| {"event": "start", "run": "spec_E_s0", "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": 2.0} |
| {"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": 2.0} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8739.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.502155, "cps": 68.273438, "mean_p": 0.50363, "boundary_entropy": 0.687415, "bpb": 8.248358}, "de": {"chunks": 12222.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.595237, "cps": 95.484375, "mean_p": 0.507356, "boundary_entropy": 0.688163, "bpb": 8.204506}, "vi": {"chunks": 6386.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.346406, "cps": 49.890625, "mean_p": 0.491233, "boundary_entropy": 0.688507, "bpb": 8.13873}, "tr": {"chunks": 11403.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.629896, "cps": 89.085938, "mean_p": 0.50659, "boundary_entropy": 0.687851, "bpb": 8.143105}, "fi": {"chunks": 7137.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.379406, "cps": 55.757812, "mean_p": 0.497164, "boundary_entropy": 0.68792, "bpb": 8.234202}, "ru": {"chunks": 10473.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.552665, "cps": 81.820312, "mean_p": 0.496692, "boundary_entropy": 0.690192, "bpb": 8.071624}, "el": {"chunks": 36555.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.781172, "cps": 285.585938, "mean_p": 0.530217, "boundary_entropy": 0.688772, "bpb": 8.142597}, "hi": {"chunks": 42584.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.530996, "cps": 332.6875, "mean_p": 0.521073, "boundary_entropy": 0.69018, "bpb": 7.880816}, "ta": {"chunks": 49758.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.661567, "cps": 388.734375, "mean_p": 0.521091, "boundary_entropy": 0.690409, "bpb": 7.814466}, "th": {"chunks": 45716.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.842151, "cps": 357.15625, "mean_p": 0.525713, "boundary_entropy": 0.689893, "bpb": 8.063302}, "zh": {"chunks": 6218.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.052828, "cps": 48.578125, "mean_p": 0.496822, "boundary_entropy": 0.686552, "bpb": 8.034403}, "my": {"chunks": 53266.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.811167, "cps": 416.140625, "mean_p": 0.521325, "boundary_entropy": 0.690575, "bpb": 8.24483}}, "bpb": {"en": 8.24836, "de": 8.20451, "vi": 8.13873, "tr": 8.1431, "fi": 8.2342, "ru": 8.07162, "el": 8.1426, "hi": 7.88082, "ta": 7.81447, "th": 8.0633, "zh": 8.0344, "my": 8.24483}, "gini_cpc": 0.391191, "gini_cps": 0.409188, "bpic": 1.321301, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1853654183968407, "vi": 0.6898396505624507, "tr": 1.2543852937189204, "fi": 0.755555191784985, "ru": 1.1005867247811874, "el": 3.54705827773647, "hi": 5.040269530092341, "ta": 5.300292378412309, "th": 5.659910185771798, "zh": 2.096619669942964, "my": 5.598208760550946}, "t": 9.2} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.63905, "bpb": 8.13544, "chunk_loss": 0.057711, "F": 0.53552, "G": 0.50248, "chunks": 548.38, "bytes_per_chunk": 1.867, "boundary_entropy": 0.69172, "router_noise": 0.0, "grad_norm": 9.2841, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 9.9} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8721.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.50112, "cps": 68.132812, "mean_p": 0.503564, "boundary_entropy": 0.687414, "bpb": 8.209979}, "de": {"chunks": 12187.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.593532, "cps": 95.210938, "mean_p": 0.507262, "boundary_entropy": 0.688164, "bpb": 8.167545}, "vi": {"chunks": 6367.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.345376, "cps": 49.742188, "mean_p": 0.491184, "boundary_entropy": 0.688504, "bpb": 8.11921}, "tr": {"chunks": 11366.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.627852, "cps": 88.796875, "mean_p": 0.506498, "boundary_entropy": 0.687851, "bpb": 8.116503}, "fi": {"chunks": 7084.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.376588, "cps": 55.34375, "mean_p": 0.497024, "boundary_entropy": 0.687916, "bpb": 8.205208}, "ru": {"chunks": 10506.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.554406, "cps": 82.078125, "mean_p": 0.496728, "boundary_entropy": 0.690193, "bpb": 8.027905}, "el": {"chunks": 36547.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.780783, "cps": 285.523438, "mean_p": 0.53005, "boundary_entropy": 0.688791, "bpb": 8.124618}, "hi": {"chunks": 42563.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.529747, "cps": 332.523438, "mean_p": 0.520975, "boundary_entropy": 0.690188, "bpb": 7.857681}, "ta": {"chunks": 49697.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.658304, "cps": 388.257812, "mean_p": 0.520856, "boundary_entropy": 0.690427, "bpb": 7.780756}, "th": {"chunks": 45715.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.842089, "cps": 357.148438, "mean_p": 0.525586, "boundary_entropy": 0.689906, "bpb": 8.031105}, "zh": {"chunks": 6210.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.051473, "cps": 48.515625, "mean_p": 0.496796, "boundary_entropy": 0.686551, "bpb": 8.029641}, "my": {"chunks": 53259.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.810798, "cps": 416.085938, "mean_p": 0.521207, "boundary_entropy": 0.690584, "bpb": 8.216883}}, "bpb": {"en": 8.20998, "de": 8.16755, "vi": 8.11921, "tr": 8.1165, "fi": 8.20521, "ru": 8.0279, "el": 8.12462, "hi": 7.85768, "ta": 7.78076, "th": 8.0311, "zh": 8.02964, "my": 8.21688}, "gini_cpc": 0.391544, "gini_cps": 0.40948, "bpic": 1.32237, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.184410469585478, "vi": 0.6892067807856777, "tr": 1.2528957440549502, "fi": 0.7514922441887808, "ru": 1.1063333796232004, "el": 3.553601477537154, "hi": 5.048181859537694, "ta": 5.304720876916054, "th": 5.671468085513631, "zh": 2.098244006462442, "my": 5.609026158823028}, "t": 16.3} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8647.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.496868, "cps": 67.554688, "mean_p": 0.503388, "boundary_entropy": 0.687412, "bpb": 8.133911}, "de": {"chunks": 12053.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.587006, "cps": 94.164062, "mean_p": 0.507025, "boundary_entropy": 0.688166, "bpb": 8.095631}, "vi": {"chunks": 6328.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.34326, "cps": 49.4375, "mean_p": 0.491077, "boundary_entropy": 0.688497, "bpb": 8.079286}, "tr": {"chunks": 11287.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.623488, "cps": 88.179688, "mean_p": 0.506283, "boundary_entropy": 0.687853, "bpb": 8.064897}, "fi": {"chunks": 6961.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.370049, "cps": 54.382812, "mean_p": 0.496717, "boundary_entropy": 0.687908, "bpb": 8.148054}, "ru": {"chunks": 10555.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.556992, "cps": 82.460938, "mean_p": 0.496761, "boundary_entropy": 0.690194, "bpb": 7.949542}, "el": {"chunks": 36519.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.779418, "cps": 285.304688, "mean_p": 0.529732, "boundary_entropy": 0.688827, "bpb": 8.089722}, "hi": {"chunks": 42523.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.52737, "cps": 332.210938, "mean_p": 0.520726, "boundary_entropy": 0.690207, "bpb": 7.805685}, "ta": {"chunks": 49565.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.651244, "cps": 387.226562, "mean_p": 0.520379, "boundary_entropy": 0.690465, "bpb": 7.71433}, "th": {"chunks": 45709.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.841716, "cps": 357.101562, "mean_p": 0.525309, "boundary_entropy": 0.689933, "bpb": 7.952337}, "zh": {"chunks": 6194.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.048764, "cps": 48.390625, "mean_p": 0.496744, "boundary_entropy": 0.68655, "bpb": 8.018816}, "my": {"chunks": 53244.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.810006, "cps": 415.96875, "mean_p": 0.520993, "boundary_entropy": 0.690601, "bpb": 8.175895}}, "bpb": {"en": 8.13391, "de": 8.09563, "vi": 8.07929, "tr": 8.0649, "fi": 8.14805, "ru": 7.94954, "el": 8.08972, "hi": 7.80568, "ta": 7.71433, "th": 7.95234, "zh": 8.01882, "my": 8.17589}, "gini_cpc": 0.392606, "gini_cps": 0.410345, "bpic": 1.325279, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.181412089228333, "vi": 0.6908471809456579, "tr": 1.2548350325212356, "fi": 0.7447635468562069, "ru": 1.121005348141851, "el": 3.5812669334127776, "hi": 5.0865988032163285, "ta": 5.335907621919715, "th": 5.719253097674589, "zh": 2.1107481611904415, "my": 5.655434279556627}, "t": 23.3} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8434.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.484629, "cps": 65.890625, "mean_p": 0.502847, "boundary_entropy": 0.687406, "bpb": 7.880243}, "de": {"chunks": 11670.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.568353, "cps": 91.171875, "mean_p": 0.506293, "boundary_entropy": 0.688173, "bpb": 7.857328}, "vi": {"chunks": 6156.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.33393, "cps": 48.09375, "mean_p": 0.490665, "boundary_entropy": 0.688473, "bpb": 7.947298}, "tr": {"chunks": 11010.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.608186, "cps": 86.015625, "mean_p": 0.505596, "boundary_entropy": 0.687857, "bpb": 7.887249}, "fi": {"chunks": 6503.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.345702, "cps": 50.804688, "mean_p": 0.495724, "boundary_entropy": 0.687877, "bpb": 7.953905}, "ru": {"chunks": 10725.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.565963, "cps": 83.789062, "mean_p": 0.49689, "boundary_entropy": 0.690197, "bpb": 7.714065}, "el": {"chunks": 36341.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.770745, "cps": 283.914062, "mean_p": 0.528228, "boundary_entropy": 0.688991, "bpb": 7.945168}, "hi": {"chunks": 42405.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.520357, "cps": 331.289062, "mean_p": 0.519957, "boundary_entropy": 0.690263, "bpb": 7.626611}, "ta": {"chunks": 49032.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.622733, "cps": 383.0625, "mean_p": 0.518699, "boundary_entropy": 0.69059, "bpb": 7.463956}, "th": {"chunks": 45684.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.840162, "cps": 356.90625, "mean_p": 0.524343, "boundary_entropy": 0.690024, "bpb": 7.684668}, "zh": {"chunks": 6137.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.039113, "cps": 47.945312, "mean_p": 0.496565, "boundary_entropy": 0.686548, "bpb": 7.983365}, "my": {"chunks": 53171.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.806154, "cps": 415.398438, "mean_p": 0.520204, "boundary_entropy": 0.690662, "bpb": 8.006459}}, "bpb": {"en": 7.88024, "de": 7.85733, "vi": 7.9473, "tr": 7.88725, "fi": 7.9539, "ru": 7.71406, "el": 7.94517, "hi": 7.62661, "ta": 7.46396, "th": 7.68467, "zh": 7.98336, "my": 8.00646}, "gini_cpc": 0.396266, "gini_cps": 0.413345, "bpic": 1.335969, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.172759534409874, "vi": 0.6890424727067569, "tr": 1.2549525086674018, "fi": 0.7133331287765762, "ru": 1.1678272640643508, "el": 3.6538149803783537, "hi": 5.200588821655501, "ta": 5.41183643707575, "th": 5.860485302728001, "zh": 2.144140322325877, "my": 5.79031213642179}, "t": 30.9} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 7617.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.437683, "cps": 59.507812, "mean_p": 0.500778, "boundary_entropy": 0.68737, "bpb": 7.154396}, "de": {"chunks": 10355.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.50431, "cps": 80.898438, "mean_p": 0.503632, "boundary_entropy": 0.688184, "bpb": 7.186435}, "vi": {"chunks": 5625.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.305126, "cps": 43.945312, "mean_p": 0.489195, "boundary_entropy": 0.688385, "bpb": 7.546882}, "tr": {"chunks": 10041.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.554659, "cps": 78.445312, "mean_p": 0.503187, "boundary_entropy": 0.687852, "bpb": 7.361735}, "fi": {"chunks": 5055.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.268726, "cps": 39.492188, "mean_p": 0.4921, "boundary_entropy": 0.687733, "bpb": 7.360859}, "ru": {"chunks": 12134.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.640317, "cps": 94.796875, "mean_p": 0.498125, "boundary_entropy": 0.690223, "bpb": 6.973027}, "el": {"chunks": 35198.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.715051, "cps": 274.984375, "mean_p": 0.52275, "boundary_entropy": 0.689513, "bpb": 7.476359}, "hi": {"chunks": 41446.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.463358, "cps": 323.796875, "mean_p": 0.516814, "boundary_entropy": 0.690472, "bpb": 7.075459}, "ta": {"chunks": 44747.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.393528, "cps": 349.585938, "mean_p": 0.512876, "boundary_entropy": 0.690936, "bpb": 6.797428}, "th": {"chunks": 45452.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.825738, "cps": 355.09375, "mean_p": 0.520859, "boundary_entropy": 0.690323, "bpb": 6.953516}, "zh": {"chunks": 5984.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.013207, "cps": 46.75, "mean_p": 0.496043, "boundary_entropy": 0.686543, "bpb": 7.87701}, "my": {"chunks": 52940.