| {"event": "start", "run": "spec_B_s2", "condition": "B", "params": 82470528, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "parity_a", "p3_objective": "none", "freeze_in_p3": true, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 1.9} |
| {"event": "probe_schedule", "n": 21, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 11000], "t": 1.9} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10185.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.585244, "cps": 79.570312, "mean_p": 0.50663, "boundary_entropy": 0.687329, "bpb": 8.017475}, "de": {"chunks": 10966.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.534067, "cps": 85.671875, "mean_p": 0.503879, "boundary_entropy": 0.688251, "bpb": 8.111776}, "vi": {"chunks": 5741.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.311418, "cps": 44.851562, "mean_p": 0.489692, "boundary_entropy": 0.688475, "bpb": 7.989146}, "tr": {"chunks": 6370.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.351875, "cps": 49.765625, "mean_p": 0.495802, "boundary_entropy": 0.687878, "bpb": 8.107657}, "fi": {"chunks": 2115.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.112434, "cps": 16.523438, "mean_p": 0.479404, "boundary_entropy": 0.686744, "bpb": 8.107682}, "ru": {"chunks": 23602.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.245488, "cps": 184.390625, "mean_p": 0.507156, "boundary_entropy": 0.690197, "bpb": 8.281387}, "el": {"chunks": 1455.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.070896, "cps": 11.367188, "mean_p": 0.480429, "boundary_entropy": 0.689544, "bpb": 7.890102}, "hi": {"chunks": 7663.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.455453, "cps": 59.867188, "mean_p": 0.494285, "boundary_entropy": 0.690846, "bpb": 7.810752}, "ta": {"chunks": 357.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.019096, "cps": 2.789062, "mean_p": 0.475379, "boundary_entropy": 0.689908, "bpb": 7.609497}, "th": {"chunks": 27234.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.69313, "cps": 212.765625, "mean_p": 0.50318, "boundary_entropy": 0.691076, "bpb": 7.876604}, "zh": {"chunks": 14702.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.489333, "cps": 114.859375, "mean_p": 0.531852, "boundary_entropy": 0.685234, "bpb": 8.024454}, "my": {"chunks": 789.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.04164, "cps": 6.164062, "mean_p": 0.475663, "boundary_entropy": 0.690063, "bpb": 8.245358}}, "bpb": {"en": 8.01748, "de": 8.11178, "vi": 7.98915, "tr": 8.10766, "fi": 8.10768, "ru": 8.28139, "el": 7.8901, "hi": 7.81075, "ta": 7.6095, "th": 7.8766, "zh": 8.02445, "my": 8.24536}, "gini_cpc": 0.562508, "gini_cps": 0.489131, "bpic": 3.45192, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9125547315729478, "vi": 0.5321174384544991, "tr": 0.6012456783662331, "fi": 0.19211513271032077, "ru": 2.1281521716321925, "el": 0.1211393488838307, "hi": 0.7782279770749311, "ta": 0.032629155488401715, "th": 2.8930334475858417, "zh": 4.253496332079356, "my": 0.07115029646287195}, "t": 7.4} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.5908, "bpb": 8.06582, "chunk_loss": 0.059712, "F": 0.56091, "G": 0.50224, "chunks": 574.38, "bytes_per_chunk": 1.783, "boundary_entropy": 0.6917, "router_noise": 0.0, "grad_norm": 9.2827, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 8.2} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10111.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.580992, "cps": 78.992188, "mean_p": 0.506452, "boundary_entropy": 0.687331, "bpb": 7.98283}, "de": {"chunks": 10912.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.531437, "cps": 85.25, "mean_p": 0.503747, "boundary_entropy": 0.688251, "bpb": 8.080507}, "vi": {"chunks": 5695.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.308923, "cps": 44.492188, "mean_p": 0.489579, "boundary_entropy": 0.688469, "bpb": 7.968876}, "tr": {"chunks": 6315.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.348837, "cps": 49.335938, "mean_p": 0.495629, "boundary_entropy": 0.687872, "bpb": 8.083766}, "fi": {"chunks": 2105.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.111903, "cps": 16.445312, "mean_p": 0.479274, "boundary_entropy": 0.686729, "bpb": 8.086404}, "ru": {"chunks": 23593.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.245013, "cps": 184.320312, "mean_p": 0.507134, "boundary_entropy": 0.690198, "bpb": 8.228871}, "el": {"chunks": 1445.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.070409, "cps": 11.289062, "mean_p": 0.480401, "boundary_entropy": 0.689541, "bpb": 7.879932}, "hi": {"chunks": 7627.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.453314, "cps": 59.585938, "mean_p": 0.494253, "boundary_entropy": 0.690845, "bpb": 7.788147}, "ta": {"chunks": 357.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.019096, "cps": 2.789062, "mean_p": 0.475324, "boundary_entropy": 0.689903, "bpb": 7.596212}, "th": {"chunks": 26972.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.676842, "cps": 210.71875, "mean_p": 0.50308, "boundary_entropy": 0.691076, "bpb": 7.839512}, "zh": {"chunks": 14694.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.487978, "cps": 114.796875, "mean_p": 0.53176, "boundary_entropy": 0.685245, "bpb": 8.019137}, "my": {"chunks": 786.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.041482, "cps": 6.140625, "mean_p": 0.475631, "boundary_entropy": 0.69006, "bpb": 8.245367}}, "bpb": {"en": 7.98283, "de": 8.08051, "vi": 7.96888, "tr": 8.08377, "fi": 8.0864, "ru": 8.22887, "el": 7.87993, "hi": 7.78815, "ta": 7.59621, "th": 7.83951, "zh": 8.01914, "my": 8.24537}, "gini_cpc": 0.563041, "gini_cps": 0.489222, "bpic": 3.469615, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9147069101184132, "vi": 0.5317170618502403, "tr": 0.6004167692105996, "fi": 0.19260618384667633, "ru": 2.1429101563672264, "el": 0.12118727266936478, "hi": 0.7802408481310646, "ta": 0.03286796050334997, "th": 2.886171245522232, "zh": 4.282295205909964, "my": 0.07139851511427713}, "t": 13.3} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 9970.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.57289, "cps": 77.890625, "mean_p": 0.506054, "boundary_entropy": 0.687335, "bpb": 7.915954}, "de": {"chunks": 10768.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.524424, "cps": 84.125, "mean_p": 0.503441, "boundary_entropy": 0.68825, "bpb": 8.020561}, "vi": {"chunks": 5576.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.302468, "cps": 43.5625, "mean_p": 0.489305, "boundary_entropy": 0.688455, "bpb": 7.931217}, "tr": {"chunks": 6162.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.340386, "cps": 48.140625, "mean_p": 0.495259, "boundary_entropy": 0.687858, "bpb": 8.037834}, "fi": {"chunks": 2072.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.110148, "cps": 16.1875, "mean_p": 0.478978, "boundary_entropy": 0.686696, "bpb": 8.047534}, "ru": {"chunks": 23560.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.243272, "cps": 184.0625, "mean_p": 0.5071, "boundary_entropy": 0.6902, "bpb": 8.121227}, "el": {"chunks": 1432.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.069775, "cps": 11.1875, "mean_p": 0.48036, "boundary_entropy": 0.689538, "bpb": 7.856415}, "hi": {"chunks": 7614.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.452541, "cps": 59.484375, "mean_p": 0.494229, "boundary_entropy": 0.690844, "bpb": 7.745884}, "ta": {"chunks": 352.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.018829, "cps": 2.75, "mean_p": 0.475301, "boundary_entropy": 0.689901, "bpb": 7.564107}, "th": {"chunks": 26704.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.66018, "cps": 208.625, "mean_p": 0.502965, "boundary_entropy": 0.691077, "bpb": 7.776915}, "zh": {"chunks": 14681.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.485777, "cps": 114.695312, "mean_p": 0.531544, "boundary_entropy": 0.685271, "bpb": 8.006242}, "my": {"chunks": 676.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.035677, "cps": 5.28125, "mean_p": 0.475492, "boundary_entropy": 0.690046, "bpb": 8.235555}}, "bpb": {"en": 7.91595, "de": 8.02056, "vi": 7.93122, "tr": 8.03783, "fi": 8.04753, "ru": 8.12123, "el": 7.85641, "hi": 7.74588, "ta": 7.56411, "th": 7.77692, "zh": 8.00624, "my": 8.23555}, "gini_cpc": 0.565727, "gini_cps": 0.491849, "bpic": 3.502706, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9154014617758697, "vi": 0.5279691964420519, "tr": 0.5941554764276256, "fi": 0.19226792066167134, "ru": 2.1701763866798283, "el": 0.1217954697303429, "hi": 0.7899266417379924, "ta": 0.03286594740090826, "th": 2.8979054789485024, "zh": 4.339015064156234, "my": 0.06227479144209068}, "t": 19.0} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 9433.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.542033, "cps": 73.695312, "mean_p": 0.50457, "boundary_entropy": 0.687346, "bpb": 7.681071}, "de": {"chunks": 10207.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.497102, "cps": 79.742188, "mean_p": 0.502294, "boundary_entropy": 0.688245, "bpb": 7.806253}, "vi": {"chunks": 5258.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.285218, "cps": 41.078125, "mean_p": 0.488413, "boundary_entropy": 0.688405, "bpb": 7.800229}, "tr": {"chunks": 5693.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.314478, "cps": 44.476562, "mean_p": 0.493944, "boundary_entropy": 0.687805, "bpb": 7.874022}, "fi": {"chunks": 1985.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.105523, "cps": 15.507812, "mean_p": 0.477974, "boundary_entropy": 0.686583, "bpb": 7.910002}, "ru": {"chunks": 23392.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.234406, "cps": 182.75, "mean_p": 0.50697, "boundary_entropy": 0.690208, "bpb": 7.802236}, "el": {"chunks": 1390.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.067729, "cps": 10.859375, "mean_p": 0.480081, "boundary_entropy": 0.689513, "bpb": 7.777203}, "hi": {"chunks": 7478.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.444458, "cps": 58.421875, "mean_p": 0.494055, "boundary_entropy": 0.690838, "bpb": 7.594888}, "ta": {"chunks": 349.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.018668, "cps": 2.726562, "mean_p": 0.475234, "boundary_entropy": 0.689896, "bpb": 7.4683}, "th": {"chunks": 25546.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.588188, "cps": 199.578125, "mean_p": 0.502552, "boundary_entropy": 0.691079, "bpb": 7.561634}, "zh": {"chunks": 14610.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.473756, "cps": 114.140625, "mean_p": 0.530851, "boundary_entropy": 0.685353, "bpb": 7.968461}, "my": {"chunks": 652.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.03441, "cps": 5.09375, "mean_p": 0.475189, "boundary_entropy": 0.690014, "bpb": 8.209746}}, "bpb": {"en": 7.68107, "de": 7.80625, "vi": 7.80023, "tr": 7.87402, "fi": 7.91, "ru": 7.80224, "el": 7.7772, "hi": 7.59489, "ta": 7.4683, "th": 7.56163, "zh": 7.96846, "my": 8.20975}, "gini_cpc": 0.570877, "gini_cps": 0.495148, "bpic": 3.620815, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.917106968472855, "vi": 0.5262010684406326, "tr": 0.580182891520206, "fi": 0.19468070640339968, "ru": 2.2773638719064913, "el": 0.12495344690338943, "hi": 0.8199826696736142, "ta": 0.03444088028725386, "th": 2.930057400390121, "zh": 4.563846816134135, "my": 0.06348315545918426}, "t": 25.0} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 7419.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.426306, "cps": 57.960938, "mean_p": 0.499508, "boundary_entropy": 0.687322, "bpb": 7.072525}, "de": {"chunks": 8366.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.407442, "cps": 65.359375, "mean_p": 0.498487, "boundary_entropy": 0.688197, "bpb": 7.217478}, "vi": {"chunks": 4140.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.224573, "cps": 32.34375, "mean_p": 0.485408, "boundary_entropy": 0.688222, "bpb": 7.448121}, "tr": {"chunks": 4328.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.239076, "cps": 33.8125, "mean_p": 0.489654, "boundary_entropy": 0.687588, "bpb": 7.395666}, "fi": {"chunks": 1718.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.09133, "cps": 13.421875, "mean_p": 0.474886, "boundary_entropy": 0.686208, "bpb": 7.502653}, "ru": {"chunks": 22010.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.161478, "cps": 171.953125, "mean_p": 0.505828, "boundary_entropy": 0.690249, "bpb": 6.813976}, "el": {"chunks": 1241.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.060469, "cps": 9.695312, "mean_p": 0.479046, "boundary_entropy": 0.68942, "bpb": 7.514336}, "hi": {"chunks": 7264.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.431738, "cps": 56.75, "mean_p": 0.493695, "boundary_entropy": 0.690822, "bpb": 7.038682}, "ta": {"chunks": 334.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.017866, "cps": 2.609375, "mean_p": 0.475412, "boundary_entropy": 0.689923, "bpb": 7.122198}, "th": {"chunks": 21109.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.312341, "cps": 164.914062, "mean_p": 0.50095, "boundary_entropy": 0.691079, "bpb": 6.852993}, "zh": {"chunks": 14329.