "step": 3400, "phase": 2, "objective": "parity_a", "lm_loss": 1.36088, "bpb": 1.96334, "chunk_loss": 0.030694, "F": 0.1571, "G": 0.17178, "chunks": 160.88, "bytes_per_chunk": 6.365, "boundary_entropy": 0.10607, "router_noise": 0.0, "grad_norm": 0.7034, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.085, "n_target_mean": 8.9964, "alpha": 0.03, "t": 206.0}
{"event": "train", "step": 3450, "phase": 2, "objective": "parity_a", "lm_loss": 1.37143, "bpb": 1.97855, "chunk_loss": 0.030674, "F": 0.14883, "G": 0.16504, "chunks": 152.41, "bytes_per_chunk": 6.719, "boundary_entropy": 0.11245, "router_noise": 0.0, "grad_norm": 0.7227, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.522, "alpha": 0.03, "t": 208.0}
{"event": "train", "step": 3500, "phase": 2, "objective": "parity_a", "lm_loss": 1.43659, "bpb": 2.07257, "chunk_loss": 0.031174, "F": 0.16318, "G": 0.1758, "chunks": 167.09, "bytes_per_chunk": 6.128, "boundary_entropy": 0.10578, "router_noise": 0.0, "grad_norm": 0.7167, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 9.4057, "alpha": 0.03, "t": 210.2}
{"event": "train", "step": 3550, "phase": 2, "objective": "parity_a", "lm_loss": 1.39906, "bpb": 2.01842, "chunk_loss": 0.030588, "F": 0.15546, "G": 0.17132, "chunks": 159.19, "bytes_per_chunk": 6.433, "boundary_entropy": 0.10919, "router_noise": 0.0, "grad_norm": 0.6551, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.046, "n_target_mean": 8.9203, "alpha": 0.03, "t": 212.5}
{"event": "probe", "step": 3600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2897.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166466, "cps": 22.632812, "mean_p": 0.179924, "boundary_entropy": 0.078542, "bpb": 2.116236}, "de": {"chunks": 3280.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159743, "cps": 25.625, "mean_p": 0.17529, "boundary_entropy": 0.103135, "bpb": 2.351919}, "vi": {"chunks": 4155.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225386, "cps": 32.460938, "mean_p": 0.17742, "boundary_entropy": 0.081182, "bpb": 1.956308}, "tr": {"chunks": 3049.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168425, "cps": 23.820312, "mean_p": 0.174089, "boundary_entropy": 0.114, "bpb": 2.352425}, "fi": {"chunks": 2858.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.151932, "cps": 22.328125, "mean_p": 0.167332, "boundary_entropy": 0.122889, "bpb": 2.531181}, "ru": {"chunks": 6261.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.330396, "cps": 48.914062, "mean_p": 0.180297, "boundary_entropy": 0.108103, "bpb": 1.408688}, "el": {"chunks": 6321.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.307996, "cps": 49.382812, "mean_p": 0.171735, "boundary_entropy": 0.1023, "bpb": 1.469625}, "hi": {"chunks": 3338.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.198395, "cps": 26.078125, "mean_p": 0.083791, "boundary_entropy": 0.147683, "bpb": 1.257582}, "ta": {"chunks": 3217.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.172078, "cps": 25.132812, "mean_p": 0.07536, "boundary_entropy": 0.145263, "bpb": 1.021759}, "th": {"chunks": 2744.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170594, "cps": 21.4375, "mean_p": 0.088118, "boundary_entropy": 0.167989, "bpb": 1.150032}, "zh": {"chunks": 3023.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.511852, "cps": 23.617188, "mean_p": 0.212749, "boundary_entropy": 0.133237, "bpb": 3.178562}, "my": {"chunks": 3090.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.163078, "cps": 24.140625, "mean_p": 0.089595, "boundary_entropy": 0.168073, "bpb": 1.152299}}, "bpb": {"en": 2.11624, "de": 2.35192, "vi": 1.95631, "tr": 2.35243, "fi": 2.53118, "ru": 1.40869, "el": 1.46963, "hi": 1.25758, "ta": 1.02176, "th": 1.15003, "zh": 3.17856, "my": 1.1523}, "gini_cpc": 0.210765, "gini_cps": 0.152931, "bpic": 8.67635, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.959615074286825, "vi": 1.3539527577278254, "tr": 1.0117716306401574, "fi": 0.912695619252148, "ru": 1.9847696689232697, "el": 1.8502080669875234, "hi": 1.1918096139830054, "ta": 1.0337159476334816, "th": 1.024798938243275, "zh": 3.0748244765741406, "my": 0.9796495162277239}, "t": 215.8}
{"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.12941, "de": 2.35856, "vi": 1.95889, "tr": 2.35653, "fi": 2.54076, "ru": 1.38653, "el": 1.45858, "hi": 1.2546, "ta": 1.02902, "th": 1.16927, "zh": 3.17866, "my": 1.16108}, "bytes_per_chunk": {"en": 5.674, "de": 6.296, "vi": 5.834, "tr": 6.243, "fi": 6.625, "ru": 5.62, "el": 5.928, "hi": 11.279, "ta": 14.81, "th": 15.376, "zh": 5.198, "my": 16.774}, "bpb_mean": 1.83182, "t": 217.4}
{"event": "train", "step": 3600, "phase": 2, "objective": "parity_a", "lm_loss": 1.44655, "bpb": 2.08692, "chunk_loss": 0.031667, "F": 0.16879, "G": 0.17729, "chunks": 172.84, "bytes_per_chunk": 5.924, "boundary_entropy": 0.10306, "router_noise": 0.0, "grad_norm": 1.1933, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.099, "n_target_mean": 9.551, "alpha": 0.03, "t": 217.4}
{"event": "train", "step": 3650, "phase": 2, "objective": "parity_a", "lm_loss": 1.32823, "bpb": 1.91624, "chunk_loss": 0.031086, "F": 0.16458, "G": 0.17604, "chunks": 168.53, "bytes_per_chunk": 6.076, "boundary_entropy": 0.09435, "router_noise": 0.0, "grad_norm": 0.6572, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.046, "n_target_mean": 8.7039, "alpha": 0.03, "t": 219.7}
{"event": "train", "step": 3700, "phase": 2, "objective": "parity_a", "lm_loss": 1.58479, "bpb": 2.28638, "chunk_loss": 0.030599, "F": 0.16742, "G": 0.18217, "chunks": 171.44, "bytes_per_chunk": 5.973, "boundary_entropy": 0.09818, "router_noise": 0.0, "grad_norm": 0.7243, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.047, "n_target_mean": 7.5805, "alpha": 0.03, "t": 222.0}
{"event": "train", "step": 3750, "phase": 2, "objective": "parity_a", "lm_loss": 1.30228, "bpb": 1.87879, "chunk_loss": 0.030866, "F": 0.14908, "G": 0.16352, "chunks": 152.66, "bytes_per_chunk": 6.708, "boundary_entropy": 0.10414, "router_noise": 0.0, "grad_norm": 0.6721, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.048, "n_target_mean": 10.2592, "alpha": 0.03, "t": 224.4}
{"event": "probe", "step": 3800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2893.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166236, "cps": 22.601562, "mean_p": 0.178609, "boundary_entropy": 0.07214, "bpb": 2.100971}, "de": {"chunks": 3283.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159889, "cps": 25.648438, "mean_p": 0.17273, "boundary_entropy": 0.091242, "bpb": 2.33017}, "vi": {"chunks": 4150.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225115, "cps": 32.421875, "mean_p": 0.177226, "boundary_entropy": 0.072617, "bpb": 1.951941}, "tr": {"chunks": 3061.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169088, "cps": 23.914062, "mean_p": 0.175568, "boundary_entropy": 0.112637, "bpb": 2.333366}, "fi": {"chunks": 2871.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152623, "cps": 22.429688, "mean_p": 0.169515, "boundary_entropy": 0.124593, "bpb": 2.535389}, "ru": {"chunks": 6314.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.333193, "cps": 49.328125, "mean_p": 0.182206, "boundary_entropy": 0.103443, "bpb": 1.381134}, "el": {"chunks": 6298.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.306875, "cps": 49.203125, "mean_p": 0.173183, "boundary_entropy": 0.100775, "bpb": 1.442219}, "hi": {"chunks": 3166.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.188172, "cps": 24.734375, "mean_p": 0.082614, "boundary_entropy": 0.145598, "bpb": 1.240737}, "ta": {"chunks": 3838.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.205296, "cps": 29.984375, "mean_p": 0.089007, "boundary_entropy": 0.151195, "bpb": 1.012688}, "th": {"chunks": 3004.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.186758, "cps": 23.46875, "mean_p": 0.09423, "boundary_entropy": 0.167102, "bpb": 1.127417}, "zh": {"chunks": 3007.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.509143, "cps": 23.492188, "mean_p": 0.21098, "boundary_entropy": 0.12707, "bpb": 3.169001}, "my": {"chunks": 2436.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.128562, "cps": 19.03125, "mean_p": 0.076374, "boundary_entropy": 0.154684, "bpb": 1.13459}}, "bpb": {"en": 2.10097, "de": 2.33017, "vi": 1.95194, "tr": 2.33337, "fi": 2.53539, "ru": 1.38113, "el": 1.44222, "hi": 1.24074, "ta": 1.01269, "th": 1.12742, "zh": 3.169, "my": 1.13459}, "gini_cpc": 0.221072, "gini_cps": 0.164337, "bpic": 8.659123, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9618207941894991, "vi": 1.354193239369911, "tr": 1.0171581066279738, "fi": 0.9181148155162833, "ru": 2.00433841307423, "el": 1.8460246577575388, "hi": 1.131961156431803, "ta": 1.234966529773197, "th": 1.1234520722041272, "zh": 3.0627790798790455, "my": 0.7733740480340942}, "t": 228.0}
