| {"event": "start", "run": "pilot_D_s2", "condition": "D", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "parity_a", "p2_objective": "parity_a", "p3_objective": "parity_a", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 1.4} |
| {"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 1.4} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 10933.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.628225, "cps": 85.414062, "mean_p": 0.513249, "boundary_entropy": 0.686295, "bpb": 8.12598}, "de": {"chunks": 16287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.793211, "cps": 127.242188, "mean_p": 0.524631, "boundary_entropy": 0.686391, "bpb": 8.12701}, "vi": {"chunks": 5099.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276593, "cps": 39.835938, "mean_p": 0.481643, "boundary_entropy": 0.68728, "bpb": 7.869214}, "tr": {"chunks": 8187.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.452245, "cps": 63.960938, "mean_p": 0.498002, "boundary_entropy": 0.686826, "bpb": 8.097536}, "fi": {"chunks": 13540.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.719792, "cps": 105.78125, "mean_p": 0.517169, "boundary_entropy": 0.686661, "bpb": 8.065117}, "ru": {"chunks": 33477.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.766596, "cps": 261.539062, "mean_p": 0.530376, "boundary_entropy": 0.688543, "bpb": 8.240733}, "el": {"chunks": 20696.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.00843, "cps": 161.6875, "mean_p": 0.503648, "boundary_entropy": 0.690112, "bpb": 8.286272}, "hi": {"chunks": 875.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.052006, "cps": 6.835938, "mean_p": 0.461978, "boundary_entropy": 0.68743, "bpb": 8.022688}, "ta": {"chunks": 1261.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.067451, "cps": 9.851562, "mean_p": 0.477518, "boundary_entropy": 0.690053, "bpb": 8.091383}, "th": {"chunks": 3630.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.225676, "cps": 28.359375, "mean_p": 0.484845, "boundary_entropy": 0.690398, "bpb": 7.721507}, "zh": {"chunks": 11722.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.984761, "cps": 91.578125, "mean_p": 0.520316, "boundary_entropy": 0.68562, "bpb": 8.130734}, "my": {"chunks": 1628.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085919, "cps": 12.71875, "mean_p": 0.474628, "boundary_entropy": 0.689685, "bpb": 7.8029}}, "bpb": {"en": 8.12598, "de": 8.12701, "vi": 7.86921, "tr": 8.09754, "fi": 8.06512, "ru": 8.24073, "el": 8.28627, "hi": 8.02269, "ta": 8.09138, "th": 7.72151, "zh": 8.13073, "my": 7.8029}, "gini_cpc": 0.489538, "gini_cps": 0.463551, "bpic": 3.013947, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2626223171148012, "vi": 0.440277652395235, "tr": 0.7198781817505053, "fi": 1.1457544508645294, "ru": 2.8120438593717987, "el": 1.6052044042347315, "hi": 0.08278235025738971, "ta": 0.10736788276924242, "th": 0.3592281253258146, "zh": 3.159315851384177, "my": 0.13676526035979295}, "t": 3.2} |
| {"event": "train", "step": 0, "phase": 1, "objective": "parity_a", "lm_loss": 5.64089, "bpb": 8.13808, "chunk_loss": 0.088392, "F": 0.62439, "G": 0.5091, "chunks": 639.38, "bytes_per_chunk": 1.602, "boundary_entropy": 0.69066, "router_noise": 0.0, "grad_norm": 7.8727, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "n_target_mean": 8.4968, "alpha": 0.03, "t": 3.5} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 10921.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.627535, "cps": 85.320312, "mean_p": 0.513208, "boundary_entropy": 0.686301, "bpb": 8.100839}, "de": {"chunks": 16285.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.793114, "cps": 127.226562, "mean_p": 0.52459, "boundary_entropy": 0.686398, "bpb": 8.104407}, "vi": {"chunks": 5100.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276648, "cps": 39.84375, "mean_p": 0.481626, "boundary_entropy": 0.687278, "bpb": 7.854785}, "tr": {"chunks": 8207.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.45335, "cps": 64.117188, "mean_p": 0.498087, "boundary_entropy": 0.686828, "bpb": 8.081346}, "fi": {"chunks": 13521.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.718782, "cps": 105.632812, "mean_p": 0.517089, "boundary_entropy": 0.686669, "bpb": 8.042906}, "ru": {"chunks": 33458.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.765594, "cps": 261.390625, "mean_p": 0.530219, "boundary_entropy": 0.688561, "bpb": 8.19577}, "el": {"chunks": 20658.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.006578, "cps": 161.390625, "mean_p": 0.503601, "boundary_entropy": 0.690113, "bpb": 8.279542}, "hi": {"chunks": 875.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.052006, "cps": 6.835938, "mean_p": 0.462019, "boundary_entropy": 0.687436, "bpb": 8.01737}, "ta": {"chunks": 1278.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.068361, "cps": 9.984375, "mean_p": 0.477585, "boundary_entropy": 0.69006, "bpb": 8.085566}, "th": {"chunks": 3635.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.225987, "cps": 28.398438, "mean_p": 0.484851, "boundary_entropy": 0.690398, "bpb": 7.718648}, "zh": {"chunks": 11708.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.982391, "cps": 91.46875, "mean_p": 0.520264, "boundary_entropy": 0.685623, "bpb": 8.12678}, "my": {"chunks": 1620.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085497, "cps": 12.65625, "mean_p": 0.474597, "boundary_entropy": 0.689682, "bpb": 7.801136}}, "bpb": {"en": 8.10084, "de": 8.10441, "vi": 7.85479, "tr": 8.08135, "fi": 8.04291, "ru": 8.19577, "el": 8.27954, "hi": 8.01737, "ta": 8.08557, "th": 7.71865, "zh": 8.12678, "my": 7.80114}, "gini_cpc": 0.489222, "gini_cps": 0.46324, "bpic": 3.015582, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2638544701121932, "vi": 0.4408478704482351, "tr": 0.7224297050265782, "fi": 1.1454038602617196, "ru": 2.8135359945340235, "el": 1.6040176435029043, "hi": 0.08287331154326909, "ta": 0.10893491463492375, "th": 0.360118193632815, "zh": 3.159009880180949, "my": 0.13624273446600405}, "t": 4.9} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 10903.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.626501, "cps": 85.179688, "mean_p": 0.513147, "boundary_entropy": 0.686312, "bpb": 8.0526}, "de": {"chunks": 16261.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.791945, "cps": 127.039062, "mean_p": 0.524514, "boundary_entropy": 0.686412, "bpb": 8.06017}, "vi": {"chunks": 5087.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.275943, "cps": 39.742188, "mean_p": 0.481569, "boundary_entropy": 0.687273, "bpb": 7.831417}, "tr": {"chunks": 8253.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.455891, "cps": 64.476562, "mean_p": 0.498257, "boundary_entropy": 0.686833, "bpb": 8.048594}, "fi": {"chunks": 13502.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.717772, "cps": 105.484375, "mean_p": 0.516953, "boundary_entropy": 0.686682, "bpb": 7.998111}, "ru": {"chunks": 33428.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.764011, "cps": 261.15625, "mean_p": 0.529943, "boundary_entropy": 0.68859, "bpb": 8.105313}, "el": {"chunks": 20589.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.003216, "cps": 160.851562, "mean_p": 0.503533, "boundary_entropy": 0.690113, "bpb": 8.273413}, "hi": {"chunks": 903.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.05367, "cps": 7.054688, "mean_p": 0.462245, "boundary_entropy": 0.687473, "bpb": 7.989887}, "ta": {"chunks": 1315.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.07034, "cps": 10.273438, "mean_p": 0.477777, "boundary_entropy": 0.690078, "bpb": 8.070684}, "th": {"chunks": 3686.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.229158, "cps": 28.796875, "mean_p": 0.484951, "boundary_entropy": 0.690404, "bpb": 7.702839}, "zh": {"chunks": 11684.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.978327, "cps": 91.28125, "mean_p": 0.520164, "boundary_entropy": 0.685629, "bpb": 8.118672}, "my": {"chunks": 1619.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085444, "cps": 12.648438, "mean_p": 0.474592, "boundary_entropy": 0.689681, "bpb": 7.800515}}, "bpb": {"en": 8.0526, "de": 8.06017, "vi": 7.83142, "tr": 8.04859, "fi": 7.99811, "ru": 8.10531, "el": 8.27341, "hi": 7.98989, "ta": 8.07068, "th": 7.70284, "zh": 8.11867, "my": 7.80052}, "gini_cpc": 0.487947, "gini_cps": 0.46193, "bpic": 3.016435, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2640753158804874, "vi": 0.4404500907365169, "tr": 0.7276782626693873, "fi": 1.145682629050223, "ru": 2.8156540101644705, "el": 1.6012993102067623, "hi": 0.08566645302170879, "ta": 0.11227379211126676, "th": 0.36577361433805, "zh": 3.1577388751388, "my": 0.1363834212649667}, "t": 6.4} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 10866.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.624375, "cps": 84.890625, "mean_p": 0.512827, "boundary_entropy": 0.686352, "bpb": 7.898092}, "de": {"chunks": 16248.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.791312, "cps": 126.9375, "mean_p": 0.524166, "boundary_entropy": 0.686468, "bpb": 7.92194}, "vi": {"chunks": 5065.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.274749, "cps": 39.570312, "mean_p": 0.481487, "boundary_entropy": 0.687265, "bpb": 7.752845}, "tr": {"chunks": 8390.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.463459, "cps": 65.546875, "mean_p": 0.498758, "boundary_entropy": 0.686848, "bpb": 7.941238}, "fi": {"chunks": 13408.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.712774, "cps": 104.75, "mean_p": 0.516411, "boundary_entropy": 0.686731, "bpb": 7.856397}, "ru": {"chunks": 33264.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.755356, "cps": 259.875, "mean_p": 0.528984, "boundary_entropy": 0.688691, "bpb": 7.817429}, "el": {"chunks": 20039.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.976417, "cps": 156.554688, "mean_p": 0.502891, "boundary_entropy": 0.69012, "bpb": 8.2182}, "hi": {"chunks": 989.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.058782, "cps": 7.726562, "mean_p": 0.463001, "boundary_entropy": 0.687591, "bpb": 7.903021}, "ta": {"chunks": 1432.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.076598, "cps": 11.1875, "mean_p": 0.478483, "boundary_entropy": 0.690143, "bpb": 7.987464}, "th": {"chunks": 3889.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.241778, "cps": 30.382812, "mean_p": 0.485177, "boundary_entropy": 0.690414, "bpb": 7.619674}, "zh": {"chunks": 11603.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.964612, "cps": 90.648438, "mean_p": 0.519718, "boundary_entropy": 0.685656, "bpb": 8.088012}, "my": {"chunks": 1602.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.084547, "cps": 12.515625, "mean_p": 0.474492, "boundary_entropy": 0.689668, "bpb": 7.794086}}, "bpb": {"en": 7.89809, "de": 7.92194, "vi": 7.75285, "tr": 7.94124, "fi": 7.8564, "ru": 7.81743, "el": 8.2182, "hi": 7.90302, "ta": 7.98746, "th": 7.61967, "zh": 8.08801, "my": 7.79409}, "gini_cpc": 0.483282, "gini_cps": 0.456574, "bpic": 3.026783, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2673656227737318, "vi": 0.4400385523359718, "tr": 0.7422767010285003, "fi": 1.1415804884773033, "ru": 2.8113808170240793, "el": 1.5638302858440625, "hi": 0.09414464847887641, "ta": 0.12267949172163987, "th": 0.3872320518394566, "zh": 3.146525597138157, "my": 0.13541087835264687}, "t": 7.9} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 10165.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.584095, "cps": 79.414062, "mean_p": 0.509553, "boundary_entropy": 0.686564, "bpb": 7.422785}, "de": {"chunks": 15636.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.761506, "cps": 122.15625, "mean_p": 0.520677, "boundary_entropy": 0.686853, "bpb": 7.489088}, "vi": {"chunks": 4860.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.263629, "cps": 37.96875, "mean_p": 0.480732, "boundary_entropy": 0.687196, "bpb": 7.511777}, "tr": {"chunks": 8675.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.479202, "cps": 67.773438, "mean_p": 0.49982, "boundary_entropy": 0.686899, "bpb": 7.589607}, "fi": {"chunks": 12537.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.666472, "cps": 97.945312, "mean_p": 0.512654, "boundary_entropy": 0.686985, "bpb": 7.426389}, "ru": {"chunks": 32572.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.718839, "cps": 254.46875, "mean_p": 0.5255, "boundary_entropy": 0.689027, "bpb": 7.200721}, "el": {"chunks": 17645.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.859767, "cps": 137.851562, "mean_p": 0.499933, "boundary_entropy": 0.690127, "bpb": 7.948304}, "hi": {"chunks": 1230.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.073105, "cps": 9.609375, "mean_p": 0.464975, "boundary_entropy": 0.687894, "bpb": 7.609637}, "ta": {"chunks": 1853.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.099117, "cps": 14.476562, "mean_p": 0.480197, "boundary_entropy": 0.690292, "bpb": 7.731336}, "th": {"chunks": 4298.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.267205, "cps": 33.578125, "mean_p": 0.485605, "boundary_entropy": 0.690429, "bpb": 7.33699}, "zh": {"chunks": 11257.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.906028, "cps": 87.945312, "mean_p": 0.517943, "boundary_entropy": 0.685758, "bpb": 7.997714}, "my": {"chunks": 1448.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.07642, "cps": 11.3125, "mean_p": 0.473896, "boundary_entropy": 0.689609, "bpb": 7.664098}}, "bpb": {"en": 7.42279, "de": 7.48909, "vi": 7.51178, "tr": 7.58961, "fi": 7.42639, "ru": 7.20072, "el": 7.9483, "hi": 7.60964, "ta": 7.73134, "th": 7.33699, "zh": 7.99771, "my": 7.6641}, "gini_cpc": 0.47344, "gini_cps": 0.442594, "bpic": 3.141214, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.303737002584996, "vi": 0.45134627707112546, "tr": 0.820418923806467, "fi": 1.1410337218573878, "ru": 2.942740382631177, "el": 1.4719652412578799, "hi": 0.12516035196377165, "ta": 0.16969407819914972, "th": 0.4574694353854592, "zh": 3.263217044069168, "my": 0.1308343927476561}, "t": 9.7} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 4846.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.278458, "cps": 37.859375, "mean_p": 0.488824, "boundary_entropy": 0.686425, "bpb": 6.856665}, "de": {"chunks": 8530.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.415429, "cps": 66.640625, "mean_p": 0.497545, "boundary_entropy": 0.687716, "bpb": 6.930522}, "vi": {"chunks": 3468.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.18812, "cps": 27.09375, "mean_p": 0.475102, "boundary_entropy": 0.686655, "bpb": 7.16775}, "tr": {"chunks": 7552.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.417168, "cps": 59.0, "mean_p": 0.495271, "boundary_entropy": 0.686956, "bpb": 7.150403}, "fi": {"chunks": 6411.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.340811, "cps": 50.085938, "mean_p": 0.492308, "boundary_entropy": 0.687153, "bpb": 6.916448}, "ru": {"chunks": 12402.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.654459, "cps": 96.890625, "mean_p": 0.496361, "boundary_entropy": 0.690059, "bpb": 6.387331}, "el": {"chunks": 10738.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.523218, "cps": 83.890625, "mean_p": 0.491205, "boundary_entropy": 0.689941, "bpb": 7.328564}, "hi": {"chunks": 2894.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.172006, "cps": 22.609375, "mean_p": 0.473492, "boundary_entropy": 0.689041, "bpb": 6.587031}, "ta": {"chunks": 4873.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.260658, "cps": 38.070312, "mean_p": 0.486832, "boundary_entropy": 0.69076, "bpb": 6.771387}, "th": {"chunks": 5247.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.326205, "cps": 40.992188, "mean_p": 0.486664, "boundary_entropy": 0.690481, "bpb": 6.496658}, "zh": {"chunks": 9261.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.568066, "cps": 72.351562, "mean_p": 0.508713, "boundary_entropy": 0.686103, "bpb": 7.74809}, "my": {"chunks": 1131.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.05969, "cps": 8.835938, "mean_p": 0.472691, "boundary_entropy": 0.689505, "bpb": 7.218873}}, "bpb": {"en": 6.85667, "de": 6.93052, "vi": 7.16775, "tr": 7.1504, "fi": 6.91645, "ru": 6.38733, "el": 7.32856, "hi": 6.58703, "ta": 6.77139, "th": 6.49666, "zh": 7.74809, "my": 7.21887}, "gini_cpc": 0.392112, "gini_cps": 0.283063, "bpic": 4.961424, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.4918918257251512, "vi": 0.6755798025902433, "tr": 1.4981391049141488, "fi": 1.223924430812549, "ru": 2.350299580645899, "el": 1.8789847912096609, "hi": 0.6177093345778047, "ta": 0.9360771676709241, "th": 1.1714687539409332, "zh": 5.631254455726034, "my": 0.21435811989648496}, "t": 11.6} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 5280.