| {"event": "start", "run": "spec_E_s2", "condition": "E", "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_b", "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": 2.1} |
| {"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": 2.1} |
| {"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": 8.2} |
| {"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": 9.2835, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 9.0} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10110.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.580934, "cps": 78.984375, "mean_p": 0.506452, "boundary_entropy": 0.687331, "bpb": 7.982856}, "de": {"chunks": 10910.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.53134, "cps": 85.234375, "mean_p": 0.503747, "boundary_entropy": 0.688251, "bpb": 8.080532}, "vi": {"chunks": 5697.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.309032, "cps": 44.507812, "mean_p": 0.489579, "boundary_entropy": 0.688469, "bpb": 7.968692}, "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": 8.083754}, "fi": {"chunks": 2104.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.111849, "cps": 16.4375, "mean_p": 0.479274, "boundary_entropy": 0.686729, "bpb": 8.086849}, "ru": {"chunks": 23586.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.244644, "cps": 184.265625, "mean_p": 0.507134, "boundary_entropy": 0.690198, "bpb": 8.228921}, "el": {"chunks": 1445.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.070409, "cps": 11.289062, "mean_p": 0.4804, "boundary_entropy": 0.689541, "bpb": 7.87989}, "hi": {"chunks": 7636.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.453848, "cps": 59.65625, "mean_p": 0.494253, "boundary_entropy": 0.690845, "bpb": 7.788207}, "ta": {"chunks": 357.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.019096, "cps": 2.789062, "mean_p": 0.475325, "boundary_entropy": 0.689903, "bpb": 7.595515}, "th": {"chunks": 26982.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.677463, "cps": 210.796875, "mean_p": 0.50308, "boundary_entropy": 0.691076, "bpb": 7.83949}, "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": 8.019155}, "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": 8.23623}}, "bpb": {"en": 7.98286, "de": 8.08053, "vi": 7.96869, "tr": 8.08375, "fi": 8.08685, "ru": 8.22892, "el": 7.87989, "hi": 7.78821, "ta": 7.59552, "th": 7.83949, "zh": 8.01916, "my": 8.23623}, "gini_cpc": 0.563022, "gini_cps": 0.48921, "bpic": 3.469301, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9146297174368364, "vi": 0.5319564046907934, "tr": 0.6004761576150713, "fi": 0.19253372647830433, "ru": 2.1424862554121025, "el": 0.12119925954104324, "hi": 0.7812388128800137, "ta": 0.03287121153801893, "th": 2.8875268902501454, "zh": 4.282718776157828, "my": 0.0714055772819442}, "t": 14.5} |
| {"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": 7.916026}, "de": {"chunks": 10768.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.524424, "cps": 84.125, "mean_p": 0.503442, "boundary_entropy": 0.68825, "bpb": 8.020665}, "vi": {"chunks": 5575.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.302414, "cps": 43.554688, "mean_p": 0.489306, "boundary_entropy": 0.688455, "bpb": 7.931398}, "tr": {"chunks": 6161.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.34033, "cps": 48.132812, "mean_p": 0.49526, "boundary_entropy": 0.687858, "bpb": 8.037807}, "fi": {"chunks": 2072.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.110148, "cps": 16.1875, "mean_p": 0.478979, "boundary_entropy": 0.686697, "bpb": 8.047363}, "ru": {"chunks": 23556.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.243061, "cps": 184.03125, "mean_p": 0.5071, "boundary_entropy": 0.6902, "bpb": 8.121331}, "el": {"chunks": 1431.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.069727, "cps": 11.179688, "mean_p": 0.480359, "boundary_entropy": 0.689538, "bpb": 7.855193}, "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": 7.745901}, "ta": {"chunks": 353.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.018882, "cps": 2.757812, "mean_p": 0.475301, "boundary_entropy": 0.689901, "bpb": 7.563421}, "th": {"chunks": 26704.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.66018, "cps": 208.625, "mean_p": 0.502966, "boundary_entropy": 0.691077, "bpb": 7.776861}, "zh": {"chunks": 14681.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.485777, "cps": 114.695312, "mean_p": 0.531545, "boundary_entropy": 0.685271, "bpb": 8.00628}, "my": {"chunks": 675.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.035624, "cps": 5.273438, "mean_p": 0.475493, "boundary_entropy": 0.690046, "bpb": 8.246032}}, "bpb": {"en": 7.91603, "de": 8.02066, "vi": 7.9314, "tr": 8.03781, "fi": 8.04736, "ru": 8.12133, "el": 7.85519, "hi": 7.7459, "ta": 7.56342, "th": 7.77686, "zh": 8.00628, "my": 8.24603}, "gini_cpc": 0.565743, "gini_cps": 0.49186, "bpic": 3.50293, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9154014617758697, "vi": 0.5278745104312121, "tr": 0.5940590539225254, "fi": 0.19226792066167134, "ru": 2.169807935680392, "el": 0.12171041702801723, "hi": 0.7899266417379924, "ta": 0.03295931656966084, "th": 2.8979054789485024, "zh": 4.339015064156234, "my": 0.062182668969543206}, "t": 20.3} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 9435.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.542148, "cps": 73.710938, "mean_p": 0.504568, "boundary_entropy": 0.687346, "bpb": 7.680963}, "de": {"chunks": 10208.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.497151, "cps": 79.75, "mean_p": 0.502291, "boundary_entropy": 0.688245, "bpb": 7.806156}, "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": 7.800307}, "tr": {"chunks": 5696.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.314644, "cps": 44.5, "mean_p": 0.493942, "boundary_entropy": 0.687805, "bpb": 7.8741}, "fi": {"chunks": 1985.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.105523, "cps": 15.507812, "mean_p": 0.477972, "boundary_entropy": 0.686582, "bpb": 7.910136}, "ru": {"chunks": 23396.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.234617, "cps": 182.78125, "mean_p": 0.506972, "boundary_entropy": 0.690208, "bpb": 7.802257}, "el": {"chunks": 1390.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.067729, "cps": 10.859375, "mean_p": 0.480079, "boundary_entropy": 0.689513, "bpb": 7.776559}, "hi": {"chunks": 7470.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.443982, "cps": 58.359375, "mean_p": 0.494051, "boundary_entropy": 0.690837, "bpb": 7.594347}, "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": 7.467579}, "th": {"chunks": 25532.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.587317, "cps": 199.46875, "mean_p": 0.502548, "boundary_entropy": 0.691079, "bpb": 7.561715}, "zh": {"chunks": 14609.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.473586, "cps": 114.132812, "mean_p": 0.530851, "boundary_entropy": 0.685353, "bpb": 7.968484}, "my": {"chunks": 650.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.034304, "cps": 5.078125, "mean_p": 0.475185, "boundary_entropy": 0.690014, "bpb": 8.209547}}, "bpb": {"en": 7.68096, "de": 7.80616, "vi": 7.80031, "tr": 7.8741, "fi": 7.91014, "ru": 7.80226, "el": 7.77656, "hi": 7.59435, "ta": 7.46758, "th": 7.56171, "zh": 7.96848, "my": 8.20955}, "gini_cpc": 0.570879, "gini_cps": 0.495134, "bpic": 3.621327, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9170023949194352, "vi": 0.5260895260837826, "tr": 0.5803655763473929, "fi": 0.19463943863309693, "ru": 2.2772704674253643, "el": 0.1249269596862398, "hi": 0.8189318170543588, "ta": 0.03443357962370596, "th": 2.9278308746311454, "zh": 4.562567074953283, "my": 0.06327500618377198}, "t": 26.3} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 7418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.426248, "cps": 57.953125, "mean_p": 0.499506, "boundary_entropy": 0.687322, "bpb": 7.072563}, "de": {"chunks": 8366.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.407442, "cps": 65.359375, "mean_p": 0.498483, "boundary_entropy": 0.688196, "bpb": 7.217209}, "vi": {"chunks": 4141.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.224627, "cps": 32.351562, "mean_p": 0.485405, "boundary_entropy": 0.688222, "bpb": 7.448256}, "tr": {"chunks": 4327.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.239021, "cps": 33.804688, "mean_p": 0.489651, "boundary_entropy": 0.687588, "bpb": 7.395606}, "fi": {"chunks": 1719.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.091383, "cps": 13.429688, "mean_p": 0.474883, "boundary_entropy": 0.686208, "bpb": 7.50119}, "ru": {"chunks": 22021.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.162058, "cps": 172.039062, "mean_p": 0.505836, "boundary_entropy": 0.690249, "bpb": 6.814214}, "el": {"chunks": 1238.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.060323, "cps": 9.671875, "mean_p": 0.479049, "boundary_entropy": 0.68942, "bpb": 7.513823}, "hi": {"chunks": 7246.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.430669, "cps": 56.609375, "mean_p": 0.493686, "boundary_entropy": 0.690821, "bpb": 7.038836}, "ta": {"chunks": 334.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.017866, "cps": 2.609375, "mean_p": 0.475421, "boundary_entropy": 0.689924, "bpb": 7.122755}, "th": {"chunks": 21100.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.311781, "cps": 164.84375, "mean_p": 0.500949, "boundary_entropy": 0.691079, "bpb": 6.852822}, "zh": {"chunks": 14328.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.426007, "cps": 111.9375, "mean_p": 0.528148, "boundary_entropy": 0.685652, "bpb": 7.848818}, "my": {"chunks": 592.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.031243, "cps": 4.625, "mean_p": 0.474325, "boundary_entropy": 0.689923, "bpb": 8.106453}}, "bpb": {"en": 7.07256, "de": 7.21721, "vi": 7.44826, "tr": 7.39561, "fi": 7.50119, "ru": 6.81421, "el": 7.51382, "hi": 7.03884, "ta": 7.12276, "th": 6.85282, "zh": 7.84882, "my": 8.10645}, "gini_cpc": 0.588161, "gini_cps": 0.503379, "bpic": 4.134235, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9558786120185543, "vi": 0.5269863663715971, "tr": 0.5607556201536735, "fi": 0.21438840066487602, "ru": 2.726246454641103, "el": 0.14151976239246114, "hi": 1.0103702451408305, "ta": 0.041913920809209665, "th": 3.0775043909776625, "zh": 5.691535137939337, "my": 0.07329859023572226}, "t": 33.0} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4287.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.246337, "cps": 33.492188, "mean_p": 0.491823, "boundary_entropy": 0.687181, "bpb": 6.36551}, "de": {"chunks": 6159.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.299956, "cps": 48.117188, "mean_p": 0.494687, "boundary_entropy": 0.688166, "bpb": 6.434304}, "vi": {"chunks": 2407.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.130567, "cps": 18.804688, "mean_p": 0.480057, "boundary_entropy": 0.687872, "bpb": 6.979908}, "tr": {"chunks": 2878.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.158979, "cps": 22.484375, "mean_p": 0.484705, "boundary_entropy": 0.687316, "bpb": 6.72223}, "fi": {"chunks": 1384.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.073574, "cps": 10.8125, "mean_p": 0.473379, "boundary_entropy": 0.68608, "bpb": 6.725671}, "ru": {"chunks": 10592.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.558945, "cps": 82.75, "mean_p": 0.496817, "boundary_entropy": 0.690275, "bpb": 5.685447}, "el": {"chunks": 1319.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.064269, "cps": 10.304688, "mean_p": 0.478555, "boundary_entropy": 0.689366, "bpb": 6.744521}, "hi": {"chunks": 7925.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.471025, "cps": 61.914062, "mean_p": 0.493777, "boundary_entropy": 0.690813, "bpb": 5.965986}, "ta": {"chunks": 359.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.019203, "cps": 2.804688, "mean_p": 0.478803, "boundary_entropy": 0.690277, "bpb": 6.475352}, "th": {"chunks": 4185.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.26018, "cps": 32.695312, "mean_p": 0.491329, "boundary_entropy": 0.690888, "bpb": 5.581559}, "zh": {"chunks": 12176.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.061632, "cps": 95.125, "mean_p": 0.517053, "boundary_entropy": 0.686555, "bpb": 7.484686}, "my": {"chunks": 611.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.032246, "cps": 4.773438, "mean_p": 0.472691, "boundary_entropy": 0.689738, "bpb": 7.43153}}, "bpb": {"en": 6.36551, "de": 6.4343, "vi": 6.97991, "tr": 6.72223, "fi": 6.72567, "ru": 5.68545, "el": 6.74452, "hi": 5.96599, "ta": 6.47535, "th": 5.58156, "zh": 7.48469, "my": 7.43153}, "gini_cpc": 0.60283, "gini_cps": 0.455768, "bpic": 7.070134, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2176667118697488, "vi": 0.5300338241353739, "tr": 0.6453731228584803, "fi": 0.2986722291749029, "ru": 2.2690255931136223, "el": 0.260900305165479, "hi": 1.912118637811331, "ta": 0.07795421737212455, "th": 1.0561972533511939, "zh": 8.369159282692673, "my": 0.13090265974883492}, "t": 40.8} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3710.