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.426177, "cps": 111.945312, "mean_p": 0.528148, "boundary_entropy": 0.685652, "bpb": 7.848755}, "my": {"chunks": 592.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.031243, "cps": 4.625, "mean_p": 0.474322, "boundary_entropy": 0.689923, "bpb": 8.105449}}, "bpb": {"en": 7.07253, "de": 7.21748, "vi": 7.44812, "tr": 7.39567, "fi": 7.50265, "ru": 6.81398, "el": 7.51434, "hi": 7.03868, "ta": 7.1222, "th": 6.85299, "zh": 7.84876, "my": 8.10545}, "gini_cpc": 0.588137, "gini_cps": 0.503247, "bpic": 4.133344, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9557497700436225, "vi": 0.5267880908769963, "tr": 0.5608096135160803, "fi": 0.21423480333656328, "ru": 2.724517346711236, "el": 0.1418435805032804, "hi": 1.012743610187649, "ta": 0.041908271271427044, "th": 3.078402079731853, "zh": 5.691165159407342, "my": 0.07328871038800211}, "t": 31.5} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4288.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.246394, "cps": 33.5, "mean_p": 0.491829, "boundary_entropy": 0.687182, "bpb": 6.365488}, "de": {"chunks": 6165.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.300248, "cps": 48.164062, "mean_p": 0.494703, "boundary_entropy": 0.688167, "bpb": 6.434539}, "vi": {"chunks": 2405.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.130458, "cps": 18.789062, "mean_p": 0.480047, "boundary_entropy": 0.687872, "bpb": 6.979368}, "tr": {"chunks": 2880.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.15909, "cps": 22.5, "mean_p": 0.48471, "boundary_entropy": 0.687317, "bpb": 6.722095}, "fi": {"chunks": 1383.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.073521, "cps": 10.804688, "mean_p": 0.473382, "boundary_entropy": 0.68608, "bpb": 6.72596}, "ru": {"chunks": 10614.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.560106, "cps": 82.921875, "mean_p": 0.496832, "boundary_entropy": 0.690276, "bpb": 5.685295}, "el": {"chunks": 1312.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.063928, "cps": 10.25, "mean_p": 0.478506, "boundary_entropy": 0.689361, "bpb": 6.743536}, "hi": {"chunks": 7906.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.469896, "cps": 61.765625, "mean_p": 0.493758, "boundary_entropy": 0.690812, "bpb": 5.96578}, "ta": {"chunks": 358.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.01915, "cps": 2.796875, "mean_p": 0.478773, "boundary_entropy": 0.690274, "bpb": 6.475433}, "th": {"chunks": 4154.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.258253, "cps": 32.453125, "mean_p": 0.491304, "boundary_entropy": 0.690887, "bpb": 5.581939}, "zh": {"chunks": 12179.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.06214, "cps": 95.148438, "mean_p": 0.517051, "boundary_entropy": 0.686555, "bpb": 7.484813}, "my": {"chunks": 614.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.032404, "cps": 4.796875, "mean_p": 0.472691, "boundary_entropy": 0.689737, "bpb": 7.435623}}, "bpb": {"en": 6.36549, "de": 6.43454, "vi": 6.97937, "tr": 6.7221, "fi": 6.72596, "ru": 5.6853, "el": 6.74354, "hi": 5.96578, "ta": 6.47543, "th": 5.58194, "zh": 7.48481, "my": 7.43562}, "gini_cpc": 0.603125, "gini_cps": 0.456188, "bpic": 7.073261, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2185686960233653, "vi": 0.529469907885309, "tr": 0.6456709987047583, "fi": 0.2983868228299257, "ru": 2.273208192218328, "el": 0.25945517260939854, "hi": 1.9070895245170878, "ta": 0.07771894572337563, "th": 1.0481290798880947, "zh": 8.36926908538496, "my": 0.13151471210288015}, "t": 38.9} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3712.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.213297, "cps": 29.0, "mean_p": 0.493854, "boundary_entropy": 0.687518, "bpb": 5.830766}, "de": {"chunks": 6575.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.320216, "cps": 51.367188, "mean_p": 0.497335, "boundary_entropy": 0.688449, "bpb": 5.948263}, "vi": {"chunks": 1774.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.09623, "cps": 13.859375, "mean_p": 0.48205, "boundary_entropy": 0.688259, "bpb": 6.570486}, "tr": {"chunks": 3446.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190355, "cps": 26.921875, "mean_p": 0.49071, "boundary_entropy": 0.687922, "bpb": 6.25555}, "fi": {"chunks": 2381.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.126575, "cps": 18.601562, "mean_p": 0.488369, "boundary_entropy": 0.687784, "bpb": 5.981356}, "ru": {"chunks": 6841.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.361003, "cps": 53.445312, "mean_p": 0.494115, "boundary_entropy": 0.690272, "bpb": 5.153195}, "el": {"chunks": 12459.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.607075, "cps": 97.335938, "mean_p": 0.496894, "boundary_entropy": 0.690463, "bpb": 5.523912}, "hi": {"chunks": 10986.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.652957, "cps": 85.828125, "mean_p": 0.495567, "boundary_entropy": 0.690859, "bpb": 5.249137}, "ta": {"chunks": 26229.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.402995, "cps": 204.914062, "mean_p": 0.50129, "boundary_entropy": 0.691301, "bpb": 4.997163}, "th": {"chunks": 5602.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.348275, "cps": 43.765625, "mean_p": 0.494362, "boundary_entropy": 0.691025, "bpb": 4.856678}, "zh": {"chunks": 3268.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.553336, "cps": 25.53125, "mean_p": 0.489537, "boundary_entropy": 0.686627, "bpb": 7.018122}, "my": {"chunks": 7570.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.399514, "cps": 59.140625, "mean_p": 0.493483, "boundary_entropy": 0.691251, "bpb": 4.993469}}, "bpb": {"en": 5.83077, "de": 5.94826, "vi": 6.57049, "tr": 6.25555, "fi": 5.98136, "ru": 5.15319, "el": 5.52391, "hi": 5.24914, "ta": 4.99716, "th": 4.85668, "zh": 7.01812, "my": 4.99347}, "gini_cpc": 0.379054, "gini_cps": 0.40795, "bpic": 4.224662, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.50127240768435, "vi": 0.45115588437473697, "tr": 0.8924437950368295, "fi": 0.5934220296405808, "ru": 1.6924916266718222, "el": 2.84615464283554, "hi": 3.061263225649434, "ta": 6.577675073894919, "th": 1.6328195510381271, "zh": 2.5942077825589407, "my": 1.8730469176530322}, "t": 48.8} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4968.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.285468, "cps": 38.8125, "mean_p": 0.497916, "boundary_entropy": 0.687757, "bpb": 5.414332}, "de": {"chunks": 4836.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.235523, "cps": 37.78125, "mean_p": 0.495876, "boundary_entropy": 0.688518, "bpb": 5.569721}, "vi": {"chunks": 2555.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.138595, "cps": 19.960938, "mean_p": 0.488593, "boundary_entropy": 0.688829, "bpb": 5.969205}, "tr": {"chunks": 2678.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.147931, "cps": 20.921875, "mean_p": 0.491384, "boundary_entropy": 0.688084, "bpb": 5.890114}, "fi": {"chunks": 3454.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183616, "cps": 26.984375, "mean_p": 0.49349, "boundary_entropy": 0.688146, "bpb": 5.570872}, "ru": {"chunks": 5081.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.268127, "cps": 39.695312, "mean_p": 0.495177, "boundary_entropy": 0.690402, "bpb": 4.844542}, "el": {"chunks": 4530.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.220728, "cps": 35.390625, "mean_p": 0.4911, "boundary_entropy": 0.690366, "bpb": 5.094242}, "hi": {"chunks": 1368.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.081308, "cps": 10.6875, "mean_p": 0.483874, "boundary_entropy": 0.690297, "bpb": 4.910716}, "ta": {"chunks": 882.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.047178, "cps": 6.890625, "mean_p": 0.488954, "boundary_entropy": 0.69103, "bpb": 4.565489}, "th": {"chunks": 5398.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.335592, "cps": 42.171875, "mean_p": 0.491152, "boundary_entropy": 0.690846, "bpb": 4.410236}, "zh": {"chunks": 5931.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.004233, "cps": 46.335938, "mean_p": 0.499947, "boundary_entropy": 0.687294, "bpb": 6.673935}, "my": {"chunks": 13200.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.696643, "cps": 103.125, "mean_p": 0.496877, "boundary_entropy": 0.691372, "bpb": 4.277871}}, "bpb": {"en": 5.41433, "de": 5.56972, "vi": 5.9692, "tr": 5.89011, "fi": 5.57087, "ru": 4.84454, "el": 5.09424, "hi": 4.91072, "ta": 4.56549, "th": 4.41024, "zh": 6.67394, "my": 4.27787}, "gini_cpc": 0.420688, "gini_cps": 0.323927, "bpic": 6.992967, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8250426849108928, "vi": 0.48550118643779094, "tr": 0.5182061397996062, "fi": 0.6432102888714769, "ru": 0.9392528329063262, "el": 0.773214322283871, "hi": 0.28482201692615805, "ta": 0.16526681163296394, "th": 1.1755858445246345, "zh": 3.517847546341966, "my": 2.4403554503094664}, "t": 63.2} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.25061, "bpb": 4.68964, "chunk_loss": 0.032146, "F": 0.19919, "G": 0.4788, "chunks": 203.97, "bytes_per_chunk": 5.02, "boundary_entropy": 0.68987, "router_noise": 0.0, "grad_norm": 2.5908, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.651, "alpha": 0.03, "t": 73.3} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4655.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.267483, "cps": 36.367188, "mean_p": 0.450072, "boundary_entropy": 0.671358, "bpb": 4.337783}, "de": {"chunks": 6239.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.303852, "cps": 48.742188, "mean_p": 0.452374, "boundary_entropy": 0.672247, "bpb": 4.569444}, "vi": {"chunks": 5275.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.28614, "cps": 41.210938, "mean_p": 0.467645, "boundary_entropy": 0.682834, "bpb": 4.759263}, "tr": {"chunks": 4480.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.247473, "cps": 35.0, "mean_p": 0.4554, "boundary_entropy": 0.677164, "bpb": 4.797178}, "fi": {"chunks": 4881.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.259476, "cps": 38.132812, "mean_p": 0.461645, "boundary_entropy": 0.678775, "bpb": 4.574221}, "ru": {"chunks": 7392.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.390079, "cps": 57.75, "mean_p": 0.474187, "boundary_entropy": 0.687613, "bpb": 3.243776}, "el": {"chunks": 11934.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.581494, "cps": 93.234375, "mean_p": 0.490914, "boundary_entropy": 0.689403, "bpb": 4.003275}, "hi": {"chunks": 13208.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.785022, "cps": 103.1875, "mean_p": 0.493505, "boundary_entropy": 0.690067, "bpb": 3.019558}, "ta": {"chunks": 15494.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.828778, "cps": 121.046875, "mean_p": 0.493081, "boundary_entropy": 0.690946, "bpb": 3.117804}, "th": {"chunks": 11311.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.703202, "cps": 88.367188, "mean_p": 0.484532, "boundary_entropy": 0.689939, "bpb": 2.799166}, "zh": {"chunks": 6273.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.06214, "cps": 49.007812, "mean_p": 0.501002, "boundary_entropy": 0.687329, "bpb": 6.238526}, "my": {"chunks": 23145.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.221501, "cps": 180.820312, "mean_p": 0.498034, "boundary_entropy": 0.69119, "bpb": 3.344278}}, "bpb": {"en": 4.33778, "de": 4.56944, "vi": 4.75926, "tr": 4.79718, "fi": 4.57422, "ru": 3.24378, "el": 4.00327, "hi": 3.01956, "ta": 3.1178, "th": 2.79917, "zh": 6.23853, "my": 3.34428}, "gini_cpc": 0.309362, "gini_cps": 0.299527, "bpic": 3.358046, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1359703954070843, "vi": 1.06975360055375, "tr": 0.9251920609677539, "fi": 0.9700661695475961, "ru": 1.4583345963854226, "el": 2.1739503602701284, "hi": 2.934853465653095, "ta": 3.0984359075199177, "th": 2.6289623833107347, "zh": 3.9708755828422584, "my": 4.566655431213517}, "t": 83.5} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.54243, "bpb": 3.66795, "chunk_loss": 0.032953, "F": 0.23032, "G": 0.39462, "chunks": 235.84, "bytes_per_chunk": 4.342, "boundary_entropy": 0.62757, "router_noise": 0.0, "grad_norm": 1.5613, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.249, "alpha": 0.03, "t": 98.2} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3055.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.175544, "cps": 23.867188, "mean_p": 0.326973, "boundary_entropy": 0.539401, "bpb": 3.563619}, "de": {"chunks": 4340.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.211367, "cps": 33.90625, "mean_p": 0.336419, "boundary_entropy": 0.546607, "bpb": 3.77072}, "vi": {"chunks": 5126.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.278058, "cps": 40.046875, "mean_p": 0.387758, "boundary_entropy": 0.618212, "bpb": 3.056666}, "tr": {"chunks": 3446.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190355, "cps": 26.921875, "mean_p": 0.350922, "boundary_entropy": 0.570746, "bpb": 3.746724}, "fi": {"chunks": 3521.