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.793962, "cps": 413.59375, "mean_p": 0.51819, "boundary_entropy": 0.690804, "bpb": 7.572345}}, "bpb": {"en": 7.1544, "de": 7.18643, "vi": 7.54688, "tr": 7.36173, "fi": 7.36086, "ru": 6.97303, "el": 7.47636, "hi": 7.07546, "ta": 6.79743, "th": 6.95352, "zh": 7.87701, "my": 7.57234}, "gini_cpc": 0.408437, "gini_cps": 0.422675, "bpic": 1.387525, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1522265035772634, "vi": 0.6971392733888363, "tr": 1.2672624899439267, "fi": 0.6139732375254362, "ru": 1.4629683844947579, "el": 3.918477040054025, "hi": 5.628176576169556, "ta": 5.4686309880388535, "th": 6.456127482457989, "zh": 2.3149323649609936, "my": 6.3835273802982115}, "t": 39.3} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4511.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.259208, "cps": 35.242188, "mean_p": 0.492306, "boundary_entropy": 0.687145, "bpb": 6.329881}, "de": {"chunks": 6199.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.301904, "cps": 48.429688, "mean_p": 0.49467, "boundary_entropy": 0.6881, "bpb": 6.404737}, "vi": {"chunks": 3432.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.186168, "cps": 26.8125, "mean_p": 0.481712, "boundary_entropy": 0.687844, "bpb": 7.023443}, "tr": {"chunks": 6314.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.348782, "cps": 49.328125, "mean_p": 0.493769, "boundary_entropy": 0.687686, "bpb": 6.709007}, "fi": {"chunks": 2446.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.13003, "cps": 19.109375, "mean_p": 0.482989, "boundary_entropy": 0.687182, "bpb": 6.607137}, "ru": {"chunks": 14753.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.778522, "cps": 115.257812, "mean_p": 0.500097, "boundary_entropy": 0.690296, "bpb": 5.842038}, "el": {"chunks": 19915.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.970375, "cps": 155.585938, "mean_p": 0.50323, "boundary_entropy": 0.690397, "bpb": 6.52067}, "hi": {"chunks": 27446.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.631263, "cps": 214.421875, "mean_p": 0.504117, "boundary_entropy": 0.690918, "bpb": 5.801911}, "ta": {"chunks": 7765.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.415352, "cps": 60.664062, "mean_p": 0.490868, "boundary_entropy": 0.691014, "bpb": 5.849646}, "th": {"chunks": 26102.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.622754, "cps": 203.921875, "mean_p": 0.502674, "boundary_entropy": 0.691089, "bpb": 5.339468}, "zh": {"chunks": 5223.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.884355, "cps": 40.804688, "mean_p": 0.493447, "boundary_entropy": 0.686483, "bpb": 7.501185}, "my": {"chunks": 49048.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.588558, "cps": 383.1875, "mean_p": 0.510261, "boundary_entropy": 0.691206, "bpb": 6.227835}}, "bpb": {"en": 6.32988, "de": 6.40474, "vi": 7.02344, "tr": 6.70901, "fi": 6.60714, "ru": 5.84204, "el": 6.52067, "hi": 5.80191, "ta": 5.84965, "th": 5.33947, "zh": 7.50119, "my": 6.22783}, "gini_cpc": 0.452252, "gini_cps": 0.471862, "bpic": 2.216414, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1647172893377955, "vi": 0.7182165861985682, "tr": 1.345566974310025, "fi": 0.5016442774554444, "ru": 3.003463933662013, "el": 3.743611379106721, "hi": 6.2932542706404, "ta": 1.6023865231161487, "th": 6.260427827201246, "zh": 3.4117553111508365, "my": 9.98640603947651}, "t": 48.4} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5963.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.342642, "cps": 46.585938, "mean_p": 0.498478, "boundary_entropy": 0.687578, "bpb": 5.765791}, "de": {"chunks": 8123.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.395607, "cps": 63.460938, "mean_p": 0.499943, "boundary_entropy": 0.688533, "bpb": 5.889612}, "vi": {"chunks": 3478.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.188663, "cps": 27.171875, "mean_p": 0.484959, "boundary_entropy": 0.688277, "bpb": 6.443155}, "tr": {"chunks": 5525.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.305198, "cps": 43.164062, "mean_p": 0.494109, "boundary_entropy": 0.687971, "bpb": 6.245065}, "fi": {"chunks": 4403.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.234065, "cps": 34.398438, "mean_p": 0.493004, "boundary_entropy": 0.688004, "bpb": 5.942957}, "ru": {"chunks": 2274.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.12, "cps": 17.765625, "mean_p": 0.48876, "boundary_entropy": 0.690077, "bpb": 5.230939}, "el": {"chunks": 489.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.023827, "cps": 3.820312, "mean_p": 0.466873, "boundary_entropy": 0.688049, "bpb": 6.167507}, "hi": {"chunks": 6171.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.366776, "cps": 48.210938, "mean_p": 0.492965, "boundary_entropy": 0.690801, "bpb": 4.970178}, "ta": {"chunks": 4258.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.227761, "cps": 33.265625, "mean_p": 0.489036, "boundary_entropy": 0.690961, "bpb": 5.234655}, "th": {"chunks": 2586.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.160771, "cps": 20.203125, "mean_p": 0.488684, "boundary_entropy": 0.690772, "bpb": 4.935033}, "zh": {"chunks": 2796.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.473417, "cps": 21.84375, "mean_p": 0.484969, "boundary_entropy": 0.686152, "bpb": 7.083276}, "my": {"chunks": 4938.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.260608, "cps": 38.578125, "mean_p": 0.491034, "boundary_entropy": 0.691175, "bpb": 4.780113}}, "bpb": {"en": 5.76579, "de": 5.88961, "vi": 6.44316, "tr": 6.24506, "fi": 5.94296, "ru": 5.23094, "el": 6.16751, "hi": 4.97018, "ta": 5.23466, "th": 4.93503, "zh": 7.08328, "my": 4.78011}, "gini_cpc": 0.266098, "gini_cps": 0.262816, "bpic": 7.524527, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1545782099730428, "vi": 0.5506120943540224, "tr": 0.8907196673767266, "fi": 0.6831186037428031, "ru": 0.3502196880764716, "el": 0.06953882126938662, "hi": 1.070433727152009, "ta": 0.6647211517471766, "th": 0.46920946708411326, "zh": 1.3816658876027408, "my": 0.7605837114390334}, "t": 58.2} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6278.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.360742, "cps": 49.046875, "mean_p": 0.498933, "boundary_entropy": 0.687615, "bpb": 5.367533}, "de": {"chunks": 6359.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.309697, "cps": 49.679688, "mean_p": 0.496434, "boundary_entropy": 0.688458, "bpb": 5.511446}, "vi": {"chunks": 4580.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.24844, "cps": 35.78125, "mean_p": 0.489383, "boundary_entropy": 0.688666, "bpb": 5.902446}, "tr": {"chunks": 4849.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.267856, "cps": 37.882812, "mean_p": 0.493172, "boundary_entropy": 0.688013, "bpb": 5.846451}, "fi": {"chunks": 6335.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.336771, "cps": 49.492188, "mean_p": 0.497324, "boundary_entropy": 0.688182, "bpb": 5.549653}, "ru": {"chunks": 5339.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.281741, "cps": 41.710938, "mean_p": 0.495136, "boundary_entropy": 0.690389, "bpb": 4.832255}, "el": {"chunks": 1397.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.06807, "cps": 10.914062, "mean_p": 0.488623, "boundary_entropy": 0.69031, "bpb": 5.109341}, "hi": {"chunks": 9606.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.570936, "cps": 75.046875, "mean_p": 0.495962, "boundary_entropy": 0.690916, "bpb": 4.440451}, "ta": {"chunks": 13969.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.747205, "cps": 109.132812, "mean_p": 0.494263, "boundary_entropy": 0.69114, "bpb": 4.371307}, "th": {"chunks": 4690.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.291576, "cps": 36.640625, "mean_p": 0.491511, "boundary_entropy": 0.690892, "bpb": 4.441501}, "zh": {"chunks": 1604.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.271588, "cps": 12.53125, "mean_p": 0.484307, "boundary_entropy": 0.686404, "bpb": 6.715434}, "my": {"chunks": 20104.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.061009, "cps": 157.0625, "mean_p": 0.49907, "boundary_entropy": 0.691398, "bpb": 4.24241}}, "bpb": {"en": 5.36753, "de": 5.51145, "vi": 5.90245, "tr": 5.84645, "fi": 5.54965, "ru": 4.83226, "el": 5.10934, "hi": 4.44045, "ta": 4.37131, "th": 4.4415, "zh": 6.71543, "my": 4.24241}, "gini_cpc": 0.314285, "gini_cps": 0.358346, "bpic": 4.509235, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8584978792404466, "vi": 0.6886921692526188, "tr": 0.742513649329928, "fi": 0.9335499168816467, "ru": 0.7810044625407988, "el": 0.188694121174734, "hi": 1.5826698104231527, "ta": 2.0712983379258993, "th": 0.8082665134242732, "zh": 0.7528591449756497, "my": 2.941182060417461}, "t": 73.2} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.2982, "bpb": 4.75829, "chunk_loss": 0.030763, "F": 0.17767, "G": 0.47269, "chunks": 181.94, "bytes_per_chunk": 5.628, "boundary_entropy": 0.68858, "router_noise": 0.0, "grad_norm": 2.3862, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.732, "alpha": 0.03, "t": 83.2} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3732.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.214446, "cps": 29.15625, "mean_p": 0.458767, "boundary_entropy": 0.662117, "bpb": 4.291413}, "de": {"chunks": 4339.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.211318, "cps": 33.898438, "mean_p": 0.45932, "boundary_entropy": 0.666917, "bpb": 4.527998}, "vi": {"chunks": 7340.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.398156, "cps": 57.34375, "mean_p": 0.492384, "boundary_entropy": 0.683585, "bpb": 4.684704}, "tr": {"chunks": 4260.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.23532, "cps": 33.28125, "mean_p": 0.477719, "boundary_entropy": 0.677827, "bpb": 4.833567}, "fi": {"chunks": 4271.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.227048, "cps": 33.367188, "mean_p": 0.471951, "boundary_entropy": 0.677222, "bpb": 4.556088}, "ru": {"chunks": 5165.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.272559, "cps": 40.351562, "mean_p": 0.482529, "boundary_entropy": 0.688097, "bpb": 3.213937}, "el": {"chunks": 6354.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.309604, "cps": 49.640625, "mean_p": 0.49126, "boundary_entropy": 0.690299, "bpb": 4.144032}, "hi": {"chunks": 8651.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.514175, "cps": 67.585938, "mean_p": 0.470733, "boundary_entropy": 0.687952, "bpb": 2.865444}, "ta": {"chunks": 11804.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.631399, "cps": 92.21875, "mean_p": 0.437869, "boundary_entropy": 0.678841, "bpb": 2.945545}, "th": {"chunks": 7947.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.494063, "cps": 62.085938, "mean_p": 0.478204, "boundary_entropy": 0.689027, "bpb": 2.740424}, "zh": {"chunks": 2170.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.367423, "cps": 16.953125, "mean_p": 0.486613, "boundary_entropy": 0.686562, "bpb": 6.221186}, "my": {"chunks": 11082.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.584864, "cps": 86.578125, "mean_p": 0.476001, "boundary_entropy": 0.689399, "bpb": 2.909095}}, "bpb": {"en": 4.29141, "de": 4.528, "vi": 4.6847, "tr": 4.83357, "fi": 4.55609, "ru": 3.21394, "el": 4.14403, "hi": 2.86544, "ta": 2.94554, "th": 2.74042, "zh": 6.22119, "my": 2.90909}, "gini_cpc": 0.218458, "gini_cps": 0.250095, "bpic": 4.976736, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9854162689720868, "vi": 1.8566729196263572, "tr": 1.0973408148762183, "fi": 1.058766455147592, "ru": 1.2709942817874078, "el": 1.4437395390286765, "hi": 2.3976938218754626, "ta": 2.9443281858295585, "th": 2.3039053480954776, "zh": 1.7133606022759567, "my": 2.7273272696912247}, "t": 93.0} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.44623, "bpb": 3.52917, "chunk_loss": 0.032049, "F": 0.20293, "G": 0.41058, "chunks": 207.8, "bytes_per_chunk": 4.928, "boundary_entropy": 0.59341, "router_noise": 0.0, "grad_norm": 1.8597, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.247, "alpha": 0.03, "t": 107.8} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2952.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.169626, "cps": 23.0625, "mean_p": 0.334738, "boundary_entropy": 0.45096, "bpb": 3.520911}, "de": {"chunks": 3305.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.16096, "cps": 25.820312, "mean_p": 0.343204, "boundary_entropy": 0.494653, "bpb": 3.703863}, "vi": {"chunks": 5555.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.301329, "cps": 43.398438, "mean_p": 0.319074, "boundary_entropy": 0.40809, "bpb": 3.090574}, "tr": {"chunks": 3560.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.196652, "cps": 27.8125, "mean_p": 0.337199, "boundary_entropy": 0.498816, "bpb": 3.684948}, "fi": {"chunks": 2933.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155919, "cps": 22.914062, "mean_p": 0.327033, "boundary_entropy": 0.500968, "bpb": 3.75938}, "ru": {"chunks": 8080.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.426385, "cps": 63.125, "mean_p": 0.293318, "boundary_entropy": 0.489856, "bpb": 2.607481}, "el": {"chunks": 8019.