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.213182, "cps": 28.984375, "mean_p": 0.493843, "boundary_entropy": 0.687518, "bpb": 5.831107}, "de": {"chunks": 6572.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.32007, "cps": 51.34375, "mean_p": 0.49732, "boundary_entropy": 0.688448, "bpb": 5.948489}, "vi": {"chunks": 1787.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.096935, "cps": 13.960938, "mean_p": 0.482107, "boundary_entropy": 0.688263, "bpb": 6.570415}, "tr": {"chunks": 3452.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190687, "cps": 26.96875, "mean_p": 0.490716, "boundary_entropy": 0.687922, "bpb": 6.255984}, "fi": {"chunks": 2386.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.126841, "cps": 18.640625, "mean_p": 0.488372, "boundary_entropy": 0.687784, "bpb": 5.982234}, "ru": {"chunks": 6866.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.362322, "cps": 53.640625, "mean_p": 0.494152, "boundary_entropy": 0.690273, "bpb": 5.152432}, "el": {"chunks": 12408.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.60459, "cps": 96.9375, "mean_p": 0.496844, "boundary_entropy": 0.690461, "bpb": 5.520684}, "hi": {"chunks": 10915.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.648737, "cps": 85.273438, "mean_p": 0.495536, "boundary_entropy": 0.690858, "bpb": 5.247904}, "ta": {"chunks": 26117.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.397005, "cps": 204.039062, "mean_p": 0.501254, "boundary_entropy": 0.691301, "bpb": 5.000775}, "th": {"chunks": 5504.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.342182, "cps": 43.0, "mean_p": 0.494295, "boundary_entropy": 0.691023, "bpb": 4.853771}, "zh": {"chunks": 3265.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.552828, "cps": 25.507812, "mean_p": 0.489506, "boundary_entropy": 0.686625, "bpb": 7.01763}, "my": {"chunks": 7746.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.408803, "cps": 60.515625, "mean_p": 0.493606, "boundary_entropy": 0.691255, "bpb": 4.995568}}, "bpb": {"en": 5.83111, "de": 5.94849, "vi": 6.57042, "tr": 6.25598, "fi": 5.98223, "ru": 5.15243, "el": 5.52068, "hi": 5.2479, "ta": 5.00077, "th": 4.85377, "zh": 7.01763, "my": 4.99557}, "gini_cpc": 0.378022, "gini_cps": 0.407247, "bpic": 4.230017, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.501396358475207, "vi": 0.45470698023287187, "tr": 0.8944796122457701, "fi": 0.5949887668237931, "ru": 1.6995924585197248, "el": 2.836032152736927, "hi": 3.0431185943776713, "ta": 6.553118632264707, "th": 1.6051202380882559, "zh": 2.5932235298198276, "my": 1.9176278445705002}, "t": 51.7} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4933.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.283457, "cps": 38.539062, "mean_p": 0.49783, "boundary_entropy": 0.687753, "bpb": 5.414203}, "de": {"chunks": 4829.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.235182, "cps": 37.726562, "mean_p": 0.495825, "boundary_entropy": 0.688514, "bpb": 5.569849}, "vi": {"chunks": 2527.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.137076, "cps": 19.742188, "mean_p": 0.488543, "boundary_entropy": 0.688826, "bpb": 5.967136}, "tr": {"chunks": 2691.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.148649, "cps": 21.023438, "mean_p": 0.49137, "boundary_entropy": 0.688081, "bpb": 5.888769}, "fi": {"chunks": 3449.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18335, "cps": 26.945312, "mean_p": 0.493438, "boundary_entropy": 0.688141, "bpb": 5.570892}, "ru": {"chunks": 5111.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.26971, "cps": 39.929688, "mean_p": 0.495181, "boundary_entropy": 0.690401, "bpb": 4.84416}, "el": {"chunks": 4764.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.23213, "cps": 37.21875, "mean_p": 0.491423, "boundary_entropy": 0.690379, "bpb": 5.089984}, "hi": {"chunks": 1503.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.089331, "cps": 11.742188, "mean_p": 0.484369, "boundary_entropy": 0.690332, "bpb": 4.902951}, "ta": {"chunks": 892.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.047713, "cps": 6.96875, "mean_p": 0.489, "boundary_entropy": 0.691032, "bpb": 4.56477}, "th": {"chunks": 5801.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.360647, "cps": 45.320312, "mean_p": 0.491505, "boundary_entropy": 0.690861, "bpb": 4.410352}, "zh": {"chunks": 5937.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.005249, "cps": 46.382812, "mean_p": 0.499985, "boundary_entropy": 0.687294, "bpb": 6.673303}, "my": {"chunks": 12658.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.668039, "cps": 98.890625, "mean_p": 0.496647, "boundary_entropy": 0.691368, "bpb": 4.281943}}, "bpb": {"en": 5.4142, "de": 5.56985, "vi": 5.96714, "tr": 5.88877, "fi": 5.57089, "ru": 4.84416, "el": 5.08998, "hi": 4.90295, "ta": 4.56477, "th": 4.41035, "zh": 6.6733, "my": 4.28194}, "gini_cpc": 0.415295, "gini_cps": 0.315919, "bpic": 6.965805, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8296937201326391, "vi": 0.4835875425576607, "tr": 0.5244162626009496, "fi": 0.6468361978964242, "ru": 0.9515019252709258, "el": 0.8189245908327559, "hi": 0.3151497104288099, "ta": 0.1683264598250957, "th": 1.272315461753868, "zh": 3.546390958175824, "my": 2.3567565350362365}, "t": 67.6} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.23144, "bpb": 4.66198, "chunk_loss": 0.030793, "F": 0.17912, "G": 0.47705, "chunks": 183.42, "bytes_per_chunk": 5.583, "boundary_entropy": 0.68964, "router_noise": 0.0, "grad_norm": 2.1335, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.698, "alpha": 0.03, "t": 78.9} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4377.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.251508, "cps": 34.195312, "mean_p": 0.448472, "boundary_entropy": 0.672038, "bpb": 4.336462}, "de": {"chunks": 6032.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.293771, "cps": 47.125, "mean_p": 0.451358, "boundary_entropy": 0.673027, "bpb": 4.554649}, "vi": {"chunks": 5146.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.279143, "cps": 40.203125, "mean_p": 0.469042, "boundary_entropy": 0.683535, "bpb": 4.735316}, "tr": {"chunks": 3993.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.220571, "cps": 31.195312, "mean_p": 0.453445, "boundary_entropy": 0.677431, "bpb": 4.783187}, "fi": {"chunks": 4484.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.238371, "cps": 35.03125, "mean_p": 0.460013, "boundary_entropy": 0.679038, "bpb": 4.561036}, "ru": {"chunks": 6557.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.346016, "cps": 51.226562, "mean_p": 0.466888, "boundary_entropy": 0.686226, "bpb": 3.219574}, "el": {"chunks": 12702.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.618915, "cps": 99.234375, "mean_p": 0.493955, "boundary_entropy": 0.689533, "bpb": 3.889968}, "hi": {"chunks": 15769.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.937236, "cps": 123.195312, "mean_p": 0.493193, "boundary_entropy": 0.6899, "bpb": 2.97509}, "ta": {"chunks": 18851.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.008344, "cps": 147.273438, "mean_p": 0.49317, "boundary_entropy": 0.69085, "bpb": 3.011291}, "th": {"chunks": 12916.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.802984, "cps": 100.90625, "mean_p": 0.486014, "boundary_entropy": 0.689896, "bpb": 2.793028}, "zh": {"chunks": 5541.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.938198, "cps": 43.289062, "mean_p": 0.499546, "boundary_entropy": 0.687316, "bpb": 6.227708}, "my": {"chunks": 23096.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.218915, "cps": 180.4375, "mean_p": 0.498488, "boundary_entropy": 0.691206, "bpb": 3.370278}}, "bpb": {"en": 4.33646, "de": 4.55465, "vi": 4.73532, "tr": 4.78319, "fi": 4.56104, "ru": 3.21957, "el": 3.88997, "hi": 2.97509, "ta": 3.01129, "th": 2.79303, "zh": 6.22771, "my": 3.37028}, "gini_cpc": 0.323939, "gini_cps": 0.335005, "bpic": 3.212524, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.168036728652287, "vi": 1.1098753693069383, "tr": 0.8769934148426507, "fi": 0.9477663693268005, "ru": 1.375762739144352, "el": 2.4608142715688652, "hi": 3.7264616302765945, "ta": 4.009188699004858, "th": 3.19267377321857, "zh": 3.7302872951079364, "my": 4.84641910915817}, "t": 90.0} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.54693, "bpb": 3.67444, "chunk_loss": 0.034179, "F": 0.24968, "G": 0.40196, "chunks": 255.67, "bytes_per_chunk": 4.005, "boundary_entropy": 0.62998, "router_noise": 0.0, "grad_norm": 1.748, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.27, "alpha": 0.03, "t": 105.9} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3046.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.175027, "cps": 23.796875, "mean_p": 0.312996, "boundary_entropy": 0.52721, "bpb": 3.549532}, "de": {"chunks": 3881.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189013, "cps": 30.320312, "mean_p": 0.309772, "boundary_entropy": 0.526661, "bpb": 3.775852}, "vi": {"chunks": 4897.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.265636, "cps": 38.257812, "mean_p": 0.378387, "boundary_entropy": 0.615461, "bpb": 3.050614}, "tr": {"chunks": 3816.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210794, "cps": 29.8125, "mean_p": 0.352985, "boundary_entropy": 0.570258, "bpb": 3.746405}, "fi": {"chunks": 3536.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.187975, "cps": 27.625, "mean_p": 0.329991, "boundary_entropy": 0.562016, "bpb": 3.814961}, "ru": {"chunks": 5620.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.29657, "cps": 43.90625, "mean_p": 0.312171, "boundary_entropy": 0.554016, "bpb": 2.614087}, "el": {"chunks": 8115.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.39541, "cps": 63.398438, "mean_p": 0.333238, "boundary_entropy": 0.587975, "bpb": 2.656704}, "hi": {"chunks": 8153.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.484577, "cps": 63.695312, "mean_p": 0.425781, "boundary_entropy": 0.642744, "bpb": 2.029981}, "ta": {"chunks": 10531.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.563306, "cps": 82.273438, "mean_p": 0.424273, "boundary_entropy": 0.63883, "bpb": 1.86111}, "th": {"chunks": 8159.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.507243, "cps": 63.742188, "mean_p": 0.379678, "boundary_entropy": 0.635154, "bpb": 2.058396}, "zh": {"chunks": 3962.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.670843, "cps": 30.953125, "mean_p": 0.45701, "boundary_entropy": 0.671822, "bpb": 4.900401}, "my": {"chunks": 10952.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.578003, "cps": 85.5625, "mean_p": 0.471361, "boundary_entropy": 0.681518, "bpb": 2.121436}}, "bpb": {"en": 3.54953, "de": 3.77585, "vi": 3.05061, "tr": 3.7464, "fi": 3.81496, "ru": 2.61409, "el": 2.6567, "hi": 2.02998, "ta": 1.86111, "th": 2.0584, "zh": 4.9004, "my": 2.12144}, "gini_cpc": 0.254586, "gini_cps": 0.239799, "bpic": 5.139832, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0799047632721908, "vi": 1.5176833975596322, "tr": 1.2043481110443055, "fi": 1.0739760438990988, "ru": 1.6944209889868107, "el": 2.25913352375115, "hi": 2.7685768976730256, "ta": 3.218387734684907, "th": 2.8980781268911397, "zh": 3.832791985803413, "my": 3.302358973676583}, "t": 122.0} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.16223, "bpb": 3.11943, "chunk_loss": 0.031613, "F": 0.21413, "G": 0.3285, "chunks": 219.27, "bytes_per_chunk": 4.67, "boundary_entropy": 0.52867, "router_noise": 0.0, "grad_norm": 1.478, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.251, "alpha": 0.03, "t": 131.0} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.05535, "bpb": 2.96524, "chunk_loss": 0.031514, "F": 0.21564, "G": 0.30283, "chunks": 220.81, "bytes_per_chunk": 4.637, "boundary_entropy": 0.45891, "router_noise": 0.0, "grad_norm": 1.3134, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.207, "alpha": 0.03, "t": 151.7} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.81134, "bpb": 2.6132, "chunk_loss": 0.030526, "F": 0.18655, "G": 0.28574, "chunks": 191.03, "bytes_per_chunk": 5.36, "boundary_entropy": 0.42143, "router_noise": 0.0, "grad_norm": 0.8467, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.206, "alpha": 0.03, "t": 172.3} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.78781, "bpb": 2.57927, "chunk_loss": 0.030529, "F": 0.19138, "G": 0.27961, "chunks": 195.97, "bytes_per_chunk": 5.225, "boundary_entropy": 0.36821, "router_noise": 0.0, "grad_norm": 0.865, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.202, "alpha": 0.03, "t": 192.5} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.68128, "bpb": 2.42557, "chunk_loss": 0.030437, "F": 0.19148, "G": 0.27835, "chunks": 196.08, "bytes_per_chunk": 5.222, "boundary_entropy": 0.36407, "router_noise": 0.0, "grad_norm": 0.7457, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.201, "alpha": 0.03, "t": 212.6} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.74308, "bpb": 2.51473, "chunk_loss": 0.030508, "F": 0.19423, "G": 0.2658, "chunks": 198.89, "bytes_per_chunk": 5.149, "boundary_entropy": 0.31529, "router_noise": 0.0, "grad_norm": 0.9015, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.19, "alpha": 0.03, "t": 231.5} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.63246, "bpb": 2.35514, "chunk_loss": 0.030579, "F": 0.1963, "G": 0.27769, "chunks": 201.02, "bytes_per_chunk": 5.094, "boundary_entropy": 0.34085, "router_noise": 0.0, "grad_norm": 0.6555, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.197, "alpha": 0.03, "t": 251.2} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3363.