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.187178, "cps": 27.507812, "mean_p": 0.339947, "boundary_entropy": 0.571032, "bpb": 3.814681}, "ru": {"chunks": 5834.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.307863, "cps": 45.578125, "mean_p": 0.35031, "boundary_entropy": 0.60548, "bpb": 2.61955}, "el": {"chunks": 8500.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.414169, "cps": 66.40625, "mean_p": 0.330216, "boundary_entropy": 0.57954, "bpb": 2.658371}, "hi": {"chunks": 8504.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.505438, "cps": 66.4375, "mean_p": 0.434142, "boundary_entropy": 0.644386, "bpb": 2.022856}, "ta": {"chunks": 10841.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.579888, "cps": 84.695312, "mean_p": 0.432762, "boundary_entropy": 0.640598, "bpb": 1.842594}, "th": {"chunks": 10468.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.650793, "cps": 81.78125, "mean_p": 0.391908, "boundary_entropy": 0.642771, "bpb": 2.054915}, "zh": {"chunks": 3662.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.620047, "cps": 28.609375, "mean_p": 0.447405, "boundary_entropy": 0.668249, "bpb": 4.904357}, "my": {"chunks": 9644.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.508972, "cps": 75.34375, "mean_p": 0.461558, "boundary_entropy": 0.679538, "bpb": 2.106773}}, "bpb": {"en": 3.56362, "de": 3.77072, "vi": 3.05667, "tr": 3.74672, "fi": 3.81468, "ru": 2.61955, "el": 2.65837, "hi": 2.02286, "ta": 1.84259, "th": 2.05492, "zh": 4.90436, "my": 2.10677}, "gini_cpc": 0.254279, "gini_cps": 0.248137, "bpic": 4.987991, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2040658193991023, "vi": 1.5839751557757895, "tr": 1.0843703329547336, "fi": 1.066269648787347, "ru": 1.7537598210468586, "el": 2.3593424741695337, "hi": 2.8792613283463395, "ta": 3.3033666545708105, "th": 3.707281417713894, "zh": 3.532139137818181, "my": 2.899390631332474}, "t": 112.8} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.1528, "bpb": 3.10584, "chunk_loss": 0.031308, "F": 0.20551, "G": 0.32887, "chunks": 210.44, "bytes_per_chunk": 4.866, "boundary_entropy": 0.53386, "router_noise": 0.0, "grad_norm": 1.3036, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.234, "alpha": 0.03, "t": 121.6} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.06336, "bpb": 2.9768, "chunk_loss": 0.031229, "F": 0.20947, "G": 0.29507, "chunks": 214.5, "bytes_per_chunk": 4.774, "boundary_entropy": 0.44178, "router_noise": 0.0, "grad_norm": 1.4491, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.192, "alpha": 0.03, "t": 140.9} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.81011, "bpb": 2.61144, "chunk_loss": 0.030475, "F": 0.18675, "G": 0.27864, "chunks": 191.23, "bytes_per_chunk": 5.355, "boundary_entropy": 0.40468, "router_noise": 0.0, "grad_norm": 0.7841, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.189, "alpha": 0.03, "t": 159.7} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.79132, "bpb": 2.58432, "chunk_loss": 0.030596, "F": 0.19466, "G": 0.27386, "chunks": 199.33, "bytes_per_chunk": 5.137, "boundary_entropy": 0.36172, "router_noise": 0.0, "grad_norm": 1.004, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.185, "alpha": 0.03, "t": 178.2} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.67804, "bpb": 2.4209, "chunk_loss": 0.030353, "F": 0.19188, "G": 0.27352, "chunks": 196.48, "bytes_per_chunk": 5.212, "boundary_entropy": 0.34985, "router_noise": 0.0, "grad_norm": 0.6942, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 196.5} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.73607, "bpb": 2.50462, "chunk_loss": 0.030722, "F": 0.19836, "G": 0.25828, "chunks": 203.12, "bytes_per_chunk": 5.041, "boundary_entropy": 0.29506, "router_noise": 0.0, "grad_norm": 0.8163, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 214.8} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.63275, "bpb": 2.35556, "chunk_loss": 0.03046, "F": 0.19464, "G": 0.26965, "chunks": 199.31, "bytes_per_chunk": 5.138, "boundary_entropy": 0.32455, "router_noise": 0.0, "grad_norm": 0.6288, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.184, "alpha": 0.03, "t": 233.2} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3465.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199104, "cps": 27.070312, "mean_p": 0.212496, "boundary_entropy": 0.164987, "bpb": 2.309635}, "de": {"chunks": 4748.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.231238, "cps": 37.09375, "mean_p": 0.237531, "boundary_entropy": 0.161929, "bpb": 2.58971}, "vi": {"chunks": 4099.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222349, "cps": 32.023438, "mean_p": 0.246113, "boundary_entropy": 0.329994, "bpb": 2.109931}, "tr": {"chunks": 3760.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.2077, "cps": 29.375, "mean_p": 0.20781, "boundary_entropy": 0.231033, "bpb": 2.606616}, "fi": {"chunks": 3841.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.204189, "cps": 30.007812, "mean_p": 0.222772, "boundary_entropy": 0.218643, "bpb": 2.872663}, "ru": {"chunks": 5828.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.307546, "cps": 45.53125, "mean_p": 0.287593, "boundary_entropy": 0.472033, "bpb": 1.731181}, "el": {"chunks": 6395.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.311602, "cps": 49.960938, "mean_p": 0.288015, "boundary_entropy": 0.458078, "bpb": 1.797684}, "hi": {"chunks": 5513.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.327667, "cps": 43.070312, "mean_p": 0.372228, "boundary_entropy": 0.59251, "bpb": 1.411132}, "ta": {"chunks": 5853.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.313078, "cps": 45.726562, "mean_p": 0.300244, "boundary_entropy": 0.524136, "bpb": 1.149768}, "th": {"chunks": 6722.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.417905, "cps": 52.515625, "mean_p": 0.35446, "boundary_entropy": 0.574952, "bpb": 1.383368}, "zh": {"chunks": 3303.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.559262, "cps": 25.804688, "mean_p": 0.248341, "boundary_entropy": 0.223346, "bpb": 3.445654}, "my": {"chunks": 12119.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.639593, "cps": 94.679688, "mean_p": 0.443758, "boundary_entropy": 0.629732, "bpb": 1.319807}}, "bpb": {"en": 2.30964, "de": 2.58971, "vi": 2.10993, "tr": 2.60662, "fi": 2.87266, "ru": 1.73118, "el": 1.79768, "hi": 1.41113, "ta": 1.14977, "th": 1.38337, "zh": 3.44565, "my": 1.31981}, "gini_cpc": 0.217543, "gini_cps": 0.204057, "bpic": 5.846221, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1613929472206392, "vi": 1.1167492194994229, "tr": 1.0431774121770254, "fi": 1.0255416548436591, "ru": 1.5446539880524048, "el": 1.5650224971961562, "hi": 1.6457118671323725, "ta": 1.5724394674595263, "th": 2.098931784203777, "zh": 2.808898280858999, "my": 3.2123606004859964}, "t": 254.3} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.32073, "de": 2.58027, "vi": 2.10958, "tr": 2.60053, "fi": 2.88073, "ru": 1.72939, "el": 1.79341, "hi": 1.41829, "ta": 1.15032, "th": 1.37754, "zh": 3.45097, "my": 1.33007}, "bytes_per_chunk": {"en": 4.837, "de": 4.458, "vi": 5.773, "tr": 5.117, "fi": 5.159, "ru": 5.754, "el": 5.784, "hi": 7.493, "ta": 7.606, "th": 6.404, "zh": 4.873, "my": 4.58}, "bpb_mean": 2.06182, "t": 260.1} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.56354, "bpb": 2.25572, "chunk_loss": 0.030332, "F": 0.19185, "G": 0.25553, "chunks": 196.45, "bytes_per_chunk": 5.212, "boundary_entropy": 0.28996, "router_noise": 0.0, "grad_norm": 0.6965, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.271, "alpha": 0.03, "t": 260.2} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.52115, "bpb": 2.19456, "chunk_loss": 0.030235, "F": 0.18712, "G": 0.24717, "chunks": 191.61, "bytes_per_chunk": 5.344, "boundary_entropy": 0.28053, "router_noise": 0.0, "grad_norm": 0.5981, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 278.3} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.51827, "bpb": 2.1904, "chunk_loss": 0.030384, "F": 0.19426, "G": 0.25523, "chunks": 198.92, "bytes_per_chunk": 5.148, "boundary_entropy": 0.28081, "router_noise": 0.0, "grad_norm": 0.5683, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 296.4} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.56758, "bpb": 2.26154, "chunk_loss": 0.030671, "F": 0.20242, "G": 0.24889, "chunks": 207.28, "bytes_per_chunk": 4.94, "boundary_entropy": 0.24223, "router_noise": 0.0, "grad_norm": 0.5629, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 314.4} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.39648, "bpb": 2.01469, "chunk_loss": 0.03027, "F": 0.18839, "G": 0.23758, "chunks": 192.91, "bytes_per_chunk": 5.308, "boundary_entropy": 0.26309, "router_noise": 0.0, "grad_norm": 0.5403, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 332.2} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.50786, "bpb": 2.17539, "chunk_loss": 0.030347, "F": 0.19205, "G": 0.24474, "chunks": 196.66, "bytes_per_chunk": 5.207, "boundary_entropy": 0.2451, "router_noise": 0.0, "grad_norm": 0.5589, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 350.2} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.53762, "bpb": 2.21832, "chunk_loss": 0.030444, "F": 0.19679, "G": 0.24086, "chunks": 201.52, "bytes_per_chunk": 5.081, "boundary_entropy": 0.22693, "router_noise": 0.0, "grad_norm": 0.6018, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 368.3} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.40866, "bpb": 2.03226, "chunk_loss": 0.03039, "F": 0.18907, "G": 0.25495, "chunks": 193.61, "bytes_per_chunk": 5.289, "boundary_entropy": 0.27866, "router_noise": 0.0, "grad_norm": 0.5183, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 386.4} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.41489, "bpb": 2.04125, "chunk_loss": 0.030489, "F": 0.19508, "G": 0.25077, "chunks": 199.77, "bytes_per_chunk": 5.126, "boundary_entropy": 0.25524, "router_noise": 0.0, "grad_norm": 0.4853, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 404.0} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.40971, "bpb": 2.03379, "chunk_loss": 0.030555, "F": 0.19383, "G": 0.24874, "chunks": 198.48, "bytes_per_chunk": 5.159, "boundary_entropy": 0.25345, "router_noise": 0.0, "grad_norm": 0.4888, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 421.8} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3470.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199391, "cps": 27.109375, "mean_p": 0.203598, "boundary_entropy": 0.11779, "bpb": 2.080389}, "de": {"chunks": 4572.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.222666, "cps": 35.71875, "mean_p": 0.226497, "boundary_entropy": 0.117925, "bpb": 2.270448}, "vi": {"chunks": 3940.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.213724, "cps": 30.78125, "mean_p": 0.23336, "boundary_entropy": 0.295049, "bpb": 1.970165}, "tr": {"chunks": 3342.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.18461, "cps": 26.109375, "mean_p": 0.192084, "boundary_entropy": 0.182949, "bpb": 2.266233}, "fi": {"chunks": 3635.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.193238, "cps": 28.398438, "mean_p": 0.20267, "boundary_entropy": 0.149924, "bpb": 2.438006}, "ru": {"chunks": 5785.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305277, "cps": 45.195312, "mean_p": 0.261879, "boundary_entropy": 0.389844, "bpb": 1.440506}, "el": {"chunks": 5911.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.288018, "cps": 46.179688, "mean_p": 0.253137, "boundary_entropy": 0.396203, "bpb": 1.514638}, "hi": {"chunks": 6425.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.381872, "cps": 50.195312, "mean_p": 0.36237, "boundary_entropy": 0.543812, "bpb": 1.216953}, "ta": {"chunks": 8634.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.461835, "cps": 67.453125, "mean_p": 0.277364, "boundary_entropy": 0.426319, "bpb": 1.002791}, "th": {"chunks": 8106.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.503948, "cps": 63.328125, "mean_p": 0.295937, "boundary_entropy": 0.443641, "bpb": 1.133558}, "zh": {"chunks": 3147.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.532848, "cps": 24.585938, "mean_p": 0.230079, "boundary_entropy": 0.177078, "bpb": 3.164023}, "my": {"chunks": 9751.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.514619, "cps": 76.179688, "mean_p": 0.389987, "boundary_entropy": 0.59927, "bpb": 1.164398}}, "bpb": {"en": 2.08039, "de": 2.27045, "vi": 1.97017, "tr": 2.26623, "fi": 2.43801, "ru": 1.44051, "el": 1.51464, "hi": 1.21695, "ta": 1.00279, "th": 1.13356, "zh": 3.16402, "my": 1.1644}, "gini_cpc": 0.22052, "gini_cps": 0.217083, "bpic": 5.752286, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1167307115515601, "vi": 1.0718838445539238, "tr": 0.925871123944779, "fi": 0.9691414689761662, "ru": 1.5310479572361668, "el": 1.4444907316329287, "hi": 1.9151937001939785, "ta": 2.316227535448844, "th": 2.5274360722351266, "zh": 2.6723783560117145, "my": 2.5809549600027744}, "t": 442.8} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.0884, "de": 2.26124, "vi": 1.97286, "tr": 2.25914, "fi": 2.45293, "ru": 1.44057, "el": 1.50973, "hi": 1.22526, "ta": 1.00415, "th": 1.13128, "zh": 3.17164, "my": 1.17442}, "bytes_per_chunk": {"en": 4.877, "de": 4.633, "vi": 5.884, "tr": 5.759, "fi": 5.341, "ru": 5.892, "el": 6.16, "hi": 6.807, "ta": 5.515, "th": 5.668, "zh": 5.151, "my": 5.35}, "bpb_mean": 1.80763, "t": 448.4} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.40538, "bpb": 2.02754, "chunk_loss": 0.030425, "F": 0.19461, "G": 0.23879, "chunks": 199.28, "bytes_per_chunk": 5.138, "boundary_entropy": 0.22042, "router_noise": 0.0, "grad_norm": 0.4565, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.268, "alpha": 0.03, "t": 448.6} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.45388, "bpb": 2.09751, "chunk_loss": 0.030465, "F": 0.19612, "G": 0.24766, "chunks": 200.83, "bytes_per_chunk": 5.099, "boundary_entropy": 0.23629, "router_noise": 0.0, "grad_norm": 0.4392, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 466.4} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.41327, "bpb": 2.03892, "chunk_loss": 0.03031, "F": 0.19202, "G": 0.23556, "chunks": 196.62, "bytes_per_chunk": 5.208, "boundary_entropy": 0.2245, "router_noise": 0.0, "grad_norm": 0.4346, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 484.1} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.34531, "bpb": 1.94087, "chunk_loss": 0.030189, "F": 0.18309, "G": 0.24039, "chunks": 187.48, "bytes_per_chunk": 5.462, "boundary_entropy": 0.24554, "router_noise": 0.0, "grad_norm": 0.4143, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 502.0} |