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.303396, "cps": 41.25, "mean_p": 0.494509, "boundary_entropy": 0.687151, "bpb": 6.380641}, "de": {"chunks": 6930.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.337505, "cps": 54.140625, "mean_p": 0.495815, "boundary_entropy": 0.688108, "bpb": 6.461395}, "vi": {"chunks": 10246.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.555791, "cps": 80.046875, "mean_p": 0.497823, "boundary_entropy": 0.688439, "bpb": 6.774463}, "tr": {"chunks": 10404.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.574711, "cps": 81.28125, "mean_p": 0.504459, "boundary_entropy": 0.687567, "bpb": 6.735745}, "fi": {"chunks": 4945.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.262878, "cps": 38.632812, "mean_p": 0.492202, "boundary_entropy": 0.687684, "bpb": 6.468772}, "ru": {"chunks": 160.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.008443, "cps": 1.25, "mean_p": 0.45523, "boundary_entropy": 0.685889, "bpb": 6.208371}, "el": {"chunks": 5490.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.267505, "cps": 42.890625, "mean_p": 0.485809, "boundary_entropy": 0.68978, "bpb": 6.26912}, "hi": {"chunks": 5421.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.322199, "cps": 42.351562, "mean_p": 0.486093, "boundary_entropy": 0.690254, "bpb": 5.749721}, "ta": {"chunks": 2890.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.154587, "cps": 22.578125, "mean_p": 0.481806, "boundary_entropy": 0.690403, "bpb": 5.965152}, "th": {"chunks": 807.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.050171, "cps": 6.304688, "mean_p": 0.469732, "boundary_entropy": 0.688932, "bpb": 5.863124}, "zh": {"chunks": 3963.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.671013, "cps": 30.960938, "mean_p": 0.485639, "boundary_entropy": 0.685618, "bpb": 7.347065}, "my": {"chunks": 1358.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.07167, "cps": 10.609375, "mean_p": 0.479536, "boundary_entropy": 0.690341, "bpb": 5.975501}}, "bpb": {"en": 6.38064, "de": 6.46139, "vi": 6.77446, "tr": 6.73574, "fi": 6.46877, "ru": 6.20837, "el": 6.26912, "hi": 5.74972, "ta": 5.96515, "th": 5.86312, "zh": 7.34706, "my": 5.9755}, "gini_cpc": 0.384707, "gini_cps": 0.366112, "bpic": 6.629029, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1124257293137874, "vi": 1.8318985008753113, "tr": 1.8942617496848837, "fi": 0.866452202773039, "ru": 0.02782921563924202, "el": 0.8817017381828813, "hi": 1.0619755842226124, "ta": 0.5095215662914245, "th": 0.16536464662164072, "zh": 2.2116725479481576, "my": 0.2362253991786132}, "t": 13.8} |
| {"event": "train", "step": 50, "phase": 1, "objective": "parity_a", "lm_loss": 4.27196, "bpb": 6.16313, "chunk_loss": 0.030331, "F": 0.12454, "G": 0.48398, "chunks": 127.53, "bytes_per_chunk": 8.029, "boundary_entropy": 0.69145, "router_noise": 0.0, "grad_norm": 1.581, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.233, "n_target_mean": 8.8017, "alpha": 0.03, "t": 15.2} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3043.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.174855, "cps": 23.773438, "mean_p": 0.488207, "boundary_entropy": 0.687001, "bpb": 5.811097}, "de": {"chunks": 4294.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.209127, "cps": 33.546875, "mean_p": 0.489642, "boundary_entropy": 0.687979, "bpb": 5.942107}, "vi": {"chunks": 3038.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.164795, "cps": 23.734375, "mean_p": 0.47833, "boundary_entropy": 0.687448, "bpb": 6.261936}, "tr": {"chunks": 3153.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.17417, "cps": 24.632812, "mean_p": 0.484003, "boundary_entropy": 0.687126, "bpb": 6.246337}, "fi": {"chunks": 1151.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.061188, "cps": 8.992188, "mean_p": 0.477344, "boundary_entropy": 0.686753, "bpb": 5.931646}, "ru": {"chunks": 4680.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.246966, "cps": 36.5625, "mean_p": 0.491627, "boundary_entropy": 0.690185, "bpb": 5.503375}, "el": {"chunks": 4535.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.220972, "cps": 35.429688, "mean_p": 0.488657, "boundary_entropy": 0.69019, "bpb": 5.670446}, "hi": {"chunks": 1047.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.062229, "cps": 8.179688, "mean_p": 0.472222, "boundary_entropy": 0.689092, "bpb": 5.429053}, "ta": {"chunks": 228.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.012196, "cps": 1.78125, "mean_p": 0.450515, "boundary_entropy": 0.686024, "bpb": 5.607345}, "th": {"chunks": 2614.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.162512, "cps": 20.421875, "mean_p": 0.477447, "boundary_entropy": 0.689712, "bpb": 5.080296}, "zh": {"chunks": 4833.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.81832, "cps": 37.757812, "mean_p": 0.493807, "boundary_entropy": 0.686772, "bpb": 6.96459}, "my": {"chunks": 1696.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.089508, "cps": 13.25, "mean_p": 0.485492, "boundary_entropy": 0.690869, "bpb": 5.235904}}, "bpb": {"en": 5.8111, "de": 5.94211, "vi": 6.26194, "tr": 6.24634, "fi": 5.93165, "ru": 5.50337, "el": 5.67045, "hi": 5.42905, "ta": 5.60734, "th": 5.0803, "zh": 6.96459, "my": 5.2359}, "gini_cpc": 0.427476, "gini_cps": 0.295455, "bpic": 11.185037, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.196001710510844, "vi": 0.9424683950974464, "tr": 0.996083135410077, "fi": 0.3499335704825731, "ru": 1.412403569939053, "el": 1.2637425666504123, "hi": 0.3558883540964843, "ta": 0.06974796570542734, "th": 0.9294085979930599, "zh": 4.679996414414055, "my": 0.5118994226098915}, "t": 16.9} |
| {"event": "train", "step": 100, "phase": 1, "objective": "parity_a", "lm_loss": 3.44901, "bpb": 4.97586, "chunk_loss": 0.034541, "F": 0.14859, "G": 0.47054, "chunks": 152.16, "bytes_per_chunk": 6.73, "boundary_entropy": 0.68882, "router_noise": 0.0, "grad_norm": 2.2537, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.075, "n_target_mean": 9.954, "alpha": 0.03, "t": 18.9} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3105.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.178418, "cps": 24.257812, "mean_p": 0.441543, "boundary_entropy": 0.672397, "bpb": 4.461601}, "de": {"chunks": 3487.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.169824, "cps": 27.242188, "mean_p": 0.445346, "boundary_entropy": 0.675634, "bpb": 4.663916}, "vi": {"chunks": 3882.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.210578, "cps": 30.328125, "mean_p": 0.455792, "boundary_entropy": 0.680509, "bpb": 5.026579}, "tr": {"chunks": 3997.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.220792, "cps": 31.226562, "mean_p": 0.457018, "boundary_entropy": 0.678591, "bpb": 5.024892}, "fi": {"chunks": 2016.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.107171, "cps": 15.75, "mean_p": 0.443337, "boundary_entropy": 0.676517, "bpb": 4.705446}, "ru": {"chunks": 4062.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.214354, "cps": 31.734375, "mean_p": 0.485825, "boundary_entropy": 0.689435, "bpb": 3.782782}, "el": {"chunks": 5168.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.251815, "cps": 40.375, "mean_p": 0.481687, "boundary_entropy": 0.689327, "bpb": 4.591913}, "hi": {"chunks": 5761.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.342407, "cps": 45.007812, "mean_p": 0.488463, "boundary_entropy": 0.690414, "bpb": 3.686722}, "ta": {"chunks": 216.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.011554, "cps": 1.6875, "mean_p": 0.456643, "boundary_entropy": 0.686783, "bpb": 3.896657}, "th": {"chunks": 2721.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.169164, "cps": 21.257812, "mean_p": 0.446581, "boundary_entropy": 0.682615, "bpb": 3.449175}, "zh": {"chunks": 2976.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503894, "cps": 23.25, "mean_p": 0.485961, "boundary_entropy": 0.686238, "bpb": 6.406848}, "my": {"chunks": 2102.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.110935, "cps": 16.421875, "mean_p": 0.479435, "boundary_entropy": 0.690355, "bpb": 4.133107}}, "bpb": {"en": 4.4616, "de": 4.66392, "vi": 5.02658, "tr": 5.02489, "fi": 4.70545, "ru": 3.78278, "el": 4.59191, "hi": 3.68672, "ta": 3.89666, "th": 3.44917, "zh": 6.40685, "my": 4.13311}, "gini_cpc": 0.299186, "gini_cps": 0.237517, "bpic": 9.717697, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9518358452553459, "vi": 1.1802524341301006, "tr": 1.2375025785087532, "fi": 0.6006772170769646, "ru": 1.2014154716836833, "el": 1.4113807037750268, "hi": 1.9191340813001128, "ta": 0.06475760782352873, "th": 0.9481345929459889, "zh": 2.8242426027081278, "my": 0.6217729886727876}, "t": 21.3} |
| {"event": "train", "step": 150, "phase": 1, "objective": "parity_a", "lm_loss": 2.71629, "bpb": 3.91877, "chunk_loss": 0.031727, "F": 0.11981, "G": 0.43652, "chunks": 122.69, "bytes_per_chunk": 8.346, "boundary_entropy": 0.67488, "router_noise": 0.0, "grad_norm": 1.8441, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.072, "n_target_mean": 10.9936, "alpha": 0.03, "t": 22.6} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2890.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166063, "cps": 22.578125, "mean_p": 0.360702, "boundary_entropy": 0.610536, "bpb": 3.880748}, "de": {"chunks": 3547.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.172746, "cps": 27.710938, "mean_p": 0.368318, "boundary_entropy": 0.620952, "bpb": 4.112309}, "vi": {"chunks": 2967.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.160944, "cps": 23.179688, "mean_p": 0.385009, "boundary_entropy": 0.63843, "bpb": 3.663684}, "tr": {"chunks": 3967.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.219135, "cps": 30.992188, "mean_p": 0.384026, "boundary_entropy": 0.629206, "bpb": 4.222893}, "fi": {"chunks": 3204.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.170326, "cps": 25.03125, "mean_p": 0.374038, "boundary_entropy": 0.628132, "bpb": 4.114929}, "ru": {"chunks": 2789.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.147177, "cps": 21.789062, "mean_p": 0.390705, "boundary_entropy": 0.65474, "bpb": 2.900054}, "el": {"chunks": 6808.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.331725, "cps": 53.1875, "mean_p": 0.481301, "boundary_entropy": 0.688816, "bpb": 3.087578}, "hi": {"chunks": 3404.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.202318, "cps": 26.59375, "mean_p": 0.409023, "boundary_entropy": 0.670345, "bpb": 2.575926}, "ta": {"chunks": 2101.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.112383, "cps": 16.414062, "mean_p": 0.454075, "boundary_entropy": 0.685225, "bpb": 2.365708}, "th": {"chunks": 754.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.046876, "cps": 5.890625, "mean_p": 0.445961, "boundary_entropy": 0.682645, "bpb": 2.424666}, "zh": {"chunks": 3219.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545039, "cps": 25.148438, "mean_p": 0.472594, "boundary_entropy": 0.682816, "bpb": 5.413964}, "my": {"chunks": 970.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.051193, "cps": 7.578125, "mean_p": 0.446169, "boundary_entropy": 0.684518, "bpb": 2.54516}}, "bpb": {"en": 3.88075, "de": 4.11231, "vi": 3.66368, "tr": 4.22289, "fi": 4.11493, "ru": 2.90005, "el": 3.08758, "hi": 2.57593, "ta": 2.36571, "th": 2.42467, "zh": 5.41396, "my": 2.54516}, "gini_cpc": 0.324861, "gini_cps": 0.24574, "bpic": 10.480093, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.040243614928589, "vi": 0.9691716054631301, "tr": 1.3195866903094118, "fi": 1.0256682251182268, "ru": 0.8862690379892452, "el": 1.9975836714213964, "hi": 1.2183182671198014, "ta": 0.6767478120364141, "th": 0.282277692148007, "zh": 3.2821151324616222, "my": 0.30827239438039494}, "t": 25.7} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.88333, "de": 4.11482, "vi": 3.66992, "tr": 4.23252, "fi": 4.11556, "ru": 2.8839, "el": 3.0803, "hi": 2.57046, "ta": 2.38275, "th": 2.45841, "zh": 5.42213, "my": 2.56083}, "bytes_per_chunk": {"en": 5.744, "de": 6.126, "vi": 7.933, "tr": 5.047, "fi": 5.992, "ru": 11.416, "el": 5.61, "hi": 12.651, "ta": 23.814, "th": 49.686, "zh": 5.401, "my": 45.197}, "bpb_mean": 3.44791, "t": 27.0} |
| {"event": "train", "step": 200, "phase": 1, "objective": "parity_a", "lm_loss": 2.54134, "bpb": 3.66639, "chunk_loss": 0.029494, "F": 0.13306, "G": 0.39214, "chunks": 136.25, "bytes_per_chunk": 7.516, "boundary_entropy": 0.64331, "router_noise": 0.0, "grad_norm": 1.4741, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.091, "n_target_mean": 9.1136, "alpha": 0.03, "t": 27.1} |
| {"event": "train", "step": 250, "phase": 1, "objective": "parity_a", "lm_loss": 2.48656, "bpb": 3.58735, "chunk_loss": 0.030154, "F": 0.1452, "G": 0.39784, "chunks": 148.69, "bytes_per_chunk": 6.887, "boundary_entropy": 0.64597, "router_noise": 0.0, "grad_norm": 1.5397, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.045, "n_target_mean": 9.1181, "alpha": 0.03, "t": 29.3} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2805.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.161179, "cps": 21.914062, "mean_p": 0.354354, "boundary_entropy": 0.599956, "bpb": 3.645661}, "de": {"chunks": 3339.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162616, "cps": 26.085938, "mean_p": 0.356012, "boundary_entropy": 0.60557, "bpb": 3.853344}, "vi": {"chunks": 3424.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.185734, "cps": 26.75, "mean_p": 0.385625, "boundary_entropy": 0.633934, "bpb": 3.363483}, "tr": {"chunks": 3335.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.184224, "cps": 26.054688, "mean_p": 0.349094, "boundary_entropy": 0.594453, "bpb": 3.905167}, "fi": {"chunks": 2845.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.151241, "cps": 22.226562, "mean_p": 0.334021, "boundary_entropy": 0.592259, "bpb": 3.893653}, "ru": {"chunks": 2692.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.142058, "cps": 21.03125, "mean_p": 0.406758, "boundary_entropy": 0.662553, "bpb": 2.664173}, "el": {"chunks": 4046.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.197145, "cps": 31.609375, "mean_p": 0.458371, "boundary_entropy": 0.683749, "bpb": 2.765123}, "hi": {"chunks": 2965.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.176226, "cps": 23.164062, "mean_p": 0.368922, "boundary_entropy": 0.647816, "bpb": 2.154831}, "ta": {"chunks": 2463.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.131746, "cps": 19.242188, "mean_p": 0.451846, "boundary_entropy": 0.683282, "bpb": 2.018769}, "th": {"chunks": 670.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.041654, "cps": 5.234375, "mean_p": 0.408239, "boundary_entropy": 0.669014, "bpb": 2.108186}, "zh": {"chunks": 3826.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.647816, "cps": 29.890625, "mean_p": 0.41787, "boundary_entropy": 0.648436, "bpb": 5.001401}, "my": {"chunks": 4403.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.232373, "cps": 34.398438, "mean_p": 0.434187, "boundary_entropy": 0.678578, "bpb": 2.281587}}, "bpb": {"en": 3.64566, "de": 3.85334, "vi": 3.36348, "tr": 3.90517, "fi": 3.89365, "ru": 2.66417, "el": 2.76512, "hi": 2.15483, "ta": 2.01877, "th": 2.10819, "zh": 5.0014, "my": 2.28159}, "gini_cpc": 0.283561, "gini_cps": 0.154373, "bpic": 10.425149, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0089165972813816, "vi": 1.1523432670649443, "tr": 1.1429745003413327, "fi": 0.9383430505020813, "ru": 0.8813676294216416, "el": 1.2231403347906784, "hi": 1.0933542045752518, "ta": 0.81739165008803, "th": 0.25843122848370703, "zh": 4.019229243894091, "my": 1.4417055290076188}, "t": 30.5} |
| {"event": "train", "step": 300, "phase": 1, "objective": "parity_a", "lm_loss": 2.32558, "bpb": 3.3551, "chunk_loss": 0.030349, "F": 0.14273, "G": 0.37338, "chunks": 146.16, "bytes_per_chunk": 7.006, "boundary_entropy": 0.61832, "router_noise": 0.0, "grad_norm": 1.6187, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.067, "n_target_mean": 8.5595, "alpha": 0.03, "t": 32.7} |