{"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.11327, "de": 2.3389, "vi": 1.95385, "tr": 2.33686, "fi": 2.54471, "ru": 1.35868, "el": 1.43026, "hi": 1.23835, "ta": 1.0207, "th": 1.14683, "zh": 3.16941, "my": 1.14328}, "bytes_per_chunk": {"en": 5.676, "de": 6.292, "vi": 5.833, "tr": 6.207, "fi": 6.621, "ru": 5.574, "el": 5.987, "hi": 11.818, "ta": 12.54, "th": 14.568, "zh": 5.214, "my": 20.8}, "bpb_mean": 1.81626, "t": 229.4}
{"event": "train", "step": 3800, "phase": 2, "objective": "parity_a", "lm_loss": 1.37353, "bpb": 1.98158, "chunk_loss": 0.03066, "F": 0.16351, "G": 0.17364, "chunks": 167.44, "bytes_per_chunk": 6.116, "boundary_entropy": 0.09164, "router_noise": 0.0, "grad_norm": 0.7097, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.1, "n_target_mean": 8.2652, "alpha": 0.03, "t": 229.4}
{"event": "train", "step": 3850, "phase": 2, "objective": "parity_a", "lm_loss": 1.473, "bpb": 2.12508, "chunk_loss": 0.030469, "F": 0.15359, "G": 0.16915, "chunks": 157.28, "bytes_per_chunk": 6.511, "boundary_entropy": 0.1073, "router_noise": 0.0, "grad_norm": 0.6846, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.045, "n_target_mean": 9.011, "alpha": 0.03, "t": 231.7}
{"event": "train", "step": 3900, "phase": 2, "objective": "parity_a", "lm_loss": 1.48161, "bpb": 2.13751, "chunk_loss": 0.030712, "F": 0.15973, "G": 0.17271, "chunks": 163.56, "bytes_per_chunk": 6.261, "boundary_entropy": 0.10445, "router_noise": 0.0, "grad_norm": 0.7001, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 8.7756, "alpha": 0.03, "t": 233.9}
{"event": "train", "step": 3950, "phase": 2, "objective": "parity_a", "lm_loss": 1.41899, "bpb": 2.04718, "chunk_loss": 0.030672, "F": 0.15662, "G": 0.16999, "chunks": 160.38, "bytes_per_chunk": 6.385, "boundary_entropy": 0.10177, "router_noise": 0.0, "grad_norm": 0.6876, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.045, "n_target_mean": 8.8445, "alpha": 0.03, "t": 236.1}
{"event": "probe", "step": 4000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2901.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166695, "cps": 22.664062, "mean_p": 0.179226, "boundary_entropy": 0.073871, "bpb": 2.084828}, "de": {"chunks": 3292.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160327, "cps": 25.71875, "mean_p": 0.17455, "boundary_entropy": 0.097527, "bpb": 2.312926}, "vi": {"chunks": 4153.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225278, "cps": 32.445312, "mean_p": 0.175258, "boundary_entropy": 0.07443, "bpb": 1.933234}, "tr": {"chunks": 3040.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167928, "cps": 23.75, "mean_p": 0.172316, "boundary_entropy": 0.108684, "bpb": 2.302224}, "fi": {"chunks": 2882.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.153208, "cps": 22.515625, "mean_p": 0.168294, "boundary_entropy": 0.11778, "bpb": 2.497615}, "ru": {"chunks": 6310.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.332982, "cps": 49.296875, "mean_p": 0.181633, "boundary_entropy": 0.101303, "bpb": 1.369356}, "el": {"chunks": 6244.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.304244, "cps": 48.78125, "mean_p": 0.172405, "boundary_entropy": 0.095364, "bpb": 1.426313}, "hi": {"chunks": 2473.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.146984, "cps": 19.320312, "mean_p": 0.07789, "boundary_entropy": 0.150893, "bpb": 1.244218}, "ta": {"chunks": 2582.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.138112, "cps": 20.171875, "mean_p": 0.070994, "boundary_entropy": 0.145137, "bpb": 1.027125}, "th": {"chunks": 2885.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.17936, "cps": 22.539062, "mean_p": 0.090651, "boundary_entropy": 0.166312, "bpb": 1.128468}, "zh": {"chunks": 2975.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503725, "cps": 23.242188, "mean_p": 0.206294, "boundary_entropy": 0.119436, "bpb": 3.141713}, "my": {"chunks": 2081.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.109827, "cps": 16.257812, "mean_p": 0.074732, "boundary_entropy": 0.16354, "bpb": 1.136234}}, "bpb": {"en": 2.08483, "de": 2.31293, "vi": 1.93323, "tr": 2.30222, "fi": 2.49761, "ru": 1.36936, "el": 1.42631, "hi": 1.24422, "ta": 1.02712, "th": 1.12847, "zh": 3.14171, "my": 1.13623}, "gini_cpc": 0.244856, "gini_cps": 0.187407, "bpic": 9.177412, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9617978696084407, "vi": 1.3514350569049869, "tr": 1.007394147735916, "fi": 0.9190909387271864, "ru": 1.9975448371455935, "el": 1.8251494867337028, "hi": 0.8817499320058205, "ta": 0.8285279425646495, "th": 1.075972430588661, "zh": 3.0218292371594835, "my": 0.6588477930891905}, "t": 239.6}
{"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.09845, "de": 2.32149, "vi": 1.93477, "tr": 2.30646, "fi": 2.50732, "ru": 1.34576, "el": 1.41596, "hi": 1.24092, "ta": 1.03451, "th": 1.14611, "zh": 3.13895, "my": 1.14408}, "bytes_per_chunk": {"en": 5.666, "de": 6.284, "vi": 5.836, "tr": 6.269, "fi": 6.581, "ru": 5.576, "el": 5.997, "hi": 15.067, "ta": 18.513, "th": 15.222, "zh": 5.258, "my": 23.704}, "bpb_mean": 1.8029, "t": 241.0}
{"event": "train", "step": 4000, "phase": 2, "objective": "parity_a", "lm_loss": 1.39745, "bpb": 2.01609, "chunk_loss": 0.030875, "F": 0.16397, "G": 0.17691, "chunks": 167.91, "bytes_per_chunk": 6.099, "boundary_entropy": 0.10228, "router_noise": 0.0, "grad_norm": 0.748, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.097, "n_target_mean": 8.8695, "alpha": 0.03, "t": 241.0}
{"event": "train", "step": 4050, "phase": 2, "objective": "parity_a", "lm_loss": 1.296, "bpb": 1.86974, "chunk_loss": 0.030983, "F": 0.15726, "G": 0.16777, "chunks": 161.03, "bytes_per_chunk": 6.359, "boundary_entropy": 0.09681, "router_noise": 0.0, "grad_norm": 0.709, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.3813, "alpha": 0.03, "t": 243.0}
{"event": "train", "step": 4100, "phase": 2, "objective": "parity_a", "lm_loss": 1.44216, "bpb": 2.0806, "chunk_loss": 0.030663, "F": 0.1608, "G": 0.17266, "chunks": 164.66, "bytes_per_chunk": 6.219, "boundary_entropy": 0.09626, "router_noise": 0.0, "grad_norm": 0.6973, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.3701, "alpha": 0.03, "t": 245.0}
{"event": "train", "step": 4150, "phase": 2, "objective": "parity_a", "lm_loss": 1.23678, "bpb": 1.7843, "chunk_loss": 0.031066, "F": 0.15393, "G": 0.16403, "chunks": 157.62, "bytes_per_chunk": 6.496, "boundary_entropy": 0.10396, "router_noise": 0.0, "grad_norm": 0.6652, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 10.0005, "alpha": 0.03, "t": 247.1}