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.390732, "cps": 62.648438, "mean_p": 0.353554, "boundary_entropy": 0.569074, "bpb": 2.757733}, "hi": {"chunks": 7346.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.436612, "cps": 57.390625, "mean_p": 0.422035, "boundary_entropy": 0.651895, "bpb": 2.061176}, "ta": {"chunks": 11827.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.632629, "cps": 92.398438, "mean_p": 0.395905, "boundary_entropy": 0.645123, "bpb": 1.932285}, "th": {"chunks": 5167.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.321231, "cps": 40.367188, "mean_p": 0.451679, "boundary_entropy": 0.672064, "bpb": 2.069784}, "zh": {"chunks": 3945.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.667965, "cps": 30.820312, "mean_p": 0.371657, "boundary_entropy": 0.605627, "bpb": 4.884461}, "my": {"chunks": 12894.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.680494, "cps": 100.734375, "mean_p": 0.306195, "boundary_entropy": 0.556552, "bpb": 2.136331}}, "bpb": {"en": 3.52091, "de": 3.70386, "vi": 3.09057, "tr": 3.68495, "fi": 3.75938, "ru": 2.60748, "el": 2.75773, "hi": 2.06118, "ta": 1.93228, "th": 2.06978, "zh": 4.88446, "my": 2.13633}, "gini_cpc": 0.279315, "gini_cps": 0.281983, "bpic": 5.07761, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9489139335093736, "vi": 1.7764324113949033, "tr": 1.1593303719591912, "fi": 0.9191956207080013, "ru": 2.5136795589591783, "el": 2.303494265208695, "hi": 2.5739708138537556, "ta": 3.7295539544724146, "th": 1.8937609776315358, "zh": 3.9378696117218537, "my": 4.01173333178867}, "t": 121.7} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.13998, "bpb": 3.08734, "chunk_loss": 0.031005, "F": 0.20123, "G": 0.31156, "chunks": 206.06, "bytes_per_chunk": 4.969, "boundary_entropy": 0.43723, "router_noise": 0.0, "grad_norm": 1.343, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.232, "alpha": 0.03, "t": 131.0} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.06561, "bpb": 2.98005, "chunk_loss": 0.031004, "F": 0.19702, "G": 0.30989, "chunks": 201.75, "bytes_per_chunk": 5.076, "boundary_entropy": 0.41303, "router_noise": 0.0, "grad_norm": 1.0733, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 150.4} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.95868, "bpb": 2.82578, "chunk_loss": 0.03093, "F": 0.20061, "G": 0.28002, "chunks": 205.42, "bytes_per_chunk": 4.985, "boundary_entropy": 0.34434, "router_noise": 0.0, "grad_norm": 0.9973, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.186, "alpha": 0.03, "t": 169.0} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.82797, "bpb": 2.6372, "chunk_loss": 0.030647, "F": 0.19881, "G": 0.24575, "chunks": 203.58, "bytes_per_chunk": 5.03, "boundary_entropy": 0.26118, "router_noise": 0.0, "grad_norm": 0.8066, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 187.3} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.7077, "bpb": 2.46368, "chunk_loss": 0.030441, "F": 0.1936, "G": 0.24172, "chunks": 198.25, "bytes_per_chunk": 5.165, "boundary_entropy": 0.24782, "router_noise": 0.0, "grad_norm": 0.7431, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 205.4} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.58913, "bpb": 2.29263, "chunk_loss": 0.030357, "F": 0.19032, "G": 0.22894, "chunks": 194.89, "bytes_per_chunk": 5.254, "boundary_entropy": 0.21168, "router_noise": 0.0, "grad_norm": 0.6689, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 223.5} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.56298, "bpb": 2.2549, "chunk_loss": 0.030316, "F": 0.1907, "G": 0.23526, "chunks": 195.28, "bytes_per_chunk": 5.244, "boundary_entropy": 0.21702, "router_noise": 0.0, "grad_norm": 0.6357, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 241.6} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3601.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.206918, "cps": 28.132812, "mean_p": 0.22023, "boundary_entropy": 0.123755, "bpb": 2.323567}, "de": {"chunks": 4040.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196756, "cps": 31.5625, "mean_p": 0.214797, "boundary_entropy": 0.140896, "bpb": 2.60014}, "vi": {"chunks": 4356.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23629, "cps": 34.03125, "mean_p": 0.222912, "boundary_entropy": 0.177931, "bpb": 2.096501}, "tr": {"chunks": 3850.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212672, "cps": 30.078125, "mean_p": 0.212729, "boundary_entropy": 0.173675, "bpb": 2.593663}, "fi": {"chunks": 3399.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180692, "cps": 26.554688, "mean_p": 0.211117, "boundary_entropy": 0.200005, "bpb": 2.841025}, "ru": {"chunks": 6488.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.342375, "cps": 50.6875, "mean_p": 0.208224, "boundary_entropy": 0.151012, "bpb": 1.613231}, "el": {"chunks": 6965.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.339375, "cps": 54.414062, "mean_p": 0.208171, "boundary_entropy": 0.164925, "bpb": 1.71604}, "hi": {"chunks": 8783.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.522021, "cps": 68.617188, "mean_p": 0.306691, "boundary_entropy": 0.417472, "bpb": 1.403107}, "ta": {"chunks": 8805.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.470982, "cps": 68.789062, "mean_p": 0.26315, "boundary_entropy": 0.450113, "bpb": 1.144169}, "th": {"chunks": 6662.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.414175, "cps": 52.046875, "mean_p": 0.274677, "boundary_entropy": 0.465504, "bpb": 1.387147}, "zh": {"chunks": 3406.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.576702, "cps": 26.609375, "mean_p": 0.245625, "boundary_entropy": 0.196459, "bpb": 3.460919}, "my": {"chunks": 7578.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.399937, "cps": 59.203125, "mean_p": 0.305395, "boundary_entropy": 0.527552, "bpb": 1.291081}}, "bpb": {"en": 2.32357, "de": 2.60014, "vi": 2.0965, "tr": 2.59366, "fi": 2.84102, "ru": 1.61323, "el": 1.71604, "hi": 1.40311, "ta": 1.14417, "th": 1.38715, "zh": 3.46092, "my": 1.29108}, "gini_cpc": 0.216184, "gini_cps": 0.197831, "bpic": 5.649405, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9508892919009697, "vi": 1.1419464216272985, "tr": 1.0278060669313611, "fi": 0.8732533894247976, "ru": 1.6546365968409538, "el": 1.6401413370816897, "hi": 2.522834774722062, "ta": 2.276171019921883, "th": 2.0016335047331255, "zh": 2.787097744324135, "my": 1.9328236175335856}, "t": 262.3} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.33156, "de": 2.59175, "vi": 2.10157, "tr": 2.58384, "fi": 2.85058, "ru": 1.61236, "el": 1.71181, "hi": 1.41169, "ta": 1.143, "th": 1.38443, "zh": 3.4665, "my": 1.30308}, "bytes_per_chunk": {"en": 4.713, "de": 5.164, "vi": 5.67, "tr": 5.133, "fi": 5.762, "ru": 5.404, "el": 5.455, "hi": 4.807, "ta": 5.897, "th": 6.812, "zh": 4.768, "my": 7.893}, "bpb_mean": 2.04101, "t": 268.0} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.48054, "bpb": 2.13597, "chunk_loss": 0.030328, "F": 0.19232, "G": 0.22537, "chunks": 196.94, "bytes_per_chunk": 5.2, "boundary_entropy": 0.18705, "router_noise": 0.0, "grad_norm": 0.6506, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.265, "alpha": 0.03, "t": 268.1} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.40808, "bpb": 2.03144, "chunk_loss": 0.030075, "F": 0.17889, "G": 0.2245, "chunks": 183.19, "bytes_per_chunk": 5.59, "boundary_entropy": 0.2089, "router_noise": 0.0, "grad_norm": 0.5464, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 286.1} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.44013, "bpb": 2.07767, "chunk_loss": 0.030443, "F": 0.19536, "G": 0.23019, "chunks": 200.05, "bytes_per_chunk": 5.119, "boundary_entropy": 0.19452, "router_noise": 0.0, "grad_norm": 0.5733, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 304.0} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.45125, "bpb": 2.0937, "chunk_loss": 0.030323, "F": 0.19168, "G": 0.22411, "chunks": 196.28, "bytes_per_chunk": 5.217, "boundary_entropy": 0.18286, "router_noise": 0.0, "grad_norm": 0.595, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 321.7} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.41637, "bpb": 2.04339, "chunk_loss": 0.030308, "F": 0.19138, "G": 0.22146, "chunks": 195.97, "bytes_per_chunk": 5.225, "boundary_entropy": 0.17086, "router_noise": 0.0, "grad_norm": 0.4777, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 339.5} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.41739, "bpb": 2.04487, "chunk_loss": 0.03034, "F": 0.18631, "G": 0.21835, "chunks": 190.78, "bytes_per_chunk": 5.367, "boundary_entropy": 0.17546, "router_noise": 0.0, "grad_norm": 0.5028, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 357.2} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.39938, "bpb": 2.01887, "chunk_loss": 0.030284, "F": 0.19067, "G": 0.21527, "chunks": 195.25, "bytes_per_chunk": 5.245, "boundary_entropy": 0.15523, "router_noise": 0.0, "grad_norm": 0.5065, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.176, "alpha": 0.03, "t": 374.8} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.34149, "bpb": 1.93536, "chunk_loss": 0.030208, "F": 0.1879, "G": 0.21199, "chunks": 192.41, "bytes_per_chunk": 5.322, "boundary_entropy": 0.14642, "router_noise": 0.0, "grad_norm": 0.4626, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 392.5} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.35375, "bpb": 1.95305, "chunk_loss": 0.030228, "F": 0.189, "G": 0.22524, "chunks": 193.53, "bytes_per_chunk": 5.291, "boundary_entropy": 0.18987, "router_noise": 0.0, "grad_norm": 0.4684, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 410.5} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.36174, "bpb": 1.96458, "chunk_loss": 0.03045, "F": 0.1944, "G": 0.22121, "chunks": 199.06, "bytes_per_chunk": 5.144, "boundary_entropy": 0.16075, "router_noise": 0.0, "grad_norm": 0.4997, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 428.6} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3552.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.204103, "cps": 27.75, "mean_p": 0.208496, "boundary_entropy": 0.095349, "bpb": 2.070667}, "de": {"chunks": 3990.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.194321, "cps": 31.171875, "mean_p": 0.203154, "boundary_entropy": 0.098109, "bpb": 2.289216}, "vi": {"chunks": 4273.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.231787, "cps": 33.382812, "mean_p": 0.197301, "boundary_entropy": 0.108063, "bpb": 1.937617}, "tr": {"chunks": 3444.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190245, "cps": 26.90625, "mean_p": 0.190114, "boundary_entropy": 0.144065, "bpb": 2.287232}, "fi": {"chunks": 3466.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184254, "cps": 27.078125, "mean_p": 0.193535, "boundary_entropy": 0.129117, "bpb": 2.476042}, "ru": {"chunks": 6456.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.340686, "cps": 50.4375, "mean_p": 0.198518, "boundary_entropy": 0.1164, "bpb": 1.359407}, "el": {"chunks": 7111.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.346489, "cps": 55.554688, "mean_p": 0.203027, "boundary_entropy": 0.131101, "bpb": 1.413299}, "hi": {"chunks": 8041.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.47792, "cps": 62.820312, "mean_p": 0.257018, "boundary_entropy": 0.344226, "bpb": 1.198525}, "ta": {"chunks": 7886.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.421824, "cps": 61.609375, "mean_p": 0.274331, "boundary_entropy": 0.46337, "bpb": 1.047019}, "th": {"chunks": 7401.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.460118, "cps": 57.820312, "mean_p": 0.301539, "boundary_entropy": 0.486074, "bpb": 1.157475}, "zh": {"chunks": 3254.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.550965, "cps": 25.421875, "mean_p": 0.220893, "boundary_entropy": 0.125716, "bpb": 3.17226}, "my": {"chunks": 10948.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.577792, "cps": 85.53125, "mean_p": 0.314992, "boundary_entropy": 0.491615, "bpb": 1.173406}}, "bpb": {"en": 2.07067, "de": 2.28922, "vi": 1.93762, "tr": 2.28723, "fi": 2.47604, "ru": 1.35941, "el": 1.4133, "hi": 1.19853, "ta": 1.04702, "th": 1.15748, "zh": 3.17226, "my": 1.17341}, "gini_cpc": 0.227576, "gini_cps": 0.224067, "bpic": 5.496563, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9520760746379262, "vi": 1.1356406090793805, "tr": 0.932102675232267, "fi": 0.9027507086854012, "ru": 1.6691888326321045, "el": 1.6976220071482095, "hi": 2.3415646459981527, "ta": 2.066723921701857, "th": 2.2543456321568693, "zh": 2.6994498837645637, "my": 2.830887271850698}, "t": 449.6} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.08065, "de": 2.27989, "vi": 1.94572, "tr": 2.27842, "fi": 2.49401, "ru": 1.3598, "el": 1.41086, "hi": 1.2073, "ta": 1.04623, "th": 1.15263, "zh": 3.17874, "my": 1.18441}, "bytes_per_chunk": {"en": 4.754, "de": 5.239, "vi": 5.695, "tr": 5.707, "fi": 5.478, "ru": 5.451, "el": 5.383, "hi": 5.18, "ta": 6.508, "th": 6.153, "zh": 4.88, "my": 5.19}, "bpb_mean": 1.80156, "t": 455.0} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.34364, "bpb": 1.93847, "chunk_loss": 0.030304, "F": 0.19429, "G": 0.21432, "chunks": 198.95, "bytes_per_chunk": 5.147, "boundary_entropy": 0.14016, "router_noise": 0.0, "grad_norm": 0.4492, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.266, "alpha": 0.03, "t": 455.2} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.28572, "bpb": 1.8549, "chunk_loss": 0.030524, "F": 0.19975, "G": 0.22266, "chunks": 204.55, "bytes_per_chunk": 5.006, "boundary_entropy": 0.1638, "router_noise": 0.0, "grad_norm": 0.4491, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 473.0} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.25054, "bpb": 1.80414, "chunk_loss": 0.030173, "F": 0.18494, "G": 0.219, "chunks": 189.38, "bytes_per_chunk": 5.407, "boundary_entropy": 0.17834, "router_noise": 0.0, "grad_norm": 0.4333, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 490.9} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.321, "bpb": 1.90579, "chunk_loss": 0.030358, "F": 0.19322, "G": 0.21569, "chunks": 197.86, "bytes_per_chunk": 5.175, "boundary_entropy": 0.14911, "router_noise": 0.0, "grad_norm": 0.4289, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 508.6} |