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193243, "cps": 26.273438, "mean_p": 0.209836, "boundary_entropy": 0.178749, "bpb": 2.315797}, "de": {"chunks": 4815.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.234501, "cps": 37.617188, "mean_p": 0.24203, "boundary_entropy": 0.181261, "bpb": 2.595028}, "vi": {"chunks": 3915.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.212368, "cps": 30.585938, "mean_p": 0.261398, "boundary_entropy": 0.377007, "bpb": 2.118153}, "tr": {"chunks": 3976.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.219632, "cps": 31.0625, "mean_p": 0.221327, "boundary_entropy": 0.245398, "bpb": 2.595463}, "fi": {"chunks": 4045.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.215034, "cps": 31.601562, "mean_p": 0.231294, "boundary_entropy": 0.230249, "bpb": 2.866852}, "ru": {"chunks": 5552.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.292982, "cps": 43.375, "mean_p": 0.296207, "boundary_entropy": 0.482747, "bpb": 1.724074}, "el": {"chunks": 6093.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.296886, "cps": 47.601562, "mean_p": 0.273122, "boundary_entropy": 0.462816, "bpb": 1.782578}, "hi": {"chunks": 7237.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.430134, "cps": 56.539062, "mean_p": 0.394092, "boundary_entropy": 0.586772, "bpb": 1.385742}, "ta": {"chunks": 7065.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.377909, "cps": 55.195312, "mean_p": 0.383697, "boundary_entropy": 0.601897, "bpb": 1.14451}, "th": {"chunks": 6969.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.433261, "cps": 54.445312, "mean_p": 0.383518, "boundary_entropy": 0.58787, "bpb": 1.372339}, "zh": {"chunks": 3167.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.536234, "cps": 24.742188, "mean_p": 0.245182, "boundary_entropy": 0.236643, "bpb": 3.456289}, "my": {"chunks": 10097.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.532879, "cps": 78.882812, "mean_p": 0.438952, "boundary_entropy": 0.635048, "bpb": 1.328581}}, "bpb": {"en": 2.3158, "de": 2.59503, "vi": 2.11815, "tr": 2.59546, "fi": 2.86685, "ru": 1.72407, "el": 1.78258, "hi": 1.38574, "ta": 1.14451, "th": 1.37234, "zh": 3.45629, "my": 1.32858}, "gini_cpc": 0.202759, "gini_cps": 0.194829, "bpic": 5.789076, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2135037903563173, "vi": 1.0989701018139535, "tr": 1.1365618573823504, "fi": 1.1127661226176286, "ru": 1.5161336819980276, "el": 1.5363408769682614, "hi": 2.22587490207878, "ta": 1.9556176559776057, "th": 2.242057013684826, "zh": 2.774928987077657, "my": 2.7575680151527906}, "t": 273.1} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.3264, "de": 2.58443, "vi": 2.1184, "tr": 2.58753, "fi": 2.87418, "ru": 1.72201, "el": 1.77581, "hi": 1.39281, "ta": 1.14363, "th": 1.36768, "zh": 3.46157, "my": 1.33874}, "bytes_per_chunk": {"en": 5.001, "de": 4.364, "vi": 5.945, "tr": 4.833, "fi": 5.036, "ru": 6.197, "el": 6.044, "hi": 5.719, "ta": 6.935, "th": 6.304, "zh": 5.109, "my": 4.435}, "bpb_mean": 2.05777, "t": 279.2} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.56402, "bpb": 2.25641, "chunk_loss": 0.030297, "F": 0.18971, "G": 0.26403, "chunks": 194.27, "bytes_per_chunk": 5.271, "boundary_entropy": 0.3063, "router_noise": 0.0, "grad_norm": 0.6814, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.282, "alpha": 0.03, "t": 279.4} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.52593, "bpb": 2.20146, "chunk_loss": 0.030215, "F": 0.18181, "G": 0.25424, "chunks": 186.17, "bytes_per_chunk": 5.5, "boundary_entropy": 0.3041, "router_noise": 0.0, "grad_norm": 0.5595, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 298.7} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.52247, "bpb": 2.19646, "chunk_loss": 0.03039, "F": 0.19095, "G": 0.26266, "chunks": 195.53, "bytes_per_chunk": 5.237, "boundary_entropy": 0.30523, "router_noise": 0.0, "grad_norm": 0.5842, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 318.1} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.56811, "bpb": 2.2623, "chunk_loss": 0.030666, "F": 0.19659, "G": 0.25478, "chunks": 201.31, "bytes_per_chunk": 5.087, "boundary_entropy": 0.26305, "router_noise": 0.0, "grad_norm": 0.5577, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.198, "alpha": 0.03, "t": 337.9} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.39576, "bpb": 2.01366, "chunk_loss": 0.030143, "F": 0.1814, "G": 0.25422, "chunks": 185.75, "bytes_per_chunk": 5.513, "boundary_entropy": 0.29885, "router_noise": 0.0, "grad_norm": 0.5596, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.189, "alpha": 0.03, "t": 356.8} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.51241, "bpb": 2.18194, "chunk_loss": 0.030369, "F": 0.18718, "G": 0.25233, "chunks": 191.67, "bytes_per_chunk": 5.342, "boundary_entropy": 0.27493, "router_noise": 0.0, "grad_norm": 0.5314, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 376.1} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.53963, "bpb": 2.22122, "chunk_loss": 0.030372, "F": 0.19165, "G": 0.24795, "chunks": 196.25, "bytes_per_chunk": 5.218, "boundary_entropy": 0.2548, "router_noise": 0.0, "grad_norm": 0.5959, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 395.4} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.41299, "bpb": 2.03851, "chunk_loss": 0.030408, "F": 0.18504, "G": 0.26568, "chunks": 189.48, "bytes_per_chunk": 5.404, "boundary_entropy": 0.31115, "router_noise": 0.0, "grad_norm": 0.4951, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.203, "alpha": 0.03, "t": 415.7} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.41552, "bpb": 2.04216, "chunk_loss": 0.030417, "F": 0.19228, "G": 0.26725, "chunks": 196.89, "bytes_per_chunk": 5.201, "boundary_entropy": 0.29507, "router_noise": 0.0, "grad_norm": 0.5029, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.188, "alpha": 0.03, "t": 434.5} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.41032, "bpb": 2.03466, "chunk_loss": 0.030548, "F": 0.19112, "G": 0.26195, "chunks": 195.7, "bytes_per_chunk": 5.232, "boundary_entropy": 0.28995, "router_noise": 0.0, "grad_norm": 0.5038, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.203, "alpha": 0.03, "t": 454.8} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3306.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189967, "cps": 25.828125, "mean_p": 0.201805, "boundary_entropy": 0.147214, "bpb": 2.090802}, "de": {"chunks": 4516.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.219939, "cps": 35.28125, "mean_p": 0.224899, "boundary_entropy": 0.154312, "bpb": 2.287595}, "vi": {"chunks": 4281.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.232221, "cps": 33.445312, "mean_p": 0.272801, "boundary_entropy": 0.367603, "bpb": 1.959059}, "tr": {"chunks": 3625.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200243, "cps": 28.320312, "mean_p": 0.203226, "boundary_entropy": 0.198945, "bpb": 2.270936}, "fi": {"chunks": 3652.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194142, "cps": 28.53125, "mean_p": 0.209848, "boundary_entropy": 0.186523, "bpb": 2.45632}, "ru": {"chunks": 5541.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.292401, "cps": 43.289062, "mean_p": 0.276467, "boundary_entropy": 0.420181, "bpb": 1.448342}, "el": {"chunks": 6692.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326073, "cps": 52.28125, "mean_p": 0.272032, "boundary_entropy": 0.423105, "bpb": 1.508986}, "hi": {"chunks": 7098.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.421872, "cps": 55.453125, "mean_p": 0.381814, "boundary_entropy": 0.565524, "bpb": 1.202167}, "ta": {"chunks": 7378.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.394651, "cps": 57.640625, "mean_p": 0.371976, "boundary_entropy": 0.557383, "bpb": 1.006666}, "th": {"chunks": 7651.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.475661, "cps": 59.773438, "mean_p": 0.394301, "boundary_entropy": 0.576261, "bpb": 1.132736}, "zh": {"chunks": 3057.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.517609, "cps": 23.882812, "mean_p": 0.234378, "boundary_entropy": 0.210178, "bpb": 3.166491}, "my": {"chunks": 8465.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.446749, "cps": 66.132812, "mean_p": 0.420809, "boundary_entropy": 0.629541, "bpb": 1.165992}}, "bpb": {"en": 2.0908, "de": 2.28759, "vi": 1.95906, "tr": 2.27094, "fi": 2.45632, "ru": 1.44834, "el": 1.50899, "hi": 1.20217, "ta": 1.00667, "th": 1.13274, "zh": 3.16649, "my": 1.16599}, "gini_cpc": 0.200133, "gini_cps": 0.191985, "bpic": 5.88062, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1577713464306891, "vi": 1.2224281910691266, "tr": 1.0540925173786257, "fi": 1.021974727881059, "ru": 1.5392182599160078, "el": 1.7164705562958755, "hi": 2.2207628976627665, "ta": 2.0774685235189585, "th": 2.5039082302671773, "zh": 2.724728704943694, "my": 2.351715910235942}, "t": 477.1} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.09779, "de": 2.27727, "vi": 1.96174, "tr": 2.26391, "fi": 2.47191, "ru": 1.44831, "el": 1.50559, "hi": 1.20976, "ta": 1.00704, "th": 1.12948, "zh": 3.17435, "my": 1.17569}, "bytes_per_chunk": {"en": 5.093, "de": 4.656, "vi": 5.54, "tr": 5.357, "fi": 5.306, "ru": 6.186, "el": 5.756, "hi": 6.135, "ta": 6.901, "th": 5.926, "zh": 5.297, "my": 6.186}, "bpb_mean": 1.81024, "t": 483.0} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.41234, "bpb": 2.03758, "chunk_loss": 0.030415, "F": 0.18927, "G": 0.25419, "chunks": 193.81, "bytes_per_chunk": 5.283, "boundary_entropy": 0.26142, "router_noise": 0.0, "grad_norm": 0.5192, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.285, "alpha": 0.03, "t": 483.3} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.46022, "bpb": 2.10665, "chunk_loss": 0.030473, "F": 0.19115, "G": 0.2543, "chunks": 195.73, "bytes_per_chunk": 5.232, "boundary_entropy": 0.26832, "router_noise": 0.0, "grad_norm": 0.467, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.191, "alpha": 0.03, "t": 502.3} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.41555, "bpb": 2.04221, "chunk_loss": 0.030332, "F": 0.18814, "G": 0.24416, "chunks": 192.66, "bytes_per_chunk": 5.315, "boundary_entropy": 0.25246, "router_noise": 0.0, "grad_norm": 0.4557, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.196, "alpha": 0.03, "t": 522.0} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.34954, "bpb": 1.94697, "chunk_loss": 0.03025, "F": 0.18182, "G": 0.24925, "chunks": 186.19, "bytes_per_chunk": 5.5, "boundary_entropy": 0.2751, "router_noise": 0.0, "grad_norm": 0.4211, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.19, "alpha": 0.03, "t": 540.9} |
| {"event": "train", "step": 2400, "phase": 1, "objective": "global", "lm_loss": 1.38107, "bpb": 1.99246, "chunk_loss": 0.030297, "F": 0.18404, "G": 0.25403, "chunks": 188.45, "bytes_per_chunk": 5.434, "boundary_entropy": 0.27891, "router_noise": 0.0, "grad_norm": 0.4397, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.189, "alpha": 0.03, "t": 559.8} |
| {"event": "train", "step": 2500, "phase": 1, "objective": "global", "lm_loss": 1.23631, "bpb": 1.78362, "chunk_loss": 0.030117, "F": 0.17957, "G": 0.26746, "chunks": 183.88, "bytes_per_chunk": 5.569, "boundary_entropy": 0.30988, "router_noise": 0.0, "grad_norm": 0.4227, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.191, "alpha": 0.03, "t": 579.0} |
| {"event": "train", "step": 2600, "phase": 1, "objective": "global", "lm_loss": 1.22409, "bpb": 1.76599, "chunk_loss": 0.030419, "F": 0.18373, "G": 0.26132, "chunks": 188.14, "bytes_per_chunk": 5.443, "boundary_entropy": 0.30215, "router_noise": 0.0, "grad_norm": 0.4065, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.188, "alpha": 0.03, "t": 597.8} |