| {"event": "train", "step": 350, "phase": 1, "objective": "parity_a", "lm_loss": 2.09702, "bpb": 3.02536, "chunk_loss": 0.030575, "F": 0.13278, "G": 0.35125, "chunks": 135.97, "bytes_per_chunk": 7.531, "boundary_entropy": 0.59779, "router_noise": 0.0, "grad_norm": 0.6642, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.045, "n_target_mean": 9.411, "alpha": 0.03, "t": 34.9} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2855.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.164052, "cps": 22.304688, "mean_p": 0.324622, "boundary_entropy": 0.545789, "bpb": 3.226687}, "de": {"chunks": 3245.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.158038, "cps": 25.351562, "mean_p": 0.318567, "boundary_entropy": 0.550504, "bpb": 3.523836}, "vi": {"chunks": 3636.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.197234, "cps": 28.40625, "mean_p": 0.358331, "boundary_entropy": 0.602395, "bpb": 2.821678}, "tr": {"chunks": 3643.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201237, "cps": 28.460938, "mean_p": 0.325755, "boundary_entropy": 0.539516, "bpb": 3.460696}, "fi": {"chunks": 3251.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.172824, "cps": 25.398438, "mean_p": 0.334211, "boundary_entropy": 0.554075, "bpb": 3.586771}, "ru": {"chunks": 3072.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.162111, "cps": 24.0, "mean_p": 0.264412, "boundary_entropy": 0.52484, "bpb": 2.456021}, "el": {"chunks": 3555.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.17322, "cps": 27.773438, "mean_p": 0.288387, "boundary_entropy": 0.565574, "bpb": 2.530789}, "hi": {"chunks": 1646.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.097831, "cps": 12.859375, "mean_p": 0.392685, "boundary_entropy": 0.657865, "bpb": 1.839365}, "ta": {"chunks": 2905.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.155389, "cps": 22.695312, "mean_p": 0.327399, "boundary_entropy": 0.60367, "bpb": 1.626247}, "th": {"chunks": 2188.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.136027, "cps": 17.09375, "mean_p": 0.374138, "boundary_entropy": 0.644165, "bpb": 1.93728}, "zh": {"chunks": 4007.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.678463, "cps": 31.304688, "mean_p": 0.289946, "boundary_entropy": 0.496199, "bpb": 4.147699}, "my": {"chunks": 1917.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.101172, "cps": 14.976562, "mean_p": 0.420701, "boundary_entropy": 0.673702, "bpb": 2.00594}}, "bpb": {"en": 3.22669, "de": 3.52384, "vi": 2.82168, "tr": 3.4607, "fi": 3.58677, "ru": 2.45602, "el": 2.53079, "hi": 1.83937, "ta": 1.62625, "th": 1.93728, "zh": 4.1477, "my": 2.00594}, "gini_cpc": 0.272893, "gini_cps": 0.130735, "bpic": 10.684326, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9633415706108223, "vi": 1.2022609471342727, "tr": 1.2266668459611294, "fi": 1.0534722343123564, "ru": 0.9881662222920489, "el": 1.0558853579594139, "hi": 0.5963383860332099, "ta": 0.9471934252909527, "th": 0.8291712971859763, "zh": 4.135651238937161, "my": 0.6167039703463475}, "t": 38.0} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.2319, "de": 3.5196, "vi": 2.82731, "tr": 3.46907, "fi": 3.57996, "ru": 2.44464, "el": 2.52539, "hi": 1.83505, "ta": 1.64385, "th": 1.96635, "zh": 4.15525, "my": 2.02076}, "bytes_per_chunk": {"en": 5.866, "de": 6.65, "vi": 7.272, "tr": 5.036, "fi": 5.804, "ru": 12.21, "el": 10.871, "hi": 22.76, "ta": 15.148, "th": 19.82, "zh": 4.856, "my": 26.975}, "bpb_mean": 2.76826, "t": 39.2} |
| {"event": "train", "step": 400, "phase": 1, "objective": "parity_a", "lm_loss": 2.22936, "bpb": 3.21628, "chunk_loss": 0.03024, "F": 0.14896, "G": 0.30906, "chunks": 152.53, "bytes_per_chunk": 6.713, "boundary_entropy": 0.5423, "router_noise": 0.0, "grad_norm": 0.9373, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 8.0193, "alpha": 0.03, "t": 39.3} |
| {"event": "train", "step": 450, "phase": 1, "objective": "parity_a", "lm_loss": 2.17204, "bpb": 3.13359, "chunk_loss": 0.030736, "F": 0.15625, "G": 0.31762, "chunks": 160.0, "bytes_per_chunk": 6.4, "boundary_entropy": 0.54031, "router_noise": 0.0, "grad_norm": 0.8715, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.2744, "alpha": 0.03, "t": 41.4} |
| {"event": "train", "step": 500, "phase": 1, "objective": "parity_a", "lm_loss": 2.06612, "bpb": 2.98078, "chunk_loss": 0.030804, "F": 0.15796, "G": 0.32261, "chunks": 161.75, "bytes_per_chunk": 6.331, "boundary_entropy": 0.5438, "router_noise": 0.0, "grad_norm": 1.0407, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 7.8003, "alpha": 0.03, "t": 43.6} |
| {"event": "train", "step": 550, "phase": 1, "objective": "parity_a", "lm_loss": 1.92316, "bpb": 2.77453, "chunk_loss": 0.030116, "F": 0.13681, "G": 0.30798, "chunks": 140.09, "bytes_per_chunk": 7.309, "boundary_entropy": 0.53212, "router_noise": 0.0, "grad_norm": 0.8534, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.6276, "alpha": 0.03, "t": 45.7} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3109.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.178647, "cps": 24.289062, "mean_p": 0.279474, "boundary_entropy": 0.45621, "bpb": 2.895781}, "de": {"chunks": 3594.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.175035, "cps": 28.078125, "mean_p": 0.287591, "boundary_entropy": 0.471567, "bpb": 3.218933}, "vi": {"chunks": 2971.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.161161, "cps": 23.210938, "mean_p": 0.331523, "boundary_entropy": 0.563233, "bpb": 2.58857}, "tr": {"chunks": 4064.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.224493, "cps": 31.75, "mean_p": 0.308527, "boundary_entropy": 0.48555, "bpb": 3.171639}, "fi": {"chunks": 3411.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18133, "cps": 26.648438, "mean_p": 0.297371, "boundary_entropy": 0.487308, "bpb": 3.373599}, "ru": {"chunks": 2566.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.135409, "cps": 20.046875, "mean_p": 0.259337, "boundary_entropy": 0.509691, "bpb": 2.139147}, "el": {"chunks": 2692.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.13117, "cps": 21.03125, "mean_p": 0.271414, "boundary_entropy": 0.54242, "bpb": 2.228115}, "hi": {"chunks": 2431.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.144487, "cps": 18.992188, "mean_p": 0.417754, "boundary_entropy": 0.664514, "bpb": 1.651334}, "ta": {"chunks": 2309.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.123509, "cps": 18.039062, "mean_p": 0.363923, "boundary_entropy": 0.635067, "bpb": 1.392496}, "th": {"chunks": 1813.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.112714, "cps": 14.164062, "mean_p": 0.345021, "boundary_entropy": 0.617399, "bpb": 1.743755}, "zh": {"chunks": 3046.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.515747, "cps": 23.796875, "mean_p": 0.305353, "boundary_entropy": 0.530733, "bpb": 3.858333}, "my": {"chunks": 3243.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.171153, "cps": 25.335938, "mean_p": 0.336809, "boundary_entropy": 0.614407, "bpb": 1.661568}}, "bpb": {"en": 2.89578, "de": 3.21893, "vi": 2.58857, "tr": 3.17164, "fi": 3.3736, "ru": 2.13915, "el": 2.22812, "hi": 1.65133, "ta": 1.3925, "th": 1.74375, "zh": 3.85833, "my": 1.66157}, "gini_cpc": 0.222157, "gini_cps": 0.113235, "bpic": 10.887713, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.979781113793442, "vi": 0.9021170852044814, "tr": 1.2566274607435133, "fi": 1.015016808445077, "ru": 0.7579679388548524, "el": 0.7342392699797059, "hi": 0.8087853650213611, "ta": 0.6913561994781562, "th": 0.6309284875210128, "zh": 2.8869539369713806, "my": 0.9580473429914296}, "t": 48.8} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.90475, "de": 3.21663, "vi": 2.59222, "tr": 3.18009, "fi": 3.36587, "ru": 2.12539, "el": 2.21822, "hi": 1.64489, "ta": 1.40634, "th": 1.77188, "zh": 3.86883, "my": 1.67418}, "bytes_per_chunk": {"en": 5.517, "de": 5.894, "vi": 7.799, "tr": 4.732, "fi": 5.922, "ru": 12.118, "el": 13.962, "hi": 17.555, "ta": 23.433, "th": 27.095, "zh": 5.878, "my": 15.716}, "bpb_mean": 2.49744, "t": 50.0} |
| {"event": "train", "step": 600, "phase": 1, "objective": "parity_a", "lm_loss": 1.90667, "bpb": 2.75074, "chunk_loss": 0.029933, "F": 0.14178, "G": 0.29896, "chunks": 145.19, "bytes_per_chunk": 7.053, "boundary_entropy": 0.51025, "router_noise": 0.0, "grad_norm": 1.0637, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 8.3285, "alpha": 0.03, "t": 50.1} |
| {"event": "train", "step": 650, "phase": 1, "objective": "parity_a", "lm_loss": 1.94467, "bpb": 2.80557, "chunk_loss": 0.030575, "F": 0.14038, "G": 0.28862, "chunks": 143.75, "bytes_per_chunk": 7.123, "boundary_entropy": 0.5017, "router_noise": 0.0, "grad_norm": 0.8329, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.8206, "alpha": 0.03, "t": 52.2} |
| {"event": "train", "step": 700, "phase": 1, "objective": "parity_a", "lm_loss": 1.78931, "bpb": 2.58143, "chunk_loss": 0.029934, "F": 0.12988, "G": 0.27895, "chunks": 133.0, "bytes_per_chunk": 7.699, "boundary_entropy": 0.4839, "router_noise": 0.0, "grad_norm": 1.0712, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.9896, "alpha": 0.03, "t": 54.4} |
| {"event": "train", "step": 750, "phase": 1, "objective": "parity_a", "lm_loss": 1.86657, "bpb": 2.69289, "chunk_loss": 0.029788, "F": 0.12979, "G": 0.27874, "chunks": 132.91, "bytes_per_chunk": 7.705, "boundary_entropy": 0.48307, "router_noise": 0.0, "grad_norm": 0.8738, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.6283, "alpha": 0.03, "t": 56.5} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2939.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168879, "cps": 22.960938, "mean_p": 0.249518, "boundary_entropy": 0.403243, "bpb": 2.694513}, "de": {"chunks": 3096.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.150782, "cps": 24.1875, "mean_p": 0.260113, "boundary_entropy": 0.430863, "bpb": 3.043408}, "vi": {"chunks": 2556.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.138649, "cps": 19.96875, "mean_p": 0.307041, "boundary_entropy": 0.534467, "bpb": 2.41756}, "tr": {"chunks": 3094.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170911, "cps": 24.171875, "mean_p": 0.272914, "boundary_entropy": 0.445386, "bpb": 3.052538}, "fi": {"chunks": 2792.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.148424, "cps": 21.8125, "mean_p": 0.250451, "boundary_entropy": 0.433438, "bpb": 3.236477}, "ru": {"chunks": 2896.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.152823, "cps": 22.625, "mean_p": 0.253764, "boundary_entropy": 0.504727, "bpb": 1.996649}, "el": {"chunks": 2741.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.133557, "cps": 21.414062, "mean_p": 0.305715, "boundary_entropy": 0.578085, "bpb": 2.087354}, "hi": {"chunks": 2260.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.134324, "cps": 17.65625, "mean_p": 0.413194, "boundary_entropy": 0.663386, "bpb": 1.571061}, "ta": {"chunks": 3793.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.202888, "cps": 29.632812, "mean_p": 0.364825, "boundary_entropy": 0.632872, "bpb": 1.340538}, "th": {"chunks": 1931.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.12005, "cps": 15.085938, "mean_p": 0.30802, "boundary_entropy": 0.579266, "bpb": 1.603339}, "zh": {"chunks": 3239.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.548425, "cps": 25.304688, "mean_p": 0.378972, "boundary_entropy": 0.606583, "bpb": 3.725282}, "my": {"chunks": 2050.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.108191, "cps": 16.015625, "mean_p": 0.260182, "boundary_entropy": 0.537259, "bpb": 1.494687}}, "bpb": {"en": 2.69451, "de": 3.04341, "vi": 2.41756, "tr": 3.05254, "fi": 3.23648, "ru": 1.99665, "el": 2.08735, "hi": 1.57106, "ta": 1.34054, "th": 1.60334, "zh": 3.72528, "my": 1.49469}, "gini_cpc": 0.232152, "gini_cps": 0.100895, "bpic": 11.494923, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8928388490948944, "vi": 0.8209982693946979, "tr": 1.012032109859733, "fi": 0.8788769299828894, "ru": 0.9049276897621918, "el": 0.7908474607125074, "hi": 0.7953859228786211, "ta": 1.2013841759350168, "th": 0.7108627260143908, "zh": 3.24744675773923, "my": 0.6406414274111925}, "t": 59.6} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.704, "de": 3.04691, "vi": 2.42213, "tr": 3.06331, "fi": 3.23332, "ru": 1.98169, "el": 2.07617, "hi": 1.56689, "ta": 1.35341, "th": 1.63283, "zh": 3.73542, "my": 1.50606}, "bytes_per_chunk": {"en": 5.696, "de": 6.895, "vi": 9.438, "tr": 6.173, "fi": 7.042, "ru": 11.244, "el": 13.365, "hi": 28.352, "ta": 14.101, "th": 23.693, "zh": 5.327, "my": 23.082}, "bpb_mean": 2.36018, "t": 60.8} |
| {"event": "train", "step": 800, "phase": 1, "objective": "parity_a", "lm_loss": 1.8447, "bpb": 2.66133, "chunk_loss": 0.029938, "F": 0.13757, "G": 0.29677, "chunks": 140.88, "bytes_per_chunk": 7.269, "boundary_entropy": 0.4965, "router_noise": 0.0, "grad_norm": 0.9357, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 9.4394, "alpha": 0.03, "t": 60.8} |
| {"event": "train", "step": 850, "phase": 1, "objective": "parity_a", "lm_loss": 1.64323, "bpb": 2.37067, "chunk_loss": 0.030076, "F": 0.12527, "G": 0.28776, "chunks": 128.28, "bytes_per_chunk": 7.982, "boundary_entropy": 0.49529, "router_noise": 0.0, "grad_norm": 0.8621, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.894, "alpha": 0.03, "t": 63.0} |
| {"event": "train", "step": 900, "phase": 1, "objective": "parity_a", "lm_loss": 1.73442, "bpb": 2.50224, "chunk_loss": 0.030353, "F": 0.13818, "G": 0.28973, "chunks": 141.5, "bytes_per_chunk": 7.237, "boundary_entropy": 0.4797, "router_noise": 0.0, "grad_norm": 1.0444, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.5699, "alpha": 0.03, "t": 65.1} |
| {"event": "train", "step": 950, "phase": 1, "objective": "parity_a", "lm_loss": 1.87669, "bpb": 2.70749, "chunk_loss": 0.030707, "F": 0.1514, "G": 0.26842, "chunks": 155.03, "bytes_per_chunk": 6.605, "boundary_entropy": 0.44111, "router_noise": 0.0, "grad_norm": 0.8648, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.0076, "alpha": 0.03, "t": 67.3} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2979.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.171177, "cps": 23.273438, "mean_p": 0.233433, "boundary_entropy": 0.345652, "bpb": 2.560585}, "de": {"chunks": 3257.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.158623, "cps": 25.445312, "mean_p": 0.243275, "boundary_entropy": 0.387313, "bpb": 2.912043}, "vi": {"chunks": 3270.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.17738, "cps": 25.546875, "mean_p": 0.282051, "boundary_entropy": 0.489213, "bpb": 2.306124}, "tr": {"chunks": 3418.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.188808, "cps": 26.703125, "mean_p": 0.263816, "boundary_entropy": 0.404652, "bpb": 2.911324}, "fi": {"chunks": 3449.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18335, "cps": 26.945312, "mean_p": 0.256696, "boundary_entropy": 0.410477, "bpb": 3.142301}, "ru": {"chunks": 3257.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.171873, "cps": 25.445312, "mean_p": 0.295471, "boundary_entropy": 0.551297, "bpb": 1.919076}, "el": {"chunks": 2954.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.143936, "cps": 23.078125, "mean_p": 0.30422, "boundary_entropy": 0.576269, "bpb": 2.00467}, "hi": {"chunks": 2800.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.166419, "cps": 21.875, "mean_p": 0.420498, "boundary_entropy": 0.663002, "bpb": 1.510697}, "ta": {"chunks": 2907.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.155496, "cps": 22.710938, "mean_p": 0.325027, "boundary_entropy": 0.600125, "bpb": 1.31177}, "th": {"chunks": 3333.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.207212, "cps": 26.039062, "mean_p": 0.321837, "boundary_entropy": 0.586783, "bpb": 1.538584}, "zh": {"chunks": 3139.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.531493, "cps": 24.523438, "mean_p": 0.373303, "boundary_entropy": 0.601929, "bpb": 3.633751}, "my": {"chunks": 2288.