| {"event": "train", "step": 2400, "phase": 1, "objective": "global", "lm_loss": 1.34373, "bpb": 1.9386, "chunk_loss": 0.030208, "F": 0.18916, "G": 0.21098, "chunks": 193.7, "bytes_per_chunk": 5.286, "boundary_entropy": 0.142, "router_noise": 0.0, "grad_norm": 0.4853, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 526.9} |
| {"event": "train", "step": 2500, "phase": 1, "objective": "global", "lm_loss": 1.24384, "bpb": 1.79448, "chunk_loss": 0.030251, "F": 0.1899, "G": 0.21557, "chunks": 194.45, "bytes_per_chunk": 5.266, "boundary_entropy": 0.15513, "router_noise": 0.0, "grad_norm": 0.3909, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 544.6} |
| {"event": "train", "step": 2600, "phase": 1, "objective": "global", "lm_loss": 1.26293, "bpb": 1.82202, "chunk_loss": 0.030202, "F": 0.1864, "G": 0.21742, "chunks": 190.88, "bytes_per_chunk": 5.365, "boundary_entropy": 0.15867, "router_noise": 0.0, "grad_norm": 0.4278, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 562.6} |
| {"event": "train", "step": 2700, "phase": 1, "objective": "global", "lm_loss": 1.26535, "bpb": 1.82552, "chunk_loss": 0.030151, "F": 0.18524, "G": 0.2125, "chunks": 189.69, "bytes_per_chunk": 5.398, "boundary_entropy": 0.15206, "router_noise": 0.0, "grad_norm": 0.4005, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 580.4} |
| {"event": "train", "step": 2800, "phase": 1, "objective": "global", "lm_loss": 1.34991, "bpb": 1.94752, "chunk_loss": 0.030315, "F": 0.19289, "G": 0.22025, "chunks": 197.52, "bytes_per_chunk": 5.184, "boundary_entropy": 0.15642, "router_noise": 0.0, "grad_norm": 0.4206, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 598.3} |
| {"event": "train", "step": 2900, "phase": 1, "objective": "global", "lm_loss": 1.26941, "bpb": 1.83137, "chunk_loss": 0.030193, "F": 0.18773, "G": 0.21555, "chunks": 192.23, "bytes_per_chunk": 5.327, "boundary_entropy": 0.15599, "router_noise": 0.0, "grad_norm": 0.3875, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 616.3} |
| {"event": "probe", "step": 3000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3504.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201345, "cps": 27.375, "mean_p": 0.20211, "boundary_entropy": 0.085069, "bpb": 1.959655}, "de": {"chunks": 3956.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.192665, "cps": 30.90625, "mean_p": 0.198373, "boundary_entropy": 0.080412, "bpb": 2.144263}, "vi": {"chunks": 4290.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23271, "cps": 33.515625, "mean_p": 0.194588, "boundary_entropy": 0.105077, "bpb": 1.805651}, "tr": {"chunks": 3392.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187372, "cps": 26.5, "mean_p": 0.187611, "boundary_entropy": 0.121526, "bpb": 2.126437}, "fi": {"chunks": 3382.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179788, "cps": 26.421875, "mean_p": 0.187547, "boundary_entropy": 0.109778, "bpb": 2.302428}, "ru": {"chunks": 6343.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.334723, "cps": 49.554688, "mean_p": 0.192334, "boundary_entropy": 0.101362, "bpb": 1.269468}, "el": {"chunks": 7144.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.348097, "cps": 55.8125, "mean_p": 0.198048, "boundary_entropy": 0.10364, "bpb": 1.302295}, "hi": {"chunks": 7065.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.419911, "cps": 55.195312, "mean_p": 0.26126, "boundary_entropy": 0.357132, "bpb": 1.113244}, "ta": {"chunks": 8821.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.471837, "cps": 68.914062, "mean_p": 0.284646, "boundary_entropy": 0.467818, "bpb": 0.956149}, "th": {"chunks": 6988.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.434442, "cps": 54.59375, "mean_p": 0.300278, "boundary_entropy": 0.489756, "bpb": 1.077624}, "zh": {"chunks": 3278.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.555029, "cps": 25.609375, "mean_p": 0.217573, "boundary_entropy": 0.107223, "bpb": 2.985687}, "my": {"chunks": 9175.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.48422, "cps": 71.679688, "mean_p": 0.291723, "boundary_entropy": 0.478421, "bpb": 1.100131}}, "bpb": {"en": 1.95966, "de": 2.14426, "vi": 1.80565, "tr": 2.12644, "fi": 2.30243, "ru": 1.26947, "el": 1.3023, "hi": 1.11324, "ta": 0.95615, "th": 1.07762, "zh": 2.98569, "my": 1.10013}, "gini_cpc": 0.216345, "gini_cps": 0.209188, "bpic": 5.699323, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9568941476767435, "vi": 1.1557773331698562, "tr": 0.9306048569865435, "fi": 0.8929388990557113, "ru": 1.6624382386959193, "el": 1.728863173225629, "hi": 2.0855332390237944, "ta": 2.343432104034849, "th": 2.157703935854472, "zh": 2.7566112822344584, "my": 2.4049315484194667}, "t": 636.9} |
| {"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 1.96882, "de": 2.13801, "vi": 1.81227, "tr": 2.11552, "fi": 2.31801, "ru": 1.26754, "el": 1.29996, "hi": 1.12165, "ta": 0.95572, "th": 1.07464, "zh": 2.99431, "my": 1.10988}, "bytes_per_chunk": {"en": 4.799, "de": 5.266, "vi": 5.677, "tr": 5.795, "fi": 5.593, "ru": 5.569, "el": 5.358, "hi": 6.14, "ta": 6.104, "th": 6.333, "zh": 4.922, "my": 6.06}, "bpb_mean": 1.68136, "t": 642.6} |
| {"event": "train", "step": 3000, "phase": 1, "objective": "global", "lm_loss": 1.31437, "bpb": 1.89624, "chunk_loss": 0.030226, "F": 0.19179, "G": 0.2109, "chunks": 196.39, "bytes_per_chunk": 5.214, "boundary_entropy": 0.13071, "router_noise": 0.0, "grad_norm": 0.4261, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.265, "alpha": 0.03, "t": 642.8} |
| {"event": "train", "step": 3100, "phase": 1, "objective": "global", "lm_loss": 1.31452, "bpb": 1.89646, "chunk_loss": 0.030231, "F": 0.1913, "G": 0.2115, "chunks": 195.89, "bytes_per_chunk": 5.227, "boundary_entropy": 0.12873, "router_noise": 0.0, "grad_norm": 0.3947, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.177, "alpha": 0.03, "t": 660.5} |
| {"event": "train", "step": 3200, "phase": 1, "objective": "global", "lm_loss": 1.18859, "bpb": 1.71477, "chunk_loss": 0.030286, "F": 0.19264, "G": 0.21878, "chunks": 197.27, "bytes_per_chunk": 5.191, "boundary_entropy": 0.15525, "router_noise": 0.0, "grad_norm": 0.3764, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 678.5} |
| {"event": "train", "step": 3300, "phase": 1, "objective": "global", "lm_loss": 1.25929, "bpb": 1.81677, "chunk_loss": 0.030362, "F": 0.19237, "G": 0.22194, "chunks": 196.98, "bytes_per_chunk": 5.198, "boundary_entropy": 0.16561, "router_noise": 0.0, "grad_norm": 0.3936, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 696.0} |
| {"event": "train", "step": 3400, "phase": 1, "objective": "global", "lm_loss": 1.21325, "bpb": 1.75034, "chunk_loss": 0.030341, "F": 0.19443, "G": 0.22014, "chunks": 199.09, "bytes_per_chunk": 5.143, "boundary_entropy": 0.15006, "router_noise": 0.0, "grad_norm": 0.4115, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.176, "alpha": 0.03, "t": 713.6} |
| {"event": "train", "step": 3500, "phase": 1, "objective": "global", "lm_loss": 1.28586, "bpb": 1.85511, "chunk_loss": 0.03021, "F": 0.1886, "G": 0.20465, "chunks": 193.12, "bytes_per_chunk": 5.302, "boundary_entropy": 0.11481, "router_noise": 0.0, "grad_norm": 0.3785, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 731.5} |
| {"event": "train", "step": 3600, "phase": 1, "objective": "global", "lm_loss": 1.20017, "bpb": 1.73148, "chunk_loss": 0.030206, "F": 0.19048, "G": 0.21235, "chunks": 195.05, "bytes_per_chunk": 5.25, "boundary_entropy": 0.13603, "router_noise": 0.0, "grad_norm": 0.4016, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 749.5} |
| {"event": "train", "step": 3700, "phase": 1, "objective": "global", "lm_loss": 1.12171, "bpb": 1.61828, "chunk_loss": 0.030143, "F": 0.18217, "G": 0.21456, "chunks": 186.55, "bytes_per_chunk": 5.489, "boundary_entropy": 0.1665, "router_noise": 0.0, "grad_norm": 0.3998, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 767.5} |
| {"event": "train", "step": 3800, "phase": 1, "objective": "global", "lm_loss": 1.19279, "bpb": 1.72083, "chunk_loss": 0.030164, "F": 0.18787, "G": 0.2114, "chunks": 192.38, "bytes_per_chunk": 5.323, "boundary_entropy": 0.13258, "router_noise": 0.0, "grad_norm": 0.3645, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.178, "alpha": 0.03, "t": 785.4} |
| {"event": "train", "step": 3900, "phase": 1, "objective": "global", "lm_loss": 1.21699, "bpb": 1.75574, "chunk_loss": 0.030214, "F": 0.18924, "G": 0.20689, "chunks": 193.78, "bytes_per_chunk": 5.284, "boundary_entropy": 0.12025, "router_noise": 0.0, "grad_norm": 0.3702, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 803.3} |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3471.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199448, "cps": 27.117188, "mean_p": 0.198775, "boundary_entropy": 0.07289, "bpb": 1.828264}, "de": {"chunks": 3940.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.191886, "cps": 30.78125, "mean_p": 0.194634, "boundary_entropy": 0.066458, "bpb": 1.990287}, "vi": {"chunks": 4243.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23016, "cps": 33.148438, "mean_p": 0.188743, "boundary_entropy": 0.088563, "bpb": 1.676333}, "tr": {"chunks": 3412.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.188477, "cps": 26.65625, "mean_p": 0.183338, "boundary_entropy": 0.111756, "bpb": 1.941201}, "fi": {"chunks": 3545.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188454, "cps": 27.695312, "mean_p": 0.18961, "boundary_entropy": 0.092997, "bpb": 2.133375}, "ru": {"chunks": 6352.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.335198, "cps": 49.625, "mean_p": 0.188665, "boundary_entropy": 0.085974, "bpb": 1.217823}, "el": {"chunks": 6971.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.339668, "cps": 54.460938, "mean_p": 0.193178, "boundary_entropy": 0.105723, "bpb": 1.233115}, "hi": {"chunks": 7459.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.443328, "cps": 58.273438, "mean_p": 0.269513, "boundary_entropy": 0.362297, "bpb": 1.052474}, "ta": {"chunks": 8872.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.474565, "cps": 69.3125, "mean_p": 0.286955, "boundary_entropy": 0.469232, "bpb": 0.908424}, "th": {"chunks": 7381.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.458875, "cps": 57.664062, "mean_p": 0.303557, "boundary_entropy": 0.494008, "bpb": 1.024565}, "zh": {"chunks": 3221.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545378, "cps": 25.164062, "mean_p": 0.21107, "boundary_entropy": 0.087848, "bpb": 2.870235}, "my": {"chunks": 9113.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.480948, "cps": 71.195312, "mean_p": 0.3326, "boundary_entropy": 0.531751, "bpb": 1.058232}}, "bpb": {"en": 1.82826, "de": 1.99029, "vi": 1.67633, "tr": 1.9412, "fi": 2.13337, "ru": 1.21782, "el": 1.23312, "hi": 1.05247, "ta": 0.90842, "th": 1.02457, "zh": 2.87023, "my": 1.05823}, "gini_cpc": 0.214207, "gini_cps": 0.211153, "bpic": 5.645499, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9620847289232027, "vi": 1.1539829609933194, "tr": 0.9449916641883708, "fi": 0.944873921440346, "ru": 1.680624853193707, "el": 1.7030356691110016, "hi": 2.222772631478911, "ta": 2.379389661127553, "th": 2.3007193579699896, "zh": 2.734429865062587, "my": 2.4113902507868965}, "t": 824.6} |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.82906, "de": 1.99278, "vi": 1.68991, "tr": 1.93346, "fi": 2.14968, "ru": 1.21554, "el": 1.23008, "hi": 1.06136, "ta": 0.91041, "th": 1.02295, "zh": 2.87959, "my": 1.06576}, "bytes_per_chunk": {"en": 4.847, "de": 5.291, "vi": 5.681, "tr": 5.753, "fi": 5.323, "ru": 5.552, "el": 5.461, "hi": 5.711, "ta": 5.83, "th": 5.87, "zh": 4.939, "my": 5.83}, "bpb_mean": 1.58171, "t": 830.0} |
| {"event": "train", "step": 4000, "phase": 1, "objective": "global", "lm_loss": 1.28119, "bpb": 1.84836, "chunk_loss": 0.030235, "F": 0.18785, "G": 0.21269, "chunks": 192.36, "bytes_per_chunk": 5.323, "boundary_entropy": 0.14332, "router_noise": 0.0, "grad_norm": 0.3816, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.268, "alpha": 0.03, "t": 830.1} |