{"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2897.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166466, "cps": 22.632812, "mean_p": 0.178849, "boundary_entropy": 0.071775, "bpb": 2.072872}, "de": {"chunks": 3276.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159548, "cps": 25.59375, "mean_p": 0.171202, "boundary_entropy": 0.085793, "bpb": 2.302232}, "vi": {"chunks": 4149.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225061, "cps": 32.414062, "mean_p": 0.174185, "boundary_entropy": 0.079675, "bpb": 1.923337}, "tr": {"chunks": 3038.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167817, "cps": 23.734375, "mean_p": 0.170872, "boundary_entropy": 0.103974, "bpb": 2.29877}, "fi": {"chunks": 2894.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.153846, "cps": 22.609375, "mean_p": 0.167529, "boundary_entropy": 0.115079, "bpb": 2.471745}, "ru": {"chunks": 6292.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.332032, "cps": 49.15625, "mean_p": 0.182937, "boundary_entropy": 0.105015, "bpb": 1.355449}, "el": {"chunks": 6256.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.304829, "cps": 48.875, "mean_p": 0.172303, "boundary_entropy": 0.091733, "bpb": 1.413032}, "hi": {"chunks": 2185.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.129866, "cps": 17.070312, "mean_p": 0.077973, "boundary_entropy": 0.15163, "bpb": 1.233498}, "ta": {"chunks": 2787.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.149077, "cps": 21.773438, "mean_p": 0.076711, "boundary_entropy": 0.14368, "bpb": 1.002114}, "th": {"chunks": 3246.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.201803, "cps": 25.359375, "mean_p": 0.099307, "boundary_entropy": 0.171165, "bpb": 1.11386}, "zh": {"chunks": 2997.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50745, "cps": 23.414062, "mean_p": 0.206609, "boundary_entropy": 0.119156, "bpb": 3.128623}, "my": {"chunks": 2521.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.133048, "cps": 19.695312, "mean_p": 0.076683, "boundary_entropy": 0.153513, "bpb": 1.122475}}, "bpb": {"en": 2.07287, "de": 2.30223, "vi": 1.92334, "tr": 2.29877, "fi": 2.47174, "ru": 1.35545, "el": 1.41303, "hi": 1.2335, "ta": 1.00211, "th": 1.11386, "zh": 3.12862, "my": 1.12248}, "gini_cpc": 0.238159, "gini_cps": 0.178907, "bpic": 9.022074, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9584448120011093, "vi": 1.3519975912906732, "tr": 1.0081214214118723, "fi": 0.9241921350999708, "ru": 1.9945968306764434, "el": 1.8311820387713886, "hi": 0.7801390073555623, "ta": 0.8955444034984498, "th": 1.21228037665367, "zh": 3.048378748360139, "my": 0.7992545082233307}, "t": 250.1}
{"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.08532, "de": 2.31065, "vi": 1.926, "tr": 2.30003, "fi": 2.48362, "ru": 1.33131, "el": 1.4027, "hi": 1.23174, "ta": 1.00988, "th": 1.13158, "zh": 3.12794, "my": 1.13104}, "bytes_per_chunk": {"en": 5.674, "de": 6.3, "vi": 5.834, "tr": 6.259, "fi": 6.548, "ru": 5.587, "el": 5.998, "hi": 17.261, "ta": 17.677, "th": 13.82, "zh": 5.24, "my": 20.068}, "bpb_mean": 1.78932, "t": 251.4}
{"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 251.4}
{"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.256, "bpb": 1.81203, "chunk_loss": 0.030391, "F": 0.15393, "G": 0.16635, "chunks": 157.62, "bytes_per_chunk": 6.496, "boundary_entropy": 0.09596, "router_noise": 0.0, "grad_norm": 0.7, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 251.4}
{"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.27437, "bpb": 1.83853, "chunk_loss": 0.03014, "F": 0.16409, "G": 0.17299, "chunks": 168.03, "bytes_per_chunk": 6.094, "boundary_entropy": 0.08869, "router_noise": 0.0, "grad_norm": 0.6934, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 253.4}
{"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.34974, "bpb": 1.94726, "chunk_loss": 0.030113, "F": 0.16385, "G": 0.1819, "chunks": 167.78, "bytes_per_chunk": 6.103, "boundary_entropy": 0.11336, "router_noise": 0.0, "grad_norm": 0.7214, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 255.5}
{"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.32464, "bpb": 1.91106, "chunk_loss": 0.030092, "F": 0.16293, "G": 0.18106, "chunks": 166.84, "bytes_per_chunk": 6.137, "boundary_entropy": 0.11331, "router_noise": 0.0, "grad_norm": 0.6931, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 257.4}
{"event": "probe", "step": 4400, "phase": 3, "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": 0.181209, "boundary_entropy": 0.081552, "bpb": 2.063351}, "de": {"chunks": 3279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159694, "cps": 25.617188, "mean_p": 0.175236, "boundary_entropy": 0.099788, "bpb": 2.28872}, "vi": {"chunks": 4174.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226417, "cps": 32.609375, "mean_p": 0.1821, "boundary_entropy": 0.090263, "bpb": 1.919503}, "tr": {"chunks": 3125.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172623, "cps": 24.414062, "mean_p": 0.182833, "boundary_entropy": 0.124704, "bpb": 2.277358}, "fi": {"chunks": 2876.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152889, "cps": 22.46875, "mean_p": 0.168409, "boundary_entropy": 0.122086, "bpb": 2.459291}, "ru": {"chunks": 6342.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.33467, "cps": 49.546875, "mean_p": 0.191562, "boundary_entropy": 0.103267, "bpb": 1.344775}, "el": {"chunks": 6450.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.314282, "cps": 50.390625, "mean_p": 0.180985, "boundary_entropy": 0.093734, "bpb": 1.398941}, "hi": {"chunks": 4874.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.289688, "cps": 38.078125, "mean_p": 0.142847, "boundary_entropy": 0.226975, "bpb": 1.217651}, "ta": {"chunks": 6170.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.330035, "cps": 48.203125, "mean_p": 0.153946, "boundary_entropy": 0.228048, "bpb": 0.986595}, "th": {"chunks": 6644.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.413056, "cps": 51.90625, "mean_p": 0.265671, "boundary_entropy": 0.426819, "bpb": 1.105255}, "zh": {"chunks": 2989.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506095, "cps": 23.351562, "mean_p": 0.206962, "boundary_entropy": 0.11927, "bpb": 3.117886}, "my": {"chunks": 8713.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.459837, "cps": 68.070312, "mean_p": 0.282312, "boundary_entropy": 0.460911, "bpb": 1.115543}}, "bpb": {"en": 2.06335, "de": 2.28872, "vi": 1.9195, "tr": 2.27736, "fi": 2.45929, "ru": 1.34478, "el": 1.39894, "hi": 1.21765, "ta": 0.9866, "th": 1.10526, "zh": 3.11789, "my": 1.11554}, "gini_cpc": 0.223646, "gini_cps": 0.210274, "bpic": 6.557445, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9616461272486168, "vi": 1.3634385848342112, "tr": 1.039503001567157, "fi": 0.9206684817228528, "ru": 2.0153166866001406, "el": 1.8925403467491198, "hi": 1.7444427831791751, "ta": 1.9874031414872324, "th": 2.487338178556178, "zh": 3.0476055082099376, "my": 2.769048837357094}, "t": 260.4}
{"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.07588, "de": 2.29696, "vi": 1.92176, "tr": 2.28121, "fi": 2.46919, "ru": 1.32258, "el": 1.38931, "hi": 1.21382, "ta": 0.99361, "th": 1.12358, "zh": 3.11342, "my": 1.12412}, "bytes_per_chunk": {"en": 5.675, "de": 6.294, "vi": 5.829, "tr": 6.138, "fi": 6.582, "ru": 5.547, "el": 5.823, "hi": 8.54, "ta": 8.206, "th": 6.815, "zh": 5.271, "my": 6.87}, "bpb_mean": 1.77712, "t": 261.7}
{"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.3176, "bpb": 1.90089, "chunk_loss": 0.0301, "F": 0.1734, "G": 0.19578, "chunks": 177.56, "bytes_per_chunk": 5.767, "boundary_entropy": 0.13134, "router_noise": 0.0, "grad_norm": 0.687, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 261.7}
{"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.16596, "bpb": 1.68212, "chunk_loss": 0.030024, "F": 0.16559, "G": 0.20593, "chunks": 169.56, "bytes_per_chunk": 6.039, "boundary_entropy": 0.16893, "router_noise": 0.0, "grad_norm": 0.8071, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 263.7}