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.120752, "cps": 17.875, "mean_p": 0.288064, "boundary_entropy": 0.56487, "bpb": 1.418736}}, "bpb": {"en": 2.56058, "de": 2.91204, "vi": 2.30612, "tr": 2.91132, "fi": 3.1423, "ru": 1.91908, "el": 2.00467, "hi": 1.5107, "ta": 1.31177, "th": 1.53858, "zh": 3.63375, "my": 1.41874}, "gini_cpc": 0.192089, "gini_cps": 0.053636, "bpic": 10.358182, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9266569095733643, "vi": 1.0362349301088087, "tr": 1.1029990133108192, "fi": 1.071112106150742, "ru": 1.0040657690907593, "el": 0.8408591662369421, "hi": 0.9722021460775203, "ta": 0.9083917456967426, "th": 1.2105085614846354, "zh": 3.1049276871728924, "my": 0.7054175513160414}, "t": 70.3} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.56975, "de": 2.91304, "vi": 2.31018, "tr": 2.92149, "fi": 3.13659, "ru": 1.90142, "el": 1.99458, "hi": 1.50804, "ta": 1.32135, "th": 1.56411, "zh": 3.64216, "my": 1.42704}, "bytes_per_chunk": {"en": 5.6, "de": 6.542, "vi": 7.668, "tr": 5.572, "fi": 5.503, "ru": 10.639, "el": 11.752, "hi": 17.795, "ta": 17.993, "th": 13.365, "zh": 5.248, "my": 21.789}, "bpb_mean": 2.26748, "t": 71.6} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "parity_a", "lm_loss": 1.55468, "bpb": 2.24293, "chunk_loss": 0.030392, "F": 0.12698, "G": 0.29787, "chunks": 130.03, "bytes_per_chunk": 7.875, "boundary_entropy": 0.50059, "router_noise": 0.0, "grad_norm": 0.8726, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 10.0711, "alpha": 0.03, "t": 71.6} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "parity_a", "lm_loss": 1.77791, "bpb": 2.56498, "chunk_loss": 0.030472, "F": 0.14285, "G": 0.28559, "chunks": 146.28, "bytes_per_chunk": 7.0, "boundary_entropy": 0.47141, "router_noise": 0.0, "grad_norm": 1.0097, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.5227, "alpha": 0.03, "t": 73.8} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "parity_a", "lm_loss": 1.6031, "bpb": 2.31279, "chunk_loss": 0.029768, "F": 0.11847, "G": 0.26894, "chunks": 121.31, "bytes_per_chunk": 8.441, "boundary_entropy": 0.45748, "router_noise": 0.0, "grad_norm": 0.9049, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.9616, "alpha": 0.03, "t": 75.9} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "parity_a", "lm_loss": 1.62361, "bpb": 2.34238, "chunk_loss": 0.029803, "F": 0.12955, "G": 0.26424, "chunks": 132.66, "bytes_per_chunk": 7.719, "boundary_entropy": 0.44757, "router_noise": 0.0, "grad_norm": 0.962, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.2259, "alpha": 0.03, "t": 78.1} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3040.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.174683, "cps": 23.75, "mean_p": 0.23457, "boundary_entropy": 0.336863, "bpb": 2.476677}, "de": {"chunks": 2957.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.144012, "cps": 23.101562, "mean_p": 0.229115, "boundary_entropy": 0.369413, "bpb": 2.825039}, "vi": {"chunks": 3411.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.185028, "cps": 26.648438, "mean_p": 0.265886, "boundary_entropy": 0.458944, "bpb": 2.256346}, "tr": {"chunks": 3684.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203502, "cps": 28.78125, "mean_p": 0.264724, "boundary_entropy": 0.382457, "bpb": 2.83944}, "fi": {"chunks": 3465.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184201, "cps": 27.070312, "mean_p": 0.252333, "boundary_entropy": 0.381685, "bpb": 3.056468}, "ru": {"chunks": 2297.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.121214, "cps": 17.945312, "mean_p": 0.290008, "boundary_entropy": 0.544583, "bpb": 1.845083}, "el": {"chunks": 2877.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.140184, "cps": 22.476562, "mean_p": 0.311148, "boundary_entropy": 0.579866, "bpb": 1.940919}, "hi": {"chunks": 2126.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.12636, "cps": 16.609375, "mean_p": 0.347748, "boundary_entropy": 0.624478, "bpb": 1.474028}, "ta": {"chunks": 2708.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.144852, "cps": 21.15625, "mean_p": 0.305573, "boundary_entropy": 0.589289, "bpb": 1.180173}, "th": {"chunks": 2293.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.142555, "cps": 17.914062, "mean_p": 0.27892, "boundary_entropy": 0.556952, "bpb": 1.471131}, "zh": {"chunks": 2908.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.492381, "cps": 22.71875, "mean_p": 0.375712, "boundary_entropy": 0.603101, "bpb": 3.59927}, "my": {"chunks": 3468.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.183027, "cps": 27.09375, "mean_p": 0.343205, "boundary_entropy": 0.613989, "bpb": 1.3931}}, "bpb": {"en": 2.47668, "de": 2.82504, "vi": 2.25635, "tr": 2.83944, "fi": 3.05647, "ru": 1.84508, "el": 1.94092, "hi": 1.47403, "ta": 1.18017, "th": 1.47113, "zh": 3.59927, "my": 1.3931}, "gini_cpc": 0.202429, "gini_cps": 0.094468, "bpic": 10.892348, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8244217748322983, "vi": 1.059227174425078, "tr": 1.1649830502068572, "fi": 1.054488607589624, "ru": 0.6939086758783503, "el": 0.8025083378853506, "hi": 0.72336705247517, "ta": 0.8292275587336889, "th": 0.8160814873288288, "zh": 2.8187171386814476, "my": 1.0477706993100229}, "t": 81.1} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.48374, "de": 2.82722, "vi": 2.25984, "tr": 2.84872, "fi": 3.05629, "ru": 1.82737, "el": 1.92999, "hi": 1.47102, "ta": 1.18928, "th": 1.49724, "zh": 3.60621, "my": 1.40263}, "bytes_per_chunk": {"en": 5.538, "de": 6.966, "vi": 7.354, "tr": 5.15, "fi": 5.301, "ru": 15.292, "el": 11.972, "hi": 23.58, "ta": 18.635, "th": 20.638, "zh": 5.76, "my": 16.377}, "bpb_mean": 2.19996, "t": 82.4} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "parity_a", "lm_loss": 1.71096, "bpb": 2.46839, "chunk_loss": 0.030104, "F": 0.14233, "G": 0.27858, "chunks": 145.75, "bytes_per_chunk": 7.026, "boundary_entropy": 0.45133, "router_noise": 0.0, "grad_norm": 0.9084, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 8.5108, "alpha": 0.03, "t": 82.4} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "parity_a", "lm_loss": 1.60902, "bpb": 2.32132, "chunk_loss": 0.031165, "F": 0.13541, "G": 0.2842, "chunks": 138.66, "bytes_per_chunk": 7.385, "boundary_entropy": 0.46563, "router_noise": 0.0, "grad_norm": 0.9413, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.4731, "alpha": 0.03, "t": 84.6} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "parity_a", "lm_loss": 1.71687, "bpb": 2.47693, "chunk_loss": 0.030183, "F": 0.14551, "G": 0.27252, "chunks": 149.0, "bytes_per_chunk": 6.872, "boundary_entropy": 0.44593, "router_noise": 0.0, "grad_norm": 0.9173, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 8.8148, "alpha": 0.03, "t": 86.7} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "parity_a", "lm_loss": 1.51674, "bpb": 2.18819, "chunk_loss": 0.030222, "F": 0.12622, "G": 0.26963, "chunks": 129.25, "bytes_per_chunk": 7.923, "boundary_entropy": 0.45409, "router_noise": 0.0, "grad_norm": 0.8428, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.5385, "alpha": 0.03, "t": 88.9} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2874.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165144, "cps": 22.453125, "mean_p": 0.214505, "boundary_entropy": 0.302239, "bpb": 2.40998}, "de": {"chunks": 3023.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.147226, "cps": 23.617188, "mean_p": 0.215014, "boundary_entropy": 0.330138, "bpb": 2.745969}, "vi": {"chunks": 2874.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.155899, "cps": 22.453125, "mean_p": 0.235779, "boundary_entropy": 0.42527, "bpb": 2.20002}, "tr": {"chunks": 3042.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168038, "cps": 23.765625, "mean_p": 0.236413, "boundary_entropy": 0.355052, "bpb": 2.737566}, "fi": {"chunks": 2739.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.145606, "cps": 21.398438, "mean_p": 0.19624, "boundary_entropy": 0.316721, "bpb": 2.987479}, "ru": {"chunks": 3795.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.200264, "cps": 29.648438, "mean_p": 0.317168, "boundary_entropy": 0.561326, "bpb": 1.815605}, "el": {"chunks": 2194.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.106904, "cps": 17.140625, "mean_p": 0.313096, "boundary_entropy": 0.582971, "bpb": 1.886221}, "hi": {"chunks": 3991.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.237207, "cps": 31.179688, "mean_p": 0.301571, "boundary_entropy": 0.571921, "bpb": 1.440993}, "ta": {"chunks": 3180.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.170099, "cps": 24.84375, "mean_p": 0.288261, "boundary_entropy": 0.566464, "bpb": 1.159266}, "th": {"chunks": 3597.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.223624, "cps": 28.101562, "mean_p": 0.285079, "boundary_entropy": 0.555928, "bpb": 1.43284}, "zh": {"chunks": 2852.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482899, "cps": 22.28125, "mean_p": 0.374179, "boundary_entropy": 0.601497, "bpb": 3.559129}, "my": {"chunks": 3634.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.191788, "cps": 28.390625, "mean_p": 0.376717, "boundary_entropy": 0.642547, "bpb": 1.365485}}, "bpb": {"en": 2.40998, "de": 2.74597, "vi": 2.20002, "tr": 2.73757, "fi": 2.98748, "ru": 1.8156, "el": 1.88622, "hi": 1.44099, "ta": 1.15927, "th": 1.43284, "zh": 3.55913, "my": 1.36549}, "gini_cpc": 0.201798, "gini_cps": 0.086645, "bpic": 10.15428, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8915035900760392, "vi": 0.9440195280716028, "tr": 1.01752717026221, "fi": 0.8816932281771348, "ru": 1.2126624288728165, "el": 0.6473410299783826, "hi": 1.4363623447809484, "ta": 1.0300042267338403, "th": 1.3541186792762614, "zh": 2.9241081749072277, "my": 1.1613387127411223}, "t": 91.9} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.41697, "de": 2.74932, "vi": 2.20341, "tr": 2.74734, "fi": 2.98909, "ru": 1.79522, "el": 1.87592, "hi": 1.43753, "ta": 1.16675, "th": 1.45463, "zh": 3.56253, "my": 1.37617}, "bytes_per_chunk": {"en": 5.93, "de": 6.906, "vi": 8.846, "tr": 6.205, "fi": 6.881, "ru": 9.728, "el": 15.436, "hi": 11.277, "ta": 15.802, "th": 13.38, "zh": 5.969, "my": 15.669}, "bpb_mean": 2.14791, "t": 93.2} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "parity_a", "lm_loss": 1.70856, "bpb": 2.46493, "chunk_loss": 0.03012, "F": 0.13464, "G": 0.28003, "chunks": 137.88, "bytes_per_chunk": 7.427, "boundary_entropy": 0.46445, "router_noise": 0.0, "grad_norm": 0.8276, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 8.7969, "alpha": 0.03, "t": 93.2} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "parity_a", "lm_loss": 1.71018, "bpb": 2.46727, "chunk_loss": 0.0293, "F": 0.12991, "G": 0.291, "chunks": 133.03, "bytes_per_chunk": 7.697, "boundary_entropy": 0.47817, "router_noise": 0.0, "grad_norm": 0.8118, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 8.8901, "alpha": 0.03, "t": 95.4} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "parity_a", "lm_loss": 1.68244, "bpb": 2.42725, "chunk_loss": 0.030291, "F": 0.14255, "G": 0.25607, "chunks": 145.97, "bytes_per_chunk": 7.015, "boundary_entropy": 0.40311, "router_noise": 0.0, "grad_norm": 0.9477, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 8.1667, "alpha": 0.03, "t": 97.6} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "parity_a", "lm_loss": 1.57562, "bpb": 2.27315, "chunk_loss": 0.030258, "F": 0.14038, "G": 0.26855, "chunks": 143.75, "bytes_per_chunk": 7.123, "boundary_entropy": 0.42017, "router_noise": 0.0, "grad_norm": 0.8868, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 9.122, "alpha": 0.03, "t": 99.8} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2861.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.164397, "cps": 22.351562, "mean_p": 0.215343, "boundary_entropy": 0.297021, "bpb": 2.355947}, "de": {"chunks": 2904.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.141431, "cps": 22.6875, "mean_p": 0.208345, "boundary_entropy": 0.315649, "bpb": 2.676093}, "vi": {"chunks": 2971.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.161161, "cps": 23.210938, "mean_p": 0.229205, "boundary_entropy": 0.411119, "bpb": 2.160611}, "tr": {"chunks": 3067.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169419, "cps": 23.960938, "mean_p": 0.234089, "boundary_entropy": 0.348555, "bpb": 2.679677}, "fi": {"chunks": 2918.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155122, "cps": 22.796875, "mean_p": 0.208502, "boundary_entropy": 0.319915, "bpb": 2.906889}, "ru": {"chunks": 3285.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.173351, "cps": 25.664062, "mean_p": 0.301547, "boundary_entropy": 0.547746, "bpb": 1.743833}, "el": {"chunks": 2612.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.127272, "cps": 20.40625, "mean_p": 0.333607, "boundary_entropy": 0.594692, "bpb": 1.841995}, "hi": {"chunks": 2971.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.176582, "cps": 23.210938, "mean_p": 0.282558, "boundary_entropy": 0.550048, "bpb": 1.411213}, "ta": {"chunks": 3104.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.166034, "cps": 24.25, "mean_p": 0.304578, "boundary_entropy": 0.581603, "bpb": 1.129801}, "th": {"chunks": 2889.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.179608, "cps": 22.570312, "mean_p": 0.295442, "boundary_entropy": 0.57088, "bpb": 1.377063}, "zh": {"chunks": 2791.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.47257, "cps": 21.804688, "mean_p": 0.38402, "boundary_entropy": 0.611609, "bpb": 3.511116}, "my": {"chunks": 1764.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.093097, "cps": 13.78125, "mean_p": 0.316294, "boundary_entropy": 0.595639, "bpb": 1.330762}}, "bpb": {"en": 2.35595, "de": 2.67609, "vi": 2.16061, "tr": 2.67968, "fi": 2.90689, "ru": 1.74383, "el": 1.84199, "hi": 1.41121, "ta": 1.1298, "th": 1.37706, "zh": 3.51112, "my": 1.33076}, "gini_cpc": 0.191466, "gini_cps": 0.058346, "bpic": 11.242376, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8603010783248586, "vi": 0.9803152806363973, "tr": 1.0305509950121718, "fi": 0.9435820411250299, "ru": 1.0544656139014443, "el": 0.7741740143278435, "hi": 1.0741225675204744, "ta": 1.009956120617215, "th": 1.0925284096187275, "zh": 2.8745684605836304, "my": 0.5662933570395609}, "t": 102.9} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.36483, "de": 2.68083, "vi": 2.16495, "tr": 2.68815, "fi": 2.91155, "ru": 1.72469, "el": 1.83229, "hi": 1.40846, "ta": 1.13772, "th": 1.39844, "zh": 3.51979, "my": 1.33927}, "bytes_per_chunk": {"en": 5.855, "de": 6.991, "vi": 8.547, "tr": 6.004, "fi": 6.286, "ru": 10.853, "el": 14.384, "hi": 14.241, "ta": 16.603, "th": 15.991, "zh": 5.81, "my": 28.482}, "bpb_mean": 2.09758, "t": 104.1} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "parity_a", "lm_loss": 1.79537, "bpb": 2.59018, "chunk_loss": 0.030101, "F": 0.14932, "G": 0.25785, "chunks": 152.91, "bytes_per_chunk": 6.697, "boundary_entropy": 0.40092, "router_noise": 0.0, "grad_norm": 0.9408, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 7.4124, "alpha": 0.03, "t": 104.1} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "parity_a", "lm_loss": 1.58229, "bpb": 2.28276, "chunk_loss": 0.029876, "F": 0.13693, "G": 0.26985, "chunks": 140.22, "bytes_per_chunk": 7.303, "boundary_entropy": 0.43128, "router_noise": 0.0, "grad_norm": 0.8345, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 9.4189, "alpha": 0.03, "t": 106.3} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "parity_a", "lm_loss": 1.56547, "bpb": 2.25849, "chunk_loss": 0.030817, "F": 0.1329, "G": 0.27161, "chunks": 136.09, "bytes_per_chunk": 7.524, "boundary_entropy": 0.44832, "router_noise": 0.0, "grad_norm": 0.8133, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.3926, "alpha": 0.03, "t": 108.5} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "parity_a", "lm_loss": 1.51848, "bpb": 2.1907, "chunk_loss": 0.02989, "F": 0.12555, "G": 0.26967, "chunks": 128.56, "bytes_per_chunk": 7.965, "boundary_entropy": 0.44338, "router_noise": 0.0, "grad_norm": 0.8534, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 9.9495, "alpha": 0.03, "t": 110.7} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3087.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.177383, "cps": 24.117188, "mean_p": 0.226402, "boundary_entropy": 0.29046, "bpb": 2.314632}, "de": {"chunks": 3246.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.158087, "cps": 25.359375, "mean_p": 0.222909, "boundary_entropy": 0.324228, "bpb": 2.618527}, "vi": {"chunks": 2806.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.15221, "cps": 21.921875, "mean_p": 0.224827, "boundary_entropy": 0.40012, "bpb": 2.132503}, "tr": {"chunks": 3254.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179749, "cps": 25.421875, "mean_p": 0.241297, "boundary_entropy": 0.343755, "bpb": 2.631606}, "fi": {"chunks": 2976.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158205, "cps": 23.25, "mean_p": 0.217771, "boundary_entropy": 0.315737, "bpb": 2.83784}, "ru": {"chunks": 3713.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.195937, "cps": 29.007812, "mean_p": 0.318152, "boundary_entropy": 0.563004, "bpb": 1.700371}, "el": {"chunks": 3426.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.166935, "cps": 26.765625, "mean_p": 0.3508, "boundary_entropy": 0.609004, "bpb": 1.795872}, "hi": {"chunks": 2774.