| {"event": "train", "step": 4100, "phase": 1, "objective": "global", "lm_loss": 1.25836, "bpb": 1.81543, "chunk_loss": 0.030144, "F": 0.186, "G": 0.20482, "chunks": 190.47, "bytes_per_chunk": 5.376, "boundary_entropy": 0.11985, "router_noise": 0.0, "grad_norm": 0.3923, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 848.1} |
| {"event": "train", "step": 4200, "phase": 1, "objective": "global", "lm_loss": 1.20048, "bpb": 1.73193, "chunk_loss": 0.03018, "F": 0.18791, "G": 0.2054, "chunks": 192.42, "bytes_per_chunk": 5.322, "boundary_entropy": 0.11347, "router_noise": 0.0, "grad_norm": 0.364, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.179, "alpha": 0.03, "t": 866.0} |
| {"event": "train", "step": 4300, "phase": 1, "objective": "global", "lm_loss": 1.18014, "bpb": 1.70259, "chunk_loss": 0.030219, "F": 0.18738, "G": 0.21087, "chunks": 191.88, "bytes_per_chunk": 5.337, "boundary_entropy": 0.13814, "router_noise": 0.0, "grad_norm": 0.3767, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.18, "alpha": 0.03, "t": 884.0} |
| {"event": "phase_transition", "step": 4400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 901.7} |
| {"event": "train", "step": 4400, "phase": 2, "objective": "parity_b", "lm_loss": 1.13399, "bpb": 1.636, "chunk_loss": 0.044939, "F": 0.18654, "G": 0.2108, "chunks": 191.02, "bytes_per_chunk": 5.361, "boundary_entropy": 0.13743, "router_noise": 0.0, "grad_norm": 0.406, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.179, "worst_lang": "my", "n_target_mean": 21.145, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 901.9} |
| {"event": "train", "step": 4500, "phase": 2, "objective": "parity_b", "lm_loss": 1.22533, "bpb": 1.76778, "chunk_loss": 0.030556, "F": 0.16539, "G": 0.18899, "chunks": 169.36, "bytes_per_chunk": 6.046, "boundary_entropy": 0.13281, "router_noise": 0.0, "grad_norm": 0.3704, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 920.0} |
| {"event": "train", "step": 4600, "phase": 2, "objective": "parity_b", "lm_loss": 1.18036, "bpb": 1.7029, "chunk_loss": 0.031646, "F": 0.15964, "G": 0.17084, "chunks": 163.47, "bytes_per_chunk": 6.264, "boundary_entropy": 0.10968, "router_noise": 0.0, "grad_norm": 0.3869, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 938.0} |
| {"event": "train", "step": 4700, "phase": 2, "objective": "parity_b", "lm_loss": 1.10498, "bpb": 1.59415, "chunk_loss": 0.030513, "F": 0.14703, "G": 0.1642, "chunks": 150.56, "bytes_per_chunk": 6.801, "boundary_entropy": 0.11915, "router_noise": 0.0, "grad_norm": 0.357, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.182, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 956.2} |
| {"event": "train", "step": 4800, "phase": 2, "objective": "parity_b", "lm_loss": 1.23272, "bpb": 1.77844, "chunk_loss": 0.030482, "F": 0.17026, "G": 0.17941, "chunks": 174.34, "bytes_per_chunk": 5.873, "boundary_entropy": 0.10113, "router_noise": 0.0, "grad_norm": 0.3808, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 974.9} |
| {"event": "train", "step": 4900, "phase": 2, "objective": "parity_b", "lm_loss": 1.21224, "bpb": 1.7489, "chunk_loss": 0.030249, "F": 0.17729, "G": 0.18668, "chunks": 181.55, "bytes_per_chunk": 5.64, "boundary_entropy": 0.09603, "router_noise": 0.0, "grad_norm": 0.4036, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.182, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 993.1} |
| {"event": "probe", "step": 5000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3472.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199506, "cps": 27.125, "mean_p": 0.203123, "boundary_entropy": 0.069739, "bpb": 1.767218}, "de": {"chunks": 3959.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.192812, "cps": 30.929688, "mean_p": 0.199692, "boundary_entropy": 0.078922, "bpb": 1.927509}, "vi": {"chunks": 4043.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.219311, "cps": 31.585938, "mean_p": 0.169801, "boundary_entropy": 0.069875, "bpb": 1.598447}, "tr": {"chunks": 3394.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187483, "cps": 26.515625, "mean_p": 0.183041, "boundary_entropy": 0.108874, "bpb": 1.861217}, "fi": {"chunks": 3482.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185104, "cps": 27.203125, "mean_p": 0.187221, "boundary_entropy": 0.094815, "bpb": 2.04524}, "ru": {"chunks": 4007.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.211451, "cps": 31.304688, "mean_p": 0.10771, "boundary_entropy": 0.107965, "bpb": 1.280185}, "el": {"chunks": 3917.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.190859, "cps": 30.601562, "mean_p": 0.100416, "boundary_entropy": 0.105635, "bpb": 1.302786}, "hi": {"chunks": 3662.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.217652, "cps": 28.609375, "mean_p": 0.101648, "boundary_entropy": 0.121913, "bpb": 1.030127}, "ta": {"chunks": 1515.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.081038, "cps": 11.835938, "mean_p": 0.090342, "boundary_entropy": 0.221828, "bpb": 0.887746}, "th": {"chunks": 3676.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.228536, "cps": 28.71875, "mean_p": 0.175875, "boundary_entropy": 0.345333, "bpb": 0.998672}, "zh": {"chunks": 3251.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.550457, "cps": 25.398438, "mean_p": 0.214677, "boundary_entropy": 0.09618, "bpb": 2.744304}, "my": {"chunks": 2490.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.131412, "cps": 19.453125, "mean_p": 0.089682, "boundary_entropy": 0.202379, "bpb": 1.019355}}, "bpb": {"en": 1.76722, "de": 1.92751, "vi": 1.59845, "tr": 1.86122, "fi": 2.04524, "ru": 1.28018, "el": 1.30279, "hi": 1.03013, "ta": 0.88775, "th": 0.99867, "zh": 2.7443, "my": 1.01935}, "gini_cpc": 0.208572, "gini_cps": 0.101595, "bpic": 9.390746, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9664457894126146, "vi": 1.0992715875557288, "tr": 0.9397356231660561, "fi": 0.9278147811347758, "ru": 1.0598747157812822, "el": 0.9566589306007355, "hi": 1.0909570942406583, "ta": 0.4061921883300076, "th": 1.1455098848427636, "zh": 2.759103108843463, "my": 0.6586889451995225}, "t": 1013.9} |
| {"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.76742, "de": 1.93202, "vi": 1.61294, "tr": 1.85331, "fi": 2.05844, "ru": 1.27835, "el": 1.30062, "hi": 1.03751, "ta": 0.89081, "th": 0.99758, "zh": 2.75223, "my": 1.02552}, "bytes_per_chunk": {"en": 4.845, "de": 5.306, "vi": 5.945, "tr": 5.783, "fi": 5.399, "ru": 9.059, "el": 9.731, "hi": 11.511, "ta": 35.844, "th": 11.785, "zh": 4.914, "my": 21.247}, "bpb_mean": 1.54223, "t": 1018.9} |
| {"event": "train", "step": 5000, "phase": 2, "objective": "parity_b", "lm_loss": 1.19085, "bpb": 1.71803, "chunk_loss": 0.030767, "F": 0.15787, "G": 0.16739, "chunks": 161.66, "bytes_per_chunk": 6.334, "boundary_entropy": 0.10256, "router_noise": 0.0, "grad_norm": 0.4021, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.26, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 1019.1} |
| {"event": "train", "step": 5100, "phase": 2, "objective": "parity_b", "lm_loss": 1.19372, "bpb": 1.72217, "chunk_loss": 0.030309, "F": 0.1579, "G": 0.16667, "chunks": 161.69, "bytes_per_chunk": 6.333, "boundary_entropy": 0.09595, "router_noise": 0.0, "grad_norm": 0.373, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 1037.1} |
| {"event": "train", "step": 5200, "phase": 2, "objective": "parity_b", "lm_loss": 1.11983, "bpb": 1.61557, "chunk_loss": 0.030443, "F": 0.15134, "G": 0.16439, "chunks": 154.97, "bytes_per_chunk": 6.608, "boundary_entropy": 0.10991, "router_noise": 0.0, "grad_norm": 0.3719, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.183, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 1055.4} |
| {"event": "train", "step": 5300, "phase": 2, "objective": "parity_b", "lm_loss": 1.14445, "bpb": 1.65109, "chunk_loss": 0.030257, "F": 0.15538, "G": 0.16382, "chunks": 159.11, "bytes_per_chunk": 6.436, "boundary_entropy": 0.09948, "router_noise": 0.0, "grad_norm": 0.3864, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.177, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.2031, "fallback": 0.0, "alpha": 0.03, "t": 1073.1} |
| {"event": "train", "step": 5400, "phase": 2, "objective": "parity_b", "lm_loss": 1.20183, "bpb": 1.73387, "chunk_loss": 0.031054, "F": 0.16454, "G": 0.17274, "chunks": 168.48, "bytes_per_chunk": 6.078, "boundary_entropy": 0.08912, "router_noise": 0.0, "grad_norm": 0.3951, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.174, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1090.4} |
| {"event": "train", "step": 5500, "phase": 2, "objective": "parity_b", "lm_loss": 1.15303, "bpb": 1.66348, "chunk_loss": 0.030307, "F": 0.15646, "G": 0.16929, "chunks": 160.22, "bytes_per_chunk": 6.391, "boundary_entropy": 0.10117, "router_noise": 0.0, "grad_norm": 0.3868, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.181, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1108.5} |
| {"event": "train", "step": 5600, "phase": 2, "objective": "parity_b", "lm_loss": 1.10991, "bpb": 1.60127, "chunk_loss": 0.030137, "F": 0.15366, "G": 0.16208, "chunks": 157.34, "bytes_per_chunk": 6.508, "boundary_entropy": 0.09108, "router_noise": 0.0, "grad_norm": 0.3764, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.182, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1126.6} |
| {"event": "train", "step": 5700, "phase": 2, "objective": "parity_b", "lm_loss": 1.10135, "bpb": 1.58891, "chunk_loss": 0.030356, "F": 0.15199, "G": 0.16157, "chunks": 155.64, "bytes_per_chunk": 6.579, "boundary_entropy": 0.09793, "router_noise": 0.0, "grad_norm": 0.3876, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.173, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1144.0} |
| {"event": "train", "step": 5800, "phase": 2, "objective": "parity_b", "lm_loss": 1.08664, "bpb": 1.56769, "chunk_loss": 0.030066, "F": 0.15285, "G": 0.16189, "chunks": 156.52, "bytes_per_chunk": 6.542, "boundary_entropy": 0.08992, "router_noise": 0.0, "grad_norm": 0.3862, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1162.0} |
| {"event": "train", "step": 5900, "phase": 2, "objective": "parity_b", "lm_loss": 1.13987, "bpb": 1.64449, "chunk_loss": 0.030309, "F": 0.16328, "G": 0.16837, "chunks": 167.2, "bytes_per_chunk": 6.124, "boundary_entropy": 0.07891, "router_noise": 0.0, "grad_norm": 0.4002, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.179, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1179.9} |
| {"event": "probe", "step": 6000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3468.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199276, "cps": 27.09375, "mean_p": 0.200084, "boundary_entropy": 0.049621, "bpb": 1.728146}, "de": {"chunks": 3852.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.1876, "cps": 30.09375, "mean_p": 0.183978, "boundary_entropy": 0.055809, "bpb": 1.872255}, "vi": {"chunks": 3806.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.206455, "cps": 29.734375, "mean_p": 0.148703, "boundary_entropy": 0.070241, "bpb": 1.539374}, "tr": {"chunks": 3623.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200133, "cps": 28.304688, "mean_p": 0.188854, "boundary_entropy": 0.088147, "bpb": 1.804123}, "fi": {"chunks": 3413.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181436, "cps": 26.664062, "mean_p": 0.180044, "boundary_entropy": 0.073709, "bpb": 1.990354}, "ru": {"chunks": 4072.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.214881, "cps": 31.8125, "mean_p": 0.114066, "boundary_entropy": 0.087477, "bpb": 1.188097}, "el": {"chunks": 3955.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.192711, "cps": 30.898438, "mean_p": 0.102042, "boundary_entropy": 0.09222, "bpb": 1.217558}, "hi": {"chunks": 3517.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.209034, "cps": 27.476562, "mean_p": 0.089916, "boundary_entropy": 0.096091, "bpb": 0.999387}, "ta": {"chunks": 1618.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.086547, "cps": 12.640625, "mean_p": 0.094365, "boundary_entropy": 0.222933, "bpb": 0.861054}, "th": {"chunks": 2773.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.172397, "cps": 21.664062, "mean_p": 0.12918, "boundary_entropy": 0.26594, "bpb": 0.962949}, "zh": {"chunks": 3240.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.548595, "cps": 25.3125, "mean_p": 0.213267, "boundary_entropy": 0.081967, "bpb": 2.636505}, "my": {"chunks": 2791.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.147298, "cps": 21.804688, "mean_p": 0.097582, "boundary_entropy": 0.210358, "bpb": 0.984208}}, "bpb": {"en": 1.72815, "de": 1.87225, "vi": 1.53937, "tr": 1.80412, "fi": 1.99035, "ru": 1.1881, "el": 1.21756, "hi": 0.99939, "ta": 0.86105, "th": 0.96295, "zh": 2.6365, "my": 0.98421}, "gini_cpc": 0.200171, "gini_cps": 0.100861, "bpic": 9.56392, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9414102069131015, "vi": 1.0360260449367131, "tr": 1.0042984999140825, "fi": 0.9104779578465739, "ru": 1.0783098848727757, "el": 0.9670538851928641, "hi": 1.0489682103065858, "ta": 0.4343082499900516, "th": 0.865114987098169, "zh": 2.752939067302386, "my": 0.7391651655633009}, "t": 1200.8} |