| {"event": "train", "step": 2400, "phase": 1, "objective": "global", "lm_loss": 1.37142, "bpb": 1.97853, "chunk_loss": 0.030462, "F": 0.19086, "G": 0.24345, "chunks": 195.44, "bytes_per_chunk": 5.24, "boundary_entropy": 0.24364, "router_noise": 0.0, "grad_norm": 0.4473, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.176, "alpha": 0.03, "t": 519.6} |
| {"event": "train", "step": 2500, "phase": 1, "objective": "global", "lm_loss": 1.22594, "bpb": 1.76865, "chunk_loss": 0.030417, "F": 0.18855, "G": 0.24943, "chunks": 193.08, "bytes_per_chunk": 5.304, "boundary_entropy": 0.26419, "router_noise": 0.0, "grad_norm": 0.3932, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 537.6} |
| {"event": "train", "step": 2600, "phase": 1, "objective": "global", "lm_loss": 1.2206, "bpb": 1.76095, "chunk_loss": 0.030178, "F": 0.18048, "G": 0.24017, "chunks": 184.81, "bytes_per_chunk": 5.541, "boundary_entropy": 0.25519, "router_noise": 0.0, "grad_norm": 0.4166, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.176, "alpha": 0.03, "t": 555.2} |
| {"event": "train", "step": 2700, "phase": 1, "objective": "global", "lm_loss": 1.36174, "bpb": 1.96457, "chunk_loss": 0.030352, "F": 0.19553, "G": 0.23513, "chunks": 200.22, "bytes_per_chunk": 5.114, "boundary_entropy": 0.2036, "router_noise": 0.0, "grad_norm": 0.4696, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 572.7} |
| {"event": "train", "step": 2800, "phase": 1, "objective": "global", "lm_loss": 1.1795, "bpb": 1.70166, "chunk_loss": 0.030285, "F": 0.18832, "G": 0.23873, "chunks": 192.84, "bytes_per_chunk": 5.31, "boundary_entropy": 0.24462, "router_noise": 0.0, "grad_norm": 0.3948, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.174, "alpha": 0.03, "t": 590.0} |
| {"event": "train", "step": 2900, "phase": 1, "objective": "global", "lm_loss": 1.23039, "bpb": 1.77507, "chunk_loss": 0.030216, "F": 0.18565, "G": 0.24098, "chunks": 190.11, "bytes_per_chunk": 5.386, "boundary_entropy": 0.24525, "router_noise": 0.0, "grad_norm": 0.4078, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.173, "alpha": 0.03, "t": 607.3} |
| {"event": "probe", "step": 3000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3430.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197092, "cps": 26.796875, "mean_p": 0.192344, "boundary_entropy": 0.103318, "bpb": 1.910679}, "de": {"chunks": 4399.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.21424, "cps": 34.367188, "mean_p": 0.21521, "boundary_entropy": 0.09672, "bpb": 2.085094}, "vi": {"chunks": 3750.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.203417, "cps": 29.296875, "mean_p": 0.21675, "boundary_entropy": 0.25954, "bpb": 1.822495}, "tr": {"chunks": 3284.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181406, "cps": 25.65625, "mean_p": 0.182991, "boundary_entropy": 0.159604, "bpb": 2.038117}, "fi": {"chunks": 3494.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185742, "cps": 27.296875, "mean_p": 0.186322, "boundary_entropy": 0.116009, "bpb": 2.251701}, "ru": {"chunks": 5804.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.30628, "cps": 45.34375, "mean_p": 0.236183, "boundary_entropy": 0.323947, "bpb": 1.316282}, "el": {"chunks": 6650.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.324027, "cps": 51.953125, "mean_p": 0.251632, "boundary_entropy": 0.346809, "bpb": 1.358618}, "hi": {"chunks": 7028.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.417712, "cps": 54.90625, "mean_p": 0.370183, "boundary_entropy": 0.53898, "bpb": 1.114843}, "ta": {"chunks": 8188.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.437978, "cps": 63.96875, "mean_p": 0.302876, "boundary_entropy": 0.465014, "bpb": 0.938344}, "th": {"chunks": 7365.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.45788, "cps": 57.539062, "mean_p": 0.351983, "boundary_entropy": 0.510335, "bpb": 1.053169}, "zh": {"chunks": 3163.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.535557, "cps": 24.710938, "mean_p": 0.219669, "boundary_entropy": 0.139811, "bpb": 2.974525}, "my": {"chunks": 8887.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.46902, "cps": 69.429688, "mean_p": 0.400182, "boundary_entropy": 0.602117, "bpb": 1.065192}}, "bpb": {"en": 1.91068, "de": 2.08509, "vi": 1.82249, "tr": 2.03812, "fi": 2.2517, "ru": 1.31628, "el": 1.35862, "hi": 1.11484, "ta": 0.93834, "th": 1.05317, "zh": 2.97452, "my": 1.06519}, "gini_cpc": 0.213544, "gini_cps": 0.206824, "bpic": 5.864445, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0870050330698546, "vi": 1.0320913207779916, "tr": 0.9204126887140301, "fi": 0.942412455462679, "ru": 1.5539898920744326, "el": 1.6440340205662736, "hi": 2.119369499954514, "ta": 2.22219601599093, "th": 2.323173719063539, "zh": 2.717288491517743, "my": 2.379697773436707}, "t": 627.8} |
| {"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 1.91323, "de": 2.08221, "vi": 1.83189, "tr": 2.03098, "fi": 2.26478, "ru": 1.31631, "el": 1.35469, "hi": 1.12083, "ta": 0.9403, "th": 1.05077, "zh": 2.98443, "my": 1.07451}, "bytes_per_chunk": {"en": 4.916, "de": 4.781, "vi": 6.04, "tr": 5.712, "fi": 5.498, "ru": 5.916, "el": 5.506, "hi": 6.174, "ta": 5.904, "th": 6.254, "zh": 5.16, "my": 6.058}, "bpb_mean": 1.66374, "t": 633.4} |
| {"event": "train", "step": 3000, "phase": 1, "objective": "global", "lm_loss": 1.26057, "bpb": 1.81861, "chunk_loss": 0.030208, "F": 0.18677, "G": 0.2264, "chunks": 191.25, "bytes_per_chunk": 5.354, "boundary_entropy": 0.20854, "router_noise": 0.0, "grad_norm": 0.4032, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.263, "alpha": 0.03, "t": 633.6} |
| {"event": "train", "step": 3100, "phase": 1, "objective": "global", "lm_loss": 1.1887, "bpb": 1.71493, "chunk_loss": 0.030159, "F": 0.18558, "G": 0.24053, "chunks": 190.03, "bytes_per_chunk": 5.389, "boundary_entropy": 0.22908, "router_noise": 0.0, "grad_norm": 0.408, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 651.3} |
| {"event": "train", "step": 3200, "phase": 1, "objective": "global", "lm_loss": 1.256, "bpb": 1.81202, "chunk_loss": 0.030392, "F": 0.1907, "G": 0.24657, "chunks": 195.28, "bytes_per_chunk": 5.244, "boundary_entropy": 0.24378, "router_noise": 0.0, "grad_norm": 0.445, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 669.1} |
| {"event": "train", "step": 3300, "phase": 1, "objective": "global", "lm_loss": 1.25377, "bpb": 1.80881, "chunk_loss": 0.030265, "F": 0.18658, "G": 0.22775, "chunks": 191.06, "bytes_per_chunk": 5.36, "boundary_entropy": 0.20914, "router_noise": 0.0, "grad_norm": 0.4111, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 686.8} |
| {"event": "train", "step": 3400, "phase": 1, "objective": "global", "lm_loss": 1.31324, "bpb": 1.8946, "chunk_loss": 0.030318, "F": 0.18965, "G": 0.23112, "chunks": 194.2, "bytes_per_chunk": 5.273, "boundary_entropy": 0.19922, "router_noise": 0.0, "grad_norm": 0.4144, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 704.3} |
| {"event": "train", "step": 3500, "phase": 1, "objective": "global", "lm_loss": 1.16988, "bpb": 1.68778, "chunk_loss": 0.030172, "F": 0.18524, "G": 0.23262, "chunks": 189.69, "bytes_per_chunk": 5.398, "boundary_entropy": 0.21925, "router_noise": 0.0, "grad_norm": 0.4117, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.176, "alpha": 0.03, "t": 721.8} |
| {"event": "train", "step": 3600, "phase": 1, "objective": "global", "lm_loss": 1.2192, "bpb": 1.75894, "chunk_loss": 0.030205, "F": 0.18587, "G": 0.23665, "chunks": 190.33, "bytes_per_chunk": 5.38, "boundary_entropy": 0.21571, "router_noise": 0.0, "grad_norm": 0.4059, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 739.5} |
| {"event": "train", "step": 3700, "phase": 1, "objective": "global", "lm_loss": 1.20639, "bpb": 1.74046, "chunk_loss": 0.030208, "F": 0.18199, "G": 0.24165, "chunks": 186.36, "bytes_per_chunk": 5.495, "boundary_entropy": 0.24149, "router_noise": 0.0, "grad_norm": 0.4025, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.174, "alpha": 0.03, "t": 756.9} |
| {"event": "train", "step": 3800, "phase": 1, "objective": "global", "lm_loss": 1.22726, "bpb": 1.77056, "chunk_loss": 0.030362, "F": 0.19272, "G": 0.23947, "chunks": 197.34, "bytes_per_chunk": 5.189, "boundary_entropy": 0.20701, "router_noise": 0.0, "grad_norm": 0.388, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.176, "alpha": 0.03, "t": 774.5} |
| {"event": "train", "step": 3900, "phase": 1, "objective": "global", "lm_loss": 1.20674, "bpb": 1.74096, "chunk_loss": 0.03026, "F": 0.18755, "G": 0.22844, "chunks": 192.05, "bytes_per_chunk": 5.332, "boundary_entropy": 0.20186, "router_noise": 0.0, "grad_norm": 0.3946, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 792.0} |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3439.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19761, "cps": 26.867188, "mean_p": 0.193962, "boundary_entropy": 0.089585, "bpb": 1.830332}, "de": {"chunks": 4424.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.215458, "cps": 34.5625, "mean_p": 0.21513, "boundary_entropy": 0.085951, "bpb": 1.979359}, "vi": {"chunks": 3610.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.195823, "cps": 28.203125, "mean_p": 0.212633, "boundary_entropy": 0.251327, "bpb": 1.706511}, "tr": {"chunks": 3200.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176766, "cps": 25.0, "mean_p": 0.184655, "boundary_entropy": 0.156821, "bpb": 1.920425}, "fi": {"chunks": 3528.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18755, "cps": 27.5625, "mean_p": 0.187134, "boundary_entropy": 0.103084, "bpb": 2.108229}, "ru": {"chunks": 5817.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306966, "cps": 45.445312, "mean_p": 0.227922, "boundary_entropy": 0.298272, "bpb": 1.256189}, "el": {"chunks": 6006.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.292647, "cps": 46.921875, "mean_p": 0.225055, "boundary_entropy": 0.314859, "bpb": 1.279926}, "hi": {"chunks": 7111.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.422645, "cps": 55.554688, "mean_p": 0.400735, "boundary_entropy": 0.545654, "bpb": 1.054453}, "ta": {"chunks": 7920.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.423643, "cps": 61.875, "mean_p": 0.325328, "boundary_entropy": 0.502827, "bpb": 0.911585}, "th": {"chunks": 7733.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.480758, "cps": 60.414062, "mean_p": 0.354583, "boundary_entropy": 0.515751, "bpb": 1.00814}, "zh": {"chunks": 3197.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.541314, "cps": 24.976562, "mean_p": 0.217516, "boundary_entropy": 0.127595, "bpb": 2.838993}, "my": {"chunks": 8933.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.471448, "cps": 69.789062, "mean_p": 0.414416, "boundary_entropy": 0.605904, "bpb": 1.018816}}, "bpb": {"en": 1.83033, "de": 1.97936, "vi": 1.70651, "tr": 1.92043, "fi": 2.10823, "ru": 1.25619, "el": 1.27993, "hi": 1.05445, "ta": 0.91159, "th": 1.00814, "zh": 2.83899, "my": 1.01882}, "gini_cpc": 0.220319, "gini_cps": 0.208122, "bpic": 5.911781, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.090321699604635, "vi": 0.9909597256000252, "tr": 0.894522710880425, "fi": 0.9490927095934739, "ru": 1.5533946098646523, "el": 1.4809364657673771, "hi": 2.1387870709504258, "ta": 2.143836478833599, "th": 2.4328699227343686, "zh": 2.739309716724172, "my": 2.385755327492767}, "t": 812.9} |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.83017, "de": 1.98245, "vi": 1.71968, "tr": 1.91385, "fi": 2.12005, "ru": 1.25548, "el": 1.27875, "hi": 1.06098, "ta": 0.91364, "th": 1.00629, "zh": 2.84983, "my": 1.02646}, "bytes_per_chunk": {"en": 4.903, "de": 4.784, "vi": 6.228, "tr": 5.84, "fi": 5.482, "ru": 5.93, "el": 6.138, "hi": 6.123, "ta": 6.153, "th": 6.044, "zh": 5.153, "my": 6.225}, "bpb_mean": 1.5798, "t": 818.5} |