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.164874, "cps": 21.671875, "mean_p": 0.276811, "boundary_entropy": 0.540298, "bpb": 1.385999}, "ta": {"chunks": 2243.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.119979, "cps": 17.523438, "mean_p": 0.273866, "boundary_entropy": 0.549871, "bpb": 1.12499}, "th": {"chunks": 3246.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.201803, "cps": 25.359375, "mean_p": 0.283121, "boundary_entropy": 0.551855, "bpb": 1.344344}, "zh": {"chunks": 3020.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.511344, "cps": 23.59375, "mean_p": 0.387282, "boundary_entropy": 0.609849, "bpb": 3.489709}, "my": {"chunks": 2095.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.110566, "cps": 16.367188, "mean_p": 0.344693, "boundary_entropy": 0.616993, "bpb": 1.294728}}, "bpb": {"en": 2.31463, "de": 2.61853, "vi": 2.1325, "tr": 2.63161, "fi": 2.83784, "ru": 1.70037, "el": 1.79587, "hi": 1.386, "ta": 1.12499, "th": 1.34434, "zh": 3.48971, "my": 1.29473}, "gini_cpc": 0.203845, "gini_cps": 0.081002, "bpic": 10.694449, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8912172810347675, "vi": 0.8580883692157169, "tr": 1.0133383709146886, "fi": 0.8918843310617989, "ru": 1.1045953880986996, "el": 0.9410961661587391, "hi": 0.9294774857512523, "ta": 0.6763808369792726, "th": 1.137666424089952, "zh": 2.8827102414668158, "my": 0.623315809028156}, "t": 113.7} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.32458, "de": 2.62504, "vi": 2.13835, "tr": 2.6351, "fi": 2.84225, "ru": 1.67882, "el": 1.78693, "hi": 1.38246, "ta": 1.13171, "th": 1.36996, "zh": 3.49214, "my": 1.3038}, "bytes_per_chunk": {"en": 5.337, "de": 6.308, "vi": 8.817, "tr": 5.618, "fi": 6.066, "ru": 9.57, "el": 11.38, "hi": 15.175, "ta": 22.949, "th": 14.02, "zh": 5.227, "my": 24.731}, "bpb_mean": 2.05926, "t": 115.0} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "parity_a", "lm_loss": 1.46009, "bpb": 2.10647, "chunk_loss": 0.030243, "F": 0.12842, "G": 0.28768, "chunks": 131.5, "bytes_per_chunk": 7.787, "boundary_entropy": 0.46711, "router_noise": 0.0, "grad_norm": 0.7719, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 10.3327, "alpha": 0.03, "t": 115.0} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "parity_a", "lm_loss": 1.44704, "bpb": 2.08764, "chunk_loss": 0.030385, "F": 0.1199, "G": 0.25729, "chunks": 122.78, "bytes_per_chunk": 8.34, "boundary_entropy": 0.43349, "router_noise": 0.0, "grad_norm": 0.7846, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 10.6748, "alpha": 0.03, "t": 117.2} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "parity_a", "lm_loss": 1.55464, "bpb": 2.24287, "chunk_loss": 0.030271, "F": 0.14307, "G": 0.26061, "chunks": 146.5, "bytes_per_chunk": 6.99, "boundary_entropy": 0.3983, "router_noise": 0.0, "grad_norm": 0.7686, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 8.603, "alpha": 0.03, "t": 119.4} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "parity_a", "lm_loss": 1.62053, "bpb": 2.33793, "chunk_loss": 0.030189, "F": 0.14087, "G": 0.28033, "chunks": 144.25, "bytes_per_chunk": 7.099, "boundary_entropy": 0.4483, "router_noise": 0.0, "grad_norm": 0.7989, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 9.0359, "alpha": 0.03, "t": 121.6} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2935.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168649, "cps": 22.929688, "mean_p": 0.211038, "boundary_entropy": 0.26552, "bpb": 2.285299}, "de": {"chunks": 3039.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.148006, "cps": 23.742188, "mean_p": 0.203219, "boundary_entropy": 0.290107, "bpb": 2.591868}, "vi": {"chunks": 3077.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.166911, "cps": 24.039062, "mean_p": 0.225363, "boundary_entropy": 0.385185, "bpb": 2.110466}, "tr": {"chunks": 3069.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16953, "cps": 23.976562, "mean_p": 0.226848, "boundary_entropy": 0.321592, "bpb": 2.577696}, "fi": {"chunks": 3112.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.165435, "cps": 24.3125, "mean_p": 0.222947, "boundary_entropy": 0.312998, "bpb": 2.832496}, "ru": {"chunks": 2413.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.127335, "cps": 18.851562, "mean_p": 0.274447, "boundary_entropy": 0.516358, "bpb": 1.67074}, "el": {"chunks": 2743.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.133655, "cps": 21.429688, "mean_p": 0.33055, "boundary_entropy": 0.590311, "bpb": 1.776292}, "hi": {"chunks": 2944.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.174978, "cps": 23.0, "mean_p": 0.275073, "boundary_entropy": 0.532035, "bpb": 1.360567}, "ta": {"chunks": 2887.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.154426, "cps": 22.554688, "mean_p": 0.288162, "boundary_entropy": 0.55673, "bpb": 1.116099}, "th": {"chunks": 2884.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.179297, "cps": 22.53125, "mean_p": 0.233366, "boundary_entropy": 0.4888, "bpb": 1.319996}, "zh": {"chunks": 3063.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.518625, "cps": 23.929688, "mean_p": 0.3837, "boundary_entropy": 0.600428, "bpb": 3.443266}, "my": {"chunks": 1239.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.065389, "cps": 9.679688, "mean_p": 0.3291, "boundary_entropy": 0.606599, "bpb": 1.267611}}, "bpb": {"en": 2.2853, "de": 2.59187, "vi": 2.11047, "tr": 2.5777, "fi": 2.8325, "ru": 1.67074, "el": 1.77629, "hi": 1.36057, "ta": 1.1161, "th": 1.32, "zh": 3.44327, "my": 1.26761}, "gini_cpc": 0.226122, "gini_cps": 0.07539, "bpic": 11.488729, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8775953380733934, "vi": 0.9896927045575203, "tr": 1.005222848886379, "fi": 0.9809428804796063, "ru": 0.7550298508133572, "el": 0.7925031134304711, "hi": 1.0375254232997764, "ta": 0.9156665070464504, "th": 1.063139380389212, "zh": 3.075173112435539, "my": 0.38772512523290675}, "t": 124.7} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.29174, "de": 2.59893, "vi": 2.11494, "tr": 2.58159, "fi": 2.83516, "ru": 1.65089, "el": 1.76695, "hi": 1.35749, "ta": 1.12286, "th": 1.34061, "zh": 3.44721, "my": 1.27574}, "bytes_per_chunk": {"en": 5.702, "de": 6.748, "vi": 8.067, "tr": 5.955, "fi": 5.794, "ru": 14.77, "el": 13.583, "hi": 14.324, "ta": 17.956, "th": 15.952, "zh": 5.175, "my": 40.523}, "bpb_mean": 2.03201, "t": 125.9} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "parity_a", "lm_loss": 1.42436, "bpb": 2.05492, "chunk_loss": 0.030931, "F": 0.11139, "G": 0.26672, "chunks": 114.06, "bytes_per_chunk": 8.978, "boundary_entropy": 0.45241, "router_noise": 0.0, "grad_norm": 1.0994, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.088, "n_target_mean": 11.6437, "alpha": 0.03, "t": 126.0} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "parity_a", "lm_loss": 1.59798, "bpb": 2.30539, "chunk_loss": 0.029884, "F": 0.13834, "G": 0.26173, "chunks": 141.66, "bytes_per_chunk": 7.229, "boundary_entropy": 0.41252, "router_noise": 0.0, "grad_norm": 0.8897, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 8.4524, "alpha": 0.03, "t": 128.1} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "parity_a", "lm_loss": 1.65238, "bpb": 2.38389, "chunk_loss": 0.030498, "F": 0.14011, "G": 0.27075, "chunks": 143.47, "bytes_per_chunk": 7.137, "boundary_entropy": 0.42664, "router_noise": 0.0, "grad_norm": 0.9597, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.1155, "alpha": 0.03, "t": 130.3} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "parity_a", "lm_loss": 1.38864, "bpb": 2.00338, "chunk_loss": 0.029739, "F": 0.11255, "G": 0.25567, "chunks": 115.25, "bytes_per_chunk": 8.885, "boundary_entropy": 0.43097, "router_noise": 0.0, "grad_norm": 0.8958, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 10.1747, "alpha": 0.03, "t": 132.5} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2885.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165776, "cps": 22.539062, "mean_p": 0.208271, "boundary_entropy": 0.255658, "bpb": 2.251733}, "de": {"chunks": 3010.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.146593, "cps": 23.515625, "mean_p": 0.201648, "boundary_entropy": 0.282328, "bpb": 2.542049}, "vi": {"chunks": 2882.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.156333, "cps": 22.515625, "mean_p": 0.208257, "boundary_entropy": 0.36358, "bpb": 2.090267}, "tr": {"chunks": 2962.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.163619, "cps": 23.140625, "mean_p": 0.217126, "boundary_entropy": 0.305345, "bpb": 2.542063}, "fi": {"chunks": 3005.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159747, "cps": 23.476562, "mean_p": 0.206985, "boundary_entropy": 0.287901, "bpb": 2.745974}, "ru": {"chunks": 2716.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.143325, "cps": 21.21875, "mean_p": 0.284419, "boundary_entropy": 0.527968, "bpb": 1.644754}, "el": {"chunks": 2334.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.113726, "cps": 18.234375, "mean_p": 0.319202, "boundary_entropy": 0.582508, "bpb": 1.746611}, "hi": {"chunks": 3073.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.182645, "cps": 24.007812, "mean_p": 0.245669, "boundary_entropy": 0.491855, "bpb": 1.353077}, "ta": {"chunks": 2286.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.122279, "cps": 17.859375, "mean_p": 0.260534, "boundary_entropy": 0.530956, "bpb": 1.095007}, "th": {"chunks": 2882.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.179173, "cps": 22.515625, "mean_p": 0.244947, "boundary_entropy": 0.504996, "bpb": 1.284734}, "zh": {"chunks": 2980.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.504572, "cps": 23.28125, "mean_p": 0.375279, "boundary_entropy": 0.595462, "bpb": 3.435299}, "my": {"chunks": 1820.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.096052, "cps": 14.21875, "mean_p": 0.332922, "boundary_entropy": 0.609774, "bpb": 1.267933}}, "bpb": {"en": 2.25173, "de": 2.54205, "vi": 2.09027, "tr": 2.54206, "fi": 2.74597, "ru": 1.64475, "el": 1.74661, "hi": 1.35308, "ta": 1.09501, "th": 1.28473, "zh": 3.4353, "my": 1.26793}, "gini_cpc": 0.221783, "gini_cps": 0.067405, "bpic": 11.688168, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8842852706734671, "vi": 0.9430378786489981, "tr": 0.9869901203693748, "fi": 0.9636312946801748, "ru": 0.8645673966426287, "el": 0.6860224201261786, "hi": 1.101756927902431, "ta": 0.7376138440523166, "th": 1.0808146280941422, "zh": 3.0436949528752297, "my": 0.5794104379074134}, "t": 135.6} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.26067, "de": 2.54888, "vi": 2.0949, "tr": 2.54597, "fi": 2.75148, "ru": 1.62354, "el": 1.73839, "hi": 1.35, "ta": 1.10069, "th": 1.31178, "zh": 3.43822, "my": 1.27711}, "bytes_per_chunk": {"en": 5.803, "de": 6.788, "vi": 8.293, "tr": 6.178, "fi": 6.015, "ru": 12.595, "el": 15.267, "hi": 13.367, "ta": 20.525, "th": 15.039, "zh": 5.198, "my": 31.814}, "bpb_mean": 2.00347, "t": 136.9} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "parity_a", "lm_loss": 1.29699, "bpb": 1.87116, "chunk_loss": 0.030197, "F": 0.10986, "G": 0.25556, "chunks": 112.5, "bytes_per_chunk": 9.102, "boundary_entropy": 0.4351, "router_noise": 0.0, "grad_norm": 0.7813, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.088, "n_target_mean": 11.3206, "alpha": 0.03, "t": 136.9} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "parity_a", "lm_loss": 1.53752, "bpb": 2.21817, "chunk_loss": 0.02976, "F": 0.1362, "G": 0.25413, "chunks": 139.47, "bytes_per_chunk": 7.342, "boundary_entropy": 0.39482, "router_noise": 0.0, "grad_norm": 0.8208, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.5994, "alpha": 0.03, "t": 139.1} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "parity_a", "lm_loss": 1.49077, "bpb": 2.15072, "chunk_loss": 0.030686, "F": 0.14032, "G": 0.26287, "chunks": 143.69, "bytes_per_chunk": 7.127, "boundary_entropy": 0.40366, "router_noise": 0.0, "grad_norm": 0.8609, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.3667, "alpha": 0.03, "t": 141.1} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "parity_a", "lm_loss": 1.46528, "bpb": 2.11395, "chunk_loss": 0.029952, "F": 0.1268, "G": 0.27909, "chunks": 129.84, "bytes_per_chunk": 7.886, "boundary_entropy": 0.45596, "router_noise": 0.0, "grad_norm": 0.7773, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.8086, "alpha": 0.03, "t": 143.3} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2867.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.164742, "cps": 22.398438, "mean_p": 0.203303, "boundary_entropy": 0.251208, "bpb": 2.229406}, "de": {"chunks": 2900.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.141236, "cps": 22.65625, "mean_p": 0.193718, "boundary_entropy": 0.272971, "bpb": 2.499923}, "vi": {"chunks": 2914.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.158069, "cps": 22.765625, "mean_p": 0.216936, "boundary_entropy": 0.372202, "bpb": 2.0691}, "tr": {"chunks": 2794.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.154339, "cps": 21.828125, "mean_p": 0.199488, "boundary_entropy": 0.287872, "bpb": 2.50468}, "fi": {"chunks": 2920.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155228, "cps": 22.8125, "mean_p": 0.202617, "boundary_entropy": 0.280582, "bpb": 2.697368}, "ru": {"chunks": 2668.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.140792, "cps": 20.84375, "mean_p": 0.28859, "boundary_entropy": 0.53277, "bpb": 1.611149}, "el": {"chunks": 2566.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.12503, "cps": 20.046875, "mean_p": 0.312069, "boundary_entropy": 0.573389, "bpb": 1.724047}, "hi": {"chunks": 2771.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.164695, "cps": 21.648438, "mean_p": 0.243306, "boundary_entropy": 0.489176, "bpb": 1.319225}, "ta": {"chunks": 2335.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.1249, "cps": 18.242188, "mean_p": 0.293563, "boundary_entropy": 0.56376, "bpb": 1.079604}, "th": {"chunks": 3120.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.19397, "cps": 24.375, "mean_p": 0.262854, "boundary_entropy": 0.519928, "bpb": 1.255729}, "zh": {"chunks": 3397.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.575178, "cps": 26.539062, "mean_p": 0.379451, "boundary_entropy": 0.59165, "bpb": 3.398806}, "my": {"chunks": 2425.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.127982, "cps": 18.945312, "mean_p": 0.372422, "boundary_entropy": 0.641945, "bpb": 1.233313}}, "bpb": {"en": 2.22941, "de": 2.49992, "vi": 2.0691, "tr": 2.50468, "fi": 2.69737, "ru": 1.61115, "el": 1.72405, "hi": 1.31923, "ta": 1.0796, "th": 1.25573, "zh": 3.39881, "my": 1.23331}, "gini_cpc": 0.224667, "gini_cps": 0.054699, "bpic": 11.395938, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8573181497194636, "vi": 0.9594952580399899, "tr": 0.9368547813473886, "fi": 0.9422527101551744, "ru": 0.854619972703671, "el": 0.7589483727420184, "hi": 0.9997188481314502, "ta": 0.7581547192541442, "th": 1.1774160619285907, "zh": 3.4913913490162294, "my": 0.7768636377392235}, "t": 146.3} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.2376, "de": 2.50887, "vi": 2.07468, "tr": 2.50965, "fi": 2.70661, "ru": 1.59049, "el": 1.71536, "hi": 1.31758, "ta": 1.08765, "th": 1.27708, "zh": 3.40268, "my": 1.24105}, "bytes_per_chunk": {"en": 5.95, "de": 6.922, "vi": 8.029, "tr": 6.565, "fi": 6.169, "ru": 12.362, "el": 13.95, "hi": 15.458, "ta": 21.242, "th": 15.014, "zh": 4.901, "my": 19.669}, "bpb_mean": 1.97244, "t": 147.6} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 147.6} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "parity_a", "lm_loss": 1.32778, "bpb": 1.91558, "chunk_loss": 0.029975, "F": 0.11646, "G": 0.25926, "chunks": 119.25, "bytes_per_chunk": 8.587, "boundary_entropy": 0.42139, "router_noise": 0.0, "grad_norm": 0.8165, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 10.8309, "alpha": 0.03, "t": 147.6} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "parity_a", "lm_loss": 1.48709, "bpb": 2.14541, "chunk_loss": 0.029985, "F": 0.1243, "G": 0.25103, "chunks": 127.28, "bytes_per_chunk": 8.045, "boundary_entropy": 0.39909, "router_noise": 0.0, "grad_norm": 0.9506, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.4427, "alpha": 0.03, "t": 149.7} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "parity_a", "lm_loss": 1.47721, "bpb": 2.13117, "chunk_loss": 0.031048, "F": 0.13406, "G": 0.26124, "chunks": 137.28, "bytes_per_chunk": 7.459, "boundary_entropy": 0.4176, "router_noise": 0.0, "grad_norm": 0.8336, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.044, "n_target_mean": 9.5095, "alpha": 0.03, "t": 151.9} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "parity_a", "lm_loss": 1.53824, "bpb": 2.21921, "chunk_loss": 0.029895, "F": 0.14478, "G": 0.23151, "chunks": 148.25, "bytes_per_chunk": 6.907, "boundary_entropy": 0.33593, "router_noise": 0.0, "grad_norm": 0.8816, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 7.5354, "alpha": 0.03, "t": 154.0} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2915.