{"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.29162, "bpb": 1.86342, "chunk_loss": 0.030025, "F": 0.16653, "G": 0.18893, "chunks": 170.53, "bytes_per_chunk": 6.005, "boundary_entropy": 0.12453, "router_noise": 0.0, "grad_norm": 0.7004, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 265.7}
{"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.29291, "bpb": 1.86527, "chunk_loss": 0.030008, "F": 0.17197, "G": 0.19347, "chunks": 176.09, "bytes_per_chunk": 5.815, "boundary_entropy": 0.112, "router_noise": 0.0, "grad_norm": 0.6666, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 267.7}
{"event": "probe", "step": 4600, "phase": 3, "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": 0.182256, "boundary_entropy": 0.08458, "bpb": 2.052075}, "de": {"chunks": 3288.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160132, "cps": 25.6875, "mean_p": 0.17715, "boundary_entropy": 0.103256, "bpb": 2.28347}, "vi": {"chunks": 4170.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.2262, "cps": 32.578125, "mean_p": 0.185953, "boundary_entropy": 0.091833, "bpb": 1.90894}, "tr": {"chunks": 3133.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.173065, "cps": 24.476562, "mean_p": 0.184882, "boundary_entropy": 0.127428, "bpb": 2.271394}, "fi": {"chunks": 2882.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.153208, "cps": 22.515625, "mean_p": 0.169883, "boundary_entropy": 0.124572, "bpb": 2.449114}, "ru": {"chunks": 6384.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.336887, "cps": 49.875, "mean_p": 0.192707, "boundary_entropy": 0.100102, "bpb": 1.335492}, "el": {"chunks": 6587.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.320957, "cps": 51.460938, "mean_p": 0.184319, "boundary_entropy": 0.092524, "bpb": 1.381578}, "hi": {"chunks": 5793.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.344309, "cps": 45.257812, "mean_p": 0.202378, "boundary_entropy": 0.314818, "bpb": 1.207272}, "ta": {"chunks": 7038.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.376464, "cps": 54.984375, "mean_p": 0.198102, "boundary_entropy": 0.292947, "bpb": 0.977407}, "th": {"chunks": 5781.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.359403, "cps": 45.164062, "mean_p": 0.28469, "boundary_entropy": 0.463927, "bpb": 1.094701}, "zh": {"chunks": 3003.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.508466, "cps": 23.460938, "mean_p": 0.209324, "boundary_entropy": 0.126551, "bpb": 3.111909}, "my": {"chunks": 9246.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.487967, "cps": 72.234375, "mean_p": 0.339112, "boundary_entropy": 0.510994, "bpb": 1.105424}}, "bpb": {"en": 2.05208, "de": 2.28347, "vi": 1.90894, "tr": 2.27139, "fi": 2.44911, "ru": 1.33549, "el": 1.38158, "hi": 1.20727, "ta": 0.97741, "th": 1.0947, "zh": 3.11191, "my": 1.10542}, "gini_cpc": 0.22214, "gini_cps": 0.218415, "bpic": 6.375311, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9632856446223188, "vi": 1.360719471848802, "tr": 1.041083419818831, "fi": 0.9216324968017863, "ru": 2.0265594597823697, "el": 1.9307342681246282, "hi": 2.071210037653012, "ta": 2.2646415936799795, "th": 2.162009463852217, "zh": 3.0587048809034827, "my": 2.935392630592461}, "t": 270.6}
{"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.06388, "de": 2.29181, "vi": 1.90989, "tr": 2.27435, "fi": 2.46019, "ru": 1.31301, "el": 1.37363, "hi": 1.20449, "ta": 0.98456, "th": 1.11226, "zh": 3.10697, "my": 1.11403}, "bytes_per_chunk": {"en": 5.674, "de": 6.29, "vi": 5.829, "tr": 6.132, "fi": 6.581, "ru": 5.515, "el": 5.675, "hi": 7.276, "ta": 7.392, "th": 7.776, "zh": 5.26, "my": 7.062}, "bpb_mean": 1.76742, "t": 271.9}
{"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.43863, "bpb": 2.07551, "chunk_loss": 0.030119, "F": 0.17752, "G": 0.22072, "chunks": 181.78, "bytes_per_chunk": 5.633, "boundary_entropy": 0.16726, "router_noise": 0.0, "grad_norm": 0.6777, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 271.9}
{"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.33563, "bpb": 1.92691, "chunk_loss": 0.029958, "F": 0.16245, "G": 0.19893, "chunks": 166.34, "bytes_per_chunk": 6.156, "boundary_entropy": 0.16263, "router_noise": 0.0, "grad_norm": 0.7194, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 273.9}
{"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.39881, "bpb": 2.01805, "chunk_loss": 0.030173, "F": 0.17746, "G": 0.21358, "chunks": 181.72, "bytes_per_chunk": 5.635, "boundary_entropy": 0.15865, "router_noise": 0.0, "grad_norm": 0.7332, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 275.9}
{"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.36683, "bpb": 1.97193, "chunk_loss": 0.030106, "F": 0.17368, "G": 0.2033, "chunks": 177.84, "bytes_per_chunk": 5.758, "boundary_entropy": 0.13901, "router_noise": 0.0, "grad_norm": 0.6872, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 277.9}
{"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2897.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166466, "cps": 22.632812, "mean_p": 0.185334, "boundary_entropy": 0.09475, "bpb": 2.043804}, "de": {"chunks": 3288.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160132, "cps": 25.6875, "mean_p": 0.180712, "boundary_entropy": 0.115823, "bpb": 2.267759}, "vi": {"chunks": 4163.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22582, "cps": 32.523438, "mean_p": 0.188295, "boundary_entropy": 0.098141, "bpb": 1.900577}, "tr": {"chunks": 3122.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172458, "cps": 24.390625, "mean_p": 0.186166, "boundary_entropy": 0.133892, "bpb": 2.252608}, "fi": {"chunks": 2865.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152305, "cps": 22.382812, "mean_p": 0.171938, "boundary_entropy": 0.134477, "bpb": 2.428392}, "ru": {"chunks": 6401.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.337784, "cps": 50.007812, "mean_p": 0.193224, "boundary_entropy": 0.099241, "bpb": 1.325931}, "el": {"chunks": 6626.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.322857, "cps": 51.765625, "mean_p": 0.186087, "boundary_entropy": 0.098628, "bpb": 1.371985}, "hi": {"chunks": 5785.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.343834, "cps": 45.195312, "mean_p": 0.228636, "boundary_entropy": 0.369366, "bpb": 1.1917}, "ta": {"chunks": 7812.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.417866, "cps": 61.03125, "mean_p": 0.234023, "boundary_entropy": 0.350203, "bpb": 0.970837}, "th": {"chunks": 6750.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.419646, "cps": 52.734375, "mean_p": 0.329417, "boundary_entropy": 0.513032, "bpb": 1.087587}, "zh": {"chunks": 2965.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.502032, "cps": 23.164062, "mean_p": 0.205368, "boundary_entropy": 0.119651, "bpb": 3.088046}, "my": {"chunks": 9762.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.515199, "cps": 76.265625, "mean_p": 0.34441, "boundary_entropy": 0.522592, "bpb": 1.102427}}, "bpb": {"en": 2.0438, "de": 2.26776, "vi": 1.90058, "tr": 2.25261, "fi": 2.42839, "ru": 1.32593, "el": 1.37199, "hi": 1.1917, "ta": 0.97084, "th": 1.08759, "zh": 3.08805, "my": 1.10243}, "gini_cpc": 0.232782, "gini_cps": 0.233677, "bpic": 6.14679, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9619555988582561, "vi": 1.3565596463106948, "tr": 1.0359957464278686, "fi": 0.9149310528892247, "ru": 2.029150399421474, "el": 1.939484045540157, "hi": 2.0654938936164426, "ta": 2.510223494843879, "th": 2.5209157555182604, "zh": 3.0158301597890596, "my": 3.0949315784514693}, "t": 280.8}