| {"event": "train", "step": 4000, "phase": 1, "objective": "global", "lm_loss": 1.14057, "bpb": 1.64549, "chunk_loss": 0.030138, "F": 0.18228, "G": 0.24134, "chunks": 186.66, "bytes_per_chunk": 5.486, "boundary_entropy": 0.23128, "router_noise": 0.0, "grad_norm": 0.3546, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.266, "alpha": 0.03, "t": 818.6} |
| {"event": "train", "step": 4100, "phase": 1, "objective": "global", "lm_loss": 1.21562, "bpb": 1.75376, "chunk_loss": 0.030198, "F": 0.1853, "G": 0.22001, "chunks": 189.75, "bytes_per_chunk": 5.397, "boundary_entropy": 0.18553, "router_noise": 0.0, "grad_norm": 0.3838, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 836.1} |
| {"event": "train", "step": 4200, "phase": 1, "objective": "global", "lm_loss": 1.19293, "bpb": 1.72103, "chunk_loss": 0.030361, "F": 0.19124, "G": 0.2398, "chunks": 195.83, "bytes_per_chunk": 5.229, "boundary_entropy": 0.21948, "router_noise": 0.0, "grad_norm": 0.3868, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 853.6} |
| {"event": "train", "step": 4300, "phase": 1, "objective": "global", "lm_loss": 1.12543, "bpb": 1.62366, "chunk_loss": 0.030146, "F": 0.18521, "G": 0.24567, "chunks": 189.66, "bytes_per_chunk": 5.399, "boundary_entropy": 0.23398, "router_noise": 0.0, "grad_norm": 0.374, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 871.1} |
| {"event": "phase_transition", "step": 4400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 888.4} |
| {"event": "train", "step": 4400, "phase": 2, "objective": "parity_a", "lm_loss": 1.15967, "bpb": 1.67306, "chunk_loss": 0.033992, "F": 0.19189, "G": 0.23286, "chunks": 196.5, "bytes_per_chunk": 5.211, "boundary_entropy": 0.19941, "router_noise": 0.0, "grad_norm": 0.3812, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 8.9436, "alpha": 0.03, "t": 888.6} |
| {"event": "train", "step": 4500, "phase": 2, "objective": "parity_a", "lm_loss": 1.26775, "bpb": 1.82897, "chunk_loss": 0.030561, "F": 0.17792, "G": 0.18811, "chunks": 182.19, "bytes_per_chunk": 5.621, "boundary_entropy": 0.12521, "router_noise": 0.0, "grad_norm": 0.3934, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 7.6421, "alpha": 0.03, "t": 906.2} |
| {"event": "train", "step": 4600, "phase": 2, "objective": "parity_a", "lm_loss": 1.11559, "bpb": 1.60946, "chunk_loss": 0.030597, "F": 0.15718, "G": 0.17095, "chunks": 160.95, "bytes_per_chunk": 6.362, "boundary_entropy": 0.13291, "router_noise": 0.0, "grad_norm": 0.3797, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.3412, "alpha": 0.03, "t": 923.6} |
| {"event": "train", "step": 4700, "phase": 2, "objective": "parity_a", "lm_loss": 1.18615, "bpb": 1.71125, "chunk_loss": 0.030504, "F": 0.16396, "G": 0.17277, "chunks": 167.89, "bytes_per_chunk": 6.099, "boundary_entropy": 0.11963, "router_noise": 0.0, "grad_norm": 0.379, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 8.9439, "alpha": 0.03, "t": 941.0} |
| {"event": "train", "step": 4800, "phase": 2, "objective": "parity_a", "lm_loss": 1.16216, "bpb": 1.67665, "chunk_loss": 0.030349, "F": 0.16257, "G": 0.17178, "chunks": 166.47, "bytes_per_chunk": 6.151, "boundary_entropy": 0.1071, "router_noise": 0.0, "grad_norm": 0.3851, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.9799, "alpha": 0.03, "t": 958.6} |
| {"event": "train", "step": 4900, "phase": 2, "objective": "parity_a", "lm_loss": 1.23254, "bpb": 1.77818, "chunk_loss": 0.030406, "F": 0.17261, "G": 0.1786, "chunks": 176.75, "bytes_per_chunk": 5.793, "boundary_entropy": 0.1041, "router_noise": 0.0, "grad_norm": 0.3879, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.1677, "alpha": 0.03, "t": 976.1} |
| {"event": "probe", "step": 5000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3416.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196288, "cps": 26.6875, "mean_p": 0.189845, "boundary_entropy": 0.069811, "bpb": 1.778485}, "de": {"chunks": 4403.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.214435, "cps": 34.398438, "mean_p": 0.208432, "boundary_entropy": 0.068663, "bpb": 1.921576}, "vi": {"chunks": 3222.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.174776, "cps": 25.171875, "mean_p": 0.157106, "boundary_entropy": 0.13373, "bpb": 1.639399}, "tr": {"chunks": 3072.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169696, "cps": 24.0, "mean_p": 0.16409, "boundary_entropy": 0.106969, "bpb": 1.860066}, "fi": {"chunks": 3461.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183988, "cps": 27.039062, "mean_p": 0.179213, "boundary_entropy": 0.071394, "bpb": 2.041366}, "ru": {"chunks": 3748.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.197784, "cps": 29.28125, "mean_p": 0.134977, "boundary_entropy": 0.191237, "bpb": 1.225374}, "el": {"chunks": 4144.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.20192, "cps": 32.375, "mean_p": 0.14154, "boundary_entropy": 0.182533, "bpb": 1.245441}, "hi": {"chunks": 3474.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.206478, "cps": 27.140625, "mean_p": 0.099385, "boundary_entropy": 0.110944, "bpb": 1.05347}, "ta": {"chunks": 3211.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.171757, "cps": 25.085938, "mean_p": 0.087788, "boundary_entropy": 0.145224, "bpb": 0.899561}, "th": {"chunks": 4037.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.250979, "cps": 31.539062, "mean_p": 0.10659, "boundary_entropy": 0.123554, "bpb": 1.007095}, "zh": {"chunks": 3180.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.538435, "cps": 24.84375, "mean_p": 0.213382, "boundary_entropy": 0.103702, "bpb": 2.694244}, "my": {"chunks": 2847.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.150253, "cps": 22.242188, "mean_p": 0.095417, "boundary_entropy": 0.197877, "bpb": 1.012172}}, "bpb": {"en": 1.77849, "de": 1.92158, "vi": 1.6394, "tr": 1.86007, "fi": 2.04137, "ru": 1.22537, "el": 1.24544, "hi": 1.05347, "ta": 0.89956, "th": 1.00709, "zh": 2.69424, "my": 1.01217}, "gini_cpc": 0.172956, "gini_cps": 0.071556, "bpic": 9.091105, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0924524336274355, "vi": 0.8904071778239768, "tr": 0.8645237291010159, "fi": 0.9373374608782519, "ru": 1.0076196449426262, "el": 1.0286915200299065, "hi": 1.0519158509383342, "ta": 0.8750262753569722, "th": 1.2786272102465024, "zh": 2.743089233272875, "my": 0.7654738328589672}, "t": 996.7} |
| {"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.77764, "de": 1.92441, "vi": 1.65398, "tr": 1.85336, "fi": 2.05206, "ru": 1.22494, "el": 1.24353, "hi": 1.05896, "ta": 0.90287, "th": 1.00535, "zh": 2.6993, "my": 1.01898}, "bytes_per_chunk": {"en": 4.934, "de": 4.814, "vi": 7.195, "tr": 6.083, "fi": 5.529, "ru": 9.057, "el": 8.863, "hi": 11.931, "ta": 14.364, "th": 12.012, "zh": 5.135, "my": 18.144}, "bpb_mean": 1.53461, "t": 1001.9} |
| {"event": "train", "step": 5000, "phase": 2, "objective": "parity_a", "lm_loss": 1.16845, "bpb": 1.68572, "chunk_loss": 0.03035, "F": 0.15984, "G": 0.1683, "chunks": 163.67, "bytes_per_chunk": 6.256, "boundary_entropy": 0.10661, "router_noise": 0.0, "grad_norm": 0.376, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.26, "n_target_mean": 9.2014, "alpha": 0.03, "t": 1002.1} |
| {"event": "train", "step": 5100, "phase": 2, "objective": "parity_a", "lm_loss": 1.17302, "bpb": 1.69231, "chunk_loss": 0.030613, "F": 0.16841, "G": 0.17418, "chunks": 172.45, "bytes_per_chunk": 5.938, "boundary_entropy": 0.10286, "router_noise": 0.0, "grad_norm": 0.3935, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.5306, "alpha": 0.03, "t": 1019.8} |
| {"event": "train", "step": 5200, "phase": 2, "objective": "parity_a", "lm_loss": 1.17559, "bpb": 1.69602, "chunk_loss": 0.030571, "F": 0.16692, "G": 0.17531, "chunks": 170.92, "bytes_per_chunk": 5.991, "boundary_entropy": 0.10674, "router_noise": 0.0, "grad_norm": 0.3905, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.802, "alpha": 0.03, "t": 1037.2} |
| {"event": "train", "step": 5300, "phase": 2, "objective": "parity_a", "lm_loss": 1.13449, "bpb": 1.63672, "chunk_loss": 0.03041, "F": 0.15848, "G": 0.16848, "chunks": 162.28, "bytes_per_chunk": 6.31, "boundary_entropy": 0.10241, "router_noise": 0.0, "grad_norm": 0.3847, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 8.9067, "alpha": 0.03, "t": 1054.7} |
| {"event": "train", "step": 5400, "phase": 2, "objective": "parity_a", "lm_loss": 1.09697, "bpb": 1.58259, "chunk_loss": 0.030391, "F": 0.16008, "G": 0.1665, "chunks": 163.92, "bytes_per_chunk": 6.247, "boundary_entropy": 0.0978, "router_noise": 0.0, "grad_norm": 0.3743, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.9318, "alpha": 0.03, "t": 1072.1} |
| {"event": "train", "step": 5500, "phase": 2, "objective": "parity_a", "lm_loss": 1.10463, "bpb": 1.59365, "chunk_loss": 0.030507, "F": 0.15401, "G": 0.16279, "chunks": 157.7, "bytes_per_chunk": 6.493, "boundary_entropy": 0.10532, "router_noise": 0.0, "grad_norm": 0.3549, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 9.5417, "alpha": 0.03, "t": 1089.5} |
| {"event": "train", "step": 5600, "phase": 2, "objective": "parity_a", "lm_loss": 1.12045, "bpb": 1.61646, "chunk_loss": 0.030515, "F": 0.15852, "G": 0.16591, "chunks": 162.33, "bytes_per_chunk": 6.308, "boundary_entropy": 0.10164, "router_noise": 0.0, "grad_norm": 0.3675, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.1849, "alpha": 0.03, "t": 1107.1} |
| {"event": "train", "step": 5700, "phase": 2, "objective": "parity_a", "lm_loss": 1.1564, "bpb": 1.66834, "chunk_loss": 0.03035, "F": 0.16504, "G": 0.16975, "chunks": 169.0, "bytes_per_chunk": 6.059, "boundary_entropy": 0.09076, "router_noise": 0.0, "grad_norm": 0.374, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.6518, "alpha": 0.03, "t": 1124.5} |
| {"event": "train", "step": 5800, "phase": 2, "objective": "parity_a", "lm_loss": 1.08285, "bpb": 1.56222, "chunk_loss": 0.03028, "F": 0.14337, "G": 0.15208, "chunks": 146.81, "bytes_per_chunk": 6.975, "boundary_entropy": 0.09448, "router_noise": 0.0, "grad_norm": 0.3639, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.9423, "alpha": 0.03, "t": 1142.1} |
| {"event": "train", "step": 5900, "phase": 2, "objective": "parity_a", "lm_loss": 1.11208, "bpb": 1.60439, "chunk_loss": 0.030314, "F": 0.15034, "G": 0.15772, "chunks": 153.95, "bytes_per_chunk": 6.651, "boundary_entropy": 0.09551, "router_noise": 0.0, "grad_norm": 0.3743, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.2794, "alpha": 0.03, "t": 1159.7} |
| {"event": "probe", "step": 6000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3437.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197495, "cps": 26.851562, "mean_p": 0.19064, "boundary_entropy": 0.056142, "bpb": 1.741215}, "de": {"chunks": 4402.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.214387, "cps": 34.390625, "mean_p": 0.208359, "boundary_entropy": 0.054381, "bpb": 1.874952}, "vi": {"chunks": 3138.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.17022, "cps": 24.515625, "mean_p": 0.144913, "boundary_entropy": 0.093406, "bpb": 1.590034}, "tr": {"chunks": 2933.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.162017, "cps": 22.914062, "mean_p": 0.154921, "boundary_entropy": 0.081825, "bpb": 1.806537}, "fi": {"chunks": 3476.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184785, "cps": 27.15625, "mean_p": 0.176626, "boundary_entropy": 0.061033, "bpb": 1.990392}, "ru": {"chunks": 3373.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.177995, "cps": 26.351562, "mean_p": 0.119673, "boundary_entropy": 0.160936, "bpb": 1.182685}, "el": {"chunks": 3619.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.176339, "cps": 28.273438, "mean_p": 0.11687, "boundary_entropy": 0.139252, "bpb": 1.205121}, "hi": {"chunks": 3560.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.21159, "cps": 27.8125, "mean_p": 0.093612, "boundary_entropy": 0.090499, "bpb": 1.002948}, "ta": {"chunks": 2758.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.147526, "cps": 21.546875, "mean_p": 0.080142, "boundary_entropy": 0.134504, "bpb": 0.876081}, "th": {"chunks": 3224.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.200435, "cps": 25.1875, "mean_p": 0.081744, "boundary_entropy": 0.092593, "bpb": 0.96319}, "zh": {"chunks": 3191.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.540298, "cps": 24.929688, "mean_p": 0.210264, "boundary_entropy": 0.083848, "bpb": 2.615079}, "my": {"chunks": 2525.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.133259, "cps": 19.726562, "mean_p": 0.071579, "boundary_entropy": 0.145502, "bpb": 0.972655}}, "bpb": {"en": 1.74122, "de": 1.87495, "vi": 1.59003, "tr": 1.80654, "fi": 1.99039, "ru": 1.18268, "el": 1.20512, "hi": 1.00295, "ta": 0.87608, "th": 0.96319, "zh": 2.61508, "my": 0.97266}, "gini_cpc": 0.187028, "gini_cps": 0.074347, "bpic": 9.682637, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0855309720136426, "vi": 0.8618950477418359, "tr": 0.8203630661524633, "fi": 0.9356479543297694, "ru": 0.9012633586233315, "el": 0.8928784187588253, "hi": 1.0713700785222953, "ta": 0.7469875790438102, "th": 1.0148890246923417, "zh": 2.7357597117718857, "my": 0.6747495357952232}, "t": 1180.0} |