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.1675, "cps": 22.773438, "mean_p": 0.203501, "boundary_entropy": 0.241237, "bpb": 2.204582}, "de": {"chunks": 2937.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.143038, "cps": 22.945312, "mean_p": 0.192187, "boundary_entropy": 0.261242, "bpb": 2.471459}, "vi": {"chunks": 2818.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.152861, "cps": 22.015625, "mean_p": 0.21522, "boundary_entropy": 0.368277, "bpb": 2.058033}, "tr": {"chunks": 2976.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164393, "cps": 23.25, "mean_p": 0.209852, "boundary_entropy": 0.286633, "bpb": 2.45899}, "fi": {"chunks": 2934.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155973, "cps": 22.921875, "mean_p": 0.198921, "boundary_entropy": 0.267989, "bpb": 2.667028}, "ru": {"chunks": 2230.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.117678, "cps": 17.421875, "mean_p": 0.28125, "boundary_entropy": 0.52992, "bpb": 1.586327}, "el": {"chunks": 2152.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.104858, "cps": 16.8125, "mean_p": 0.309154, "boundary_entropy": 0.573043, "bpb": 1.680154}, "hi": {"chunks": 2592.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.154056, "cps": 20.25, "mean_p": 0.229709, "boundary_entropy": 0.473559, "bpb": 1.315022}, "ta": {"chunks": 2413.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.129072, "cps": 18.851562, "mean_p": 0.284019, "boundary_entropy": 0.557017, "bpb": 1.083466}, "th": {"chunks": 2457.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.152751, "cps": 19.195312, "mean_p": 0.243626, "boundary_entropy": 0.505245, "bpb": 1.244249}, "zh": {"chunks": 2831.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.479343, "cps": 22.117188, "mean_p": 0.36285, "boundary_entropy": 0.585458, "bpb": 3.392409}, "my": {"chunks": 2588.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.136584, "cps": 20.21875, "mean_p": 0.373456, "boundary_entropy": 0.642617, "bpb": 1.221247}}, "bpb": {"en": 2.20458, "de": 2.47146, "vi": 2.05803, "tr": 2.45899, "fi": 2.66703, "ru": 1.58633, "el": 1.68015, "hi": 1.31502, "ta": 1.08347, "th": 1.24425, "zh": 3.39241, "my": 1.22125}, "gini_cpc": 0.200228, "gini_cps": 0.058409, "bpic": 12.052288, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8539591582441151, "vi": 0.9126061852850007, "tr": 0.981449451295896, "fi": 0.9311803164693335, "ru": 0.7025564250057703, "el": 0.6260182338202489, "hi": 0.9197408699685746, "ta": 0.7705794332328518, "th": 0.9119471120137392, "zh": 2.8617519526904927, "my": 0.8154295714442452}, "t": 157.1} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.21363, "de": 2.47869, "vi": 2.062, "tr": 2.46387, "fi": 2.67393, "ru": 1.56517, "el": 1.67149, "hi": 1.31399, "ta": 1.08968, "th": 1.26471, "zh": 3.3939, "my": 1.22992}, "bytes_per_chunk": {"en": 5.845, "de": 6.882, "vi": 8.691, "tr": 6.276, "fi": 6.171, "ru": 15.574, "el": 16.701, "hi": 16.893, "ta": 19.879, "th": 19.061, "zh": 5.609, "my": 20.006}, "bpb_mean": 1.95175, "t": 158.3} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "parity_a", "lm_loss": 1.5073, "bpb": 2.17457, "chunk_loss": 0.029941, "F": 0.13147, "G": 0.24819, "chunks": 134.62, "bytes_per_chunk": 7.606, "boundary_entropy": 0.38242, "router_noise": 0.0, "grad_norm": 0.8028, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 9.2882, "alpha": 0.03, "t": 158.3} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "parity_a", "lm_loss": 1.45751, "bpb": 2.10274, "chunk_loss": 0.030193, "F": 0.14139, "G": 0.23989, "chunks": 144.78, "bytes_per_chunk": 7.073, "boundary_entropy": 0.36454, "router_noise": 0.0, "grad_norm": 0.8923, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.0661, "alpha": 0.03, "t": 160.5} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "parity_a", "lm_loss": 1.50754, "bpb": 2.17493, "chunk_loss": 0.030318, "F": 0.1427, "G": 0.23072, "chunks": 146.12, "bytes_per_chunk": 7.008, "boundary_entropy": 0.33797, "router_noise": 0.0, "grad_norm": 0.9136, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 8.2992, "alpha": 0.03, "t": 162.6} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "parity_a", "lm_loss": 1.54008, "bpb": 2.22186, "chunk_loss": 0.030089, "F": 0.13968, "G": 0.24056, "chunks": 143.03, "bytes_per_chunk": 7.159, "boundary_entropy": 0.36477, "router_noise": 0.0, "grad_norm": 0.9127, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 8.0733, "alpha": 0.03, "t": 164.7} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2969.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.170603, "cps": 23.195312, "mean_p": 0.207974, "boundary_entropy": 0.243919, "bpb": 2.187451}, "de": {"chunks": 3084.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.150197, "cps": 24.09375, "mean_p": 0.199998, "boundary_entropy": 0.267568, "bpb": 2.436945}, "vi": {"chunks": 2829.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.153458, "cps": 22.101562, "mean_p": 0.211588, "boundary_entropy": 0.36419, "bpb": 2.042844}, "tr": {"chunks": 2956.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.163288, "cps": 23.09375, "mean_p": 0.212527, "boundary_entropy": 0.298084, "bpb": 2.428789}, "fi": {"chunks": 2861.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152092, "cps": 22.351562, "mean_p": 0.20061, "boundary_entropy": 0.273619, "bpb": 2.636257}, "ru": {"chunks": 3053.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.161108, "cps": 23.851562, "mean_p": 0.293696, "boundary_entropy": 0.534838, "bpb": 1.564186}, "el": {"chunks": 2745.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.133752, "cps": 21.445312, "mean_p": 0.318411, "boundary_entropy": 0.578018, "bpb": 1.675295}, "hi": {"chunks": 3011.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.17896, "cps": 23.523438, "mean_p": 0.247501, "boundary_entropy": 0.491908, "bpb": 1.288664}, "ta": {"chunks": 3019.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.161487, "cps": 23.585938, "mean_p": 0.286289, "boundary_entropy": 0.556843, "bpb": 1.071731}, "th": {"chunks": 2827.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.175754, "cps": 22.085938, "mean_p": 0.261507, "boundary_entropy": 0.514556, "bpb": 1.217714}, "zh": {"chunks": 2943.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.498307, "cps": 22.992188, "mean_p": 0.364062, "boundary_entropy": 0.582889, "bpb": 3.334185}, "my": {"chunks": 3336.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.176061, "cps": 26.0625, "mean_p": 0.388866, "boundary_entropy": 0.650324, "bpb": 1.213147}}, "bpb": {"en": 2.18745, "de": 2.43694, "vi": 2.04284, "tr": 2.42879, "fi": 2.63626, "ru": 1.56419, "el": 1.6753, "hi": 1.28866, "ta": 1.07173, "th": 1.21771, "zh": 3.33418, "my": 1.21315}, "gini_cpc": 0.167706, "gini_cps": 0.026635, "bpic": 10.770381, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8803915890753586, "vi": 0.899505303103592, "tr": 0.9571231064366961, "fi": 0.8914970224007192, "ru": 0.9443467990697187, "el": 0.7839988510619087, "hi": 1.0489857902636306, "ta": 0.9465674499920774, "th": 1.0301931690152717, "zh": 2.920860005358452, "my": 1.0319926127083514}, "t": 167.7} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.19722, "de": 2.44381, "vi": 2.04657, "tr": 2.43134, "fi": 2.64176, "ru": 1.5425, "el": 1.66584, "hi": 1.28525, "ta": 1.07772, "th": 1.23805, "zh": 3.33745, "my": 1.22157}, "bytes_per_chunk": {"en": 5.695, "de": 6.643, "vi": 8.667, "tr": 6.23, "fi": 6.333, "ru": 11.612, "el": 13.313, "hi": 14.542, "ta": 16.181, "th": 16.964, "zh": 5.65, "my": 17.209}, "bpb_mean": 1.92742, "t": 169.0} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "parity_a", "lm_loss": 1.41592, "bpb": 2.04274, "chunk_loss": 0.03, "F": 0.13382, "G": 0.2456, "chunks": 137.03, "bytes_per_chunk": 7.473, "boundary_entropy": 0.37797, "router_noise": 0.0, "grad_norm": 0.844, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 8.7824, "alpha": 0.03, "t": 169.0} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "parity_a", "lm_loss": 1.34041, "bpb": 1.93381, "chunk_loss": 0.029845, "F": 0.12048, "G": 0.24693, "chunks": 123.38, "bytes_per_chunk": 8.3, "boundary_entropy": 0.39388, "router_noise": 0.0, "grad_norm": 0.9213, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.986, "alpha": 0.03, "t": 171.1} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "parity_a", "lm_loss": 1.41779, "bpb": 2.04544, "chunk_loss": 0.030141, "F": 0.13965, "G": 0.242, "chunks": 143.0, "bytes_per_chunk": 7.161, "boundary_entropy": 0.36693, "router_noise": 0.0, "grad_norm": 0.8615, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.5507, "alpha": 0.03, "t": 173.2} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "parity_a", "lm_loss": 1.31365, "bpb": 1.8952, "chunk_loss": 0.02991, "F": 0.12283, "G": 0.25282, "chunks": 125.78, "bytes_per_chunk": 8.141, "boundary_entropy": 0.41201, "router_noise": 0.0, "grad_norm": 0.7335, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.6732, "alpha": 0.03, "t": 175.4} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2969.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.170603, "cps": 23.195312, "mean_p": 0.206992, "boundary_entropy": 0.241403, "bpb": 2.173678}, "de": {"chunks": 3069.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.149467, "cps": 23.976562, "mean_p": 0.19766, "boundary_entropy": 0.265907, "bpb": 2.424929}, "vi": {"chunks": 3350.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.18172, "cps": 26.171875, "mean_p": 0.223287, "boundary_entropy": 0.365386, "bpb": 2.031526}, "tr": {"chunks": 3092.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.1708, "cps": 24.15625, "mean_p": 0.213402, "boundary_entropy": 0.288595, "bpb": 2.415625}, "fi": {"chunks": 2909.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.154644, "cps": 22.726562, "mean_p": 0.209253, "boundary_entropy": 0.286181, "bpb": 2.626267}, "ru": {"chunks": 3421.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.180528, "cps": 26.726562, "mean_p": 0.309695, "boundary_entropy": 0.554283, "bpb": 1.544447}, "el": {"chunks": 3272.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.159431, "cps": 25.5625, "mean_p": 0.325831, "boundary_entropy": 0.582224, "bpb": 1.640358}, "hi": {"chunks": 3079.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.183001, "cps": 24.054688, "mean_p": 0.2528, "boundary_entropy": 0.503643, "bpb": 1.273005}, "ta": {"chunks": 2408.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.128804, "cps": 18.8125, "mean_p": 0.28606, "boundary_entropy": 0.558949, "bpb": 1.054289}, "th": {"chunks": 3176.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.197451, "cps": 24.8125, "mean_p": 0.24445, "boundary_entropy": 0.498652, "bpb": 1.200887}, "zh": {"chunks": 2951.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.499661, "cps": 23.054688, "mean_p": 0.36737, "boundary_entropy": 0.585958, "bpb": 3.336419}, "my": {"chunks": 2289.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.120804, "cps": 17.882812, "mean_p": 0.373538, "boundary_entropy": 0.643268, "bpb": 1.198538}}, "bpb": {"en": 2.17368, "de": 2.42493, "vi": 2.03153, "tr": 2.41562, "fi": 2.62627, "ru": 1.54445, "el": 1.64036, "hi": 1.273, "ta": 1.05429, "th": 1.20089, "zh": 3.33642, "my": 1.19854}, "gini_cpc": 0.189343, "gini_cps": 0.057837, "bpic": 10.665027, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.87610952881721, "vi": 1.065161811734547, "tr": 1.0011585402916998, "fi": 0.9064539804836392, "ru": 1.0581756959114013, "el": 0.9345152060745228, "hi": 1.0726759376359079, "ta": 0.7549964953895072, "th": 1.1573730119534853, "zh": 2.9287998218867797, "my": 0.7081028448709283}, "t": 178.4} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.18024, "de": 2.43218, "vi": 2.0347, "tr": 2.41651, "fi": 2.63448, "ru": 1.52179, "el": 1.63175, "hi": 1.27071, "ta": 1.06096, "th": 1.2213, "zh": 3.33407, "my": 1.20656}, "bytes_per_chunk": {"en": 5.743, "de": 6.639, "vi": 7.152, "tr": 6.059, "fi": 6.08, "ru": 10.092, "el": 11.611, "hi": 14.147, "ta": 20.168, "th": 15.152, "zh": 5.398, "my": 22.981}, "bpb_mean": 1.9121, "t": 179.7} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "parity_a", "lm_loss": 1.52997, "bpb": 2.20728, "chunk_loss": 0.030735, "F": 0.14301, "G": 0.26842, "chunks": 146.44, "bytes_per_chunk": 6.993, "boundary_entropy": 0.41916, "router_noise": 0.0, "grad_norm": 0.7929, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.088, "n_target_mean": 9.0106, "alpha": 0.03, "t": 179.7} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "parity_a", "lm_loss": 1.48793, "bpb": 2.14662, "chunk_loss": 0.030055, "F": 0.13425, "G": 0.27038, "chunks": 137.47, "bytes_per_chunk": 7.449, "boundary_entropy": 0.42257, "router_noise": 0.0, "grad_norm": 0.8891, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.0983, "alpha": 0.03, "t": 181.9} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "parity_a", "lm_loss": 1.42121, "bpb": 2.05037, "chunk_loss": 0.03056, "F": 0.12604, "G": 0.27035, "chunks": 129.06, "bytes_per_chunk": 7.934, "boundary_entropy": 0.43631, "router_noise": 0.0, "grad_norm": 0.7646, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.8865, "alpha": 0.03, "t": 184.0} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "parity_a", "lm_loss": 1.33981, "bpb": 1.93294, "chunk_loss": 0.03032, "F": 0.12375, "G": 0.2528, "chunks": 126.72, "bytes_per_chunk": 8.081, "boundary_entropy": 0.40695, "router_noise": 0.0, "grad_norm": 0.8201, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.6709, "alpha": 0.03, "t": 186.2} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2967.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.170488, "cps": 23.179688, "mean_p": 0.204072, "boundary_entropy": 0.225132, "bpb": 2.151326}, "de": {"chunks": 3027.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.147421, "cps": 23.648438, "mean_p": 0.193812, "boundary_entropy": 0.258726, "bpb": 2.392114}, "vi": {"chunks": 2762.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.149824, "cps": 21.578125, "mean_p": 0.202619, "boundary_entropy": 0.347536, "bpb": 2.020746}, "tr": {"chunks": 3004.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.165939, "cps": 23.46875, "mean_p": 0.209948, "boundary_entropy": 0.280666, "bpb": 2.373472}, "fi": {"chunks": 2805.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.149115, "cps": 21.914062, "mean_p": 0.19169, "boundary_entropy": 0.25192, "bpb": 2.5702}, "ru": {"chunks": 2682.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.14153, "cps": 20.953125, "mean_p": 0.286089, "boundary_entropy": 0.527733, "bpb": 1.531528}, "el": {"chunks": 2447.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.119232, "cps": 19.117188, "mean_p": 0.304018, "boundary_entropy": 0.564776, "bpb": 1.63303}, "hi": {"chunks": 3090.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.183655, "cps": 24.140625, "mean_p": 0.229426, "boundary_entropy": 0.466109, "bpb": 1.267607}, "ta": {"chunks": 2720.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.145493, "cps": 21.25, "mean_p": 0.291019, "boundary_entropy": 0.556724, "bpb": 1.048817}, "th": {"chunks": 2860.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177805, "cps": 22.34375, "mean_p": 0.248599, "boundary_entropy": 0.500123, "bpb": 1.186005}, "zh": {"chunks": 2942.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.498137, "cps": 22.984375, "mean_p": 0.355009, "boundary_entropy": 0.57485, "bpb": 3.291855}, "my": {"chunks": 2780.