| {"event": "train", "step": 2700, "phase": 1, "objective": "global", "lm_loss": 1.37092, "bpb": 1.97782, "chunk_loss": 0.030297, "F": 0.18874, "G": 0.24406, "chunks": 193.27, "bytes_per_chunk": 5.298, "boundary_entropy": 0.24038, "router_noise": 0.0, "grad_norm": 0.4358, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.188, "alpha": 0.03, "t": 616.6} |
| {"event": "train", "step": 2800, "phase": 1, "objective": "global", "lm_loss": 1.18895, "bpb": 1.7153, "chunk_loss": 0.030114, "F": 0.17802, "G": 0.25486, "chunks": 182.3, "bytes_per_chunk": 5.617, "boundary_entropy": 0.28603, "router_noise": 0.0, "grad_norm": 0.4256, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.174, "alpha": 0.03, "t": 634.0} |
| {"event": "train", "step": 2900, "phase": 1, "objective": "global", "lm_loss": 1.2391, "bpb": 1.78764, "chunk_loss": 0.03019, "F": 0.18074, "G": 0.26263, "chunks": 185.08, "bytes_per_chunk": 5.533, "boundary_entropy": 0.2934, "router_noise": 0.0, "grad_norm": 0.4233, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.174, "alpha": 0.03, "t": 651.4} |
| {"event": "probe", "step": 3000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3108.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.17859, "cps": 24.28125, "mean_p": 0.184586, "boundary_entropy": 0.126869, "bpb": 1.925497}, "de": {"chunks": 4185.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.203818, "cps": 32.695312, "mean_p": 0.207914, "boundary_entropy": 0.136151, "bpb": 2.107364}, "vi": {"chunks": 3747.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.203255, "cps": 29.273438, "mean_p": 0.249756, "boundary_entropy": 0.345869, "bpb": 1.827016}, "tr": {"chunks": 3191.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176269, "cps": 24.929688, "mean_p": 0.184347, "boundary_entropy": 0.182587, "bpb": 2.048925}, "fi": {"chunks": 3384.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179895, "cps": 26.4375, "mean_p": 0.190412, "boundary_entropy": 0.153789, "bpb": 2.282304}, "ru": {"chunks": 5887.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.31066, "cps": 45.992188, "mean_p": 0.286531, "boundary_entropy": 0.423053, "bpb": 1.329792}, "el": {"chunks": 6978.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.340009, "cps": 54.515625, "mean_p": 0.266419, "boundary_entropy": 0.388248, "bpb": 1.368655}, "hi": {"chunks": 8211.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.488024, "cps": 64.148438, "mean_p": 0.383945, "boundary_entropy": 0.5445, "bpb": 1.110322}, "ta": {"chunks": 8917.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.476972, "cps": 69.664062, "mean_p": 0.404319, "boundary_entropy": 0.559421, "bpb": 0.936825}, "th": {"chunks": 7595.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.472179, "cps": 59.335938, "mean_p": 0.40653, "boundary_entropy": 0.577004, "bpb": 1.054225}, "zh": {"chunks": 2983.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50508, "cps": 23.304688, "mean_p": 0.218628, "boundary_entropy": 0.169467, "bpb": 2.968094}, "my": {"chunks": 9699.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.511875, "cps": 75.773438, "mean_p": 0.41046, "boundary_entropy": 0.623576, "bpb": 1.070443}}, "bpb": {"en": 1.9255, "de": 2.10736, "vi": 1.82702, "tr": 2.04893, "fi": 2.2823, "ru": 1.32979, "el": 1.36866, "hi": 1.11032, "ta": 0.93682, "th": 1.05423, "zh": 2.96809, "my": 1.07044}, "gini_cpc": 0.22632, "gini_cps": 0.237837, "bpic": 5.653399, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1412641238409513, "vi": 1.1381084850979075, "tr": 0.9870050229521036, "fi": 1.0073063715081683, "ru": 1.739514012693432, "el": 1.9038521993499335, "hi": 2.7326504959640174, "ta": 2.670769495281397, "th": 2.6439292105285266, "zh": 2.828153131361733, "my": 2.866201330933756}, "t": 673.3} |
| {"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 1.92891, "de": 2.1044, "vi": 1.8357, "tr": 2.04322, "fi": 2.29476, "ru": 1.3295, "el": 1.36519, "hi": 1.11672, "ta": 0.93823, "th": 1.05202, "zh": 2.97663, "my": 1.08016}, "bytes_per_chunk": {"en": 5.332, "de": 4.923, "vi": 6.238, "tr": 5.929, "fi": 5.766, "ru": 5.79, "el": 5.571, "hi": 5.231, "ta": 5.728, "th": 5.982, "zh": 5.395, "my": 5.566}, "bpb_mean": 1.67212, "t": 679.0} |
| {"event": "train", "step": 3000, "phase": 1, "objective": "global", "lm_loss": 1.26825, "bpb": 1.8297, "chunk_loss": 0.03028, "F": 0.18268, "G": 0.23898, "chunks": 187.06, "bytes_per_chunk": 5.474, "boundary_entropy": 0.25324, "router_noise": 0.0, "grad_norm": 0.4247, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.278, "alpha": 0.03, "t": 679.2} |
| {"event": "train", "step": 3100, "phase": 1, "objective": "global", "lm_loss": 1.19401, "bpb": 1.7226, "chunk_loss": 0.030114, "F": 0.18268, "G": 0.25669, "chunks": 187.06, "bytes_per_chunk": 5.474, "boundary_entropy": 0.26605, "router_noise": 0.0, "grad_norm": 0.4296, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.189, "alpha": 0.03, "t": 698.1} |
| {"event": "train", "step": 3200, "phase": 1, "objective": "global", "lm_loss": 1.26327, "bpb": 1.82252, "chunk_loss": 0.030336, "F": 0.18448, "G": 0.2639, "chunks": 188.91, "bytes_per_chunk": 5.421, "boundary_entropy": 0.28231, "router_noise": 0.0, "grad_norm": 0.4244, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 716.0} |
| {"event": "train", "step": 3300, "phase": 1, "objective": "global", "lm_loss": 1.25695, "bpb": 1.8134, "chunk_loss": 0.030134, "F": 0.1776, "G": 0.23519, "chunks": 181.86, "bytes_per_chunk": 5.631, "boundary_entropy": 0.24758, "router_noise": 0.0, "grad_norm": 0.3993, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.188, "alpha": 0.03, "t": 734.8} |
| {"event": "train", "step": 3400, "phase": 1, "objective": "global", "lm_loss": 1.31682, "bpb": 1.89976, "chunk_loss": 0.030226, "F": 0.18274, "G": 0.24046, "chunks": 187.12, "bytes_per_chunk": 5.472, "boundary_entropy": 0.23561, "router_noise": 0.0, "grad_norm": 0.4242, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 752.5} |
| {"event": "train", "step": 3500, "phase": 1, "objective": "global", "lm_loss": 1.17188, "bpb": 1.69066, "chunk_loss": 0.03026, "F": 0.18518, "G": 0.25149, "chunks": 189.62, "bytes_per_chunk": 5.4, "boundary_entropy": 0.26507, "router_noise": 0.0, "grad_norm": 0.4003, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.189, "alpha": 0.03, "t": 771.4} |
| {"event": "train", "step": 3600, "phase": 1, "objective": "global", "lm_loss": 1.22407, "bpb": 1.76596, "chunk_loss": 0.030185, "F": 0.18086, "G": 0.24489, "chunks": 185.2, "bytes_per_chunk": 5.529, "boundary_entropy": 0.24943, "router_noise": 0.0, "grad_norm": 0.416, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.189, "alpha": 0.03, "t": 790.3} |
| {"event": "train", "step": 3700, "phase": 1, "objective": "global", "lm_loss": 1.20905, "bpb": 1.74429, "chunk_loss": 0.03008, "F": 0.17542, "G": 0.25516, "chunks": 179.62, "bytes_per_chunk": 5.701, "boundary_entropy": 0.28334, "router_noise": 0.0, "grad_norm": 0.398, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.176, "alpha": 0.03, "t": 807.9} |
| {"event": "train", "step": 3800, "phase": 1, "objective": "global", "lm_loss": 1.23428, "bpb": 1.78069, "chunk_loss": 0.030384, "F": 0.189, "G": 0.25345, "chunks": 193.53, "bytes_per_chunk": 5.291, "boundary_entropy": 0.24957, "router_noise": 0.0, "grad_norm": 0.4147, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.19, "alpha": 0.03, "t": 826.9} |
| {"event": "train", "step": 3900, "phase": 1, "objective": "global", "lm_loss": 1.2107, "bpb": 1.74668, "chunk_loss": 0.030198, "F": 0.18228, "G": 0.24284, "chunks": 186.66, "bytes_per_chunk": 5.486, "boundary_entropy": 0.24084, "router_noise": 0.0, "grad_norm": 0.3726, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.19, "alpha": 0.03, "t": 845.9} |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3163.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.18175, "cps": 24.710938, "mean_p": 0.185914, "boundary_entropy": 0.104468, "bpb": 1.836972}, "de": {"chunks": 4236.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.206302, "cps": 33.09375, "mean_p": 0.207257, "boundary_entropy": 0.112401, "bpb": 1.986962}, "vi": {"chunks": 3730.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.202333, "cps": 29.140625, "mean_p": 0.241286, "boundary_entropy": 0.324073, "bpb": 1.713446}, "tr": {"chunks": 3180.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175661, "cps": 24.84375, "mean_p": 0.18321, "boundary_entropy": 0.163051, "bpb": 1.926257}, "fi": {"chunks": 3520.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.187125, "cps": 27.5, "mean_p": 0.19318, "boundary_entropy": 0.13437, "bpb": 2.129474}, "ru": {"chunks": 5763.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.304116, "cps": 45.023438, "mean_p": 0.275598, "boundary_entropy": 0.395784, "bpb": 1.266186}, "el": {"chunks": 6700.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326463, "cps": 52.34375, "mean_p": 0.251966, "boundary_entropy": 0.356569, "bpb": 1.287035}, "hi": {"chunks": 7511.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.446419, "cps": 58.679688, "mean_p": 0.396037, "boundary_entropy": 0.550637, "bpb": 1.044339}, "ta": {"chunks": 9486.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.507408, "cps": 74.109375, "mean_p": 0.414349, "boundary_entropy": 0.562173, "bpb": 0.911829}, "th": {"chunks": 7926.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.492757, "cps": 61.921875, "mean_p": 0.428337, "boundary_entropy": 0.580234, "bpb": 1.021533}, "zh": {"chunks": 2981.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.504741, "cps": 23.289062, "mean_p": 0.213793, "boundary_entropy": 0.161889, "bpb": 2.843801}, "my": {"chunks": 10584.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.558581, "cps": 82.6875, "mean_p": 0.449825, "boundary_entropy": 0.639141, "bpb": 1.025222}}, "bpb": {"en": 1.83697, "de": 1.98696, "vi": 1.71345, "tr": 1.92626, "fi": 2.12947, "ru": 1.26619, "el": 1.28704, "hi": 1.04434, "ta": 0.91183, "th": 1.02153, "zh": 2.8438, "my": 1.02522}, "gini_cpc": 0.235347, "gini_cps": 0.247351, "bpic": 5.579834, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1350852299650913, "vi": 1.1132446537170657, "tr": 0.9664991954476115, "fi": 1.0295695943304641, "ru": 1.673263484367818, "el": 1.7962173451317291, "hi": 2.4562220022652332, "ta": 2.791788918693062, "th": 2.7111773713213996, "zh": 2.7771124260050946, "my": 3.0733454843316794}, "t": 868.4} |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.83786, "de": 1.9897, "vi": 1.72554, "tr": 1.91985, "fi": 2.13929, "ru": 1.26811, "el": 1.28634, "hi": 1.05122, "ta": 0.91468, "th": 1.0179, "zh": 2.85308, "my": 1.03285}, "bytes_per_chunk": {"en": 5.243, "de": 4.868, "vi": 6.298, "tr": 5.766, "fi": 5.561, "ru": 5.872, "el": 5.816, "hi": 5.734, "ta": 5.489, "th": 5.921, "zh": 5.387, "my": 5.232}, "bpb_mean": 1.58637, "t": 874.1} |
| {"event": "train", "step": 4000, "phase": 1, "objective": "global", "lm_loss": 1.14556, "bpb": 1.6527, "chunk_loss": 0.03013, "F": 0.17752, "G": 0.2509, "chunks": 181.78, "bytes_per_chunk": 5.633, "boundary_entropy": 0.26235, "router_noise": 0.0, "grad_norm": 0.3687, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.284, "alpha": 0.03, "t": 874.3} |
| {"event": "train", "step": 4100, "phase": 1, "objective": "global", "lm_loss": 1.22539, "bpb": 1.76786, "chunk_loss": 0.030125, "F": 0.17957, "G": 0.23289, "chunks": 183.88, "bytes_per_chunk": 5.569, "boundary_entropy": 0.23182, "router_noise": 0.0, "grad_norm": 0.3995, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.189, "alpha": 0.03, "t": 893.2} |
| {"event": "train", "step": 4200, "phase": 1, "objective": "global", "lm_loss": 1.1998, "bpb": 1.73095, "chunk_loss": 0.030375, "F": 0.18782, "G": 0.25687, "chunks": 192.33, "bytes_per_chunk": 5.324, "boundary_entropy": 0.26797, "router_noise": 0.0, "grad_norm": 0.3983, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 910.9} |
| {"event": "train", "step": 4300, "phase": 1, "objective": "global", "lm_loss": 1.13316, "bpb": 1.63481, "chunk_loss": 0.030064, "F": 0.17784, "G": 0.25639, "chunks": 182.11, "bytes_per_chunk": 5.623, "boundary_entropy": 0.27217, "router_noise": 0.0, "grad_norm": 0.3762, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.191, "alpha": 0.03, "t": 930.0} |
| {"event": "phase_transition", "step": 4400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 948.8} |
| {"event": "train", "step": 4400, "phase": 2, "objective": "parity_b", "lm_loss": 1.16323, "bpb": 1.67818, "chunk_loss": 0.0, "F": 0.18631, "G": 0.24726, "chunks": 190.78, "bytes_per_chunk": 5.367, "boundary_entropy": 0.24119, "router_noise": 0.0, "grad_norm": 0.3828, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.191, "worst_lang": "my", "n_target_mean": 21.145, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 949.0} |