{"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.05531, "de": 2.27664, "vi": 1.90142, "tr": 2.25671, "fi": 2.44002, "ru": 1.30399, "el": 1.36229, "hi": 1.18755, "ta": 0.97859, "th": 1.10488, "zh": 3.08666, "my": 1.11103}, "bytes_per_chunk": {"en": 5.672, "de": 6.283, "vi": 5.83, "tr": 6.153, "fi": 6.6, "ru": 5.498, "el": 5.655, "hi": 7.265, "ta": 6.665, "th": 6.694, "zh": 5.299, "my": 6.374}, "bpb_mean": 1.75542, "t": 282.1}
{"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.2487, "bpb": 1.80149, "chunk_loss": 0.030026, "F": 0.17197, "G": 0.20648, "chunks": 176.09, "bytes_per_chunk": 5.815, "boundary_entropy": 0.15396, "router_noise": 0.0, "grad_norm": 0.6954, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 282.2}
{"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.32939, "bpb": 1.9179, "chunk_loss": 0.030119, "F": 0.17313, "G": 0.20167, "chunks": 177.28, "bytes_per_chunk": 5.776, "boundary_entropy": 0.14107, "router_noise": 0.0, "grad_norm": 0.7074, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 284.1}
{"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.36931, "bpb": 1.9755, "chunk_loss": 0.030082, "F": 0.17651, "G": 0.20582, "chunks": 180.75, "bytes_per_chunk": 5.665, "boundary_entropy": 0.1368, "router_noise": 0.0, "grad_norm": 0.7159, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 286.1}
{"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.26112, "bpb": 1.81941, "chunk_loss": 0.029965, "F": 0.16919, "G": 0.23177, "chunks": 173.25, "bytes_per_chunk": 5.911, "boundary_entropy": 0.22538, "router_noise": 0.0, "grad_norm": 0.7597, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 288.1}
{"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2898.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166523, "cps": 22.640625, "mean_p": 0.184791, "boundary_entropy": 0.093364, "bpb": 2.036108}, "de": {"chunks": 3292.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160327, "cps": 25.71875, "mean_p": 0.180046, "boundary_entropy": 0.113678, "bpb": 2.257726}, "vi": {"chunks": 4170.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.2262, "cps": 32.578125, "mean_p": 0.18787, "boundary_entropy": 0.096836, "bpb": 1.892742}, "tr": {"chunks": 3114.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172016, "cps": 24.328125, "mean_p": 0.185585, "boundary_entropy": 0.131459, "bpb": 2.244103}, "fi": {"chunks": 2906.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.154484, "cps": 22.703125, "mean_p": 0.174157, "boundary_entropy": 0.134605, "bpb": 2.414496}, "ru": {"chunks": 6376.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.336464, "cps": 49.8125, "mean_p": 0.19195, "boundary_entropy": 0.099169, "bpb": 1.318463}, "el": {"chunks": 6399.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.311797, "cps": 49.992188, "mean_p": 0.182957, "boundary_entropy": 0.100381, "bpb": 1.367405}, "hi": {"chunks": 5650.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.33581, "cps": 44.140625, "mean_p": 0.237533, "boundary_entropy": 0.380588, "bpb": 1.187125}, "ta": {"chunks": 7714.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.412624, "cps": 60.265625, "mean_p": 0.243891, "boundary_entropy": 0.368807, "bpb": 0.967055}, "th": {"chunks": 5997.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.372832, "cps": 46.851562, "mean_p": 0.323129, "boundary_entropy": 0.516108, "bpb": 1.080577}, "zh": {"chunks": 2998.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.507619, "cps": 23.421875, "mean_p": 0.208428, "boundary_entropy": 0.126885, "bpb": 3.080897}, "my": {"chunks": 11182.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.590141, "cps": 87.359375, "mean_p": 0.359833, "boundary_entropy": 0.529261, "bpb": 1.100619}}, "bpb": {"en": 2.03611, "de": 2.25773, "vi": 1.89274, "tr": 2.2441, "fi": 2.4145, "ru": 1.31846, "el": 1.3674, "hi": 1.18713, "ta": 0.96706, "th": 1.08058, "zh": 3.0809, "my": 1.10062}, "gini_cpc": 0.244639, "gini_cps": 0.24471, "bpic": 6.1213, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9627935195769795, "vi": 1.3583717846993044, "tr": 1.0329844783970812, "fi": 0.9277040778197595, "ru": 2.0205278137410754, "el": 1.872392979487035, "hi": 2.016596987008512, "ta": 2.477877910469745, "th": 2.2389207567715763, "zh": 3.0483436502444436, "my": 3.543903200696163}, "t": 291.0}
{"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.04829, "de": 2.2663, "vi": 1.89454, "tr": 2.24732, "fi": 2.4263, "ru": 1.29659, "el": 1.35723, "hi": 1.18341, "ta": 0.97374, "th": 1.09849, "zh": 3.07798, "my": 1.10982}, "bytes_per_chunk": {"en": 5.674, "de": 6.29, "vi": 5.829, "tr": 6.16, "fi": 6.522, "ru": 5.525, "el": 5.87, "hi": 7.399, "ta": 6.704, "th": 7.589, "zh": 5.279, "my": 5.234}, "bpb_mean": 1.74833, "t": 292.3}
{"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.45792, "bpb": 2.10334, "chunk_loss": 0.030052, "F": 0.17239, "G": 0.19642, "chunks": 176.53, "bytes_per_chunk": 5.801, "boundary_entropy": 0.11707, "router_noise": 0.0, "grad_norm": 0.7185, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 292.4}
{"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.26333, "bpb": 1.8226, "chunk_loss": 0.02999, "F": 0.17252, "G": 0.21722, "chunks": 176.66, "bytes_per_chunk": 5.797, "boundary_entropy": 0.17847, "router_noise": 0.0, "grad_norm": 0.6889, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 294.3}
{"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.43827, "bpb": 2.07498, "chunk_loss": 0.030041, "F": 0.17432, "G": 0.20294, "chunks": 178.5, "bytes_per_chunk": 5.737, "boundary_entropy": 0.13304, "router_noise": 0.0, "grad_norm": 0.7286, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 296.3}
{"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.22294, "bpb": 1.76433, "chunk_loss": 0.030082, "F": 0.17697, "G": 0.22468, "chunks": 181.22, "bytes_per_chunk": 5.651, "boundary_entropy": 0.18983, "router_noise": 0.0, "grad_norm": 0.6627, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 298.3}
{"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2896.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166408, "cps": 22.625, "mean_p": 0.185098, "boundary_entropy": 0.094008, "bpb": 2.03104}, "de": {"chunks": 3287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160084, "cps": 25.679688, "mean_p": 0.180448, "boundary_entropy": 0.114537, "bpb": 2.250118}, "vi": {"chunks": 4172.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226309, "cps": 32.59375, "mean_p": 0.190095, "boundary_entropy": 0.102826, "bpb": 1.887905}, "tr": {"chunks": 3131.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172955, "cps": 24.460938, "mean_p": 0.187538, "boundary_entropy": 0.134881, "bpb": 2.23547}, "fi": {"chunks": 2904.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.154378, "cps": 22.6875, "mean_p": 0.174851, "boundary_entropy": 0.139023, "bpb": 2.409091}, "ru": {"chunks": 6374.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.336359, "cps": 49.796875, "mean_p": 0.193064, "boundary_entropy": 0.101419, "bpb": 1.315242}, "el": {"chunks": 6453.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.314428, "cps": 50.414062, "mean_p": 0.185502, "boundary_entropy": 0.101742, "bpb": 1.355753}, "hi": {"chunks": 6266.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.372422, "cps": 48.953125, "mean_p": 0.262321, "boundary_entropy": 0.406933, "bpb": 1.180371}, "ta": {"chunks": 8047.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.430436, "cps": 62.867188, "mean_p": 0.258434, "boundary_entropy": 0.386152, "bpb": 0.957728}, "th": {"chunks": 7293.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.453404, "cps": 56.976562, "mean_p": 0.352122, "boundary_entropy": 0.539988, "bpb": 1.075469}, "zh": {"chunks": 2979.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.504402, "cps": 23.273438, "mean_p": 0.206642, "boundary_entropy": 0.121783, "bpb": 3.072328}, "my": {"chunks": 10031.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.529396, "cps": 78.367188, "mean_p": 0.361928, "boundary_entropy": 0.529445, "bpb": 1.095027}}, "bpb": {"en": 2.03104, "de": 2.25012, "vi": 1.8879, "tr": 2.23547, "fi": 2.40909, "ru": 1.31524, "el": 1.35575, "hi": 1.18037, "ta": 0.95773, "th": 1.07547, "zh": 3.07233, "my": 1.09503}, "gini_cpc": 0.239483, "gini_cps": 0.23882, "bpic": 6.012266, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9619950992513602, "vi": 1.3599618339484556, "tr": 1.0393410458508907, "fi": 0.9277058415626686, "ru": 2.0212889765156925, "el": 1.8894977653423743, "hi": 2.238004112861517, "ta": 2.586628714948749, "th": 2.7246499303596052, "zh": 3.0311164622361755, "my": 3.1813131234713907}, "t": 301.2}