| {"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.7279, "de": 1.87738, "vi": 1.55593, "tr": 1.79676, "fi": 2.00453, "ru": 1.18773, "el": 1.21475, "hi": 1.0052, "ta": 0.86472, "th": 0.9607, "zh": 2.64216, "my": 0.99125}, "bytes_per_chunk": {"en": 4.86, "de": 5.453, "vi": 6.321, "tr": 5.435, "fi": 5.513, "ru": 8.92, "el": 9.612, "hi": 12.012, "ta": 33.396, "th": 15.318, "zh": 4.896, "my": 19.628}, "bpb_mean": 1.48575, "t": 1205.8} |
| {"event": "train", "step": 6000, "phase": 2, "objective": "parity_b", "lm_loss": 1.11863, "bpb": 1.61384, "chunk_loss": 0.030358, "F": 0.15971, "G": 0.16997, "chunks": 163.55, "bytes_per_chunk": 6.261, "boundary_entropy": 0.08834, "router_noise": 0.0, "grad_norm": 0.3906, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.26, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1206.0} |
| {"event": "train", "step": 6100, "phase": 2, "objective": "parity_b", "lm_loss": 1.04976, "bpb": 1.51449, "chunk_loss": 0.030464, "F": 0.15802, "G": 0.16776, "chunks": 161.81, "bytes_per_chunk": 6.328, "boundary_entropy": 0.09884, "router_noise": 0.0, "grad_norm": 0.367, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.182, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 1224.1} |
| {"event": "train", "step": 6200, "phase": 2, "objective": "parity_b", "lm_loss": 1.16148, "bpb": 1.67566, "chunk_loss": 0.03048, "F": 0.16711, "G": 0.17442, "chunks": 171.12, "bytes_per_chunk": 5.984, "boundary_entropy": 0.08257, "router_noise": 0.0, "grad_norm": 0.3826, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.174, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1719, "fallback": 0.0, "alpha": 0.03, "t": 1241.5} |
| {"event": "train", "step": 6300, "phase": 2, "objective": "parity_b", "lm_loss": 1.08654, "bpb": 1.56754, "chunk_loss": 0.030284, "F": 0.16383, "G": 0.16711, "chunks": 167.77, "bytes_per_chunk": 6.104, "boundary_entropy": 0.07763, "router_noise": 0.0, "grad_norm": 0.3715, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.181, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 1259.5} |
| {"event": "train", "step": 6400, "phase": 2, "objective": "parity_b", "lm_loss": 1.00132, "bpb": 1.4446, "chunk_loss": 0.030091, "F": 0.14029, "G": 0.15004, "chunks": 143.66, "bytes_per_chunk": 7.128, "boundary_entropy": 0.09895, "router_noise": 0.0, "grad_norm": 0.3714, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.182, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 1277.7} |
| {"event": "train", "step": 6500, "phase": 2, "objective": "parity_b", "lm_loss": 1.05936, "bpb": 1.52833, "chunk_loss": 0.030216, "F": 0.14986, "G": 0.15877, "chunks": 153.45, "bytes_per_chunk": 6.673, "boundary_entropy": 0.09083, "router_noise": 0.0, "grad_norm": 0.3957, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.173, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1295.0} |
| {"event": "train", "step": 6600, "phase": 2, "objective": "parity_b", "lm_loss": 1.1093, "bpb": 1.60038, "chunk_loss": 0.030135, "F": 0.16869, "G": 0.17546, "chunks": 172.73, "bytes_per_chunk": 5.928, "boundary_entropy": 0.0818, "router_noise": 0.0, "grad_norm": 0.3718, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.181, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 1313.1} |
| {"event": "train", "step": 6700, "phase": 2, "objective": "parity_b", "lm_loss": 1.09621, "bpb": 1.5815, "chunk_loss": 0.030103, "F": 0.16331, "G": 0.17122, "chunks": 167.23, "bytes_per_chunk": 6.123, "boundary_entropy": 0.09061, "router_noise": 0.0, "grad_norm": 0.3745, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0156, "fallback": 0.0, "alpha": 0.03, "t": 1331.2} |
| {"event": "train", "step": 6800, "phase": 2, "objective": "parity_b", "lm_loss": 1.10919, "bpb": 1.60022, "chunk_loss": 0.03029, "F": 0.15034, "G": 0.15772, "chunks": 153.95, "bytes_per_chunk": 6.651, "boundary_entropy": 0.1024, "router_noise": 0.0, "grad_norm": 0.3897, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.174, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1348.5} |
| {"event": "train", "step": 6900, "phase": 2, "objective": "parity_b", "lm_loss": 1.09204, "bpb": 1.57548, "chunk_loss": 0.030139, "F": 0.15154, "G": 0.15847, "chunks": 155.17, "bytes_per_chunk": 6.599, "boundary_entropy": 0.08855, "router_noise": 0.0, "grad_norm": 0.3822, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.179, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1366.4} |
| {"event": "probe", "step": 7000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3461.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198874, "cps": 27.039062, "mean_p": 0.199818, "boundary_entropy": 0.048655, "bpb": 1.692968}, "de": {"chunks": 3839.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186967, "cps": 29.992188, "mean_p": 0.180586, "boundary_entropy": 0.056732, "bpb": 1.835987}, "vi": {"chunks": 3753.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.20358, "cps": 29.320312, "mean_p": 0.133885, "boundary_entropy": 0.088077, "bpb": 1.500372}, "tr": {"chunks": 3394.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187483, "cps": 26.515625, "mean_p": 0.17896, "boundary_entropy": 0.086771, "bpb": 1.761717}, "fi": {"chunks": 3429.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182287, "cps": 26.789062, "mean_p": 0.179751, "boundary_entropy": 0.068627, "bpb": 1.93921}, "ru": {"chunks": 3998.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.210976, "cps": 31.234375, "mean_p": 0.109096, "boundary_entropy": 0.086694, "bpb": 1.155022}, "el": {"chunks": 3404.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.165863, "cps": 26.59375, "mean_p": 0.087364, "boundary_entropy": 0.090764, "bpb": 1.185619}, "hi": {"chunks": 3393.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.201664, "cps": 26.507812, "mean_p": 0.08997, "boundary_entropy": 0.083827, "bpb": 0.966391}, "ta": {"chunks": 1847.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.098796, "cps": 14.429688, "mean_p": 0.094063, "boundary_entropy": 0.215613, "bpb": 0.838694}, "th": {"chunks": 2792.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.173578, "cps": 21.8125, "mean_p": 0.138274, "boundary_entropy": 0.279186, "bpb": 0.930734}, "zh": {"chunks": 3230.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.546901, "cps": 25.234375, "mean_p": 0.214999, "boundary_entropy": 0.085914, "bpb": 2.57375}, "my": {"chunks": 2549.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.134526, "cps": 19.914062, "mean_p": 0.08356, "boundary_entropy": 0.177156, "bpb": 0.945398}}, "bpb": {"en": 1.69297, "de": 1.83599, "vi": 1.50037, "tr": 1.76172, "fi": 1.93921, "ru": 1.15502, "el": 1.18562, "hi": 0.96639, "ta": 0.83869, "th": 0.93073, "zh": 2.57375, "my": 0.9454}, "gini_cpc": 0.203089, "gini_cps": 0.092249, "bpic": 9.818133, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9401306806093764, "vi": 1.0236652091455434, "tr": 0.9427223587496524, "fi": 0.9165963480674127, "ru": 1.0608551679132496, "el": 0.8340099408274827, "hi": 1.0140311765188237, "ta": 0.4967798212134, "th": 0.8728042847351773, "zh": 2.749993077524969, "my": 0.6764395318576543}, "t": 1387.3} |
| {"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.69194, "de": 1.84127, "vi": 1.51737, "tr": 1.75354, "fi": 1.95484, "ru": 1.15414, "el": 1.18468, "hi": 0.97352, "ta": 0.84248, "th": 0.93016, "zh": 2.58094, "my": 0.95154}, "bytes_per_chunk": {"en": 4.851, "de": 5.477, "vi": 6.408, "tr": 5.75, "fi": 5.458, "ru": 9.09, "el": 11.127, "hi": 12.042, "ta": 28.764, "th": 14.856, "zh": 4.891, "my": 20.278}, "bpb_mean": 1.44804, "t": 1392.3} |
| {"event": "train", "step": 7000, "phase": 2, "objective": "parity_b", "lm_loss": 1.07991, "bpb": 1.55798, "chunk_loss": 0.030427, "F": 0.15384, "G": 0.16164, "chunks": 157.53, "bytes_per_chunk": 6.5, "boundary_entropy": 0.09356, "router_noise": 0.0, "grad_norm": 0.4011, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.26, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1392.4} |
| {"event": "train", "step": 7100, "phase": 2, "objective": "parity_b", "lm_loss": 1.05694, "bpb": 1.52485, "chunk_loss": 0.030153, "F": 0.15494, "G": 0.16415, "chunks": 158.66, "bytes_per_chunk": 6.454, "boundary_entropy": 0.09701, "router_noise": 0.0, "grad_norm": 0.3736, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1410.4} |
| {"event": "train", "step": 7200, "phase": 2, "objective": "parity_b", "lm_loss": 1.11156, "bpb": 1.60364, "chunk_loss": 0.03017, "F": 0.15596, "G": 0.15989, "chunks": 159.7, "bytes_per_chunk": 6.412, "boundary_entropy": 0.08224, "router_noise": 0.0, "grad_norm": 0.3753, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.172, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1427.6} |
| {"event": "train", "step": 7300, "phase": 2, "objective": "parity_b", "lm_loss": 1.12149, "bpb": 1.61797, "chunk_loss": 0.030409, "F": 0.15923, "G": 0.16646, "chunks": 163.05, "bytes_per_chunk": 6.28, "boundary_entropy": 0.08485, "router_noise": 0.0, "grad_norm": 0.3802, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1445.6} |
| {"event": "train", "step": 7400, "phase": 2, "objective": "parity_b", "lm_loss": 1.08578, "bpb": 1.56645, "chunk_loss": 0.030036, "F": 0.158, "G": 0.1646, "chunks": 161.8, "bytes_per_chunk": 6.329, "boundary_entropy": 0.08246, "router_noise": 0.0, "grad_norm": 0.3826, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.173, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0156, "fallback": 0.0, "alpha": 0.03, "t": 1462.9} |
| {"event": "train", "step": 7500, "phase": 2, "objective": "parity_b", "lm_loss": 1.07436, "bpb": 1.54997, "chunk_loss": 0.03042, "F": 0.1524, "G": 0.15864, "chunks": 156.06, "bytes_per_chunk": 6.561, "boundary_entropy": 0.09066, "router_noise": 0.0, "grad_norm": 0.3795, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.18, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1480.9} |
| {"event": "train", "step": 7600, "phase": 2, "objective": "parity_b", "lm_loss": 1.13108, "bpb": 1.6318, "chunk_loss": 0.030259, "F": 0.16345, "G": 0.17118, "chunks": 167.38, "bytes_per_chunk": 6.118, "boundary_entropy": 0.08176, "router_noise": 0.0, "grad_norm": 0.3989, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.173, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 1498.2} |
| {"event": "phase_transition", "step": 7700, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 295680, "t": 1516.0} |
| {"event": "train", "step": 7700, "phase": 3, "objective": "none", "lm_loss": 1.08078, "bpb": 1.55923, "chunk_loss": 0.0, "F": 0.16197, "G": 0.1663, "chunks": 165.86, "bytes_per_chunk": 6.174, "boundary_entropy": 0.07403, "router_noise": 0.0, "grad_norm": 0.3748, "lr": 9.06e-05, "router_trainable": false, "sec_per_step": 0.18, "alpha": 0.0, "t": 1516.2} |
| {"event": "train", "step": 7800, "phase": 3, "objective": "none", "lm_loss": 1.11309, "bpb": 1.60585, "chunk_loss": 0.0, "F": 0.16722, "G": 0.17174, "chunks": 171.23, "bytes_per_chunk": 5.98, "boundary_entropy": 0.07142, "router_noise": 0.0, "grad_norm": 0.3796, "lr": 8.73e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1533.5} |
| {"event": "train", "step": 7900, "phase": 3, "objective": "none", "lm_loss": 1.10325, "bpb": 1.59166, "chunk_loss": 0.0, "F": 0.16782, "G": 0.17528, "chunks": 171.84, "bytes_per_chunk": 5.959, "boundary_entropy": 0.07854, "router_noise": 0.0, "grad_norm": 0.3909, "lr": 8.4e-05, "router_trainable": false, "sec_per_step": 0.179, "alpha": 0.0, "t": 1551.3} |
| {"event": "probe", "step": 8000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3497.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200942, "cps": 27.320312, "mean_p": 0.201506, "boundary_entropy": 0.041235, "bpb": 1.664686}, "de": {"chunks": 3864.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188185, "cps": 30.1875, "mean_p": 0.184496, "boundary_entropy": 0.045752, "bpb": 1.803849}, "vi": {"chunks": 3778.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.204936, "cps": 29.515625, "mean_p": 0.144613, "boundary_entropy": 0.068891, "bpb": 1.463374}, "tr": {"chunks": 3589.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.198254, "cps": 28.039062, "mean_p": 0.186767, "boundary_entropy": 0.075974, "bpb": 1.729178}, "fi": {"chunks": 3470.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184467, "cps": 27.109375, "mean_p": 0.182385, "boundary_entropy": 0.058747, "bpb": 1.897245}, "ru": {"chunks": 4158.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.21942, "cps": 32.484375, "mean_p": 0.116306, "boundary_entropy": 0.079771, "bpb": 1.118537}, "el": {"chunks": 4015.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.195634, "cps": 31.367188, "mean_p": 0.10528, "boundary_entropy": 0.086352, "bpb": 1.136979}, "hi": {"chunks": 3497.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.207845, "cps": 27.320312, "mean_p": 0.088771, "boundary_entropy": 0.081083, "bpb": 0.946192}, "ta": {"chunks": 2673.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.142979, "cps": 20.882812, "mean_p": 0.110954, "boundary_entropy": 0.234256, "bpb": 0.821029}, "th": {"chunks": 2887.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.179484, "cps": 22.554688, "mean_p": 0.138935, "boundary_entropy": 0.276586, "bpb": 0.909721}, "zh": {"chunks": 3254.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.550965, "cps": 25.421875, "mean_p": 0.215072, "boundary_entropy": 0.079658, "bpb": 2.532469}, "my": {"chunks": 2487.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.131254, "cps": 19.429688, "mean_p": 0.087811, "boundary_entropy": 0.188326, "bpb": 0.921271}}, "bpb": {"en": 1.66469, "de": 1.80385, "vi": 1.46337, "tr": 1.72918, "fi": 1.89724, "ru": 1.11854, "el": 1.13698, "hi": 0.94619, "ta": 0.82103, "th": 0.90972, "zh": 2.53247, "my": 0.92127}, "gini_cpc": 0.180676, "gini_cps": 0.081524, "bpic": 9.322087, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9365116806118933, "vi": 1.0198758298697501, "tr": 0.9866233620026199, "fi": 0.9180071827174419, "ru": 1.0919525292316143, "el": 0.9735834716557225, "hi": 1.0343536404160476, "ta": 0.7115443543363459, "th": 0.8932112955471669, "zh": 2.741906201639042, "my": 0.6531920602169996}, "t": 1571.1} |