| {"event": "train", "step": 4500, "phase": 2, "objective": "parity_b", "lm_loss": 1.28009, "bpb": 1.84677, "chunk_loss": 0.030973, "F": 0.16255, "G": 0.1831, "chunks": 166.45, "bytes_per_chunk": 6.152, "boundary_entropy": 0.15161, "router_noise": 0.0, "grad_norm": 0.4093, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.179, "worst_lang": "th", "n_target_mean": 16.5724, "frac_in_batch": 0.0156, "fallback": 0.0, "alpha": 0.03, "t": 966.9} |
| {"event": "train", "step": 4600, "phase": 2, "objective": "parity_b", "lm_loss": 1.12561, "bpb": 1.62391, "chunk_loss": 0.030134, "F": 0.14168, "G": 0.1704, "chunks": 145.08, "bytes_per_chunk": 7.058, "boundary_entropy": 0.16921, "router_noise": 0.0, "grad_norm": 0.3766, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.19, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 985.9} |
| {"event": "train", "step": 4700, "phase": 2, "objective": "parity_b", "lm_loss": 1.19364, "bpb": 1.72205, "chunk_loss": 0.030298, "F": 0.14693, "G": 0.17005, "chunks": 150.45, "bytes_per_chunk": 6.806, "boundary_entropy": 0.14951, "router_noise": 0.0, "grad_norm": 0.3737, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.186, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1004.6} |
| {"event": "train", "step": 4800, "phase": 2, "objective": "parity_b", "lm_loss": 1.16995, "bpb": 1.68789, "chunk_loss": 0.030142, "F": 0.14658, "G": 0.17138, "chunks": 150.09, "bytes_per_chunk": 6.822, "boundary_entropy": 0.14274, "router_noise": 0.0, "grad_norm": 0.3915, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.188, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1023.4} |
| {"event": "train", "step": 4900, "phase": 2, "objective": "parity_b", "lm_loss": 1.23847, "bpb": 1.78674, "chunk_loss": 0.030123, "F": 0.15953, "G": 0.177, "chunks": 163.36, "bytes_per_chunk": 6.268, "boundary_entropy": 0.13109, "router_noise": 0.0, "grad_norm": 0.3962, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.175, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1040.9} |
| {"event": "probe", "step": 5000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3165.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.181865, "cps": 24.726562, "mean_p": 0.1841, "boundary_entropy": 0.077882, "bpb": 1.785878}, "de": {"chunks": 3258.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.158671, "cps": 25.453125, "mean_p": 0.153658, "boundary_entropy": 0.091304, "bpb": 1.953561}, "vi": {"chunks": 3054.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.165663, "cps": 23.859375, "mean_p": 0.166963, "boundary_entropy": 0.193616, "bpb": 1.644183}, "tr": {"chunks": 3035.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167652, "cps": 23.710938, "mean_p": 0.167909, "boundary_entropy": 0.132227, "bpb": 1.863283}, "fi": {"chunks": 3338.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177449, "cps": 26.078125, "mean_p": 0.176297, "boundary_entropy": 0.09923, "bpb": 2.054639}, "ru": {"chunks": 3238.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.170871, "cps": 25.296875, "mean_p": 0.12956, "boundary_entropy": 0.196285, "bpb": 1.23483}, "el": {"chunks": 3348.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.163134, "cps": 26.15625, "mean_p": 0.135985, "boundary_entropy": 0.226336, "bpb": 1.262262}, "hi": {"chunks": 3389.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.201426, "cps": 26.476562, "mean_p": 0.106264, "boundary_entropy": 0.163004, "bpb": 1.066796}, "ta": {"chunks": 2764.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.147847, "cps": 21.59375, "mean_p": 0.098029, "boundary_entropy": 0.194034, "bpb": 0.909638}, "th": {"chunks": 3452.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.21461, "cps": 26.96875, "mean_p": 0.113371, "boundary_entropy": 0.188128, "bpb": 1.016376}, "zh": {"chunks": 3029.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.512868, "cps": 23.664062, "mean_p": 0.228861, "boundary_entropy": 0.178252, "bpb": 2.705173}, "my": {"chunks": 1020.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.053832, "cps": 7.96875, "mean_p": 0.11337, "boundary_entropy": 0.280499, "bpb": 1.022032}}, "bpb": {"en": 1.78588, "de": 1.95356, "vi": 1.64418, "tr": 1.86328, "fi": 2.05464, "ru": 1.23483, "el": 1.26226, "hi": 1.0668, "ta": 0.90964, "th": 1.01638, "zh": 2.70517, "my": 1.02203}, "gini_cpc": 0.218843, "gini_cps": 0.084996, "bpic": 10.633998, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8724671393616057, "vi": 0.910911734196106, "tr": 0.9218463698902515, "fi": 0.9757192083587274, "ru": 0.9395459740649503, "el": 0.8970053018772373, "hi": 1.1075590797377521, "ta": 0.8129483895255467, "th": 1.18004923488517, "zh": 2.820046285968322, "my": 0.295996926484757}, "t": 1062.8} |
| {"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.78485, "de": 1.95751, "vi": 1.65823, "tr": 1.85623, "fi": 2.06704, "ru": 1.23469, "el": 1.26118, "hi": 1.07143, "ta": 0.91241, "th": 1.01357, "zh": 2.70962, "my": 1.02997}, "bytes_per_chunk": {"en": 5.263, "de": 6.157, "vi": 7.685, "tr": 6.172, "fi": 5.886, "ru": 10.266, "el": 11.274, "hi": 12.44, "ta": 17.23, "th": 13.308, "zh": 5.298, "my": 45.781}, "bpb_mean": 1.54639, "t": 1068.1} |
| {"event": "train", "step": 5000, "phase": 2, "objective": "parity_b", "lm_loss": 1.17102, "bpb": 1.68942, "chunk_loss": 0.030075, "F": 0.14604, "G": 0.16954, "chunks": 149.55, "bytes_per_chunk": 6.847, "boundary_entropy": 0.13923, "router_noise": 0.0, "grad_norm": 0.3707, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.273, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1068.2} |
| {"event": "train", "step": 5100, "phase": 2, "objective": "parity_b", "lm_loss": 1.18238, "bpb": 1.70581, "chunk_loss": 0.030435, "F": 0.1528, "G": 0.16947, "chunks": 156.47, "bytes_per_chunk": 6.544, "boundary_entropy": 0.12276, "router_noise": 0.0, "grad_norm": 0.4057, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.189, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 1087.2} |
| {"event": "train", "step": 5200, "phase": 2, "objective": "parity_b", "lm_loss": 1.18216, "bpb": 1.7055, "chunk_loss": 0.030365, "F": 0.15009, "G": 0.16852, "chunks": 153.69, "bytes_per_chunk": 6.663, "boundary_entropy": 0.12983, "router_noise": 0.0, "grad_norm": 0.3902, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.188, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1719, "fallback": 0.0, "alpha": 0.03, "t": 1106.0} |
| {"event": "train", "step": 5300, "phase": 2, "objective": "parity_b", "lm_loss": 1.14245, "bpb": 1.64821, "chunk_loss": 0.030318, "F": 0.14803, "G": 0.1663, "chunks": 151.58, "bytes_per_chunk": 6.756, "boundary_entropy": 0.13024, "router_noise": 0.0, "grad_norm": 0.3813, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.176, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1123.6} |
| {"event": "train", "step": 5400, "phase": 2, "objective": "parity_b", "lm_loss": 1.11021, "bpb": 1.60169, "chunk_loss": 0.030265, "F": 0.14328, "G": 0.16113, "chunks": 146.72, "bytes_per_chunk": 6.979, "boundary_entropy": 0.12198, "router_noise": 0.0, "grad_norm": 0.368, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 1142.3} |
| {"event": "train", "step": 5500, "phase": 2, "objective": "parity_b", "lm_loss": 1.1139, "bpb": 1.60702, "chunk_loss": 0.030233, "F": 0.13887, "G": 0.15906, "chunks": 142.2, "bytes_per_chunk": 7.201, "boundary_entropy": 0.12991, "router_noise": 0.0, "grad_norm": 0.3645, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.188, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1161.1} |
| {"event": "train", "step": 5600, "phase": 2, "objective": "parity_b", "lm_loss": 1.12621, "bpb": 1.62477, "chunk_loss": 0.030038, "F": 0.14915, "G": 0.16794, "chunks": 152.73, "bytes_per_chunk": 6.704, "boundary_entropy": 0.1289, "router_noise": 0.0, "grad_norm": 0.3757, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.185, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 1179.5} |
| {"event": "train", "step": 5700, "phase": 2, "objective": "parity_b", "lm_loss": 1.16056, "bpb": 1.67433, "chunk_loss": 0.030142, "F": 0.15469, "G": 0.17037, "chunks": 158.41, "bytes_per_chunk": 6.464, "boundary_entropy": 0.11619, "router_noise": 0.0, "grad_norm": 0.376, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.178, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1197.3} |
| {"event": "train", "step": 5800, "phase": 2, "objective": "parity_b", "lm_loss": 1.08913, "bpb": 1.57128, "chunk_loss": 0.030197, "F": 0.13029, "G": 0.14951, "chunks": 133.42, "bytes_per_chunk": 7.675, "boundary_entropy": 0.12056, "router_noise": 0.0, "grad_norm": 0.3704, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.19, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1216.3} |
| {"event": "train", "step": 5900, "phase": 2, "objective": "parity_b", "lm_loss": 1.12112, "bpb": 1.61743, "chunk_loss": 0.030094, "F": 0.13792, "G": 0.1567, "chunks": 141.23, "bytes_per_chunk": 7.25, "boundary_entropy": 0.12275, "router_noise": 0.0, "grad_norm": 0.3945, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.175, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1233.8} |
| {"event": "probe", "step": 6000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3154.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.181233, "cps": 24.640625, "mean_p": 0.183585, "boundary_entropy": 0.062809, "bpb": 1.747007}, "de": {"chunks": 3244.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15799, "cps": 25.34375, "mean_p": 0.154348, "boundary_entropy": 0.078874, "bpb": 1.905484}, "vi": {"chunks": 3064.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.166206, "cps": 23.9375, "mean_p": 0.154898, "boundary_entropy": 0.152414, "bpb": 1.586212}, "tr": {"chunks": 2893.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.159808, "cps": 22.601562, "mean_p": 0.158406, "boundary_entropy": 0.108142, "bpb": 1.812294}, "fi": {"chunks": 3338.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177449, "cps": 26.078125, "mean_p": 0.175622, "boundary_entropy": 0.084767, "bpb": 2.013881}, "ru": {"chunks": 2966.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.156517, "cps": 23.171875, "mean_p": 0.113135, "boundary_entropy": 0.161278, "bpb": 1.191339}, "el": {"chunks": 2988.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.145593, "cps": 23.34375, "mean_p": 0.110216, "boundary_entropy": 0.174288, "bpb": 1.215763}, "hi": {"chunks": 2809.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.166954, "cps": 21.945312, "mean_p": 0.085439, "boundary_entropy": 0.126718, "bpb": 1.022455}, "ta": {"chunks": 2625.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.140412, "cps": 20.507812, "mean_p": 0.081633, "boundary_entropy": 0.154425, "bpb": 0.877419}, "th": {"chunks": 2739.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170283, "cps": 21.398438, "mean_p": 0.083366, "boundary_entropy": 0.13352, "bpb": 0.969147}, "zh": {"chunks": 3038.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.514392, "cps": 23.734375, "mean_p": 0.224038, "boundary_entropy": 0.158429, "bpb": 2.622096}, "my": {"chunks": 902.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.047604, "cps": 7.046875, "mean_p": 0.097639, "boundary_entropy": 0.246656, "bpb": 0.977996}}, "bpb": {"en": 1.74701, "de": 1.90548, "vi": 1.58621, "tr": 1.81229, "fi": 2.01388, "ru": 1.19134, "el": 1.21576, "hi": 1.02246, "ta": 0.87742, "th": 0.96915, "zh": 2.6221, "my": 0.978}, "gini_cpc": 0.222299, "gini_cps": 0.091341, "bpic": 11.367921, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8717478191608967, "vi": 0.9170817482597943, "tr": 0.8817801405618403, "fi": 0.9791221605755775, "ru": 0.863623325408285, "el": 0.8033451557559298, "hi": 0.9212109625645775, "ta": 0.7747583569169203, "th": 0.9395792091356772, "zh": 2.838289941271851, "my": 0.2626670476445966}, "t": 1255.9} |
| {"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.7459, "de": 1.90874, "vi": 1.60098, "tr": 1.80441, "fi": 2.02616, "ru": 1.1913, "el": 1.21513, "hi": 1.02717, "ta": 0.88046, "th": 0.96755, "zh": 2.62459, "my": 0.98553}, "bytes_per_chunk": {"en": 5.28, "de": 6.327, "vi": 7.611, "tr": 6.401, "fi": 5.851, "ru": 11.242, "el": 12.64, "hi": 14.823, "ta": 18.88, "th": 16.996, "zh": 5.337, "my": 47.754}, "bpb_mean": 1.49816, "t": 1261.0} |
| {"event": "train", "step": 6000, "phase": 2, "objective": "parity_b", "lm_loss": 1.12461, "bpb": 1.62247, "chunk_loss": 0.030252, "F": 0.1368, "G": 0.15746, "chunks": 140.08, "bytes_per_chunk": 7.31, "boundary_entropy": 0.12296, "router_noise": 0.0, "grad_norm": 0.3797, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.274, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1406, "fallback": 0.0, "alpha": 0.03, "t": 1261.2} |