{"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.04279, "de": 2.25907, "vi": 1.89016, "tr": 2.23889, "fi": 2.41983, "ru": 1.29203, "el": 1.34445, "hi": 1.17706, "ta": 0.96504, "th": 1.09266, "zh": 3.06928, "my": 1.10392}, "bytes_per_chunk": {"en": 5.672, "de": 6.288, "vi": 5.827, "tr": 6.132, "fi": 6.539, "ru": 5.515, "el": 5.767, "hi": 6.67, "ta": 6.408, "th": 6.206, "zh": 5.286, "my": 6.003}, "bpb_mean": 1.74127, "t": 302.5}
{"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.34271, "bpb": 1.93712, "chunk_loss": 0.030065, "F": 0.17285, "G": 0.21276, "chunks": 177.0, "bytes_per_chunk": 5.785, "boundary_entropy": 0.16326, "router_noise": 0.0, "grad_norm": 0.752, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 302.6}
{"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.31101, "bpb": 1.89139, "chunk_loss": 0.029969, "F": 0.16785, "G": 0.21005, "chunks": 171.88, "bytes_per_chunk": 5.958, "boundary_entropy": 0.16508, "router_noise": 0.0, "grad_norm": 0.7458, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 304.5}
{"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.26598, "bpb": 1.82643, "chunk_loss": 0.030045, "F": 0.16748, "G": 0.20259, "chunks": 171.5, "bytes_per_chunk": 5.971, "boundary_entropy": 0.15558, "router_noise": 0.0, "grad_norm": 0.7155, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 306.5}
{"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.35512, "bpb": 1.95503, "chunk_loss": 0.030034, "F": 0.16965, "G": 0.20572, "chunks": 173.72, "bytes_per_chunk": 5.895, "boundary_entropy": 0.14769, "router_noise": 0.0, "grad_norm": 0.7162, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 308.5}
{"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2896.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166408, "cps": 22.625, "mean_p": 0.185235, "boundary_entropy": 0.094118, "bpb": 2.023398}, "de": {"chunks": 3284.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159938, "cps": 25.65625, "mean_p": 0.181054, "boundary_entropy": 0.11612, "bpb": 2.242528}, "vi": {"chunks": 4167.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226037, "cps": 32.554688, "mean_p": 0.189615, "boundary_entropy": 0.102533, "bpb": 1.881017}, "tr": {"chunks": 3200.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176766, "cps": 25.0, "mean_p": 0.191043, "boundary_entropy": 0.139082, "bpb": 2.228517}, "fi": {"chunks": 2886.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.153421, "cps": 22.546875, "mean_p": 0.17483, "boundary_entropy": 0.140643, "bpb": 2.401864}, "ru": {"chunks": 6387.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.337045, "cps": 49.898438, "mean_p": 0.195041, "boundary_entropy": 0.102342, "bpb": 1.30926}, "el": {"chunks": 6496.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.316523, "cps": 50.75, "mean_p": 0.187427, "boundary_entropy": 0.103344, "bpb": 1.349608}, "hi": {"chunks": 6769.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.402318, "cps": 52.882812, "mean_p": 0.276851, "boundary_entropy": 0.421022, "bpb": 1.172387}, "ta": {"chunks": 8563.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.458037, "cps": 66.898438, "mean_p": 0.268786, "boundary_entropy": 0.397932, "bpb": 0.957902}, "th": {"chunks": 8475.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.526888, "cps": 66.210938, "mean_p": 0.371565, "boundary_entropy": 0.554173, "bpb": 1.074702}, "zh": {"chunks": 2997.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50745, "cps": 23.414062, "mean_p": 0.207726, "boundary_entropy": 0.122672, "bpb": 3.064156}, "my": {"chunks": 12248.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.646401, "cps": 95.6875, "mean_p": 0.370312, "boundary_entropy": 0.533926, "bpb": 1.089465}}, "bpb": {"en": 2.0234, "de": 2.24253, "vi": 1.88102, "tr": 2.22852, "fi": 2.40186, "ru": 1.30926, "el": 1.34961, "hi": 1.17239, "ta": 0.9579, "th": 1.0747, "zh": 3.06416, "my": 1.08947}, "gini_cpc": 0.2671, "gini_cps": 0.270633, "bpic": 5.61346, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9611170994650035, "vi": 1.3583319659787187, "tr": 1.0622457191705048, "fi": 0.9219555987430654, "ru": 2.025411467368329, "el": 1.9020885609273306, "hi": 2.417658767947592, "ta": 2.7524918213130523, "th": 3.1662427203890933, "zh": 3.049431365331258, "my": 3.8844305788333764}, "t": 311.5}
{"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.035, "de": 2.25105, "vi": 1.88312, "tr": 2.23045, "fi": 2.41167, "ru": 1.28657, "el": 1.33924, "hi": 1.1692, "ta": 0.96455, "th": 1.09305, "zh": 3.06047, "my": 1.0987}, "bytes_per_chunk": {"en": 5.674, "de": 6.293, "vi": 5.827, "tr": 6.018, "fi": 6.563, "ru": 5.501, "el": 5.746, "hi": 6.18, "ta": 6.117, "th": 5.351, "zh": 5.293, "my": 4.897}, "bpb_mean": 1.73526, "t": 312.8}
{"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.22292, "bpb": 1.7643, "chunk_loss": 0.030221, "F": 0.17712, "G": 0.21923, "chunks": 181.38, "bytes_per_chunk": 5.646, "boundary_entropy": 0.17932, "router_noise": 0.0, "grad_norm": 0.7138, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 312.8}
{"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.26882, "bpb": 1.83052, "chunk_loss": 0.029829, "F": 0.16599, "G": 0.21829, "chunks": 169.97, "bytes_per_chunk": 6.025, "boundary_entropy": 0.18592, "router_noise": 0.0, "grad_norm": 0.7185, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 314.8}
{"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.45713, "bpb": 2.10219, "chunk_loss": 0.030048, "F": 0.17456, "G": 0.20767, "chunks": 178.75, "bytes_per_chunk": 5.729, "boundary_entropy": 0.14001, "router_noise": 0.0, "grad_norm": 0.7646, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 316.8}
{"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.30168, "bpb": 1.87793, "chunk_loss": 0.030016, "F": 0.16995, "G": 0.20763, "chunks": 174.03, "bytes_per_chunk": 5.884, "boundary_entropy": 0.1619, "router_noise": 0.0, "grad_norm": 0.7254, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 318.7}