| {"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.66329, "de": 1.80944, "vi": 1.47971, "tr": 1.72018, "fi": 1.91278, "ru": 1.11727, "el": 1.13532, "hi": 0.95265, "ta": 0.82478, "th": 0.90945, "zh": 2.53917, "my": 0.92791}, "bytes_per_chunk": {"en": 4.828, "de": 5.46, "vi": 6.376, "tr": 5.451, "fi": 5.395, "ru": 8.755, "el": 9.536, "hi": 12.038, "ta": 21.032, "th": 14.524, "zh": 4.897, "my": 21.156}, "bpb_mean": 1.416, "t": 1576.9} |
| {"event": "train", "step": 8000, "phase": 3, "objective": "none", "lm_loss": 1.08887, "bpb": 1.57091, "chunk_loss": 0.0, "F": 0.16512, "G": 0.17452, "chunks": 169.08, "bytes_per_chunk": 6.056, "boundary_entropy": 0.08219, "router_noise": 0.0, "grad_norm": 0.3819, "lr": 8.08e-05, "router_trainable": false, "sec_per_step": 0.257, "alpha": 0.0, "t": 1577.1} |
| {"event": "train", "step": 8100, "phase": 3, "objective": "none", "lm_loss": 1.11022, "bpb": 1.6017, "chunk_loss": 0.0, "F": 0.16605, "G": 0.17131, "chunks": 170.03, "bytes_per_chunk": 6.022, "boundary_entropy": 0.07628, "router_noise": 0.0, "grad_norm": 0.395, "lr": 7.77e-05, "router_trainable": false, "sec_per_step": 0.18, "alpha": 0.0, "t": 1595.1} |
| {"event": "train", "step": 8200, "phase": 3, "objective": "none", "lm_loss": 1.07564, "bpb": 1.55182, "chunk_loss": 0.0, "F": 0.16368, "G": 0.16933, "chunks": 167.61, "bytes_per_chunk": 6.109, "boundary_entropy": 0.07647, "router_noise": 0.0, "grad_norm": 0.3947, "lr": 7.47e-05, "router_trainable": false, "sec_per_step": 0.174, "alpha": 0.0, "t": 1612.4} |
| {"event": "train", "step": 8300, "phase": 3, "objective": "none", "lm_loss": 1.01061, "bpb": 1.458, "chunk_loss": 0.0, "F": 0.15714, "G": 0.16639, "chunks": 160.91, "bytes_per_chunk": 6.364, "boundary_entropy": 0.0844, "router_noise": 0.0, "grad_norm": 0.3783, "lr": 7.17e-05, "router_trainable": false, "sec_per_step": 0.177, "alpha": 0.0, "t": 1630.1} |
| {"event": "train", "step": 8400, "phase": 3, "objective": "none", "lm_loss": 1.056, "bpb": 1.52349, "chunk_loss": 0.0, "F": 0.16017, "G": 0.16836, "chunks": 164.02, "bytes_per_chunk": 6.243, "boundary_entropy": 0.07882, "router_noise": 0.0, "grad_norm": 0.3883, "lr": 6.88e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 1647.3} |
| {"event": "train", "step": 8500, "phase": 3, "objective": "none", "lm_loss": 1.06965, "bpb": 1.54318, "chunk_loss": 0.0, "F": 0.15883, "G": 0.16536, "chunks": 162.64, "bytes_per_chunk": 6.296, "boundary_entropy": 0.08004, "router_noise": 0.0, "grad_norm": 0.3807, "lr": 6.6e-05, "router_trainable": false, "sec_per_step": 0.182, "alpha": 0.0, "t": 1665.5} |
| {"event": "train", "step": 8600, "phase": 3, "objective": "none", "lm_loss": 1.11988, "bpb": 1.61564, "chunk_loss": 0.0, "F": 0.16151, "G": 0.16943, "chunks": 165.39, "bytes_per_chunk": 6.191, "boundary_entropy": 0.08061, "router_noise": 0.0, "grad_norm": 0.4037, "lr": 6.33e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1682.8} |
| {"event": "train", "step": 8700, "phase": 3, "objective": "none", "lm_loss": 0.98975, "bpb": 1.42791, "chunk_loss": 0.0, "F": 0.15694, "G": 0.16647, "chunks": 160.7, "bytes_per_chunk": 6.372, "boundary_entropy": 0.0849, "router_noise": 0.0, "grad_norm": 0.3849, "lr": 6.07e-05, "router_trainable": false, "sec_per_step": 0.178, "alpha": 0.0, "t": 1700.6} |
| {"event": "train", "step": 8800, "phase": 3, "objective": "none", "lm_loss": 1.01336, "bpb": 1.46198, "chunk_loss": 0.0, "F": 0.15683, "G": 0.16628, "chunks": 160.59, "bytes_per_chunk": 6.376, "boundary_entropy": 0.08413, "router_noise": 0.0, "grad_norm": 0.3797, "lr": 5.82e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 1717.8} |
| {"event": "train", "step": 8900, "phase": 3, "objective": "none", "lm_loss": 1.06973, "bpb": 1.54329, "chunk_loss": 0.0, "F": 0.1683, "G": 0.17339, "chunks": 172.34, "bytes_per_chunk": 5.942, "boundary_entropy": 0.0684, "router_noise": 0.0, "grad_norm": 0.3692, "lr": 5.58e-05, "router_trainable": false, "sec_per_step": 0.171, "alpha": 0.0, "t": 1735.0} |
| {"event": "probe", "step": 9000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3496.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200885, "cps": 27.3125, "mean_p": 0.201708, "boundary_entropy": 0.039184, "bpb": 1.645165}, "de": {"chunks": 3881.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189013, "cps": 30.320312, "mean_p": 0.186073, "boundary_entropy": 0.04358, "bpb": 1.778244}, "vi": {"chunks": 3815.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.206943, "cps": 29.804688, "mean_p": 0.149044, "boundary_entropy": 0.066752, "bpb": 1.434842}, "tr": {"chunks": 3500.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.193338, "cps": 27.34375, "mean_p": 0.182807, "boundary_entropy": 0.074863, "bpb": 1.702994}, "fi": {"chunks": 3467.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184307, "cps": 27.085938, "mean_p": 0.181072, "boundary_entropy": 0.054155, "bpb": 1.866771}, "ru": {"chunks": 4152.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.219103, "cps": 32.4375, "mean_p": 0.115955, "boundary_entropy": 0.079531, "bpb": 1.099628}, "el": {"chunks": 3995.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.19466, "cps": 31.210938, "mean_p": 0.10515, "boundary_entropy": 0.08428, "bpb": 1.113853}, "hi": {"chunks": 3593.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.213551, "cps": 28.070312, "mean_p": 0.093717, "boundary_entropy": 0.081842, "bpb": 0.926848}, "ta": {"chunks": 2692.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.143996, "cps": 21.03125, "mean_p": 0.106453, "boundary_entropy": 0.223529, "bpb": 0.807585}, "th": {"chunks": 3159.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.196394, "cps": 24.679688, "mean_p": 0.135038, "boundary_entropy": 0.264425, "bpb": 0.89333}, "zh": {"chunks": 3268.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.553336, "cps": 25.53125, "mean_p": 0.21541, "boundary_entropy": 0.077393, "bpb": 2.501194}, "my": {"chunks": 2657.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.140226, "cps": 20.757812, "mean_p": 0.091706, "boundary_entropy": 0.187675, "bpb": 0.90306}}, "bpb": {"en": 1.64516, "de": 1.77824, "vi": 1.43484, "tr": 1.70299, "fi": 1.86677, "ru": 1.09963, "el": 1.11385, "hi": 0.92685, "ta": 0.80759, "th": 0.89333, "zh": 2.50119, "my": 0.90306}, "gini_cpc": 0.175121, "gini_cps": 0.072671, "bpic": 9.208902, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9409010036976811, "vi": 1.0301586097234454, "tr": 0.9624322984761365, "fi": 0.917475877410936, "ru": 1.0906887328450758, "el": 0.9690108385401387, "hi": 1.0630528117891473, "ta": 0.7168070734413624, "th": 0.9776451652370238, "zh": 2.7544906432662435, "my": 0.6980409076943812}, "t": 1756.6} |
| {"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.64418, "de": 1.7828, "vi": 1.45207, "tr": 1.69483, "fi": 1.88252, "ru": 1.09828, "el": 1.11249, "hi": 0.9326, "ta": 0.81194, "th": 0.89372, "zh": 2.5081, "my": 0.91078}, "bytes_per_chunk": {"en": 4.821, "de": 5.382, "vi": 6.297, "tr": 5.592, "fi": 5.418, "ru": 8.774, "el": 9.547, "hi": 11.706, "ta": 20.758, "th": 13.425, "zh": 4.884, "my": 18.854}, "bpb_mean": 1.39369, "t": 1761.8} |
| {"event": "train", "step": 9000, "phase": 3, "objective": "none", "lm_loss": 1.06208, "bpb": 1.53225, "chunk_loss": 0.0, "F": 0.1571, "G": 0.16547, "chunks": 160.88, "bytes_per_chunk": 6.365, "boundary_entropy": 0.08213, "router_noise": 0.0, "grad_norm": 0.404, "lr": 5.35e-05, "router_trainable": false, "sec_per_step": 0.269, "alpha": 0.0, "t": 1761.9} |
| {"event": "train", "step": 9100, "phase": 3, "objective": "none", "lm_loss": 1.04638, "bpb": 1.50961, "chunk_loss": 0.0, "F": 0.1609, "G": 0.16787, "chunks": 164.77, "bytes_per_chunk": 6.215, "boundary_entropy": 0.07614, "router_noise": 0.0, "grad_norm": 0.3822, "lr": 5.12e-05, "router_trainable": false, "sec_per_step": 0.18, "alpha": 0.0, "t": 1780.0} |
| {"event": "train", "step": 9200, "phase": 3, "objective": "none", "lm_loss": 0.97683, "bpb": 1.40927, "chunk_loss": 0.0, "F": 0.14725, "G": 0.15284, "chunks": 150.78, "bytes_per_chunk": 6.791, "boundary_entropy": 0.07582, "router_noise": 0.0, "grad_norm": 0.3784, "lr": 4.91e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 1797.5} |
| {"event": "train", "step": 9300, "phase": 3, "objective": "none", "lm_loss": 1.04085, "bpb": 1.50162, "chunk_loss": 0.0, "F": 0.1691, "G": 0.17615, "chunks": 173.16, "bytes_per_chunk": 5.914, "boundary_entropy": 0.07667, "router_noise": 0.0, "grad_norm": 0.3826, "lr": 4.71e-05, "router_trainable": false, "sec_per_step": 0.176, "alpha": 0.0, "t": 1815.2} |
| {"event": "train", "step": 9400, "phase": 3, "objective": "none", "lm_loss": 1.08874, "bpb": 1.57072, "chunk_loss": 0.0, "F": 0.17459, "G": 0.18482, "chunks": 178.78, "bytes_per_chunk": 5.728, "boundary_entropy": 0.0762, "router_noise": 0.0, "grad_norm": 0.4005, "lr": 4.52e-05, "router_trainable": false, "sec_per_step": 0.178, "alpha": 0.0, "t": 1832.9} |
| {"event": "train", "step": 9500, "phase": 3, "objective": "none", "lm_loss": 1.09853, "bpb": 1.58485, "chunk_loss": 0.0, "F": 0.16783, "G": 0.17404, "chunks": 171.86, "bytes_per_chunk": 5.958, "boundary_entropy": 0.07214, "router_noise": 0.0, "grad_norm": 0.405, "lr": 4.34e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1850.2} |
| {"event": "train", "step": 9600, "phase": 3, "objective": "none", "lm_loss": 0.95922, "bpb": 1.38386, "chunk_loss": 0.0, "F": 0.14897, "G": 0.16059, "chunks": 152.55, "bytes_per_chunk": 6.713, "boundary_entropy": 0.08947, "router_noise": 0.0, "grad_norm": 0.4004, "lr": 4.17e-05, "router_trainable": false, "sec_per_step": 0.183, "alpha": 0.0, "t": 1868.5} |
| {"event": "train", "step": 9700, "phase": 3, "objective": "none", "lm_loss": 0.97499, "bpb": 1.40661, "chunk_loss": 0.0, "F": 0.15717, "G": 0.16468, "chunks": 160.94, "bytes_per_chunk": 6.363, "boundary_entropy": 0.07945, "router_noise": 0.0, "grad_norm": 0.3889, "lr": 4.01e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 1885.7} |
| {"event": "train", "step": 9800, "phase": 3, "objective": "none", "lm_loss": 0.9987, "bpb": 1.44082, "chunk_loss": 0.0, "F": 0.15244, "G": 0.16494, "chunks": 156.09, "bytes_per_chunk": 6.56, "boundary_entropy": 0.09026, "router_noise": 0.0, "grad_norm": 0.388, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.184, "alpha": 0.0, "t": 1904.1} |
| {"event": "train", "step": 9900, "phase": 3, "objective": "none", "lm_loss": 1.06731, "bpb": 1.53981, "chunk_loss": 0.0, "F": 0.17062, "G": 0.18116, "chunks": 174.72, "bytes_per_chunk": 5.861, "boundary_entropy": 0.08038, "router_noise": 0.0, "grad_norm": 0.3947, "lr": 3.72e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 1921.3} |
| {"event": "probe", "step": 10000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3495.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200827, "cps": 27.304688, "mean_p": 0.20193, "boundary_entropy": 0.038257, "bpb": 1.631307}, "de": {"chunks": 3875.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188721, "cps": 30.273438, "mean_p": 0.185968, "boundary_entropy": 0.042084, "bpb": 1.762463}, "vi": {"chunks": 3816.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.206998, "cps": 29.8125, "mean_p": 0.150419, "boundary_entropy": 0.065143, "bpb": 1.41964}, "tr": {"chunks": 3473.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.191847, "cps": 27.132812, "mean_p": 0.181694, "boundary_entropy": 0.074701, "bpb": 1.686775}, "fi": {"chunks": 3456.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183722, "cps": 27.0, "mean_p": 0.180822, "boundary_entropy": 0.054922, "bpb": 1.846048}, "ru": {"chunks": 4166.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.219842, "cps": 32.546875, "mean_p": 0.11814, "boundary_entropy": 0.0792, "bpb": 1.085945}, "el": {"chunks": 4053.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.197486, "cps": 31.664062, "mean_p": 0.107231, "boundary_entropy": 0.087307, "bpb": 1.09707}, "hi": {"chunks": 3572.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.212303, "cps": 27.90625, "mean_p": 0.095913, "boundary_entropy": 0.087877, "bpb": 0.910003}, "ta": {"chunks": 2354.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.125916, "cps": 18.390625, "mean_p": 0.107664, "boundary_entropy": 0.231519, "bpb": 0.79637}, "th": {"chunks": 3142.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.195337, "cps": 24.546875, "mean_p": 0.134743, "boundary_entropy": 0.264872, "bpb": 0.883432}, "zh": {"chunks": 3254.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.550965, "cps": 25.421875, "mean_p": 0.215015, "boundary_entropy": 0.078747, "bpb": 2.483311}, "my": {"chunks": 2814.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.148512, "cps": 21.984375, "mean_p": 0.100437, "boundary_entropy": 0.208083, "bpb": 0.8916}}, "bpb": {"en": 1.63131, "de": 1.76246, "vi": 1.41964, "tr": 1.68678, "fi": 1.84605, "ru": 1.08595, "el": 1.09707, "hi": 0.91, "ta": 0.79637, "th": 0.88343, "zh": 2.48331, "my": 0.8916}, "gini_cpc": 0.179158, "gini_cps": 0.079162, "bpic": 9.254425, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9397151742959987, "vi": 1.0307234675597243, "tr": 0.9552810704791538, "fi": 0.9148266138682896, "ru": 1.0946795157801759, "el": 0.983360362570289, "hi": 1.0571419752692766, "ta": 0.6269861235303098, "th": 0.9726622382968163, "zh": 2.7434752466757453, "my": 0.7394991125474645}, "t": 1942.4} |