| {"event": "train", "step": 6100, "phase": 2, "objective": "parity_b", "lm_loss": 1.18374, "bpb": 1.70777, "chunk_loss": 0.030146, "F": 0.14601, "G": 0.16378, "chunks": 149.52, "bytes_per_chunk": 6.849, "boundary_entropy": 0.11497, "router_noise": 0.0, "grad_norm": 0.385, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.175, "worst_lang": "en", "n_target_mean": 6.0, "frac_in_batch": 0.3594, "fallback": 0.0, "alpha": 0.03, "t": 1278.7} |
| {"event": "train", "step": 6200, "phase": 2, "objective": "parity_b", "lm_loss": 1.16549, "bpb": 1.68144, "chunk_loss": 0.031178, "F": 0.14629, "G": 0.16271, "chunks": 149.8, "bytes_per_chunk": 6.836, "boundary_entropy": 0.1183, "router_noise": 0.0, "grad_norm": 0.384, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.195, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1719, "fallback": 0.0, "alpha": 0.03, "t": 1298.2} |
| {"event": "train", "step": 6300, "phase": 2, "objective": "parity_b", "lm_loss": 1.20037, "bpb": 1.73177, "chunk_loss": 0.030431, "F": 0.15477, "G": 0.17113, "chunks": 158.48, "bytes_per_chunk": 6.461, "boundary_entropy": 0.11064, "router_noise": 0.0, "grad_norm": 0.3799, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.198, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1719, "fallback": 0.0, "alpha": 0.03, "t": 1318.0} |
| {"event": "train", "step": 6400, "phase": 2, "objective": "parity_b", "lm_loss": 1.1114, "bpb": 1.60341, "chunk_loss": 0.03024, "F": 0.14738, "G": 0.16191, "chunks": 150.92, "bytes_per_chunk": 6.785, "boundary_entropy": 0.10918, "router_noise": 0.0, "grad_norm": 0.3674, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.177, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 1335.7} |
| {"event": "train", "step": 6500, "phase": 2, "objective": "parity_b", "lm_loss": 1.13109, "bpb": 1.63182, "chunk_loss": 0.030108, "F": 0.14197, "G": 0.1558, "chunks": 145.38, "bytes_per_chunk": 7.044, "boundary_entropy": 0.10366, "router_noise": 0.0, "grad_norm": 0.3734, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.198, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1406, "fallback": 0.0, "alpha": 0.03, "t": 1355.5} |
| {"event": "train", "step": 6600, "phase": 2, "objective": "parity_b", "lm_loss": 1.08775, "bpb": 1.56928, "chunk_loss": 0.030206, "F": 0.13364, "G": 0.14878, "chunks": 136.84, "bytes_per_chunk": 7.483, "boundary_entropy": 0.1106, "router_noise": 0.0, "grad_norm": 0.3724, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.178, "worst_lang": "en", "n_target_mean": 6.0, "frac_in_batch": 0.2656, "fallback": 0.0, "alpha": 0.03, "t": 1373.3} |
| {"event": "train", "step": 6700, "phase": 2, "objective": "parity_b", "lm_loss": 1.11094, "bpb": 1.60275, "chunk_loss": 0.029998, "F": 0.14009, "G": 0.15696, "chunks": 143.45, "bytes_per_chunk": 7.138, "boundary_entropy": 0.1119, "router_noise": 0.0, "grad_norm": 0.3801, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.193, "worst_lang": "ta", "n_target_mean": 19.1516, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 1392.6} |
| {"event": "train", "step": 6800, "phase": 2, "objective": "parity_b", "lm_loss": 1.15368, "bpb": 1.66441, "chunk_loss": 0.030055, "F": 0.15721, "G": 0.17162, "chunks": 160.98, "bytes_per_chunk": 6.361, "boundary_entropy": 0.10391, "router_noise": 0.0, "grad_norm": 0.3764, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.173, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 1409.9} |
| {"event": "train", "step": 6900, "phase": 2, "objective": "parity_b", "lm_loss": 1.13078, "bpb": 1.63137, "chunk_loss": 0.030199, "F": 0.15637, "G": 0.16872, "chunks": 160.12, "bytes_per_chunk": 6.395, "boundary_entropy": 0.09855, "router_noise": 0.0, "grad_norm": 0.3821, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.192, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1429.1} |
| {"event": "probe", "step": 7000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3154.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.181233, "cps": 24.640625, "mean_p": 0.182776, "boundary_entropy": 0.056881, "bpb": 1.715962}, "de": {"chunks": 3219.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.156772, "cps": 25.148438, "mean_p": 0.154042, "boundary_entropy": 0.06652, "bpb": 1.866422}, "vi": {"chunks": 3002.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.162842, "cps": 23.453125, "mean_p": 0.147859, "boundary_entropy": 0.129062, "bpb": 1.542996}, "tr": {"chunks": 2982.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164724, "cps": 23.296875, "mean_p": 0.16023, "boundary_entropy": 0.092058, "bpb": 1.768797}, "fi": {"chunks": 3198.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.170007, "cps": 24.984375, "mean_p": 0.167199, "boundary_entropy": 0.07058, "bpb": 1.952216}, "ru": {"chunks": 3209.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.16934, "cps": 25.070312, "mean_p": 0.11901, "boundary_entropy": 0.15516, "bpb": 1.156034}, "el": {"chunks": 3159.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153925, "cps": 24.679688, "mean_p": 0.112334, "boundary_entropy": 0.164519, "bpb": 1.174086}, "hi": {"chunks": 2912.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.173076, "cps": 22.75, "mean_p": 0.084389, "boundary_entropy": 0.113905, "bpb": 0.979931}, "ta": {"chunks": 2380.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.127307, "cps": 18.59375, "mean_p": 0.074763, "boundary_entropy": 0.13913, "bpb": 0.850619}, "th": {"chunks": 2955.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.183712, "cps": 23.085938, "mean_p": 0.085803, "boundary_entropy": 0.121111, "bpb": 0.938283}, "zh": {"chunks": 3026.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.51236, "cps": 23.640625, "mean_p": 0.219067, "boundary_entropy": 0.137615, "bpb": 2.57794}, "my": {"chunks": 1332.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.070298, "cps": 10.40625, "mean_p": 0.087685, "boundary_entropy": 0.219955, "bpb": 0.949436}}, "bpb": {"en": 1.71596, "de": 1.86642, "vi": 1.543, "tr": 1.7688, "fi": 1.95222, "ru": 1.15603, "el": 1.17409, "hi": 0.97993, "ta": 0.85062, "th": 0.93828, "zh": 2.57794, "my": 0.94944}, "gini_cpc": 0.21227, "gini_cps": 0.076701, "bpic": 11.115066, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8650296639577455, "vi": 0.8985246110561039, "tr": 0.9089071479970299, "fi": 0.938056521725793, "ru": 0.9343787091150325, "el": 0.8493197279226847, "hi": 0.9549897910245817, "ta": 0.7024475769380077, "th": 1.0136752694399145, "zh": 2.827078789430093, "my": 0.3878852632623089}, "t": 1449.6} |
| {"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.71427, "de": 1.8691, "vi": 1.5576, "tr": 1.76126, "fi": 1.96316, "ru": 1.15551, "el": 1.17373, "hi": 0.98535, "ta": 0.85394, "th": 0.93663, "zh": 2.57994, "my": 0.95476}, "bytes_per_chunk": {"en": 5.28, "de": 6.296, "vi": 7.741, "tr": 6.188, "fi": 6.091, "ru": 10.506, "el": 11.846, "hi": 14.556, "ta": 20.031, "th": 15.672, "zh": 5.352, "my": 34.12}, "bpb_mean": 1.45877, "t": 1454.7} |
| {"event": "train", "step": 7000, "phase": 2, "objective": "parity_b", "lm_loss": 1.22728, "bpb": 1.77059, "chunk_loss": 0.030147, "F": 0.15488, "G": 0.16901, "chunks": 158.59, "bytes_per_chunk": 6.457, "boundary_entropy": 0.1016, "router_noise": 0.0, "grad_norm": 0.3906, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.258, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 1454.9} |
| {"event": "train", "step": 7100, "phase": 2, "objective": "parity_b", "lm_loss": 1.14318, "bpb": 1.64926, "chunk_loss": 0.030181, "F": 0.14911, "G": 0.16358, "chunks": 152.69, "bytes_per_chunk": 6.707, "boundary_entropy": 0.10324, "router_noise": 0.0, "grad_norm": 0.3877, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.191, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1474.0} |
| {"event": "train", "step": 7200, "phase": 2, "objective": "parity_b", "lm_loss": 1.04641, "bpb": 1.50964, "chunk_loss": 0.030063, "F": 0.13849, "G": 0.15206, "chunks": 141.81, "bytes_per_chunk": 7.221, "boundary_entropy": 0.10577, "router_noise": 0.0, "grad_norm": 0.3728, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1492.7} |
| {"event": "train", "step": 7300, "phase": 2, "objective": "parity_b", "lm_loss": 1.13292, "bpb": 1.63445, "chunk_loss": 0.03019, "F": 0.15019, "G": 0.1623, "chunks": 153.8, "bytes_per_chunk": 6.658, "boundary_entropy": 0.09663, "router_noise": 0.0, "grad_norm": 0.4117, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1510.7} |
| {"event": "train", "step": 7400, "phase": 2, "objective": "parity_b", "lm_loss": 1.05363, "bpb": 1.52006, "chunk_loss": 0.030086, "F": 0.13123, "G": 0.14662, "chunks": 134.38, "bytes_per_chunk": 7.62, "boundary_entropy": 0.1065, "router_noise": 0.0, "grad_norm": 0.3938, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.194, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.2031, "fallback": 0.0, "alpha": 0.03, "t": 1530.1} |
| {"event": "train", "step": 7500, "phase": 2, "objective": "parity_b", "lm_loss": 1.07292, "bpb": 1.54789, "chunk_loss": 0.030073, "F": 0.1432, "G": 0.15667, "chunks": 146.64, "bytes_per_chunk": 6.983, "boundary_entropy": 0.09448, "router_noise": 0.0, "grad_norm": 0.3871, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.171, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1547.2} |
| {"event": "train", "step": 7600, "phase": 2, "objective": "parity_b", "lm_loss": 1.06372, "bpb": 1.53463, "chunk_loss": 0.030091, "F": 0.13664, "G": 0.15048, "chunks": 139.92, "bytes_per_chunk": 7.318, "boundary_entropy": 0.09878, "router_noise": 0.0, "grad_norm": 0.3766, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.196, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 1566.8} |
| {"event": "phase_transition", "step": 7700, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 295680, "t": 1583.8} |
| {"event": "train", "step": 7700, "phase": 3, "objective": "none", "lm_loss": 1.05073, "bpb": 1.51588, "chunk_loss": 0.0, "F": 0.1431, "G": 0.15502, "chunks": 146.53, "bytes_per_chunk": 6.988, "boundary_entropy": 0.09432, "router_noise": 0.0, "grad_norm": 0.3733, "lr": 9.06e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 1584.0} |
| {"event": "train", "step": 7800, "phase": 3, "objective": "none", "lm_loss": 1.10543, "bpb": 1.5948, "chunk_loss": 0.0, "F": 0.15088, "G": 0.16494, "chunks": 154.5, "bytes_per_chunk": 6.628, "boundary_entropy": 0.09916, "router_noise": 0.0, "grad_norm": 0.38, "lr": 8.73e-05, "router_trainable": false, "sec_per_step": 0.198, "alpha": 0.0, "t": 1603.8} |
| {"event": "train", "step": 7900, "phase": 3, "objective": "none", "lm_loss": 1.09525, "bpb": 1.58011, "chunk_loss": 0.0, "F": 0.14702, "G": 0.16294, "chunks": 150.55, "bytes_per_chunk": 6.802, "boundary_entropy": 0.10544, "router_noise": 0.0, "grad_norm": 0.3861, "lr": 8.4e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1621.2} |
| {"event": "probe", "step": 8000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3145.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.180716, "cps": 24.570312, "mean_p": 0.182787, "boundary_entropy": 0.053133, "bpb": 1.693178}, "de": {"chunks": 3364.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.163834, "cps": 26.28125, "mean_p": 0.162172, "boundary_entropy": 0.06621, "bpb": 1.832201}, "vi": {"chunks": 3102.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.168267, "cps": 24.234375, "mean_p": 0.153232, "boundary_entropy": 0.132459, "bpb": 1.512461}, "tr": {"chunks": 2980.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164614, "cps": 23.28125, "mean_p": 0.162421, "boundary_entropy": 0.09371, "bpb": 1.736084}, "fi": {"chunks": 3270.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.173834, "cps": 25.546875, "mean_p": 0.170847, "boundary_entropy": 0.067688, "bpb": 1.914126}, "ru": {"chunks": 3371.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.177889, "cps": 26.335938, "mean_p": 0.124798, "boundary_entropy": 0.159274, "bpb": 1.130167}, "el": {"chunks": 3370.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.164206, "cps": 26.328125, "mean_p": 0.119451, "boundary_entropy": 0.173098, "bpb": 1.143364}, "hi": {"chunks": 3024.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.179733, "cps": 23.625, "mean_p": 0.086971, "boundary_entropy": 0.111786, "bpb": 0.950547}, "ta": {"chunks": 2568.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.137363, "cps": 20.0625, "mean_p": 0.078606, "boundary_entropy": 0.141014, "bpb": 0.826608}, "th": {"chunks": 3149.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.195772, "cps": 24.601562, "mean_p": 0.087492, "boundary_entropy": 0.114367, "bpb": 0.906347}, "zh": {"chunks": 3047.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.515916, "cps": 23.804688, "mean_p": 0.21771, "boundary_entropy": 0.129661, "bpb": 2.537031}, "my": {"chunks": 1579.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.083333, "cps": 12.335938, "mean_p": 0.077068, "boundary_entropy": 0.193302, "bpb": 0.924476}}, "bpb": {"en": 1.69318, "de": 1.8322, "vi": 1.51246, "tr": 1.73608, "fi": 1.91413, "ru": 1.13017, "el": 1.14336, "hi": 0.95055, "ta": 0.82661, "th": 0.90635, "zh": 2.53703, "my": 0.92448}, "gini_cpc": 0.199716, "gini_cps": 0.072278, "bpic": 10.669771, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9065819131589871, "vi": 0.9311124248261088, "tr": 0.910896814000297, "fi": 0.9619208528920306, "ru": 0.9843578490798729, "el": 0.9086414118186218, "hi": 0.9945581585431249, "ta": 0.7601040040717211, "th": 1.0833157712925123, "zh": 2.854844659603529, "my": 0.4611287758346581}, "t": 1643.5} |