| {"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.73987, "de": 1.87808, "vi": 1.60524, "tr": 1.79854, "fi": 2.00307, "ru": 1.18321, "el": 1.20303, "hi": 1.00792, "ta": 0.87877, "th": 0.9622, "zh": 2.62007, "my": 0.97979}, "bytes_per_chunk": {"en": 4.923, "de": 4.818, "vi": 7.405, "tr": 6.347, "fi": 5.555, "ru": 9.986, "el": 10.423, "hi": 12.035, "ta": 17.081, "th": 15.055, "zh": 5.101, "my": 20.371}, "bpb_mean": 1.48832, "t": 1185.2} |
| {"event": "train", "step": 6000, "phase": 2, "objective": "parity_a", "lm_loss": 1.11634, "bpb": 1.61054, "chunk_loss": 0.03036, "F": 0.15242, "G": 0.16034, "chunks": 156.08, "bytes_per_chunk": 6.561, "boundary_entropy": 0.08951, "router_noise": 0.0, "grad_norm": 0.3904, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.257, "n_target_mean": 9.5014, "alpha": 0.03, "t": 1185.4} |
| {"event": "train", "step": 6100, "phase": 2, "objective": "parity_a", "lm_loss": 1.17334, "bpb": 1.69277, "chunk_loss": 0.030314, "F": 0.15952, "G": 0.16563, "chunks": 163.34, "bytes_per_chunk": 6.269, "boundary_entropy": 0.09263, "router_noise": 0.0, "grad_norm": 0.3799, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 9.0707, "alpha": 0.03, "t": 1202.8} |
| {"event": "train", "step": 6200, "phase": 2, "objective": "parity_a", "lm_loss": 1.16059, "bpb": 1.67437, "chunk_loss": 0.030525, "F": 0.15808, "G": 0.16493, "chunks": 161.88, "bytes_per_chunk": 6.326, "boundary_entropy": 0.09105, "router_noise": 0.0, "grad_norm": 0.377, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.6046, "alpha": 0.03, "t": 1220.5} |
| {"event": "train", "step": 6300, "phase": 2, "objective": "parity_a", "lm_loss": 1.1945, "bpb": 1.7233, "chunk_loss": 0.030466, "F": 0.16554, "G": 0.17063, "chunks": 169.52, "bytes_per_chunk": 6.041, "boundary_entropy": 0.08702, "router_noise": 0.0, "grad_norm": 0.3834, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.3698, "alpha": 0.03, "t": 1238.3} |
| {"event": "train", "step": 6400, "phase": 2, "objective": "parity_a", "lm_loss": 1.10774, "bpb": 1.59813, "chunk_loss": 0.030359, "F": 0.15742, "G": 0.16183, "chunks": 161.2, "bytes_per_chunk": 6.352, "boundary_entropy": 0.08773, "router_noise": 0.0, "grad_norm": 0.3613, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 8.9315, "alpha": 0.03, "t": 1256.0} |
| {"event": "train", "step": 6500, "phase": 2, "objective": "parity_a", "lm_loss": 1.12914, "bpb": 1.629, "chunk_loss": 0.030307, "F": 0.15361, "G": 0.15916, "chunks": 157.3, "bytes_per_chunk": 6.51, "boundary_entropy": 0.08406, "router_noise": 0.0, "grad_norm": 0.3719, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.0889, "alpha": 0.03, "t": 1273.6} |
| {"event": "train", "step": 6600, "phase": 2, "objective": "parity_a", "lm_loss": 1.08297, "bpb": 1.56239, "chunk_loss": 0.030327, "F": 0.14694, "G": 0.15258, "chunks": 150.47, "bytes_per_chunk": 6.805, "boundary_entropy": 0.09333, "router_noise": 0.0, "grad_norm": 0.3703, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.6297, "alpha": 0.03, "t": 1291.1} |
| {"event": "train", "step": 6700, "phase": 2, "objective": "parity_a", "lm_loss": 1.10485, "bpb": 1.59397, "chunk_loss": 0.030316, "F": 0.1519, "G": 0.15877, "chunks": 155.55, "bytes_per_chunk": 6.583, "boundary_entropy": 0.08434, "router_noise": 0.0, "grad_norm": 0.3696, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.4445, "alpha": 0.03, "t": 1308.3} |
| {"event": "train", "step": 6800, "phase": 2, "objective": "parity_a", "lm_loss": 1.14925, "bpb": 1.65802, "chunk_loss": 0.030334, "F": 0.1665, "G": 0.1701, "chunks": 170.5, "bytes_per_chunk": 6.006, "boundary_entropy": 0.0819, "router_noise": 0.0, "grad_norm": 0.3715, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.2605, "alpha": 0.03, "t": 1325.5} |
| {"event": "train", "step": 6900, "phase": 2, "objective": "parity_a", "lm_loss": 1.12952, "bpb": 1.62955, "chunk_loss": 0.030263, "F": 0.16493, "G": 0.16782, "chunks": 168.89, "bytes_per_chunk": 6.063, "boundary_entropy": 0.07833, "router_noise": 0.0, "grad_norm": 0.3725, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 8.0676, "alpha": 0.03, "t": 1342.8} |
| {"event": "probe", "step": 7000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3320.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.190772, "cps": 25.9375, "mean_p": 0.183605, "boundary_entropy": 0.057046, "bpb": 1.712981}, "de": {"chunks": 4361.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.21239, "cps": 34.070312, "mean_p": 0.203494, "boundary_entropy": 0.05854, "bpb": 1.840063}, "vi": {"chunks": 3090.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.167616, "cps": 24.140625, "mean_p": 0.143649, "boundary_entropy": 0.091059, "bpb": 1.543097}, "tr": {"chunks": 2971.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164116, "cps": 23.210938, "mean_p": 0.155595, "boundary_entropy": 0.075602, "bpb": 1.772613}, "fi": {"chunks": 3470.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184467, "cps": 27.109375, "mean_p": 0.176687, "boundary_entropy": 0.056762, "bpb": 1.937725}, "ru": {"chunks": 3388.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.178786, "cps": 26.46875, "mean_p": 0.118804, "boundary_entropy": 0.149634, "bpb": 1.148204}, "el": {"chunks": 3482.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.169663, "cps": 27.203125, "mean_p": 0.110528, "boundary_entropy": 0.124865, "bpb": 1.168417}, "hi": {"chunks": 3267.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.194175, "cps": 25.523438, "mean_p": 0.082957, "boundary_entropy": 0.077209, "bpb": 0.970904}, "ta": {"chunks": 2755.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.147366, "cps": 21.523438, "mean_p": 0.071292, "boundary_entropy": 0.106723, "bpb": 0.846523}, "th": {"chunks": 3428.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.213118, "cps": 26.78125, "mean_p": 0.084587, "boundary_entropy": 0.080323, "bpb": 0.931057}, "zh": {"chunks": 3200.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.541822, "cps": 25.0, "mean_p": 0.208889, "boundary_entropy": 0.075244, "bpb": 2.571544}, "my": {"chunks": 2301.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.121438, "cps": 17.976562, "mean_p": 0.064225, "boundary_entropy": 0.124257, "bpb": 0.943129}}, "bpb": {"en": 1.71298, "de": 1.84006, "vi": 1.5431, "tr": 1.77261, "fi": 1.93772, "ru": 1.1482, "el": 1.16842, "hi": 0.9709, "ta": 0.84652, "th": 0.93106, "zh": 2.57154, "my": 0.94313}, "gini_cpc": 0.192959, "gini_cps": 0.07473, "bpic": 9.832219, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1133192439558066, "vi": 0.8786205848618232, "tr": 0.8602766532073163, "fi": 0.9669491319165343, "ru": 0.9371739835330769, "el": 0.8893525571368943, "hi": 1.0178413684455503, "ta": 0.7724709751012608, "th": 1.1171353240128683, "zh": 2.840158466578811, "my": 0.6365599033245076}, "t": 1363.1} |
| {"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.71138, "de": 1.843, "vi": 1.55838, "tr": 1.76497, "fi": 1.94905, "ru": 1.14832, "el": 1.16617, "hi": 0.97651, "ta": 0.84985, "th": 0.93109, "zh": 2.57454, "my": 0.94823}, "bytes_per_chunk": {"en": 5.091, "de": 4.853, "vi": 7.467, "tr": 6.326, "fi": 5.551, "ru": 9.917, "el": 10.651, "hi": 13.036, "ta": 16.893, "th": 14.45, "zh": 5.124, "my": 22.255}, "bpb_mean": 1.45179, "t": 1368.3} |
| {"event": "train", "step": 7000, "phase": 2, "objective": "parity_a", "lm_loss": 1.2213, "bpb": 1.76197, "chunk_loss": 0.03043, "F": 0.16461, "G": 0.1687, "chunks": 168.56, "bytes_per_chunk": 6.075, "boundary_entropy": 0.08047, "router_noise": 0.0, "grad_norm": 0.3988, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.256, "n_target_mean": 7.9992, "alpha": 0.03, "t": 1368.4} |
| {"event": "train", "step": 7100, "phase": 2, "objective": "parity_a", "lm_loss": 1.13927, "bpb": 1.64363, "chunk_loss": 0.030286, "F": 0.15814, "G": 0.16264, "chunks": 161.94, "bytes_per_chunk": 6.323, "boundary_entropy": 0.08103, "router_noise": 0.0, "grad_norm": 0.3767, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.8248, "alpha": 0.03, "t": 1385.9} |
| {"event": "train", "step": 7200, "phase": 2, "objective": "parity_a", "lm_loss": 1.03696, "bpb": 1.49601, "chunk_loss": 0.030293, "F": 0.1525, "G": 0.1577, "chunks": 156.16, "bytes_per_chunk": 6.558, "boundary_entropy": 0.08647, "router_noise": 0.0, "grad_norm": 0.3632, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.0753, "alpha": 0.03, "t": 1403.2} |
| {"event": "train", "step": 7300, "phase": 2, "objective": "parity_a", "lm_loss": 1.1253, "bpb": 1.62346, "chunk_loss": 0.030379, "F": 0.16164, "G": 0.16384, "chunks": 165.52, "bytes_per_chunk": 6.187, "boundary_entropy": 0.07691, "router_noise": 0.0, "grad_norm": 0.3755, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 8.9955, "alpha": 0.03, "t": 1420.7} |
| {"event": "train", "step": 7400, "phase": 2, "objective": "parity_a", "lm_loss": 1.04788, "bpb": 1.51177, "chunk_loss": 0.030317, "F": 0.14644, "G": 0.15238, "chunks": 149.95, "bytes_per_chunk": 6.829, "boundary_entropy": 0.0824, "router_noise": 0.0, "grad_norm": 0.3868, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.8241, "alpha": 0.03, "t": 1438.1} |
| {"event": "train", "step": 7500, "phase": 2, "objective": "parity_a", "lm_loss": 1.06774, "bpb": 1.54043, "chunk_loss": 0.03025, "F": 0.15546, "G": 0.15843, "chunks": 159.19, "bytes_per_chunk": 6.433, "boundary_entropy": 0.0733, "router_noise": 0.0, "grad_norm": 0.3727, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.3011, "alpha": 0.03, "t": 1455.4} |
| {"event": "train", "step": 7600, "phase": 2, "objective": "parity_a", "lm_loss": 1.0599, "bpb": 1.52911, "chunk_loss": 0.030392, "F": 0.15082, "G": 0.15551, "chunks": 154.44, "bytes_per_chunk": 6.631, "boundary_entropy": 0.08323, "router_noise": 0.0, "grad_norm": 0.3856, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.7149, "alpha": 0.03, "t": 1473.0} |
| {"event": "phase_transition", "step": 7700, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 295680, "t": 1490.1} |
| {"event": "train", "step": 7700, "phase": 3, "objective": "none", "lm_loss": 1.04532, "bpb": 1.50808, "chunk_loss": 0.0, "F": 0.15526, "G": 0.1588, "chunks": 158.98, "bytes_per_chunk": 6.441, "boundary_entropy": 0.07913, "router_noise": 0.0, "grad_norm": 0.3662, "lr": 9.06e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1490.2} |
| {"event": "train", "step": 7800, "phase": 3, "objective": "none", "lm_loss": 1.09837, "bpb": 1.58462, "chunk_loss": 0.0, "F": 0.16231, "G": 0.16716, "chunks": 166.2, "bytes_per_chunk": 6.161, "boundary_entropy": 0.07631, "router_noise": 0.0, "grad_norm": 0.384, "lr": 8.73e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1507.7} |
| {"event": "train", "step": 7900, "phase": 3, "objective": "none", "lm_loss": 1.08903, "bpb": 1.57114, "chunk_loss": 0.0, "F": 0.15742, "G": 0.16357, "chunks": 161.2, "bytes_per_chunk": 6.352, "boundary_entropy": 0.08458, "router_noise": 0.0, "grad_norm": 0.3713, "lr": 8.4e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1525.2} |
| {"event": "probe", "step": 8000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3360.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19307, "cps": 26.25, "mean_p": 0.187899, "boundary_entropy": 0.050577, "bpb": 1.685778}, "de": {"chunks": 4370.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.212828, "cps": 34.140625, "mean_p": 0.205205, "boundary_entropy": 0.05381, "bpb": 1.806951}, "vi": {"chunks": 3117.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.169081, "cps": 24.351562, "mean_p": 0.144032, "boundary_entropy": 0.088546, "bpb": 1.51349}, "tr": {"chunks": 3096.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.171021, "cps": 24.1875, "mean_p": 0.162705, "boundary_entropy": 0.080642, "bpb": 1.734915}, "fi": {"chunks": 3496.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185849, "cps": 27.3125, "mean_p": 0.178717, "boundary_entropy": 0.053309, "bpb": 1.898114}, "ru": {"chunks": 3736.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.19715, "cps": 29.1875, "mean_p": 0.128859, "boundary_entropy": 0.153159, "bpb": 1.123078}, "el": {"chunks": 3786.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.184476, "cps": 29.578125, "mean_p": 0.122381, "boundary_entropy": 0.130274, "bpb": 1.132319}, "hi": {"chunks": 3912.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.232511, "cps": 30.5625, "mean_p": 0.100634, "boundary_entropy": 0.084456, "bpb": 0.93752}, "ta": {"chunks": 3356.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.179513, "cps": 26.21875, "mean_p": 0.083163, "boundary_entropy": 0.107161, "bpb": 0.819777}, "th": {"chunks": 3873.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.240783, "cps": 30.257812, "mean_p": 0.09389, "boundary_entropy": 0.08364, "bpb": 0.900499}, "zh": {"chunks": 3193.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.540637, "cps": 24.945312, "mean_p": 0.20659, "boundary_entropy": 0.065096, "bpb": 2.533782}, "my": {"chunks": 2837.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.149726, "cps": 22.164062, "mean_p": 0.072981, "boundary_entropy": 0.11981, "bpb": 0.914748}}, "bpb": {"en": 1.68578, "de": 1.80695, "vi": 1.51349, "tr": 1.73492, "fi": 1.89811, "ru": 1.12308, "el": 1.13232, "hi": 0.93752, "ta": 0.81978, "th": 0.9005, "zh": 2.53378, "my": 0.91475}, "gini_cpc": 0.175271, "gini_cps": 0.066735, "bpic": 9.109015, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1023357000229597, "vi": 0.8757466872021389, "tr": 0.8857991177468612, "fi": 0.9625967329146321, "ru": 1.0211334338484734, "el": 0.9554866143212145, "hi": 1.2042831670558267, "ta": 0.9297824093531503, "th": 1.2471287024290598, "zh": 2.8002081720768226, "my": 0.7754982150396575}, "t": 1545.5} |