{"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2897.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166466, "cps": 22.632812, "mean_p": 0.18589, "boundary_entropy": 0.095711, "bpb": 2.018604}, "de": {"chunks": 3277.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159597, "cps": 25.601562, "mean_p": 0.181491, "boundary_entropy": 0.117802, "bpb": 2.234343}, "vi": {"chunks": 4161.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225712, "cps": 32.507812, "mean_p": 0.1896, "boundary_entropy": 0.101724, "bpb": 1.877077}, "tr": {"chunks": 3165.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174833, "cps": 24.726562, "mean_p": 0.190318, "boundary_entropy": 0.140095, "bpb": 2.220087}, "fi": {"chunks": 2905.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.154431, "cps": 22.695312, "mean_p": 0.176215, "boundary_entropy": 0.142272, "bpb": 2.393076}, "ru": {"chunks": 6381.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.336728, "cps": 49.851562, "mean_p": 0.196035, "boundary_entropy": 0.102823, "bpb": 1.302649}, "el": {"chunks": 6580.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.320616, "cps": 51.40625, "mean_p": 0.188375, "boundary_entropy": 0.101937, "bpb": 1.340228}, "hi": {"chunks": 6144.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.365171, "cps": 48.0, "mean_p": 0.271614, "boundary_entropy": 0.415922, "bpb": 1.165482}, "ta": {"chunks": 8543.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.456967, "cps": 66.742188, "mean_p": 0.272858, "boundary_entropy": 0.397339, "bpb": 0.948711}, "th": {"chunks": 7661.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.476282, "cps": 59.851562, "mean_p": 0.363029, "boundary_entropy": 0.547034, "bpb": 1.06691}, "zh": {"chunks": 2997.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50745, "cps": 23.414062, "mean_p": 0.207682, "boundary_entropy": 0.121484, "bpb": 3.053723}, "my": {"chunks": 11600.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.612202, "cps": 90.625, "mean_p": 0.370043, "boundary_entropy": 0.533128, "bpb": 1.087975}}, "bpb": {"en": 2.0186, "de": 2.23434, "vi": 1.87708, "tr": 2.22009, "fi": 2.39308, "ru": 1.30265, "el": 1.34023, "hi": 1.16548, "ta": 0.94871, "th": 1.06691, "zh": 3.05372, "my": 1.08798}, "gini_cpc": 0.256544, "gini_cps": 0.260757, "bpic": 5.787592, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9587373775725382, "vi": 1.3559079241649774, "tr": 1.0502647461384382, "fi": 0.9277049593868055, "ru": 2.0228102950645876, "el": 1.9260194717256613, "hi": 2.1936723392185695, "ta": 2.7451151198734327, "th": 2.8611460152630213, "zh": 3.048378748360139, "my": 3.67764866933385}, "t": 321.7}
{"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.03093, "de": 2.24427, "vi": 1.87778, "tr": 2.22289, "fi": 2.40437, "ru": 1.27992, "el": 1.33013, "hi": 1.16159, "ta": 0.95623, "th": 1.08541, "zh": 3.05007, "my": 1.09659}, "bytes_per_chunk": {"en": 5.671, "de": 6.296, "vi": 5.829, "tr": 6.08, "fi": 6.515, "ru": 5.517, "el": 5.696, "hi": 6.785, "ta": 6.041, "th": 5.875, "zh": 5.291, "my": 5.246}, "bpb_mean": 1.72835, "t": 323.0}
{"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.30714, "bpb": 1.8858, "chunk_loss": 0.030104, "F": 0.17413, "G": 0.20949, "chunks": 178.31, "bytes_per_chunk": 5.743, "boundary_entropy": 0.14732, "router_noise": 0.0, "grad_norm": 0.7241, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 323.0}
{"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.21617, "bpb": 1.75456, "chunk_loss": 0.029946, "F": 0.16867, "G": 0.20851, "chunks": 172.72, "bytes_per_chunk": 5.929, "boundary_entropy": 0.16068, "router_noise": 0.0, "grad_norm": 0.7479, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 325.0}
{"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.27923, "bpb": 1.84553, "chunk_loss": 0.030121, "F": 0.17657, "G": 0.21953, "chunks": 180.81, "bytes_per_chunk": 5.663, "boundary_entropy": 0.16845, "router_noise": 0.0, "grad_norm": 0.755, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 326.9}
{"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.29508, "bpb": 1.8684, "chunk_loss": 0.02992, "F": 0.1683, "G": 0.2104, "chunks": 172.34, "bytes_per_chunk": 5.942, "boundary_entropy": 0.16233, "router_noise": 0.0, "grad_norm": 0.7313, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 328.9}
{"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2896.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166408, "cps": 22.625, "mean_p": 0.185762, "boundary_entropy": 0.095205, "bpb": 2.016613}, "de": {"chunks": 3288.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160132, "cps": 25.6875, "mean_p": 0.182394, "boundary_entropy": 0.119341, "bpb": 2.231505}, "vi": {"chunks": 4175.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226471, "cps": 32.617188, "mean_p": 0.190468, "boundary_entropy": 0.103105, "bpb": 1.874084}, "tr": {"chunks": 3161.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174612, "cps": 24.695312, "mean_p": 0.190826, "boundary_entropy": 0.140817, "bpb": 2.216017}, "fi": {"chunks": 2914.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.154909, "cps": 22.765625, "mean_p": 0.177739, "boundary_entropy": 0.144892, "bpb": 2.39134}, "ru": {"chunks": 6398.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.337625, "cps": 49.984375, "mean_p": 0.193514, "boundary_entropy": 0.097003, "bpb": 1.300276}, "el": {"chunks": 6486.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.316036, "cps": 50.671875, "mean_p": 0.187742, "boundary_entropy": 0.104055, "bpb": 1.337248}, "hi": {"chunks": 6308.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.374918, "cps": 49.28125, "mean_p": 0.279084, "boundary_entropy": 0.427216, "bpb": 1.164195}, "ta": {"chunks": 7824.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.418508, "cps": 61.125, "mean_p": 0.262213, "boundary_entropy": 0.396305, "bpb": 0.945415}, "th": {"chunks": 7326.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.455455, "cps": 57.234375, "mean_p": 0.360519, "boundary_entropy": 0.5481, "bpb": 1.063079}, "zh": {"chunks": 2974.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503556, "cps": 23.234375, "mean_p": 0.206036, "boundary_entropy": 0.119415, "bpb": 3.050493}, "my": {"chunks": 8734.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.460946, "cps": 68.234375, "mean_p": 0.355048, "boundary_entropy": 0.529918, "bpb": 1.08338}}, "bpb": {"en": 2.01661, "de": 2.23151, "vi": 1.87408, "tr": 2.21602, "fi": 2.39134, "ru": 1.30028, "el": 1.33725, "hi": 1.16419, "ta": 0.94542, "th": 1.06308, "zh": 3.05049, "my": 1.08338}, "gini_cpc": 0.227852, "gini_cps": 0.22267, "bpic": 6.142068, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9622877658468123, "vi": 1.3609397547302977, "tr": 1.0492995994681142, "fi": 0.9309004209068927, "ru": 2.028899728859021, "el": 1.8991604689308292, "hi": 2.2530050979780487, "ta": 2.514947566268051, "th": 2.7369786630761648, "zh": 3.026028989154208, "my": 2.769971968936211}, "t": 331.9}
{"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.02953, "de": 2.24159, "vi": 1.87575, "tr": 2.2187, "fi": 2.40195, "ru": 1.27757, "el": 1.32626, "hi": 1.16011, "ta": 0.95355, "th": 1.08082, "zh": 3.04583, "my": 1.09185}, "bytes_per_chunk": {"en": 5.671, "de": 6.286, "vi": 5.826, "tr": 6.041, "fi": 6.501, "ru": 5.5, "el": 5.744, "hi": 6.504, "ta": 6.631, "th": 6.177, "zh": 5.294, "my": 6.999}, "bpb_mean": 1.72529, "t": 333.2}
{"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.34221, "bpb": 1.93641, "chunk_loss": 0.030076, "F": 0.1687, "G": 0.21474, "chunks": 172.75, "bytes_per_chunk": 5.928, "boundary_entropy": 0.17946, "router_noise": 0.0, "grad_norm": 0.7388, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 333.2}
{"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.29944, "bpb": 1.8747, "chunk_loss": 0.03015, "F": 0.17447, "G": 0.21641, "chunks": 178.66, "bytes_per_chunk": 5.732, "boundary_entropy": 0.17535, "router_noise": 0.0, "grad_norm": 0.7073, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 335.2}
{"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.20772, "bpb": 1.74237, "chunk_loss": 0.029891, "F": 0.17014, "G": 0.22875, "chunks": 174.22, "bytes_per_chunk": 5.878, "boundary_entropy": 0.19726, "router_noise": 0.0, "grad_norm": 0.7315, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 337.2}
{"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.22971, "bpb": 1.7741, "chunk_loss": 0.030104, "F": 0.17197, "G": 0.21829, "chunks": 176.09, "bytes_per_chunk": 5.815, "boundary_entropy": 0.17933, "router_noise": 0.0, "grad_norm": 0.7753, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 339.2}