| {"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.62971, "de": 1.76821, "vi": 1.43699, "tr": 1.67913, "fi": 1.86334, "ru": 1.08442, "el": 1.09541, "hi": 0.91628, "ta": 0.80055, "th": 0.88338, "zh": 2.49053, "my": 0.89879}, "bytes_per_chunk": {"en": 4.827, "de": 5.383, "vi": 6.29, "tr": 5.629, "fi": 5.433, "ru": 8.754, "el": 9.452, "hi": 11.804, "ta": 22.819, "th": 13.663, "zh": 4.896, "my": 18.859}, "bpb_mean": 1.37889, "t": 1947.5} |
| {"event": "train", "step": 10000, "phase": 3, "objective": "none", "lm_loss": 1.03813, "bpb": 1.4977, "chunk_loss": 0.0, "F": 0.15765, "G": 0.16747, "chunks": 161.44, "bytes_per_chunk": 6.343, "boundary_entropy": 0.08189, "router_noise": 0.0, "grad_norm": 0.4116, "lr": 3.6e-05, "router_trainable": false, "sec_per_step": 0.263, "alpha": 0.0, "t": 1947.7} |
| {"event": "train", "step": 10100, "phase": 3, "objective": "none", "lm_loss": 0.99261, "bpb": 1.43203, "chunk_loss": 0.0, "F": 0.145, "G": 0.16159, "chunks": 148.48, "bytes_per_chunk": 6.896, "boundary_entropy": 0.1052, "router_noise": 0.0, "grad_norm": 0.3978, "lr": 3.49e-05, "router_trainable": false, "sec_per_step": 0.187, "alpha": 0.0, "t": 1966.3} |
| {"event": "train", "step": 10200, "phase": 3, "objective": "none", "lm_loss": 1.07292, "bpb": 1.54789, "chunk_loss": 0.0, "F": 0.16919, "G": 0.17535, "chunks": 173.25, "bytes_per_chunk": 5.911, "boundary_entropy": 0.07138, "router_noise": 0.0, "grad_norm": 0.4213, "lr": 3.38e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 1983.5} |
| {"event": "train", "step": 10300, "phase": 3, "objective": "none", "lm_loss": 1.03013, "bpb": 1.48617, "chunk_loss": 0.0, "F": 0.16245, "G": 0.17044, "chunks": 166.34, "bytes_per_chunk": 6.156, "boundary_entropy": 0.07598, "router_noise": 0.0, "grad_norm": 0.4372, "lr": 3.3e-05, "router_trainable": false, "sec_per_step": 0.183, "alpha": 0.0, "t": 2001.8} |
| {"event": "train", "step": 10400, "phase": 3, "objective": "none", "lm_loss": 1.01088, "bpb": 1.45839, "chunk_loss": 0.0, "F": 0.16122, "G": 0.16942, "chunks": 165.09, "bytes_per_chunk": 6.203, "boundary_entropy": 0.07611, "router_noise": 0.0, "grad_norm": 0.4064, "lr": 3.22e-05, "router_trainable": false, "sec_per_step": 0.172, "alpha": 0.0, "t": 2018.9} |
| {"event": "train", "step": 10500, "phase": 3, "objective": "none", "lm_loss": 1.03047, "bpb": 1.48666, "chunk_loss": 0.0, "F": 0.1628, "G": 0.17073, "chunks": 166.7, "bytes_per_chunk": 6.143, "boundary_entropy": 0.07678, "router_noise": 0.0, "grad_norm": 0.4006, "lr": 3.15e-05, "router_trainable": false, "sec_per_step": 0.184, "alpha": 0.0, "t": 2037.3} |
| {"event": "train", "step": 10600, "phase": 3, "objective": "none", "lm_loss": 1.05253, "bpb": 1.51848, "chunk_loss": 0.0, "F": 0.16405, "G": 0.17212, "chunks": 167.98, "bytes_per_chunk": 6.096, "boundary_entropy": 0.07428, "router_noise": 0.0, "grad_norm": 0.3968, "lr": 3.1e-05, "router_trainable": false, "sec_per_step": 0.171, "alpha": 0.0, "t": 2054.4} |
| {"event": "train", "step": 10700, "phase": 3, "objective": "none", "lm_loss": 1.0194, "bpb": 1.47069, "chunk_loss": 0.0, "F": 0.15916, "G": 0.17222, "chunks": 162.98, "bytes_per_chunk": 6.283, "boundary_entropy": 0.09201, "router_noise": 0.0, "grad_norm": 0.4007, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.182, "alpha": 0.0, "t": 2072.6} |
| {"event": "train", "step": 10800, "phase": 3, "objective": "none", "lm_loss": 1.04527, "bpb": 1.508, "chunk_loss": 0.0, "F": 0.16507, "G": 0.17302, "chunks": 169.03, "bytes_per_chunk": 6.058, "boundary_entropy": 0.07628, "router_noise": 0.0, "grad_norm": 0.3967, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.173, "alpha": 0.0, "t": 2089.8} |
| {"event": "train", "step": 10900, "phase": 3, "objective": "none", "lm_loss": 1.03356, "bpb": 1.49111, "chunk_loss": 0.0, "F": 0.15242, "G": 0.16262, "chunks": 156.08, "bytes_per_chunk": 6.561, "boundary_entropy": 0.08819, "router_noise": 0.0, "grad_norm": 0.4196, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.181, "alpha": 0.0, "t": 2107.9} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3499.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201057, "cps": 27.335938, "mean_p": 0.202288, "boundary_entropy": 0.038236, "bpb": 1.62233}, "de": {"chunks": 3895.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189695, "cps": 30.429688, "mean_p": 0.186891, "boundary_entropy": 0.040954, "bpb": 1.752106}, "vi": {"chunks": 3830.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.207757, "cps": 29.921875, "mean_p": 0.152054, "boundary_entropy": 0.063252, "bpb": 1.407669}, "tr": {"chunks": 3590.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.19831, "cps": 28.046875, "mean_p": 0.187253, "boundary_entropy": 0.077387, "bpb": 1.677584}, "fi": {"chunks": 3470.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184467, "cps": 27.109375, "mean_p": 0.181537, "boundary_entropy": 0.053764, "bpb": 1.834746}, "ru": {"chunks": 4199.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.221583, "cps": 32.804688, "mean_p": 0.118998, "boundary_entropy": 0.079443, "bpb": 1.076992}, "el": {"chunks": 4091.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.199337, "cps": 31.960938, "mean_p": 0.108107, "boundary_entropy": 0.086452, "bpb": 1.088298}, "hi": {"chunks": 3563.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.211768, "cps": 27.835938, "mean_p": 0.095414, "boundary_entropy": 0.085636, "bpb": 0.901968}, "ta": {"chunks": 2706.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.144745, "cps": 21.140625, "mean_p": 0.109787, "boundary_entropy": 0.232005, "bpb": 0.792269}, "th": {"chunks": 3197.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.198757, "cps": 24.976562, "mean_p": 0.141655, "boundary_entropy": 0.277403, "bpb": 0.878793}, "zh": {"chunks": 3261.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.55215, "cps": 25.476562, "mean_p": 0.215736, "boundary_entropy": 0.07932, "bpb": 2.472422}, "my": {"chunks": 2701.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.142548, "cps": 21.101562, "mean_p": 0.093033, "boundary_entropy": 0.190448, "bpb": 0.884067}}, "bpb": {"en": 1.62233, "de": 1.75211, "vi": 1.40767, "tr": 1.67758, "fi": 1.83475, "ru": 1.07699, "el": 1.0883, "hi": 0.90197, "ta": 0.79227, "th": 0.87879, "zh": 2.47242, "my": 0.88407}, "gini_cpc": 0.172419, "gini_cps": 0.073604, "bpic": 9.137208, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9434855053938572, "vi": 1.0333223185236464, "tr": 0.9863341609885714, "fi": 0.917482457262902, "ru": 1.1020894322614603, "el": 0.9914454226882707, "hi": 1.0532729410695563, "ta": 0.719917118412543, "th": 0.9885570751908067, "zh": 2.7462339634167647, "my": 0.7089920833257163}, "t": 2128.6} |
| {"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.62101, "de": 1.75781, "vi": 1.42479, "tr": 1.66946, "fi": 1.85108, "ru": 1.07597, "el": 1.08696, "hi": 0.90861, "ta": 0.79646, "th": 0.87853, "zh": 2.47911, "my": 0.89103}, "bytes_per_chunk": {"en": 4.823, "de": 5.355, "vi": 6.282, "tr": 5.483, "fi": 5.409, "ru": 8.702, "el": 9.386, "hi": 11.81, "ta": 20.258, "th": 13.281, "zh": 4.885, "my": 18.418}, "bpb_mean": 1.37007, "t": 2133.6} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3499.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201057, "cps": 27.335938, "mean_p": 0.202288, "boundary_entropy": 0.038236, "bpb": 1.62233}, "de": {"chunks": 3895.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189695, "cps": 30.429688, "mean_p": 0.186891, "boundary_entropy": 0.040954, "bpb": 1.752106}, "vi": {"chunks": 3830.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.207757, "cps": 29.921875, "mean_p": 0.152054, "boundary_entropy": 0.063252, "bpb": 1.407669}, "tr": {"chunks": 3590.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.19831, "cps": 28.046875, "mean_p": 0.187253, "boundary_entropy": 0.077387, "bpb": 1.677584}, "fi": {"chunks": 3470.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184467, "cps": 27.109375, "mean_p": 0.181537, "boundary_entropy": 0.053764, "bpb": 1.834746}, "ru": {"chunks": 4199.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.221583, "cps": 32.804688, "mean_p": 0.118998, "boundary_entropy": 0.079443, "bpb": 1.076992}, "el": {"chunks": 4091.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.199337, "cps": 31.960938, "mean_p": 0.108107, "boundary_entropy": 0.086452, "bpb": 1.088298}, "hi": {"chunks": 3563.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.211768, "cps": 27.835938, "mean_p": 0.095414, "boundary_entropy": 0.085636, "bpb": 0.901968}, "ta": {"chunks": 2706.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.144745, "cps": 21.140625, "mean_p": 0.109787, "boundary_entropy": 0.232005, "bpb": 0.792269}, "th": {"chunks": 3197.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.198757, "cps": 24.976562, "mean_p": 0.141655, "boundary_entropy": 0.277403, "bpb": 0.878793}, "zh": {"chunks": 3261.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.55215, "cps": 25.476562, "mean_p": 0.215736, "boundary_entropy": 0.07932, "bpb": 2.472422}, "my": {"chunks": 2701.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.142548, "cps": 21.101562, "mean_p": 0.093033, "boundary_entropy": 0.190448, "bpb": 0.884067}}, "bpb": {"en": 1.62233, "de": 1.75211, "vi": 1.40767, "tr": 1.67758, "fi": 1.83475, "ru": 1.07699, "el": 1.0883, "hi": 0.90197, "ta": 0.79227, "th": 0.87879, "zh": 2.47242, "my": 0.88407}, "gini_cpc": 0.172419, "gini_cps": 0.073604, "bpic": 9.137208, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9434855053938572, "vi": 1.0333223185236464, "tr": 0.9863341609885714, "fi": 0.917482457262902, "ru": 1.1020894322614603, "el": 0.9914454226882707, "hi": 1.0532729410695563, "ta": 0.719917118412543, "th": 0.9885570751908067, "zh": 2.7462339634167647, "my": 0.7089920833257163}, "t": 2136.6} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s0/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s0/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s0/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s0/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s0/th0.70"], "t": 2139.4} |
| {"event": "final", "run": "spec_E_s0", "condition": "E", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 0.20105728897316555, "cps": 27.3359375, "bpb": 1.6162082621020413, "bytes_per_chunk": 4.827097920608396}, "de": {"cpc": 0.1896946378999659, "cps": 30.4296875, "bpb": 1.7554358604545326, "bytes_per_chunk": 5.331624896400588}, "vi": {"cpc": 0.2077569839978302, "cps": 29.921875, "bpb": 1.4236625002143632, "bytes_per_chunk": 6.283827433495736}, "tr": {"cpc": 0.19830967242998399, "cps": 28.046875, "bpb": 1.6690557359941058, "bytes_per_chunk": 5.4882888352690555}, "fi": {"cpc": 0.1844665355377173, "cps": 27.109375, "bpb": 1.857616751227076, "bytes_per_chunk": 5.437036576133735}, "ru": {"cpc": 0.2215831134564644, "cps": 32.8046875, "bpb": 1.0764984662930592, "bytes_per_chunk": 8.710622284617749}, "el": {"cpc": 0.1993373288505579, "cps": 31.9609375, "bpb": 1.0852239523400056, "bytes_per_chunk": 9.413386957770756}, "hi": {"cpc": 0.21176820208023775, "cps": 27.8359375, "bpb": 0.9093263314086186, "bytes_per_chunk": 11.830937605776825}, "ta": {"cpc": 0.1447445841133993, "cps": 21.140625, "bpb": 0.7950613537705497, "bytes_per_chunk": 19.940591423409717}, "th": {"cpc": 0.19875660553310537, "cps": 24.9765625, "bpb": 0.8809366285338265, "bytes_per_chunk": 13.210406299212599}, "zh": {"cpc": 0.5521503555706062, "cps": 25.4765625, "bpb": 2.479438825275083, "bytes_per_chunk": 4.897347637440393}, "my": {"cpc": 0.1425480261769052, "cps": 21.1015625, "bpb": 0.8899300259296121, "bytes_per_chunk": 18.51430185747581}}, "gini_cpc": 0.17241895181531097, "gini_cps": 0.07360363792200375, "bpic": 9.13720775201181, "seed": 0, "mean_boundary_entropy": 0.08051501214504242, "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": 2149.1} |
|
|