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.146717, "cps": 21.71875, "mean_p": 0.341677, "boundary_entropy": 0.620367, "bpb": 1.188789}}, "bpb": {"en": 2.15133, "de": 2.39211, "vi": 2.02075, "tr": 2.37347, "fi": 2.5702, "ru": 1.53153, "el": 1.63303, "hi": 1.26761, "ta": 1.04882, "th": 1.18601, "zh": 3.29185, "my": 1.18879}, "gini_cpc": 0.187835, "gini_cps": 0.033572, "bpic": 11.259197, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8647022472936484, "vi": 0.8787940466915292, "tr": 0.9733206798285532, "fi": 0.8746364170997124, "ru": 0.8301491605550234, "el": 0.6993582430168978, "hi": 1.0772338216668644, "ta": 0.8533948317829367, "th": 1.0429213106685193, "zh": 2.9218357571621736, "my": 0.8605735499188562}, "t": 189.2} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.15936, "de": 2.40022, "vi": 2.02371, "tr": 2.37898, "fi": 2.5799, "ru": 1.50916, "el": 1.6232, "hi": 1.26311, "ta": 1.05482, "th": 1.20638, "zh": 3.29299, "my": 1.19591}, "bytes_per_chunk": {"en": 5.73, "de": 6.711, "vi": 8.761, "tr": 6.132, "fi": 6.413, "ru": 13.297, "el": 14.726, "hi": 13.81, "ta": 19.008, "th": 16.286, "zh": 5.587, "my": 17.365}, "bpb_mean": 1.89064, "t": 190.5} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "parity_a", "lm_loss": 1.41249, "bpb": 2.0378, "chunk_loss": 0.029807, "F": 0.1236, "G": 0.23725, "chunks": 126.56, "bytes_per_chunk": 8.091, "boundary_entropy": 0.37455, "router_noise": 0.0, "grad_norm": 0.7936, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 9.2811, "alpha": 0.03, "t": 190.5} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "parity_a", "lm_loss": 1.39373, "bpb": 2.01072, "chunk_loss": 0.030031, "F": 0.13351, "G": 0.24032, "chunks": 136.72, "bytes_per_chunk": 7.49, "boundary_entropy": 0.36402, "router_noise": 0.0, "grad_norm": 0.8919, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.2351, "alpha": 0.03, "t": 192.6} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "parity_a", "lm_loss": 1.29508, "bpb": 1.86841, "chunk_loss": 0.02992, "F": 0.11102, "G": 0.24084, "chunks": 113.69, "bytes_per_chunk": 9.007, "boundary_entropy": 0.40408, "router_noise": 0.0, "grad_norm": 0.8731, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 10.7221, "alpha": 0.03, "t": 194.8} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "parity_a", "lm_loss": 1.37097, "bpb": 1.9779, "chunk_loss": 0.029913, "F": 0.12149, "G": 0.25013, "chunks": 124.41, "bytes_per_chunk": 8.231, "boundary_entropy": 0.4045, "router_noise": 0.0, "grad_norm": 0.7972, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 10.0553, "alpha": 0.03, "t": 196.9} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2950.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.169511, "cps": 23.046875, "mean_p": 0.201158, "boundary_entropy": 0.222794, "bpb": 2.132762}, "de": {"chunks": 3049.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.148493, "cps": 23.820312, "mean_p": 0.189286, "boundary_entropy": 0.246046, "bpb": 2.37655}, "vi": {"chunks": 2851.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.154651, "cps": 22.273438, "mean_p": 0.202347, "boundary_entropy": 0.343037, "bpb": 2.011211}, "tr": {"chunks": 3099.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.171187, "cps": 24.210938, "mean_p": 0.210933, "boundary_entropy": 0.278088, "bpb": 2.353966}, "fi": {"chunks": 2737.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.1455, "cps": 21.382812, "mean_p": 0.187448, "boundary_entropy": 0.246563, "bpb": 2.572697}, "ru": {"chunks": 2729.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.144011, "cps": 21.320312, "mean_p": 0.295567, "boundary_entropy": 0.542541, "bpb": 1.517329}, "el": {"chunks": 2830.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.137894, "cps": 22.109375, "mean_p": 0.30589, "boundary_entropy": 0.564222, "bpb": 1.614082}, "hi": {"chunks": 2958.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.17581, "cps": 23.109375, "mean_p": 0.240332, "boundary_entropy": 0.483646, "bpb": 1.255992}, "ta": {"chunks": 2700.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.144424, "cps": 21.09375, "mean_p": 0.264557, "boundary_entropy": 0.533364, "bpb": 1.038716}, "th": {"chunks": 3272.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.203419, "cps": 25.5625, "mean_p": 0.254784, "boundary_entropy": 0.505719, "bpb": 1.175322}, "zh": {"chunks": 2673.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.452591, "cps": 20.882812, "mean_p": 0.35311, "boundary_entropy": 0.574535, "bpb": 3.268542}, "my": {"chunks": 1900.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.100274, "cps": 14.84375, "mean_p": 0.337006, "boundary_entropy": 0.61539, "bpb": 1.191041}}, "bpb": {"en": 2.13276, "de": 2.37655, "vi": 2.01121, "tr": 2.35397, "fi": 2.5727, "ru": 1.51733, "el": 1.61408, "hi": 1.25599, "ta": 1.03872, "th": 1.17532, "zh": 3.26854, "my": 1.19104}, "gini_cpc": 0.190113, "gini_cps": 0.057223, "bpic": 11.371963, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.876006081982951, "vi": 0.9123388727227593, "tr": 1.009887807714238, "fi": 0.8583511991270883, "ru": 0.8495646348553284, "el": 0.8134815258944698, "hi": 1.0371586672375148, "ta": 0.8520015775087147, "th": 1.2000361638119523, "zh": 2.669977615409781, "my": 0.5915511855819495}, "t": 200.0} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.14195, "de": 2.38281, "vi": 2.01432, "tr": 2.35776, "fi": 2.57799, "ru": 1.49584, "el": 1.60603, "hi": 1.25356, "ta": 1.045, "th": 1.19408, "zh": 3.26766, "my": 1.19955}, "bytes_per_chunk": {"en": 5.731, "de": 6.675, "vi": 8.214, "tr": 5.993, "fi": 6.623, "ru": 12.687, "el": 13.36, "hi": 14.484, "ta": 17.595, "th": 14.533, "zh": 6.051, "my": 25.693}, "bpb_mean": 1.87805, "t": 201.2} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "parity_a", "lm_loss": 1.31688, "bpb": 1.89986, "chunk_loss": 0.029872, "F": 0.11719, "G": 0.25639, "chunks": 120.0, "bytes_per_chunk": 8.533, "boundary_entropy": 0.41472, "router_noise": 0.0, "grad_norm": 0.834, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 10.5619, "alpha": 0.03, "t": 201.3} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "parity_a", "lm_loss": 1.2272, "bpb": 1.77048, "chunk_loss": 0.029952, "F": 0.11215, "G": 0.23835, "chunks": 114.84, "bytes_per_chunk": 8.916, "boundary_entropy": 0.39328, "router_noise": 0.0, "grad_norm": 0.7777, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 10.9543, "alpha": 0.03, "t": 203.4} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "parity_a", "lm_loss": 1.33986, "bpb": 1.93301, "chunk_loss": 0.029754, "F": 0.12463, "G": 0.2441, "chunks": 127.62, "bytes_per_chunk": 8.024, "boundary_entropy": 0.37752, "router_noise": 0.0, "grad_norm": 0.8355, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 10.0046, "alpha": 0.03, "t": 205.6} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "parity_a", "lm_loss": 1.38267, "bpb": 1.99477, "chunk_loss": 0.029969, "F": 0.12167, "G": 0.25162, "chunks": 124.59, "bytes_per_chunk": 8.219, "boundary_entropy": 0.4021, "router_noise": 0.0, "grad_norm": 0.8312, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.7442, "alpha": 0.03, "t": 207.8} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2893.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166236, "cps": 22.601562, "mean_p": 0.19741, "boundary_entropy": 0.21418, "bpb": 2.120007}, "de": {"chunks": 3044.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.148249, "cps": 23.78125, "mean_p": 0.191049, "boundary_entropy": 0.247643, "bpb": 2.350188}, "vi": {"chunks": 2774.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.150475, "cps": 21.671875, "mean_p": 0.198021, "boundary_entropy": 0.33933, "bpb": 1.995403}, "tr": {"chunks": 3037.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.167762, "cps": 23.726562, "mean_p": 0.207466, "boundary_entropy": 0.276234, "bpb": 2.336597}, "fi": {"chunks": 3012.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160119, "cps": 23.53125, "mean_p": 0.200859, "boundary_entropy": 0.255106, "bpb": 2.521369}, "ru": {"chunks": 3101.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.163641, "cps": 24.226562, "mean_p": 0.300298, "boundary_entropy": 0.544818, "bpb": 1.506741}, "el": {"chunks": 3005.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.146421, "cps": 23.476562, "mean_p": 0.307574, "boundary_entropy": 0.563947, "bpb": 1.60606}, "hi": {"chunks": 2882.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.171293, "cps": 22.515625, "mean_p": 0.241943, "boundary_entropy": 0.49079, "bpb": 1.251536}, "ta": {"chunks": 2745.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.146831, "cps": 21.445312, "mean_p": 0.275045, "boundary_entropy": 0.544049, "bpb": 1.035227}, "th": {"chunks": 2861.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177868, "cps": 22.351562, "mean_p": 0.258121, "boundary_entropy": 0.509399, "bpb": 1.160787}, "zh": {"chunks": 3137.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.531155, "cps": 24.507812, "mean_p": 0.364548, "boundary_entropy": 0.574536, "bpb": 3.253554}, "my": {"chunks": 2398.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.126557, "cps": 18.734375, "mean_p": 0.342351, "boundary_entropy": 0.621692, "bpb": 1.1691}}, "bpb": {"en": 2.12001, "de": 2.35019, "vi": 1.9954, "tr": 2.3366, "fi": 2.52137, "ru": 1.50674, "el": 1.60606, "hi": 1.25154, "ta": 1.03523, "th": 1.16079, "zh": 3.25355, "my": 1.1691}, "gini_cpc": 0.186608, "gini_cps": 0.034884, "bpic": 11.000057, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8918009435007113, "vi": 0.9051884448222006, "tr": 1.0091830022310213, "fi": 0.9632050938122765, "ru": 0.9843923691706029, "el": 0.8808040801145449, "hi": 1.030420736840321, "ta": 0.8832681407575369, "th": 1.0699721633075927, "zh": 3.1951905465848234, "my": 0.7613099208480124}, "t": 210.8} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.12859, "de": 2.35886, "vi": 1.99966, "tr": 2.33951, "fi": 2.52963, "ru": 1.48685, "el": 1.59922, "hi": 1.24667, "ta": 1.04188, "th": 1.17903, "zh": 3.25285, "my": 1.17672}, "bytes_per_chunk": {"en": 5.871, "de": 6.72, "vi": 8.64, "tr": 6.048, "fi": 6.001, "ru": 11.15, "el": 12.037, "hi": 14.682, "ta": 18.551, "th": 16.409, "zh": 5.244, "my": 19.559}, "bpb_mean": 1.86162, "t": 212.0} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "parity_a", "lm_loss": 1.32576, "bpb": 1.91267, "chunk_loss": 0.030329, "F": 0.12155, "G": 0.25049, "chunks": 124.47, "bytes_per_chunk": 8.227, "boundary_entropy": 0.40309, "router_noise": 0.0, "grad_norm": 0.8277, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 10.3328, "alpha": 0.03, "t": 212.1} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "parity_a", "lm_loss": 1.32152, "bpb": 1.90655, "chunk_loss": 0.029997, "F": 0.12808, "G": 0.22991, "chunks": 131.16, "bytes_per_chunk": 7.807, "boundary_entropy": 0.3456, "router_noise": 0.0, "grad_norm": 0.8059, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.3376, "alpha": 0.03, "t": 214.2} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "parity_a", "lm_loss": 1.53917, "bpb": 2.22056, "chunk_loss": 0.029998, "F": 0.13696, "G": 0.23882, "chunks": 140.25, "bytes_per_chunk": 7.301, "boundary_entropy": 0.35887, "router_noise": 0.0, "grad_norm": 0.8773, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.3595, "alpha": 0.03, "t": 216.4} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "parity_a", "lm_loss": 1.43751, "bpb": 2.07388, "chunk_loss": 0.03067, "F": 0.15552, "G": 0.23466, "chunks": 159.25, "bytes_per_chunk": 6.43, "boundary_entropy": 0.30856, "router_noise": 0.0, "grad_norm": 0.8294, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.2469, "alpha": 0.03, "t": 218.5} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2915.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.1675, "cps": 22.773438, "mean_p": 0.197737, "boundary_entropy": 0.205155, "bpb": 2.109387}, "de": {"chunks": 3079.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.149954, "cps": 24.054688, "mean_p": 0.186464, "boundary_entropy": 0.232605, "bpb": 2.334686}, "vi": {"chunks": 2892.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.156876, "cps": 22.59375, "mean_p": 0.199367, "boundary_entropy": 0.333232, "bpb": 1.988645}, "tr": {"chunks": 3014.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.166492, "cps": 23.546875, "mean_p": 0.203027, "boundary_entropy": 0.263119, "bpb": 2.310951}, "fi": {"chunks": 2875.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152836, "cps": 22.460938, "mean_p": 0.192418, "boundary_entropy": 0.240319, "bpb": 2.500097}, "ru": {"chunks": 2563.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.135251, "cps": 20.023438, "mean_p": 0.283939, "boundary_entropy": 0.528859, "bpb": 1.486291}, "el": {"chunks": 2582.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.12581, "cps": 20.171875, "mean_p": 0.28372, "boundary_entropy": 0.54142, "bpb": 1.576618}, "hi": {"chunks": 2786.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.165587, "cps": 21.765625, "mean_p": 0.235724, "boundary_entropy": 0.478704, "bpb": 1.235648}, "ta": {"chunks": 2868.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.15341, "cps": 22.40625, "mean_p": 0.277673, "boundary_entropy": 0.540367, "bpb": 1.022196}, "th": {"chunks": 2737.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170159, "cps": 21.382812, "mean_p": 0.248628, "boundary_entropy": 0.498458, "bpb": 1.154619}, "zh": {"chunks": 2990.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506265, "cps": 23.359375, "mean_p": 0.35379, "boundary_entropy": 0.564185, "bpb": 3.237876}, "my": {"chunks": 2510.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.132468, "cps": 19.609375, "mean_p": 0.343843, "boundary_entropy": 0.622995, "bpb": 1.161584}}, "bpb": {"en": 2.10939, "de": 2.33469, "vi": 1.98865, "tr": 2.31095, "fi": 2.5001, "ru": 1.48629, "el": 1.57662, "hi": 1.23565, "ta": 1.0222, "th": 1.15462, "zh": 3.23788, "my": 1.16158}, "gini_cpc": 0.186107, "gini_cps": 0.035464, "bpic": 11.350773, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8952469350472014, "vi": 0.9365710034933364, "tr": 0.9939813999347549, "fi": 0.912455149914565, "ru": 0.8074673171703092, "el": 0.7511055203177893, "hi": 0.9885794998967781, "ta": 0.9158813984715372, "th": 1.0158727088244217, "zh": 3.02247910227643, "my": 0.790853255148785}, "t": 221.5} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.11696, "de": 2.34357, "vi": 1.99241, "tr": 2.31707, "fi": 2.50888, "ru": 1.46389, "el": 1.56928, "hi": 1.23325, "ta": 1.0285, "th": 1.17287, "zh": 3.23649, "my": 1.1687}, "bytes_per_chunk": {"en": 5.817, "de": 6.679, "vi": 8.232, "tr": 6.094, "fi": 6.26, "ru": 13.691, "el": 14.479, "hi": 15.53, "ta": 17.998, "th": 17.197, "zh": 5.307, "my": 18.554}, "bpb_mean": 1.84599, "t": 222.8} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "parity_a", "lm_loss": 1.38677, "bpb": 2.00068, "chunk_loss": 0.030083, "F": 0.1315, "G": 0.25208, "chunks": 134.66, "bytes_per_chunk": 7.605, "boundary_entropy": 0.38875, "router_noise": 0.0, "grad_norm": 0.8204, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 9.6644, "alpha": 0.03, "t": 222.8} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "parity_a", "lm_loss": 1.54328, "bpb": 2.22648, "chunk_loss": 0.030215, "F": 0.14127, "G": 0.24819, "chunks": 144.66, "bytes_per_chunk": 7.079, "boundary_entropy": 0.37073, "router_noise": 0.0, "grad_norm": 0.8392, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.3671, "alpha": 0.03, "t": 225.0} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "parity_a", "lm_loss": 1.48026, "bpb": 2.13557, "chunk_loss": 0.029901, "F": 0.14569, "G": 0.23709, "chunks": 149.19, "bytes_per_chunk": 6.864, "boundary_entropy": 0.33604, "router_noise": 0.0, "grad_norm": 0.8126, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.3124, "alpha": 0.03, "t": 227.1} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "parity_a", "lm_loss": 1.43333, "bpb": 2.06785, "chunk_loss": 0.029804, "F": 0.13385, "G": 0.24771, "chunks": 137.06, "bytes_per_chunk": 7.471, "boundary_entropy": 0.37491, "router_noise": 0.0, "grad_norm": 0.8872, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 8.5064, "alpha": 0.03, "t": 229.2} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2876.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165259, "cps": 22.46875, "mean_p": 0.196913, "boundary_entropy": 0.208458, "bpb": 2.102401}, "de": {"chunks": 2981.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.145181, "cps": 23.289062, "mean_p": 0.185247, "boundary_entropy": 0.235152, "bpb": 2.32445}, "vi": {"chunks": 2948.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.159913, "cps": 23.03125, "mean_p": 0.196911, "boundary_entropy": 0.330704, "bpb": 1.979462}, "tr": {"chunks": 2908.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.160636, "cps": 22.71875, "mean_p": 0.196192, "boundary_entropy": 0.259615, "bpb": 2.298394}, "fi": {"chunks": 2833.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.150603, "cps": 22.132812, "mean_p": 0.189101, "boundary_entropy": 0.239064, "bpb": 2.485472}, "ru": {"chunks": 3019.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.159314, "cps": 23.585938, "mean_p": 0.293952, "boundary_entropy": 0.537668, "bpb": 1.477795}, "el": {"chunks": 3001.