{"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2899.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.16658, "cps": 22.648438, "mean_p": 0.185426, "boundary_entropy": 0.093752, "bpb": 2.013639}, "de": {"chunks": 3285.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159986, "cps": 25.664062, "mean_p": 0.181596, "boundary_entropy": 0.11801, "bpb": 2.227399}, "vi": {"chunks": 4162.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225766, "cps": 32.515625, "mean_p": 0.190111, "boundary_entropy": 0.102221, "bpb": 1.873508}, "tr": {"chunks": 3176.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175441, "cps": 24.8125, "mean_p": 0.190895, "boundary_entropy": 0.14011, "bpb": 2.215361}, "fi": {"chunks": 2899.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.154112, "cps": 22.648438, "mean_p": 0.176887, "boundary_entropy": 0.14547, "bpb": 2.386166}, "ru": {"chunks": 6412.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.338364, "cps": 50.09375, "mean_p": 0.194676, "boundary_entropy": 0.095502, "bpb": 1.297353}, "el": {"chunks": 6588.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.321006, "cps": 51.46875, "mean_p": 0.189012, "boundary_entropy": 0.102664, "bpb": 1.333217}, "hi": {"chunks": 6348.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.377296, "cps": 49.59375, "mean_p": 0.283505, "boundary_entropy": 0.429361, "bpb": 1.159116}, "ta": {"chunks": 8396.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449104, "cps": 65.59375, "mean_p": 0.275669, "boundary_entropy": 0.40396, "bpb": 0.941223}, "th": {"chunks": 7015.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.436121, "cps": 54.804688, "mean_p": 0.356573, "boundary_entropy": 0.545684, "bpb": 1.058997}, "zh": {"chunks": 2984.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.505249, "cps": 23.3125, "mean_p": 0.205788, "boundary_entropy": 0.118712, "bpb": 3.045477}, "my": {"chunks": 6771.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.357346, "cps": 52.898438, "mean_p": 0.346382, "boundary_entropy": 0.52832, "bpb": 1.084542}}, "bpb": {"en": 2.01364, "de": 2.2274, "vi": 1.87351, "tr": 2.21536, "fi": 2.38617, "ru": 1.29735, "el": 1.33322, "hi": 1.15912, "ta": 0.94122, "th": 1.059, "zh": 3.04548, "my": 1.08454}, "gini_cpc": 0.222125, "gini_cps": 0.208757, "bpic": 6.298203, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9604148611628424, "vi": 1.3552981289527464, "tr": 1.0531878667462566, "fi": 0.9251501780872894, "ru": 2.031235154329443, "el": 1.9270307737587915, "hi": 2.264945467182156, "ta": 2.6960184830083618, "th": 2.6180776002100776, "zh": 3.033061951252619, "my": 2.1451879194000423}, "t": 342.1}
{"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.02624, "de": 2.23615, "vi": 1.87505, "tr": 2.21756, "fi": 2.3977, "ru": 1.27614, "el": 1.3226, "hi": 1.15461, "ta": 0.94862, "th": 1.07656, "zh": 3.04044, "my": 1.09245}, "bytes_per_chunk": {"en": 5.671, "de": 6.293, "vi": 5.828, "tr": 6.05, "fi": 6.549, "ru": 5.486, "el": 5.681, "hi": 6.469, "ta": 6.094, "th": 6.445, "zh": 5.286, "my": 9.258}, "bpb_mean": 1.72201, "t": 343.4}
{"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2899.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.16658, "cps": 22.648438, "mean_p": 0.185426, "boundary_entropy": 0.093752, "bpb": 2.013639}, "de": {"chunks": 3285.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159986, "cps": 25.664062, "mean_p": 0.181596, "boundary_entropy": 0.11801, "bpb": 2.227399}, "vi": {"chunks": 4162.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225766, "cps": 32.515625, "mean_p": 0.190111, "boundary_entropy": 0.102221, "bpb": 1.873508}, "tr": {"chunks": 3176.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175441, "cps": 24.8125, "mean_p": 0.190895, "boundary_entropy": 0.14011, "bpb": 2.215361}, "fi": {"chunks": 2899.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.154112, "cps": 22.648438, "mean_p": 0.176887, "boundary_entropy": 0.14547, "bpb": 2.386166}, "ru": {"chunks": 6412.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.338364, "cps": 50.09375, "mean_p": 0.194676, "boundary_entropy": 0.095502, "bpb": 1.297353}, "el": {"chunks": 6588.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.321006, "cps": 51.46875, "mean_p": 0.189012, "boundary_entropy": 0.102664, "bpb": 1.333217}, "hi": {"chunks": 6348.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.377296, "cps": 49.59375, "mean_p": 0.283505, "boundary_entropy": 0.429361, "bpb": 1.159116}, "ta": {"chunks": 8396.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449104, "cps": 65.59375, "mean_p": 0.275669, "boundary_entropy": 0.40396, "bpb": 0.941223}, "th": {"chunks": 7015.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.436121, "cps": 54.804688, "mean_p": 0.356573, "boundary_entropy": 0.545684, "bpb": 1.058997}, "zh": {"chunks": 2984.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.505249, "cps": 23.3125, "mean_p": 0.205788, "boundary_entropy": 0.118712, "bpb": 3.045477}, "my": {"chunks": 6771.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.357346, "cps": 52.898438, "mean_p": 0.346382, "boundary_entropy": 0.52832, "bpb": 1.084542}}, "bpb": {"en": 2.01364, "de": 2.2274, "vi": 1.87351, "tr": 2.21536, "fi": 2.38617, "ru": 1.29735, "el": 1.33322, "hi": 1.15912, "ta": 0.94122, "th": 1.059, "zh": 3.04548, "my": 1.08454}, "gini_cpc": 0.222125, "gini_cps": 0.208757, "bpic": 6.298203, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9604148611628424, "vi": 1.3552981289527464, "tr": 1.0531878667462566, "fi": 0.9251501780872894, "ru": 2.031235154329443, "el": 1.9270307737587915, "hi": 2.264945467182156, "ta": 2.6960184830083618, "th": 2.6180776002100776, "zh": 3.033061951252619, "my": 2.1451879194000423}, "t": 344.4}
{"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s1/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s1/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s1/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s1/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s1/th0.70"], "t": 347.1}
{"event": "final", "run": "pilot_C_s1", "condition": "C", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.16658047463081077, "cps": 22.6484375, "bpb": 2.034134262125667, "bytes_per_chunk": 5.6753409828967305}, "de": {"cpc": 0.1599863634149905, "cps": 25.6640625, "bpb": 2.2279271615805345, "bytes_per_chunk": 6.283262625536301}, "vi": {"cpc": 0.22576620558719826, "cps": 32.515625, "bpb": 1.877921528186976, "bytes_per_chunk": 5.814632847937715}, "tr": {"cpc": 0.17544053471800253, "cps": 24.8125, "bpb": 2.2088697962851733, "bytes_per_chunk": 6.083921277385815}, "fi": {"cpc": 0.15411195577055978, "cps": 22.6484375, "bpb": 2.412586739629766, "bytes_per_chunk": 6.5237538262449295}, "ru": {"cpc": 0.3383641160949868, "cps": 50.09375, "bpb": 1.276287299019656, "bytes_per_chunk": 5.480515136310419}, "el": {"cpc": 0.321005700920918, "cps": 51.46875, "bpb": 1.3180558923937413, "bytes_per_chunk": 5.695687126561651}, "hi": {"cpc": 0.377295690936107, "cps": 49.59375, "bpb": 1.1627524737258252, "bytes_per_chunk": 6.470774091627172}, "ta": {"cpc": 0.44910403851297137, "cps": 65.59375, "bpb": 0.9491452058399927, "bytes_per_chunk": 6.1038244368123875}, "th": {"cpc": 0.43612060926328877, "cps": 54.8046875, "bpb": 1.0742587039841258, "bytes_per_chunk": 6.426988329900952}, "zh": {"cpc": 0.5052488994243143, "cps": 23.3125, "bpb": 3.0502380753653546, "bytes_per_chunk": 5.299906999312603}, "my": {"cpc": 0.35734642178594045, "cps": 52.8984375, "bpb": 1.101764856739026, "bytes_per_chunk": 9.165874125874126}}, "gini_cpc": 0.2221252116848107, "gini_cps": 0.20875659856130868, "bpic": 6.298203003200132, "seed": 1, "mean_boundary_entropy": 0.16041748225688934, "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": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 349.5}