| {"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.68428, "de": 1.81005, "vi": 1.53, "tr": 1.72788, "fi": 1.90914, "ru": 1.12263, "el": 1.13042, "hi": 0.94236, "ta": 0.82441, "th": 0.90022, "zh": 2.53793, "my": 0.92096}, "bytes_per_chunk": {"en": 5.026, "de": 4.843, "vi": 7.453, "tr": 6.117, "fi": 5.512, "ru": 9.142, "el": 9.546, "hi": 10.992, "ta": 14.281, "th": 12.554, "zh": 5.121, "my": 17.776}, "bpb_mean": 1.42002, "t": 1550.8} |
| {"event": "train", "step": 8000, "phase": 3, "objective": "none", "lm_loss": 1.07021, "bpb": 1.54399, "chunk_loss": 0.0, "F": 0.16042, "G": 0.16549, "chunks": 164.27, "bytes_per_chunk": 6.234, "boundary_entropy": 0.07929, "router_noise": 0.0, "grad_norm": 0.3849, "lr": 8.08e-05, "router_trainable": false, "sec_per_step": 0.258, "alpha": 0.0, "t": 1550.9} |
| {"event": "train", "step": 8100, "phase": 3, "objective": "none", "lm_loss": 1.07046, "bpb": 1.54435, "chunk_loss": 0.0, "F": 0.14926, "G": 0.1582, "chunks": 152.84, "bytes_per_chunk": 6.7, "boundary_entropy": 0.08977, "router_noise": 0.0, "grad_norm": 0.3952, "lr": 7.77e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1568.4} |
| {"event": "train", "step": 8200, "phase": 3, "objective": "none", "lm_loss": 1.12139, "bpb": 1.61783, "chunk_loss": 0.0, "F": 0.16844, "G": 0.17356, "chunks": 172.48, "bytes_per_chunk": 5.937, "boundary_entropy": 0.07279, "router_noise": 0.0, "grad_norm": 0.3845, "lr": 7.47e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1585.9} |
| {"event": "train", "step": 8300, "phase": 3, "objective": "none", "lm_loss": 1.05575, "bpb": 1.52312, "chunk_loss": 0.0, "F": 0.16556, "G": 0.17012, "chunks": 169.53, "bytes_per_chunk": 6.04, "boundary_entropy": 0.07527, "router_noise": 0.0, "grad_norm": 0.3904, "lr": 7.17e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1603.3} |
| {"event": "train", "step": 8400, "phase": 3, "objective": "none", "lm_loss": 1.07689, "bpb": 1.55363, "chunk_loss": 0.0, "F": 0.15485, "G": 0.16344, "chunks": 158.56, "bytes_per_chunk": 6.458, "boundary_entropy": 0.08772, "router_noise": 0.0, "grad_norm": 0.3938, "lr": 6.88e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1620.8} |
| {"event": "train", "step": 8500, "phase": 3, "objective": "none", "lm_loss": 1.07043, "bpb": 1.5443, "chunk_loss": 0.0, "F": 0.15781, "G": 0.16488, "chunks": 161.59, "bytes_per_chunk": 6.337, "boundary_entropy": 0.07871, "router_noise": 0.0, "grad_norm": 0.3876, "lr": 6.6e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1638.2} |
| {"event": "train", "step": 8600, "phase": 3, "objective": "none", "lm_loss": 1.03958, "bpb": 1.4998, "chunk_loss": 0.0, "F": 0.153, "G": 0.1611, "chunks": 156.67, "bytes_per_chunk": 6.536, "boundary_entropy": 0.08602, "router_noise": 0.0, "grad_norm": 0.3695, "lr": 6.33e-05, "router_trainable": false, "sec_per_step": 0.171, "alpha": 0.0, "t": 1655.3} |
| {"event": "train", "step": 8700, "phase": 3, "objective": "none", "lm_loss": 1.05036, "bpb": 1.51535, "chunk_loss": 0.0, "F": 0.15526, "G": 0.16383, "chunks": 158.98, "bytes_per_chunk": 6.441, "boundary_entropy": 0.08906, "router_noise": 0.0, "grad_norm": 0.397, "lr": 6.07e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1672.7} |
| {"event": "train", "step": 8800, "phase": 3, "objective": "none", "lm_loss": 1.04925, "bpb": 1.51374, "chunk_loss": 0.0, "F": 0.1548, "G": 0.16232, "chunks": 158.52, "bytes_per_chunk": 6.46, "boundary_entropy": 0.08257, "router_noise": 0.0, "grad_norm": 0.378, "lr": 5.82e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1690.1} |
| {"event": "train", "step": 8900, "phase": 3, "objective": "none", "lm_loss": 1.03892, "bpb": 1.49885, "chunk_loss": 0.0, "F": 0.1595, "G": 0.16625, "chunks": 163.33, "bytes_per_chunk": 6.27, "boundary_entropy": 0.08356, "router_noise": 0.0, "grad_norm": 0.3944, "lr": 5.58e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1707.6} |
| {"event": "probe", "step": 9000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3372.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19376, "cps": 26.34375, "mean_p": 0.190462, "boundary_entropy": 0.047082, "bpb": 1.66691}, "de": {"chunks": 4377.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.213169, "cps": 34.195312, "mean_p": 0.206904, "boundary_entropy": 0.050761, "bpb": 1.782163}, "vi": {"chunks": 3188.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.172932, "cps": 24.90625, "mean_p": 0.148789, "boundary_entropy": 0.094344, "bpb": 1.487287}, "tr": {"chunks": 3157.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174391, "cps": 24.664062, "mean_p": 0.165932, "boundary_entropy": 0.081222, "bpb": 1.713706}, "fi": {"chunks": 3479.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184945, "cps": 27.179688, "mean_p": 0.178588, "boundary_entropy": 0.049666, "bpb": 1.867926}, "ru": {"chunks": 3838.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.202533, "cps": 29.984375, "mean_p": 0.134294, "boundary_entropy": 0.154958, "bpb": 1.102104}, "el": {"chunks": 3917.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.190859, "cps": 30.601562, "mean_p": 0.127238, "boundary_entropy": 0.134938, "bpb": 1.111739}, "hi": {"chunks": 4025.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.239227, "cps": 31.445312, "mean_p": 0.104564, "boundary_entropy": 0.08509, "bpb": 0.911831}, "ta": {"chunks": 3374.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.180476, "cps": 26.359375, "mean_p": 0.085022, "boundary_entropy": 0.109677, "bpb": 0.799286}, "th": {"chunks": 4892.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.304134, "cps": 38.21875, "mean_p": 0.116847, "boundary_entropy": 0.094268, "bpb": 0.881314}, "zh": {"chunks": 3203.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.54233, "cps": 25.023438, "mean_p": 0.206877, "boundary_entropy": 0.063627, "bpb": 2.500371}, "my": {"chunks": 3281.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.173158, "cps": 25.632812, "mean_p": 0.082883, "boundary_entropy": 0.133344, "bpb": 0.889801}}, "bpb": {"en": 1.66691, "de": 1.78216, "vi": 1.48729, "tr": 1.71371, "fi": 1.86793, "ru": 1.1021, "el": 1.11174, "hi": 0.91183, "ta": 0.79929, "th": 0.88131, "zh": 2.50037, "my": 0.8898}, "gini_cpc": 0.17783, "gini_cps": 0.076002, "bpic": 8.701925, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.100172268475727, "vi": 0.8925071932064859, "tr": 0.9000374586465781, "fi": 0.9545069601321707, "ru": 1.045279204499573, "el": 0.9850295987680171, "hi": 1.234659965205988, "ta": 0.9314427421266083, "th": 1.5696467941109122, "zh": 2.798981643614733, "my": 0.8936745875770136}, "t": 1727.8} |
| {"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.66491, "de": 1.78633, "vi": 1.50403, "tr": 1.70786, "fi": 1.88075, "ru": 1.10257, "el": 1.10968, "hi": 0.9167, "ta": 0.80422, "th": 0.88083, "zh": 2.5039, "my": 0.89691}, "bytes_per_chunk": {"en": 4.992, "de": 4.824, "vi": 7.266, "tr": 6.029, "fi": 5.49, "ru": 8.749, "el": 9.285, "hi": 10.592, "ta": 14.065, "th": 9.755, "zh": 5.139, "my": 15.552}, "bpb_mean": 1.39656, "t": 1733.0} |
| {"event": "train", "step": 9000, "phase": 3, "objective": "none", "lm_loss": 1.08166, "bpb": 1.56051, "chunk_loss": 0.0, "F": 0.16589, "G": 0.17304, "chunks": 169.88, "bytes_per_chunk": 6.028, "boundary_entropy": 0.08408, "router_noise": 0.0, "grad_norm": 0.4036, "lr": 5.35e-05, "router_trainable": false, "sec_per_step": 0.256, "alpha": 0.0, "t": 1733.2} |
| {"event": "train", "step": 9100, "phase": 3, "objective": "none", "lm_loss": 1.01682, "bpb": 1.46695, "chunk_loss": 0.0, "F": 0.16103, "G": 0.16787, "chunks": 164.89, "bytes_per_chunk": 6.21, "boundary_entropy": 0.08488, "router_noise": 0.0, "grad_norm": 0.3669, "lr": 5.12e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 1750.4} |
| {"event": "train", "step": 9200, "phase": 3, "objective": "none", "lm_loss": 1.09848, "bpb": 1.58477, "chunk_loss": 0.0, "F": 0.16602, "G": 0.17214, "chunks": 170.0, "bytes_per_chunk": 6.024, "boundary_entropy": 0.07944, "router_noise": 0.0, "grad_norm": 0.4082, "lr": 4.91e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1767.8} |
| {"event": "train", "step": 9300, "phase": 3, "objective": "none", "lm_loss": 0.96473, "bpb": 1.39182, "chunk_loss": 0.0, "F": 0.15297, "G": 0.16161, "chunks": 156.64, "bytes_per_chunk": 6.537, "boundary_entropy": 0.08776, "router_noise": 0.0, "grad_norm": 0.3874, "lr": 4.71e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1785.1} |
| {"event": "train", "step": 9400, "phase": 3, "objective": "none", "lm_loss": 1.08265, "bpb": 1.56193, "chunk_loss": 0.0, "F": 0.17516, "G": 0.18011, "chunks": 179.36, "bytes_per_chunk": 5.709, "boundary_entropy": 0.07568, "router_noise": 0.0, "grad_norm": 0.385, "lr": 4.52e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1802.5} |
| {"event": "train", "step": 9500, "phase": 3, "objective": "none", "lm_loss": 1.01423, "bpb": 1.46323, "chunk_loss": 0.0, "F": 0.15749, "G": 0.16609, "chunks": 161.27, "bytes_per_chunk": 6.35, "boundary_entropy": 0.08608, "router_noise": 0.0, "grad_norm": 0.3808, "lr": 4.34e-05, "router_trainable": false, "sec_per_step": 0.177, "alpha": 0.0, "t": 1820.2} |
| {"event": "train", "step": 9600, "phase": 3, "objective": "none", "lm_loss": 1.08258, "bpb": 1.56184, "chunk_loss": 0.0, "F": 0.16719, "G": 0.17377, "chunks": 171.2, "bytes_per_chunk": 5.981, "boundary_entropy": 0.07652, "router_noise": 0.0, "grad_norm": 0.4023, "lr": 4.17e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1837.7} |
| {"event": "train", "step": 9700, "phase": 3, "objective": "none", "lm_loss": 1.06154, "bpb": 1.53148, "chunk_loss": 0.0, "F": 0.17169, "G": 0.17709, "chunks": 175.81, "bytes_per_chunk": 5.824, "boundary_entropy": 0.07698, "router_noise": 0.0, "grad_norm": 0.3862, "lr": 4.01e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 1855.3} |
| {"event": "train", "step": 9800, "phase": 3, "objective": "none", "lm_loss": 1.0289, "bpb": 1.48439, "chunk_loss": 0.0, "F": 0.16048, "G": 0.16742, "chunks": 164.33, "bytes_per_chunk": 6.231, "boundary_entropy": 0.08023, "router_noise": 0.0, "grad_norm": 0.3949, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 1872.9} |
| {"event": "train", "step": 9900, "phase": 3, "objective": "none", "lm_loss": 1.03805, "bpb": 1.49758, "chunk_loss": 0.0, "F": 0.1666, "G": 0.17303, "chunks": 170.59, "bytes_per_chunk": 6.003, "boundary_entropy": 0.08034, "router_noise": 0.0, "grad_norm": 0.3966, "lr": 3.72e-05, "router_trainable": false, "sec_per_step": 0.171, "alpha": 0.0, "t": 1890.0} |
| {"event": "probe", "step": 10000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3380.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194219, "cps": 26.40625, "mean_p": 0.190555, "boundary_entropy": 0.047825, "bpb": 1.653821}, "de": {"chunks": 4374.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.213023, "cps": 34.171875, "mean_p": 0.206731, "boundary_entropy": 0.050219, "bpb": 1.767176}, "vi": {"chunks": 3215.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.174397, "cps": 25.117188, "mean_p": 0.149657, "boundary_entropy": 0.093897, "bpb": 1.472239}, "tr": {"chunks": 3145.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.173728, "cps": 24.570312, "mean_p": 0.165619, "boundary_entropy": 0.080798, "bpb": 1.697906}, "fi": {"chunks": 3476.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184785, "cps": 27.15625, "mean_p": 0.178754, "boundary_entropy": 0.049628, "bpb": 1.848021}, "ru": {"chunks": 3953.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.208602, "cps": 30.882812, "mean_p": 0.136616, "boundary_entropy": 0.154511, "bpb": 1.089073}, "el": {"chunks": 4049.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.197291, "cps": 31.632812, "mean_p": 0.131685, "boundary_entropy": 0.139473, "bpb": 1.095236}, "hi": {"chunks": 4411.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.262169, "cps": 34.460938, "mean_p": 0.114441, "boundary_entropy": 0.092033, "bpb": 0.89838}, "ta": {"chunks": 3811.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.203851, "cps": 29.773438, "mean_p": 0.094974, "boundary_entropy": 0.121289, "bpb": 0.78772}, "th": {"chunks": 6104.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.379484, "cps": 47.6875, "mean_p": 0.142976, "boundary_entropy": 0.105936, "bpb": 0.868863}, "zh": {"chunks": 3202.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.542161, "cps": 25.015625, "mean_p": 0.206524, "boundary_entropy": 0.064031, "bpb": 2.489456}, "my": {"chunks": 3639.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.192052, "cps": 28.429688, "mean_p": 0.090182, "boundary_entropy": 0.137188, "bpb": 0.876633}}, "bpb": {"en": 1.65382, "de": 1.76718, "vi": 1.47224, "tr": 1.69791, "fi": 1.84802, "ru": 1.08907, "el": 1.09524, "hi": 0.89838, "ta": 0.78772, "th": 0.86886, "zh": 2.48946, "my": 0.87663}, "gini_cpc": 0.187837, "gini_cps": 0.100678, "bpic": 8.207639, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0968160360706694, "vi": 0.8979357345414803, "tr": 0.8944941781201391, "fi": 0.9514266328495319, "ru": 1.0740512872554684, "el": 1.0158143684658028, "hi": 1.3498621029216524, "ta": 1.049592936123477, "th": 1.9538934616948056, "zh": 2.7914850407869243, "my": 0.9888401673593129}, "t": 1910.4} |