| {"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.69044, "de": 1.835, "vi": 1.52827, "tr": 1.72915, "fi": 1.92565, "ru": 1.12932, "el": 1.14277, "hi": 0.9556, "ta": 0.83015, "th": 0.90397, "zh": 2.5397, "my": 0.93098}, "bytes_per_chunk": {"en": 5.258, "de": 6.05, "vi": 7.509, "tr": 6.142, "fi": 5.941, "ru": 10.043, "el": 11.295, "hi": 13.693, "ta": 18.565, "th": 14.783, "zh": 5.332, "my": 29.876}, "bpb_mean": 1.42842, "t": 1648.6} |
| {"event": "train", "step": 8000, "phase": 3, "objective": "none", "lm_loss": 1.0762, "bpb": 1.55263, "chunk_loss": 0.0, "F": 0.14461, "G": 0.15819, "chunks": 148.08, "bytes_per_chunk": 6.915, "boundary_entropy": 0.09758, "router_noise": 0.0, "grad_norm": 0.3995, "lr": 8.08e-05, "router_trainable": false, "sec_per_step": 0.275, "alpha": 0.0, "t": 1648.7} |
| {"event": "train", "step": 8100, "phase": 3, "objective": "none", "lm_loss": 1.07724, "bpb": 1.55412, "chunk_loss": 0.0, "F": 0.13693, "G": 0.1534, "chunks": 140.22, "bytes_per_chunk": 7.303, "boundary_entropy": 0.11141, "router_noise": 0.0, "grad_norm": 0.4018, "lr": 7.77e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1666.3} |
| {"event": "train", "step": 8200, "phase": 3, "objective": "none", "lm_loss": 1.13176, "bpb": 1.63278, "chunk_loss": 0.0, "F": 0.15457, "G": 0.16883, "chunks": 158.28, "bytes_per_chunk": 6.469, "boundary_entropy": 0.09601, "router_noise": 0.0, "grad_norm": 0.4053, "lr": 7.47e-05, "router_trainable": false, "sec_per_step": 0.191, "alpha": 0.0, "t": 1685.4} |
| {"event": "train", "step": 8300, "phase": 3, "objective": "none", "lm_loss": 1.06124, "bpb": 1.53105, "chunk_loss": 0.0, "F": 0.15175, "G": 0.16513, "chunks": 155.39, "bytes_per_chunk": 6.59, "boundary_entropy": 0.09672, "router_noise": 0.0, "grad_norm": 0.4124, "lr": 7.17e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1702.9} |
| {"event": "train", "step": 8400, "phase": 3, "objective": "none", "lm_loss": 1.08434, "bpb": 1.56437, "chunk_loss": 0.0, "F": 0.14499, "G": 0.16137, "chunks": 148.47, "bytes_per_chunk": 6.897, "boundary_entropy": 0.11131, "router_noise": 0.0, "grad_norm": 0.3993, "lr": 6.88e-05, "router_trainable": false, "sec_per_step": 0.194, "alpha": 0.0, "t": 1722.3} |
| {"event": "train", "step": 8500, "phase": 3, "objective": "none", "lm_loss": 1.07676, "bpb": 1.55344, "chunk_loss": 0.0, "F": 0.14575, "G": 0.16112, "chunks": 149.25, "bytes_per_chunk": 6.861, "boundary_entropy": 0.10366, "router_noise": 0.0, "grad_norm": 0.4117, "lr": 6.6e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 1739.9} |
| {"event": "train", "step": 8600, "phase": 3, "objective": "none", "lm_loss": 1.04106, "bpb": 1.50194, "chunk_loss": 0.0, "F": 0.14047, "G": 0.1565, "chunks": 143.84, "bytes_per_chunk": 7.119, "boundary_entropy": 0.10966, "router_noise": 0.0, "grad_norm": 0.3816, "lr": 6.33e-05, "router_trainable": false, "sec_per_step": 0.19, "alpha": 0.0, "t": 1758.9} |
| {"event": "train", "step": 8700, "phase": 3, "objective": "none", "lm_loss": 1.05977, "bpb": 1.52892, "chunk_loss": 0.0, "F": 0.14418, "G": 0.16202, "chunks": 147.64, "bytes_per_chunk": 6.936, "boundary_entropy": 0.11406, "router_noise": 0.0, "grad_norm": 0.4053, "lr": 6.07e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1776.3} |
| {"event": "train", "step": 8800, "phase": 3, "objective": "none", "lm_loss": 1.05398, "bpb": 1.52057, "chunk_loss": 0.0, "F": 0.14456, "G": 0.16179, "chunks": 148.03, "bytes_per_chunk": 6.917, "boundary_entropy": 0.10854, "router_noise": 0.0, "grad_norm": 0.3985, "lr": 5.82e-05, "router_trainable": false, "sec_per_step": 0.192, "alpha": 0.0, "t": 1795.5} |
| {"event": "train", "step": 8900, "phase": 3, "objective": "none", "lm_loss": 1.04397, "bpb": 1.50614, "chunk_loss": 0.0, "F": 0.14648, "G": 0.1623, "chunks": 150.0, "bytes_per_chunk": 6.827, "boundary_entropy": 0.10387, "router_noise": 0.0, "grad_norm": 0.3908, "lr": 5.58e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1813.0} |
| {"event": "probe", "step": 9000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3163.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.18175, "cps": 24.710938, "mean_p": 0.184018, "boundary_entropy": 0.052424, "bpb": 1.673418}, "de": {"chunks": 3418.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.166464, "cps": 26.703125, "mean_p": 0.165343, "boundary_entropy": 0.068167, "bpb": 1.807477}, "vi": {"chunks": 3263.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.177, "cps": 25.492188, "mean_p": 0.158792, "boundary_entropy": 0.133924, "bpb": 1.48578}, "tr": {"chunks": 3137.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.173286, "cps": 24.507812, "mean_p": 0.16861, "boundary_entropy": 0.095511, "bpb": 1.712859}, "fi": {"chunks": 3371.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179204, "cps": 26.335938, "mean_p": 0.176809, "boundary_entropy": 0.071641, "bpb": 1.882051}, "ru": {"chunks": 3460.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.182586, "cps": 27.03125, "mean_p": 0.127748, "boundary_entropy": 0.159899, "bpb": 1.110005}, "el": {"chunks": 3293.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.160454, "cps": 25.726562, "mean_p": 0.119437, "boundary_entropy": 0.170918, "bpb": 1.121455}, "hi": {"chunks": 3345.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.198811, "cps": 26.132812, "mean_p": 0.097502, "boundary_entropy": 0.125438, "bpb": 0.923289}, "ta": {"chunks": 3045.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.162878, "cps": 23.789062, "mean_p": 0.089325, "boundary_entropy": 0.148506, "bpb": 0.804172}, "th": {"chunks": 3470.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.215729, "cps": 27.109375, "mean_p": 0.096688, "boundary_entropy": 0.12271, "bpb": 0.888994}, "zh": {"chunks": 3028.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.512699, "cps": 23.65625, "mean_p": 0.217048, "boundary_entropy": 0.130414, "bpb": 2.507865}, "my": {"chunks": 1427.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.075311, "cps": 11.148438, "mean_p": 0.077817, "boundary_entropy": 0.194938, "bpb": 0.898679}}, "bpb": {"en": 1.67342, "de": 1.80748, "vi": 1.48578, "tr": 1.71286, "fi": 1.88205, "ru": 1.11001, "el": 1.12146, "hi": 0.92329, "ta": 0.80417, "th": 0.88899, "zh": 2.50787, "my": 0.89868}, "gini_cpc": 0.197963, "gini_cps": 0.068408, "bpic": 10.25604, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9158926619501139, "vi": 0.9738652292436422, "tr": 0.9534301811695464, "fi": 0.985988381388635, "ru": 1.004596851624614, "el": 0.8828274205251915, "hi": 1.0938706693618965, "ta": 0.8961624770630798, "th": 1.1869524953930426, "zh": 2.820897828226577, "my": 0.4143673475190199}, "t": 1835.2} |
| {"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.67028, "de": 1.8117, "vi": 1.50159, "tr": 1.70672, "fi": 1.89553, "ru": 1.10972, "el": 1.11991, "hi": 0.92886, "ta": 0.80836, "th": 0.88599, "zh": 2.50959, "my": 0.90559}, "bytes_per_chunk": {"en": 5.251, "de": 5.906, "vi": 7.322, "tr": 5.922, "fi": 5.803, "ru": 9.739, "el": 11.372, "hi": 12.507, "ta": 16.263, "th": 13.376, "zh": 5.349, "my": 32.873}, "bpb_mean": 1.40449, "t": 1840.5} |
| {"event": "train", "step": 9000, "phase": 3, "objective": "none", "lm_loss": 1.08861, "bpb": 1.57053, "chunk_loss": 0.0, "F": 0.15379, "G": 0.16921, "chunks": 157.48, "bytes_per_chunk": 6.502, "boundary_entropy": 0.10327, "router_noise": 0.0, "grad_norm": 0.397, "lr": 5.35e-05, "router_trainable": false, "sec_per_step": 0.276, "alpha": 0.0, "t": 1840.7} |
| {"event": "train", "step": 9100, "phase": 3, "objective": "none", "lm_loss": 1.02304, "bpb": 1.47594, "chunk_loss": 0.0, "F": 0.14667, "G": 0.16001, "chunks": 150.19, "bytes_per_chunk": 6.818, "boundary_entropy": 0.10067, "router_noise": 0.0, "grad_norm": 0.3781, "lr": 5.12e-05, "router_trainable": false, "sec_per_step": 0.194, "alpha": 0.0, "t": 1860.0} |
| {"event": "train", "step": 9200, "phase": 3, "objective": "none", "lm_loss": 1.10587, "bpb": 1.59543, "chunk_loss": 0.0, "F": 0.15141, "G": 0.16737, "chunks": 155.05, "bytes_per_chunk": 6.604, "boundary_entropy": 0.10348, "router_noise": 0.0, "grad_norm": 0.4312, "lr": 4.91e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1877.5} |
| {"event": "train", "step": 9300, "phase": 3, "objective": "none", "lm_loss": 0.97081, "bpb": 1.40058, "chunk_loss": 0.0, "F": 0.13995, "G": 0.15653, "chunks": 143.31, "bytes_per_chunk": 7.145, "boundary_entropy": 0.11158, "router_noise": 0.0, "grad_norm": 0.3903, "lr": 4.71e-05, "router_trainable": false, "sec_per_step": 0.193, "alpha": 0.0, "t": 1896.8} |
| {"event": "train", "step": 9400, "phase": 3, "objective": "none", "lm_loss": 1.08336, "bpb": 1.56296, "chunk_loss": 0.0, "F": 0.16437, "G": 0.17765, "chunks": 168.31, "bytes_per_chunk": 6.084, "boundary_entropy": 0.09622, "router_noise": 0.0, "grad_norm": 0.3896, "lr": 4.52e-05, "router_trainable": false, "sec_per_step": 0.175, "alpha": 0.0, "t": 1914.2} |
| {"event": "train", "step": 9500, "phase": 3, "objective": "none", "lm_loss": 1.01933, "bpb": 1.47058, "chunk_loss": 0.0, "F": 0.14629, "G": 0.16467, "chunks": 149.8, "bytes_per_chunk": 6.836, "boundary_entropy": 0.11364, "router_noise": 0.0, "grad_norm": 0.3895, "lr": 4.34e-05, "router_trainable": false, "sec_per_step": 0.196, "alpha": 0.0, "t": 1933.9} |
| {"event": "train", "step": 9600, "phase": 3, "objective": "none", "lm_loss": 1.09024, "bpb": 1.57288, "chunk_loss": 0.0, "F": 0.15521, "G": 0.16964, "chunks": 158.94, "bytes_per_chunk": 6.443, "boundary_entropy": 0.09959, "router_noise": 0.0, "grad_norm": 0.3999, "lr": 4.17e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 1951.5} |
| {"event": "train", "step": 9700, "phase": 3, "objective": "none", "lm_loss": 1.06846, "bpb": 1.54146, "chunk_loss": 0.0, "F": 0.15781, "G": 0.17175, "chunks": 161.59, "bytes_per_chunk": 6.337, "boundary_entropy": 0.09947, "router_noise": 0.0, "grad_norm": 0.3914, "lr": 4.01e-05, "router_trainable": false, "sec_per_step": 0.197, "alpha": 0.0, "t": 1971.2} |
| {"event": "train", "step": 9800, "phase": 3, "objective": "none", "lm_loss": 1.03508, "bpb": 1.49331, "chunk_loss": 0.0, "F": 0.14691, "G": 0.16259, "chunks": 150.44, "bytes_per_chunk": 6.807, "boundary_entropy": 0.10402, "router_noise": 0.0, "grad_norm": 0.3921, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 1988.8} |
| {"event": "train", "step": 9900, "phase": 3, "objective": "none", "lm_loss": 1.04474, "bpb": 1.50724, "chunk_loss": 0.0, "F": 0.15305, "G": 0.16812, "chunks": 156.72, "bytes_per_chunk": 6.534, "boundary_entropy": 0.10254, "router_noise": 0.0, "grad_norm": 0.4022, "lr": 3.72e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 2005.9} |
| {"event": "probe", "step": 10000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3189.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183244, "cps": 24.914062, "mean_p": 0.186239, "boundary_entropy": 0.05319, "bpb": 1.660316}, "de": {"chunks": 3613.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.175961, "cps": 28.226562, "mean_p": 0.173609, "boundary_entropy": 0.067993, "bpb": 1.789061}, "vi": {"chunks": 3291.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.178519, "cps": 25.710938, "mean_p": 0.162054, "boundary_entropy": 0.139623, "bpb": 1.469194}, "tr": {"chunks": 3144.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.173673, "cps": 24.5625, "mean_p": 0.169692, "boundary_entropy": 0.093986, "bpb": 1.695197}, "fi": {"chunks": 3392.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18032, "cps": 26.5, "mean_p": 0.178458, "boundary_entropy": 0.069793, "bpb": 1.86339}, "ru": {"chunks": 3592.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.189551, "cps": 28.0625, "mean_p": 0.130915, "boundary_entropy": 0.160569, "bpb": 1.096956}, "el": {"chunks": 3466.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.168884, "cps": 27.078125, "mean_p": 0.12472, "boundary_entropy": 0.176086, "bpb": 1.106279}, "hi": {"chunks": 3564.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.211828, "cps": 27.84375, "mean_p": 0.101262, "boundary_entropy": 0.122349, "bpb": 0.909221}, "ta": {"chunks": 3313.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.177213, "cps": 25.882812, "mean_p": 0.09557, "boundary_entropy": 0.154247, "bpb": 0.792766}, "th": {"chunks": 3592.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.223314, "cps": 28.0625, "mean_p": 0.098997, "boundary_entropy": 0.122895, "bpb": 0.874344}, "zh": {"chunks": 3036.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.514054, "cps": 23.71875, "mean_p": 0.217216, "boundary_entropy": 0.129061, "bpb": 2.491801}, "my": {"chunks": 1407.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.074256, "cps": 10.992188, "mean_p": 0.079015, "boundary_entropy": 0.198617, "bpb": 0.886983}}, "bpb": {"en": 1.66032, "de": 1.78906, "vi": 1.46919, "tr": 1.6952, "fi": 1.86339, "ru": 1.09696, "el": 1.10628, "hi": 0.90922, "ta": 0.79277, "th": 0.87434, "zh": 2.4918, "my": 0.88698}, "gini_cpc": 0.193221, "gini_cps": 0.075315, "bpic": 9.942771, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9602518562345477, "vi": 0.9742139438330653, "tr": 0.9477670083631923, "fi": 0.984041832571993, "ru": 1.034419537476699, "el": 0.9216315091312323, "hi": 1.1559850656766364, "ta": 0.967087048946587, "th": 1.218666474055545, "zh": 2.805291046081301, "my": 0.40522881755305556}, "t": 2026.4} |