| {"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.6514, "de": 1.77187, "vi": 1.48862, "tr": 1.69113, "fi": 1.86112, "ru": 1.08874, "el": 1.09393, "hi": 0.90261, "ta": 0.79217, "th": 0.86859, "zh": 2.49369, "my": 0.88283}, "bytes_per_chunk": {"en": 4.961, "de": 4.828, "vi": 7.212, "tr": 6.028, "fi": 5.467, "ru": 8.549, "el": 8.96, "hi": 9.626, "ta": 12.716, "th": 7.753, "zh": 5.141, "my": 14.458}, "bpb_mean": 1.38223, "t": 1915.7} |
| {"event": "train", "step": 10000, "phase": 3, "objective": "none", "lm_loss": 1.05098, "bpb": 1.51624, "chunk_loss": 0.0, "F": 0.17003, "G": 0.17569, "chunks": 174.11, "bytes_per_chunk": 5.881, "boundary_entropy": 0.07632, "router_noise": 0.0, "grad_norm": 0.3945, "lr": 3.6e-05, "router_trainable": false, "sec_per_step": 0.258, "alpha": 0.0, "t": 1915.9} |
| {"event": "train", "step": 10100, "phase": 3, "objective": "none", "lm_loss": 1.05671, "bpb": 1.52452, "chunk_loss": 0.0, "F": 0.1662, "G": 0.17252, "chunks": 170.19, "bytes_per_chunk": 6.017, "boundary_entropy": 0.07814, "router_noise": 0.0, "grad_norm": 0.4218, "lr": 3.49e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1933.1} |
| {"event": "train", "step": 10200, "phase": 3, "objective": "none", "lm_loss": 1.02032, "bpb": 1.47201, "chunk_loss": 0.0, "F": 0.1584, "G": 0.16681, "chunks": 162.2, "bytes_per_chunk": 6.313, "boundary_entropy": 0.08659, "router_noise": 0.0, "grad_norm": 0.4037, "lr": 3.38e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1950.4} |
| {"event": "train", "step": 10300, "phase": 3, "objective": "none", "lm_loss": 0.96747, "bpb": 1.39576, "chunk_loss": 0.0, "F": 0.15511, "G": 0.1633, "chunks": 158.83, "bytes_per_chunk": 6.447, "boundary_entropy": 0.08819, "router_noise": 0.0, "grad_norm": 0.3908, "lr": 3.3e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1967.7} |
| {"event": "train", "step": 10400, "phase": 3, "objective": "none", "lm_loss": 1.06432, "bpb": 1.53549, "chunk_loss": 0.0, "F": 0.16655, "G": 0.17378, "chunks": 170.55, "bytes_per_chunk": 6.004, "boundary_entropy": 0.08004, "router_noise": 0.0, "grad_norm": 0.3937, "lr": 3.22e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1985.1} |
| {"event": "train", "step": 10500, "phase": 3, "objective": "none", "lm_loss": 1.00471, "bpb": 1.4495, "chunk_loss": 0.0, "F": 0.15559, "G": 0.16479, "chunks": 159.33, "bytes_per_chunk": 6.427, "boundary_entropy": 0.0861, "router_noise": 0.0, "grad_norm": 0.4004, "lr": 3.15e-05, "router_trainable": false, "sec_per_step": 0.171, "alpha": 0.0, "t": 2002.3} |
| {"event": "train", "step": 10600, "phase": 3, "objective": "none", "lm_loss": 0.96531, "bpb": 1.39265, "chunk_loss": 0.0, "F": 0.15308, "G": 0.16326, "chunks": 156.75, "bytes_per_chunk": 6.533, "boundary_entropy": 0.0948, "router_noise": 0.0, "grad_norm": 0.398, "lr": 3.1e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 2019.8} |
| {"event": "train", "step": 10700, "phase": 3, "objective": "none", "lm_loss": 1.03004, "bpb": 1.48603, "chunk_loss": 0.0, "F": 0.16319, "G": 0.1718, "chunks": 167.11, "bytes_per_chunk": 6.128, "boundary_entropy": 0.08644, "router_noise": 0.0, "grad_norm": 0.4017, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 2037.1} |
| {"event": "train", "step": 10800, "phase": 3, "objective": "none", "lm_loss": 1.00725, "bpb": 1.45316, "chunk_loss": 0.0, "F": 0.16974, "G": 0.1764, "chunks": 173.81, "bytes_per_chunk": 5.891, "boundary_entropy": 0.07958, "router_noise": 0.0, "grad_norm": 0.4151, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 2054.6} |
| {"event": "train", "step": 10900, "phase": 3, "objective": "none", "lm_loss": 1.02273, "bpb": 1.47548, "chunk_loss": 0.0, "F": 0.1582, "G": 0.16542, "chunks": 162.0, "bytes_per_chunk": 6.321, "boundary_entropy": 0.08081, "router_noise": 0.0, "grad_norm": 0.4081, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 2071.9} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3388.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194679, "cps": 26.46875, "mean_p": 0.191613, "boundary_entropy": 0.047106, "bpb": 1.64487}, "de": {"chunks": 4387.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.213656, "cps": 34.273438, "mean_p": 0.208483, "boundary_entropy": 0.048672, "bpb": 1.759167}, "vi": {"chunks": 3215.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.174397, "cps": 25.117188, "mean_p": 0.149085, "boundary_entropy": 0.09329, "bpb": 1.463567}, "tr": {"chunks": 3233.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178589, "cps": 25.257812, "mean_p": 0.1683, "boundary_entropy": 0.082034, "bpb": 1.689242}, "fi": {"chunks": 3486.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185317, "cps": 27.234375, "mean_p": 0.179163, "boundary_entropy": 0.049773, "bpb": 1.836065}, "ru": {"chunks": 3970.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.209499, "cps": 31.015625, "mean_p": 0.137845, "boundary_entropy": 0.156641, "bpb": 1.081375}, "el": {"chunks": 3963.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.1931, "cps": 30.960938, "mean_p": 0.130375, "boundary_entropy": 0.139226, "bpb": 1.088289}, "hi": {"chunks": 4471.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.265736, "cps": 34.929688, "mean_p": 0.115192, "boundary_entropy": 0.091292, "bpb": 0.889632}, "ta": {"chunks": 4201.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.224712, "cps": 32.820312, "mean_p": 0.101104, "boundary_entropy": 0.119039, "bpb": 0.780226}, "th": {"chunks": 5720.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.355611, "cps": 44.6875, "mean_p": 0.135502, "boundary_entropy": 0.103604, "bpb": 0.858499}, "zh": {"chunks": 3206.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.542838, "cps": 25.046875, "mean_p": 0.207056, "boundary_entropy": 0.062882, "bpb": 2.4684}, "my": {"chunks": 3816.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.201393, "cps": 29.8125, "mean_p": 0.094181, "boundary_entropy": 0.139146, "bpb": 0.866496}}, "bpb": {"en": 1.64487, "de": 1.75917, "vi": 1.46357, "tr": 1.68924, "fi": 1.83607, "ru": 1.08138, "el": 1.08829, "hi": 0.88963, "ta": 0.78023, "th": 0.8585, "zh": 2.4684, "my": 0.8665}, "gini_cpc": 0.182344, "gini_cps": 0.093329, "bpic": 8.155836, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0974783098891518, "vi": 0.8958154612603906, "tr": 0.9173517034004749, "fi": 0.9519107204286572, "ru": 1.0761232411148458, "el": 0.991890990964093, "hi": 1.3649926582940226, "ta": 1.154271388122199, "th": 1.826651433738782, "zh": 2.7883725193317903, "my": 1.0344885985107068}, "t": 2092.2} |
| {"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.64305, "de": 1.76342, "vi": 1.48024, "tr": 1.68206, "fi": 1.84844, "ru": 1.08178, "el": 1.08633, "hi": 0.8941, "ta": 0.78507, "th": 0.85851, "zh": 2.4722, "my": 0.87265}, "bytes_per_chunk": {"en": 4.944, "de": 4.816, "vi": 7.252, "tr": 5.916, "fi": 5.471, "ru": 8.514, "el": 9.117, "hi": 9.486, "ta": 11.5, "th": 8.228, "zh": 5.134, "my": 13.432}, "bpb_mean": 1.37232, "t": 2097.5} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3388.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194679, "cps": 26.46875, "mean_p": 0.191613, "boundary_entropy": 0.047106, "bpb": 1.64487}, "de": {"chunks": 4387.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.213656, "cps": 34.273438, "mean_p": 0.208483, "boundary_entropy": 0.048672, "bpb": 1.759167}, "vi": {"chunks": 3215.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.174397, "cps": 25.117188, "mean_p": 0.149085, "boundary_entropy": 0.09329, "bpb": 1.463567}, "tr": {"chunks": 3233.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178589, "cps": 25.257812, "mean_p": 0.1683, "boundary_entropy": 0.082034, "bpb": 1.689242}, "fi": {"chunks": 3486.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185317, "cps": 27.234375, "mean_p": 0.179163, "boundary_entropy": 0.049773, "bpb": 1.836065}, "ru": {"chunks": 3970.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.209499, "cps": 31.015625, "mean_p": 0.137845, "boundary_entropy": 0.156641, "bpb": 1.081375}, "el": {"chunks": 3963.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.1931, "cps": 30.960938, "mean_p": 0.130375, "boundary_entropy": 0.139226, "bpb": 1.088289}, "hi": {"chunks": 4471.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.265736, "cps": 34.929688, "mean_p": 0.115192, "boundary_entropy": 0.091292, "bpb": 0.889632}, "ta": {"chunks": 4201.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.224712, "cps": 32.820312, "mean_p": 0.101104, "boundary_entropy": 0.119039, "bpb": 0.780226}, "th": {"chunks": 5720.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.355611, "cps": 44.6875, "mean_p": 0.135502, "boundary_entropy": 0.103604, "bpb": 0.858499}, "zh": {"chunks": 3206.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.542838, "cps": 25.046875, "mean_p": 0.207056, "boundary_entropy": 0.062882, "bpb": 2.4684}, "my": {"chunks": 3816.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.201393, "cps": 29.8125, "mean_p": 0.094181, "boundary_entropy": 0.139146, "bpb": 0.866496}}, "bpb": {"en": 1.64487, "de": 1.75917, "vi": 1.46357, "tr": 1.68924, "fi": 1.83607, "ru": 1.08138, "el": 1.08829, "hi": 0.88963, "ta": 0.78023, "th": 0.8585, "zh": 2.4684, "my": 0.8665}, "gini_cpc": 0.182344, "gini_cps": 0.093329, "bpic": 8.155836, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0974783098891518, "vi": 0.8958154612603906, "tr": 0.9173517034004749, "fi": 0.9519107204286572, "ru": 1.0761232411148458, "el": 0.991890990964093, "hi": 1.3649926582940226, "ta": 1.154271388122199, "th": 1.826651433738782, "zh": 2.7883725193317903, "my": 1.0344885985107068}, "t": 2100.5} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/spec_B_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_B_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_B_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_B_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_B_s2/th0.70"], "t": 2103.3} |
| {"event": "final", "run": "spec_B_s2", "condition": "B", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 0.19467907831982992, "cps": 26.46875, "bpb": 1.6383322392960211, "bytes_per_chunk": 4.946276527998566}, "de": {"cpc": 0.21365606584522476, "cps": 34.2734375, "bpb": 1.7616417039260976, "bytes_per_chunk": 4.798513643997602}, "vi": {"cpc": 0.17439652834282615, "cps": 25.1171875, "bpb": 1.4789427009496292, "bytes_per_chunk": 7.260651299344269}, "tr": {"cpc": 0.17858918411313043, "cps": 25.2578125, "bpb": 1.6804022376402012, "bytes_per_chunk": 5.933275994748993}, "fi": {"cpc": 0.18531710169581628, "cps": 27.234375, "bpb": 1.8550242954968859, "bytes_per_chunk": 5.4859630214818615}, "ru": {"cpc": 0.20949868073878627, "cps": 31.015625, "bpb": 1.0840031979863498, "bytes_per_chunk": 8.534861384688014}, "el": {"cpc": 0.19310042391463236, "cps": 30.9609375, "bpb": 1.0835786695616108, "bytes_per_chunk": 9.103881784005765}, "hi": {"cpc": 0.2657355126300149, "cps": 34.9296875, "bpb": 0.8959033864784154, "bytes_per_chunk": 9.481567216138746}, "ta": {"cpc": 0.22471248997058038, "cps": 32.8203125, "bpb": 0.7841394826638924, "bytes_per_chunk": 11.439718091663849}, "th": {"cpc": 0.35561081753186197, "cps": 44.6875, "bpb": 0.8609790361758979, "bytes_per_chunk": 8.199879571776004}, "zh": {"cpc": 0.542837792075855, "cps": 25.046875, "bpb": 2.4725468140222384, "bytes_per_chunk": 5.165271791335189}, "my": {"cpc": 0.20139328689043698, "cps": 29.8125, "bpb": 0.8723457511579275, "bytes_per_chunk": 13.462612982744453}}, "gini_cpc": 0.18234436606279525, "gini_cps": 0.09332851637764938, "bpic": 8.155835600136008, "seed": 2, "mean_boundary_entropy": 0.08016446977853775, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 7700, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 295680}], "t": 2113.3} |
|
|