| {"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.65712, "de": 1.79384, "vi": 1.48474, "tr": 1.68833, "fi": 1.87643, "ru": 1.09655, "el": 1.10545, "hi": 0.91477, "ta": 0.79708, "th": 0.87187, "zh": 2.4956, "my": 0.89329}, "bytes_per_chunk": {"en": 5.224, "de": 5.683, "vi": 7.233, "tr": 5.899, "fi": 5.767, "ru": 9.372, "el": 10.808, "hi": 11.757, "ta": 15.235, "th": 12.666, "zh": 5.331, "my": 32.482}, "bpb_mean": 1.38959, "t": 2032.5} |
| {"event": "train", "step": 10000, "phase": 3, "objective": "none", "lm_loss": 1.05633, "bpb": 1.52397, "chunk_loss": 0.0, "F": 0.15672, "G": 0.17045, "chunks": 160.48, "bytes_per_chunk": 6.381, "boundary_entropy": 0.09539, "router_noise": 0.0, "grad_norm": 0.3933, "lr": 3.6e-05, "router_trainable": false, "sec_per_step": 0.267, "alpha": 0.0, "t": 2032.6} |
| {"event": "train", "step": 10100, "phase": 3, "objective": "none", "lm_loss": 1.06594, "bpb": 1.53782, "chunk_loss": 0.0, "F": 0.15259, "G": 0.16978, "chunks": 156.25, "bytes_per_chunk": 6.554, "boundary_entropy": 0.10335, "router_noise": 0.0, "grad_norm": 0.4196, "lr": 3.49e-05, "router_trainable": false, "sec_per_step": 0.195, "alpha": 0.0, "t": 2052.2} |
| {"event": "train", "step": 10200, "phase": 3, "objective": "none", "lm_loss": 1.02512, "bpb": 1.47893, "chunk_loss": 0.0, "F": 0.14558, "G": 0.16274, "chunks": 149.08, "bytes_per_chunk": 6.869, "boundary_entropy": 0.1095, "router_noise": 0.0, "grad_norm": 0.4131, "lr": 3.38e-05, "router_trainable": false, "sec_per_step": 0.197, "alpha": 0.0, "t": 2071.9} |
| {"event": "train", "step": 10300, "phase": 3, "objective": "none", "lm_loss": 0.97147, "bpb": 1.40154, "chunk_loss": 0.0, "F": 0.14021, "G": 0.15801, "chunks": 143.58, "bytes_per_chunk": 7.132, "boundary_entropy": 0.11212, "router_noise": 0.0, "grad_norm": 0.4002, "lr": 3.3e-05, "router_trainable": false, "sec_per_step": 0.198, "alpha": 0.0, "t": 2091.6} |
| {"event": "train", "step": 10400, "phase": 3, "objective": "none", "lm_loss": 1.06824, "bpb": 1.54114, "chunk_loss": 0.0, "F": 0.15405, "G": 0.16988, "chunks": 157.75, "bytes_per_chunk": 6.491, "boundary_entropy": 0.10441, "router_noise": 0.0, "grad_norm": 0.3966, "lr": 3.22e-05, "router_trainable": false, "sec_per_step": 0.199, "alpha": 0.0, "t": 2111.5} |
| {"event": "train", "step": 10500, "phase": 3, "objective": "none", "lm_loss": 1.01261, "bpb": 1.46089, "chunk_loss": 0.0, "F": 0.14198, "G": 0.16176, "chunks": 145.39, "bytes_per_chunk": 7.043, "boundary_entropy": 0.11583, "router_noise": 0.0, "grad_norm": 0.4115, "lr": 3.15e-05, "router_trainable": false, "sec_per_step": 0.197, "alpha": 0.0, "t": 2131.2} |
| {"event": "train", "step": 10600, "phase": 3, "objective": "none", "lm_loss": 0.97458, "bpb": 1.40603, "chunk_loss": 0.0, "F": 0.13748, "G": 0.15649, "chunks": 140.78, "bytes_per_chunk": 7.274, "boundary_entropy": 0.11777, "router_noise": 0.0, "grad_norm": 0.3944, "lr": 3.1e-05, "router_trainable": false, "sec_per_step": 0.201, "alpha": 0.0, "t": 2151.3} |
| {"event": "train", "step": 10700, "phase": 3, "objective": "none", "lm_loss": 1.0375, "bpb": 1.4968, "chunk_loss": 0.0, "F": 0.14839, "G": 0.16566, "chunks": 151.95, "bytes_per_chunk": 6.739, "boundary_entropy": 0.111, "router_noise": 0.0, "grad_norm": 0.4062, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 2170.2} |
| {"event": "train", "step": 10800, "phase": 3, "objective": "none", "lm_loss": 1.01484, "bpb": 1.46411, "chunk_loss": 0.0, "F": 0.15717, "G": 0.17173, "chunks": 160.94, "bytes_per_chunk": 6.363, "boundary_entropy": 0.10134, "router_noise": 0.0, "grad_norm": 0.4061, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.182, "alpha": 0.0, "t": 2188.3} |
| {"event": "train", "step": 10900, "phase": 3, "objective": "none", "lm_loss": 1.03085, "bpb": 1.48721, "chunk_loss": 0.0, "F": 0.14719, "G": 0.16284, "chunks": 150.72, "bytes_per_chunk": 6.794, "boundary_entropy": 0.10361, "router_noise": 0.0, "grad_norm": 0.421, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.194, "alpha": 0.0, "t": 2207.8} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3199.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183819, "cps": 24.992188, "mean_p": 0.187443, "boundary_entropy": 0.054951, "bpb": 1.651314}, "de": {"chunks": 3665.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178493, "cps": 28.632812, "mean_p": 0.176327, "boundary_entropy": 0.069887, "bpb": 1.780604}, "vi": {"chunks": 3321.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.180146, "cps": 25.945312, "mean_p": 0.162964, "boundary_entropy": 0.139354, "bpb": 1.459227}, "tr": {"chunks": 3190.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176214, "cps": 24.921875, "mean_p": 0.17165, "boundary_entropy": 0.095349, "bpb": 1.685771}, "fi": {"chunks": 3418.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181702, "cps": 26.703125, "mean_p": 0.179207, "boundary_entropy": 0.071549, "bpb": 1.850378}, "ru": {"chunks": 3713.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.195937, "cps": 29.007812, "mean_p": 0.136515, "boundary_entropy": 0.166537, "bpb": 1.090048}, "el": {"chunks": 3501.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.170589, "cps": 27.351562, "mean_p": 0.127846, "boundary_entropy": 0.182975, "bpb": 1.098511}, "hi": {"chunks": 3760.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.223477, "cps": 29.375, "mean_p": 0.107167, "boundary_entropy": 0.131798, "bpb": 0.900034}, "ta": {"chunks": 3340.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.178657, "cps": 26.09375, "mean_p": 0.095724, "boundary_entropy": 0.154294, "bpb": 0.786874}, "th": {"chunks": 3911.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.243146, "cps": 30.554688, "mean_p": 0.107638, "boundary_entropy": 0.132238, "bpb": 0.864991}, "zh": {"chunks": 3030.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.513038, "cps": 23.671875, "mean_p": 0.217144, "boundary_entropy": 0.128827, "bpb": 2.475271}, "my": {"chunks": 1702.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.089825, "cps": 13.296875, "mean_p": 0.080738, "boundary_entropy": 0.196556, "bpb": 0.876911}}, "bpb": {"en": 1.65131, "de": 1.7806, "vi": 1.45923, "tr": 1.68577, "fi": 1.85038, "ru": 1.09005, "el": 1.09851, "hi": 0.90003, "ta": 0.78687, "th": 0.86499, "zh": 2.47527, "my": 0.87691}, "gini_cpc": 0.190303, "gini_cps": 0.07857, "bpic": 9.654868, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9710273265002531, "vi": 0.9800215232028113, "tr": 0.9586277833570768, "fi": 0.9884849355118335, "ru": 1.0659224642265352, "el": 0.9280281410115782, "hi": 1.2157454479413377, "ta": 0.9719208020091195, "th": 1.3227465297235508, "zh": 2.790995048295972, "my": 0.48865917765229433}, "t": 2229.7} |
| {"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.64834, "de": 1.78541, "vi": 1.47488, "tr": 1.67919, "fi": 1.86342, "ru": 1.08946, "el": 1.09785, "hi": 0.90598, "ta": 0.79102, "th": 0.86357, "zh": 2.47835, "my": 0.88305}, "bytes_per_chunk": {"en": 5.186, "de": 5.614, "vi": 7.142, "tr": 5.82, "fi": 5.726, "ru": 8.981, "el": 10.664, "hi": 11.281, "ta": 15.018, "th": 11.687, "zh": 5.346, "my": 27.324}, "bpb_mean": 1.38004, "t": 2235.6} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3199.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183819, "cps": 24.992188, "mean_p": 0.187443, "boundary_entropy": 0.054951, "bpb": 1.651314}, "de": {"chunks": 3665.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178493, "cps": 28.632812, "mean_p": 0.176327, "boundary_entropy": 0.069887, "bpb": 1.780604}, "vi": {"chunks": 3321.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.180146, "cps": 25.945312, "mean_p": 0.162964, "boundary_entropy": 0.139354, "bpb": 1.459227}, "tr": {"chunks": 3190.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176214, "cps": 24.921875, "mean_p": 0.17165, "boundary_entropy": 0.095349, "bpb": 1.685771}, "fi": {"chunks": 3418.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181702, "cps": 26.703125, "mean_p": 0.179207, "boundary_entropy": 0.071549, "bpb": 1.850378}, "ru": {"chunks": 3713.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.195937, "cps": 29.007812, "mean_p": 0.136515, "boundary_entropy": 0.166537, "bpb": 1.090048}, "el": {"chunks": 3501.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.170589, "cps": 27.351562, "mean_p": 0.127846, "boundary_entropy": 0.182975, "bpb": 1.098511}, "hi": {"chunks": 3760.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.223477, "cps": 29.375, "mean_p": 0.107167, "boundary_entropy": 0.131798, "bpb": 0.900034}, "ta": {"chunks": 3340.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.178657, "cps": 26.09375, "mean_p": 0.095724, "boundary_entropy": 0.154294, "bpb": 0.786874}, "th": {"chunks": 3911.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.243146, "cps": 30.554688, "mean_p": 0.107638, "boundary_entropy": 0.132238, "bpb": 0.864991}, "zh": {"chunks": 3030.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.513038, "cps": 23.671875, "mean_p": 0.217144, "boundary_entropy": 0.128827, "bpb": 2.475271}, "my": {"chunks": 1702.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.089825, "cps": 13.296875, "mean_p": 0.080738, "boundary_entropy": 0.196556, "bpb": 0.876911}}, "bpb": {"en": 1.65131, "de": 1.7806, "vi": 1.45923, "tr": 1.68577, "fi": 1.85038, "ru": 1.09005, "el": 1.09851, "hi": 0.90003, "ta": 0.78687, "th": 0.86499, "zh": 2.47527, "my": 0.87691}, "gini_cpc": 0.190303, "gini_cps": 0.07857, "bpic": 9.654868, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9710273265002531, "vi": 0.9800215232028113, "tr": 0.9586277833570768, "fi": 0.9884849355118335, "ru": 1.0659224642265352, "el": 0.9280281410115782, "hi": 1.2157454479413377, "ta": 0.9719208020091195, "th": 1.3227465297235508, "zh": 2.790995048295972, "my": 0.48865917765229433}, "t": 2238.7} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s2/th0.70"], "t": 2241.9} |
| {"event": "final", "run": "spec_E_s2", "condition": "E", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 0.18381888180198816, "cps": 24.9921875, "bpb": 1.6438765837477893, "bytes_per_chunk": 5.193541357107479}, "de": {"cpc": 0.1784931573564506, "cps": 28.6328125, "bpb": 1.7824091142168186, "bytes_per_chunk": 5.5738851708722486}, "vi": {"cpc": 0.18014646053702196, "cps": 25.9453125, "bpb": 1.4739747294799386, "bytes_per_chunk": 7.134480482809768}, "tr": {"cpc": 0.1762138872010164, "cps": 24.921875, "bpb": 1.6789048384437173, "bytes_per_chunk": 5.853976619287413}, "fi": {"cpc": 0.1817021955238956, "cps": 26.703125, "bpb": 1.8691479514912825, "bytes_per_chunk": 5.733808700977712}, "ru": {"cpc": 0.19593667546174143, "cps": 29.0078125, "bpb": 1.0907239198933933, "bytes_per_chunk": 9.00026608300073}, "el": {"cpc": 0.1705890951615261, "cps": 27.3515625, "bpb": 1.095154461702511, "bytes_per_chunk": 10.631410321403225}, "hi": {"cpc": 0.2234769687964339, "cps": 29.375, "bpb": 0.907511446864184, "bytes_per_chunk": 11.288362579394983}, "ta": {"cpc": 0.17865739502540787, "cps": 26.09375, "bpb": 0.7897428494116536, "bytes_per_chunk": 14.990150248030764}, "th": {"cpc": 0.2431457880012434, "cps": 30.5546875, "bpb": 0.8658079972935708, "bytes_per_chunk": 11.621403555437336}, "zh": {"cpc": 0.5130375888926515, "cps": 23.671875, "bpb": 2.479610840172693, "bytes_per_chunk": 5.378995480637533}, "my": {"cpc": 0.08982478361832383, "cps": 13.296875, "bpb": 0.8824134811779903, "bytes_per_chunk": 27.510126980795466}}, "gini_cpc": 0.19030326488159188, "gini_cps": 0.07857023060796653, "bpic": 9.654867924528302, "seed": 2, "mean_boundary_entropy": 0.10425905883312225, "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_b", "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": 2252.7} |
|
|