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.146226, "cps": 23.445312, "mean_p": 0.299998, "boundary_entropy": 0.553667, "bpb": 1.570646}, "hi": {"chunks": 2957.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.17575, "cps": 23.101562, "mean_p": 0.249985, "boundary_entropy": 0.501341, "bpb": 1.232749}, "ta": {"chunks": 2821.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.150896, "cps": 22.039062, "mean_p": 0.272115, "boundary_entropy": 0.538402, "bpb": 1.018868}, "th": {"chunks": 3659.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.227479, "cps": 28.585938, "mean_p": 0.245109, "boundary_entropy": 0.49093, "bpb": 1.142393}, "zh": {"chunks": 3011.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.509821, "cps": 23.523438, "mean_p": 0.353575, "boundary_entropy": 0.562487, "bpb": 3.205675}, "my": {"chunks": 2489.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.13136, "cps": 19.445312, "mean_p": 0.346095, "boundary_entropy": 0.625588, "bpb": 1.155178}}, "bpb": {"en": 2.1024, "de": 2.32445, "vi": 1.97946, "tr": 2.29839, "fi": 2.48547, "ru": 1.47779, "el": 1.57065, "hi": 1.23275, "ta": 1.01887, "th": 1.14239, "zh": 3.20568, "my": 1.15518}, "gini_cpc": 0.190349, "gini_cps": 0.039318, "bpic": 10.809819, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8785061524827873, "vi": 0.9676528403181799, "tr": 0.972028702701594, "fi": 0.9113179605428688, "ru": 0.9640268659564553, "el": 0.8848311203068956, "hi": 1.0634852971871533, "ta": 0.9130884623639635, "th": 1.3765011631963886, "zh": 3.0849814101969337, "my": 0.7948711949518235}, "t": 232.3} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.10861, "de": 2.33432, "vi": 1.9828, "tr": 2.30319, "fi": 2.49342, "ru": 1.45511, "el": 1.56328, "hi": 1.22769, "ta": 1.02522, "th": 1.16129, "zh": 3.20426, "my": 1.16364}, "bytes_per_chunk": {"en": 5.923, "de": 6.834, "vi": 8.062, "tr": 6.374, "fi": 6.386, "ru": 11.472, "el": 12.748, "hi": 14.95, "ta": 17.12, "th": 13.151, "zh": 5.346, "my": 18.743}, "bpb_mean": 1.83524, "t": 233.5} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "parity_a", "lm_loss": 1.32334, "bpb": 1.90918, "chunk_loss": 0.03043, "F": 0.12527, "G": 0.24014, "chunks": 128.28, "bytes_per_chunk": 7.982, "boundary_entropy": 0.3756, "router_noise": 0.0, "grad_norm": 0.8425, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 10.021, "alpha": 0.03, "t": 233.6} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "parity_a", "lm_loss": 1.47571, "bpb": 2.129, "chunk_loss": 0.030078, "F": 0.14087, "G": 0.24036, "chunks": 144.25, "bytes_per_chunk": 7.099, "boundary_entropy": 0.35778, "router_noise": 0.0, "grad_norm": 0.7778, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.3995, "alpha": 0.03, "t": 235.7} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "parity_a", "lm_loss": 1.33711, "bpb": 1.92904, "chunk_loss": 0.030392, "F": 0.13403, "G": 0.24689, "chunks": 137.25, "bytes_per_chunk": 7.461, "boundary_entropy": 0.37678, "router_noise": 0.0, "grad_norm": 0.8245, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.8635, "alpha": 0.03, "t": 237.9} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "parity_a", "lm_loss": 1.49654, "bpb": 2.15905, "chunk_loss": 0.030263, "F": 0.14792, "G": 0.24932, "chunks": 151.47, "bytes_per_chunk": 6.76, "boundary_entropy": 0.35827, "router_noise": 0.0, "grad_norm": 0.8226, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 8.0388, "alpha": 0.03, "t": 240.0} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2908.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167098, "cps": 22.71875, "mean_p": 0.194679, "boundary_entropy": 0.198766, "bpb": 2.087249}, "de": {"chunks": 3072.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.149613, "cps": 24.0, "mean_p": 0.185748, "boundary_entropy": 0.227018, "bpb": 2.307909}, "vi": {"chunks": 2843.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.154218, "cps": 22.210938, "mean_p": 0.194581, "boundary_entropy": 0.328704, "bpb": 1.972503}, "tr": {"chunks": 2962.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.163619, "cps": 23.140625, "mean_p": 0.199432, "boundary_entropy": 0.254924, "bpb": 2.283351}, "fi": {"chunks": 2950.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.156823, "cps": 23.046875, "mean_p": 0.193109, "boundary_entropy": 0.237318, "bpb": 2.471433}, "ru": {"chunks": 2776.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.146491, "cps": 21.6875, "mean_p": 0.279266, "boundary_entropy": 0.519783, "bpb": 1.46301}, "el": {"chunks": 2488.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.12123, "cps": 19.4375, "mean_p": 0.276526, "boundary_entropy": 0.532775, "bpb": 1.558012}, "hi": {"chunks": 3129.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.185973, "cps": 24.445312, "mean_p": 0.244592, "boundary_entropy": 0.490025, "bpb": 1.220637}, "ta": {"chunks": 3204.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.171383, "cps": 25.03125, "mean_p": 0.271768, "boundary_entropy": 0.530387, "bpb": 1.009577}, "th": {"chunks": 3008.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.187007, "cps": 23.5, "mean_p": 0.245015, "boundary_entropy": 0.489671, "bpb": 1.136514}, "zh": {"chunks": 2842.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.481206, "cps": 22.203125, "mean_p": 0.336105, "boundary_entropy": 0.54328, "bpb": 3.193833}, "my": {"chunks": 2005.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.105816, "cps": 15.664062, "mean_p": 0.3457, "boundary_entropy": 0.625195, "bpb": 1.147375}}, "bpb": {"en": 2.08725, "de": 2.30791, "vi": 1.9725, "tr": 2.28335, "fi": 2.47143, "ru": 1.46301, "el": 1.55801, "hi": 1.22064, "ta": 1.00958, "th": 1.13651, "zh": 3.19383, "my": 1.14738}, "gini_cpc": 0.195998, "gini_cps": 0.054258, "bpic": 11.225934, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8953617188581792, "vi": 0.9229186789228223, "tr": 0.9791838023609515, "fi": 0.9385120444833231, "ru": 0.8766777119256133, "el": 0.7255030605779902, "hi": 1.1129616715480788, "ta": 1.025644264901023, "th": 1.1191453245075702, "zh": 2.879786881221671, "my": 0.6332580494680439}, "t": 243.0} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.09573, "de": 2.31631, "vi": 1.97686, "tr": 2.28772, "fi": 2.47878, "ru": 1.44041, "el": 1.55133, "hi": 1.21569, "ta": 1.01705, "th": 1.15507, "zh": 3.19293, "my": 1.15444}, "bytes_per_chunk": {"en": 5.883, "de": 6.656, "vi": 8.486, "tr": 6.23, "fi": 6.133, "ru": 12.806, "el": 14.737, "hi": 13.861, "ta": 17.017, "th": 15.641, "zh": 5.592, "my": 23.931}, "bpb_mean": 1.82353, "t": 244.3} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 244.3} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "parity_a", "lm_loss": 1.30885, "bpb": 1.88827, "chunk_loss": 0.02993, "F": 0.12091, "G": 0.23998, "chunks": 123.81, "bytes_per_chunk": 8.271, "boundary_entropy": 0.38647, "router_noise": 0.0, "grad_norm": 0.7898, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 9.7253, "alpha": 0.03, "t": 244.3} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "parity_a", "lm_loss": 1.32127, "bpb": 1.9062, "chunk_loss": 0.030501, "F": 0.13431, "G": 0.24364, "chunks": 137.53, "bytes_per_chunk": 7.446, "boundary_entropy": 0.37359, "router_noise": 0.0, "grad_norm": 0.8114, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.6539, "alpha": 0.03, "t": 246.4} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "parity_a", "lm_loss": 1.36809, "bpb": 1.97374, "chunk_loss": 0.029863, "F": 0.13339, "G": 0.22383, "chunks": 136.59, "bytes_per_chunk": 7.497, "boundary_entropy": 0.32008, "router_noise": 0.0, "grad_norm": 0.8579, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.515, "alpha": 0.03, "t": 248.6} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "parity_a", "lm_loss": 1.49698, "bpb": 2.15968, "chunk_loss": 0.030079, "F": 0.13837, "G": 0.252, "chunks": 141.69, "bytes_per_chunk": 7.227, "boundary_entropy": 0.38592, "router_noise": 0.0, "grad_norm": 0.8746, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.7193, "alpha": 0.03, "t": 250.7} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2931.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168419, "cps": 22.898438, "mean_p": 0.197243, "boundary_entropy": 0.195353, "bpb": 2.072806}, "de": {"chunks": 3118.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151853, "cps": 24.359375, "mean_p": 0.186116, "boundary_entropy": 0.218912, "bpb": 2.298211}, "vi": {"chunks": 2796.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.151668, "cps": 21.84375, "mean_p": 0.192041, "boundary_entropy": 0.315536, "bpb": 1.961762}, "tr": {"chunks": 3018.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.166713, "cps": 23.578125, "mean_p": 0.200598, "boundary_entropy": 0.250518, "bpb": 2.268859}, "fi": {"chunks": 2856.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.151826, "cps": 22.3125, "mean_p": 0.190113, "boundary_entropy": 0.229634, "bpb": 2.451962}, "ru": {"chunks": 3571.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.188443, "cps": 27.898438, "mean_p": 0.292868, "boundary_entropy": 0.529788, "bpb": 1.452783}, "el": {"chunks": 2910.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.141792, "cps": 22.734375, "mean_p": 0.285242, "boundary_entropy": 0.539201, "bpb": 1.547212}, "hi": {"chunks": 2827.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.168024, "cps": 22.085938, "mean_p": 0.237167, "boundary_entropy": 0.481741, "bpb": 1.209885}, "ta": {"chunks": 3225.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.172506, "cps": 25.195312, "mean_p": 0.265023, "boundary_entropy": 0.521988, "bpb": 1.009578}, "th": {"chunks": 2539.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.157849, "cps": 19.835938, "mean_p": 0.238475, "boundary_entropy": 0.484459, "bpb": 1.128604}, "zh": {"chunks": 2956.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.500508, "cps": 23.09375, "mean_p": 0.336973, "boundary_entropy": 0.538828, "bpb": 3.18225}, "my": {"chunks": 2624.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.138484, "cps": 20.5, "mean_p": 0.317642, "boundary_entropy": 0.601788, "bpb": 1.146636}}, "bpb": {"en": 2.07281, "de": 2.29821, "vi": 1.96176, "tr": 2.26886, "fi": 2.45196, "ru": 1.45278, "el": 1.54721, "hi": 1.20989, "ta": 1.00958, "th": 1.1286, "zh": 3.18225, "my": 1.14664}, "gini_cpc": 0.17302, "gini_cps": 0.047805, "bpic": 10.85016, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9016375815828539, "vi": 0.9005385876793592, "tr": 0.9898673184282525, "fi": 0.9014769391393034, "ru": 1.1188939810901393, "el": 0.8418998760708397, "hi": 0.9976519076957239, "ta": 1.0242655152486873, "th": 0.9372381052679137, "zh": 2.971798018842452, "my": 0.8222592284567427}, "t": 253.8} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.08249, "de": 2.30674, "vi": 1.96496, "tr": 2.27361, "fi": 2.45922, "ru": 1.43007, "el": 1.53903, "hi": 1.20611, "ta": 1.01611, "th": 1.14782, "zh": 3.18021, "my": 1.15427}, "bytes_per_chunk": {"en": 5.775, "de": 6.609, "vi": 8.917, "tr": 6.097, "fi": 6.268, "ru": 10.148, "el": 13.1, "hi": 15.445, "ta": 17.08, "th": 18.049, "zh": 5.429, "my": 18.215}, "bpb_mean": 1.81339, "t": 255.0} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "parity_a", "lm_loss": 1.46128, "bpb": 2.10818, "chunk_loss": 0.030204, "F": 0.14661, "G": 0.22756, "chunks": 150.12, "bytes_per_chunk": 6.821, "boundary_entropy": 0.31237, "router_noise": 0.0, "grad_norm": 0.8814, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 7.7349, "alpha": 0.03, "t": 255.0} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "parity_a", "lm_loss": 1.39039, "bpb": 2.0059, "chunk_loss": 0.030096, "F": 0.1347, "G": 0.23711, "chunks": 137.94, "bytes_per_chunk": 7.424, "boundary_entropy": 0.35415, "router_noise": 0.0, "grad_norm": 0.8199, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.4136, "alpha": 0.03, "t": 257.2} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "parity_a", "lm_loss": 1.324, "bpb": 1.91013, "chunk_loss": 0.030162, "F": 0.14297, "G": 0.22412, "chunks": 146.41, "bytes_per_chunk": 6.994, "boundary_entropy": 0.30896, "router_noise": 0.0, "grad_norm": 0.8362, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.5424, "alpha": 0.03, "t": 259.3} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "parity_a", "lm_loss": 1.37769, "bpb": 1.98759, "chunk_loss": 0.030263, "F": 0.13577, "G": 0.24089, "chunks": 139.03, "bytes_per_chunk": 7.365, "boundary_entropy": 0.36672, "router_noise": 0.0, "grad_norm": 0.8483, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.9675, "alpha": 0.03, "t": 261.5} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2934.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168592, "cps": 22.921875, "mean_p": 0.197389, "boundary_entropy": 0.193766, "bpb": 2.067821}, "de": {"chunks": 3038.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.147957, "cps": 23.734375, "mean_p": 0.181366, "boundary_entropy": 0.215441, "bpb": 2.284161}, "vi": {"chunks": 2989.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.162137, "cps": 23.351562, "mean_p": 0.197376, "boundary_entropy": 0.321454, "bpb": 1.953208}, "tr": {"chunks": 2970.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164061, "cps": 23.203125, "mean_p": 0.197335, "boundary_entropy": 0.249429, "bpb": 2.256748}, "fi": {"chunks": 2856.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.151826, "cps": 22.3125, "mean_p": 0.188445, "boundary_entropy": 0.228134, "bpb": 2.443431}, "ru": {"chunks": 2796.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.147546, "cps": 21.84375, "mean_p": 0.28586, "boundary_entropy": 0.529465, "bpb": 1.441312}, "el": {"chunks": 2820.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.137407, "cps": 22.03125, "mean_p": 0.286866, "boundary_entropy": 0.540774, "bpb": 1.535103}, "hi": {"chunks": 2975.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.17682, "cps": 23.242188, "mean_p": 0.235782, "boundary_entropy": 0.478164, "bpb": 1.206157}, "ta": {"chunks": 3230.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.172773, "cps": 25.234375, "mean_p": 0.269817, "boundary_entropy": 0.526891, "bpb": 1.00204}, "th": {"chunks": 2847.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.176997, "cps": 22.242188, "mean_p": 0.23886, "boundary_entropy": 0.481592, "bpb": 1.117537}, "zh": {"chunks": 2907.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.492211, "cps": 22.710938, "mean_p": 0.325957, "boundary_entropy": 0.52693, "bpb": 3.167645}, "my": {"chunks": 2534.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.133734, "cps": 19.796875, "mean_p": 0.309299, "boundary_entropy": 0.591526, "bpb": 1.137793}}, "bpb": {"en": 2.06782, "de": 2.28416, "vi": 1.95321, "tr": 2.25675, "fi": 2.44343, "ru": 1.44131, "el": 1.5351, "hi": 1.20616, "ta": 1.00204, "th": 1.11754, "zh": 3.16764, "my": 1.13779}, "gini_cpc": 0.174755, "gini_cps": 0.028719, "bpic": 10.997851, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8776055742280097, "vi": 0.9617158723265239, "tr": 0.9731278651235314, "fi": 0.9005551835778113, "ru": 0.8751690758696603, "el": 0.8150275211202515, "hi": 1.0488077983087054, "ta": 1.0248045938160253, "th": 1.0498576389617276, "zh": 2.9195478654337172, "my": 0.7932448207191094}, "t": 264.5} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.07739, "de": 2.29309, "vi": 1.9565, "tr": 2.26137, "fi": 2.44823, "ru": 1.41922, "el": 1.52709, "hi": 1.20214, "ta": 1.00814, "th": 1.13655, "zh": 3.16633, "my": 1.14533}, "bytes_per_chunk": {"en": 5.751, "de": 6.733, "vi": 8.193, "tr": 6.195, "fi": 6.287, "ru": 12.385, "el": 13.348, "hi": 14.453, "ta": 16.724, "th": 16.194, "zh": 5.553, "my": 18.695}, "bpb_mean": 1.80345, "t": 265.8} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "parity_a", "lm_loss": 1.42364, "bpb": 2.05388, "chunk_loss": 0.030323, "F": 0.13504, "G": 0.24591, "chunks": 138.28, "bytes_per_chunk": 7.405, "boundary_entropy": 0.37827, "router_noise": 0.0, "grad_norm": 0.8152, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 8.9877, "alpha": 0.03, "t": 265.8} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "parity_a", "lm_loss": 1.52478, "bpb": 2.19979, "chunk_loss": 0.030008, "F": 0.14282, "G": 0.24141, "chunks": 146.25, "bytes_per_chunk": 7.002, "boundary_entropy": 0.36378, "router_noise": 0.0, "grad_norm": 0.9053, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 7.9514, "alpha": 0.03, "t": 268.0} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "parity_a", "lm_loss": 1.46926, "bpb": 2.11969, "chunk_loss": 0.030069, "F": 0.13324, "G": 0.24549, "chunks": 136.44, "bytes_per_chunk": 7.505, "boundary_entropy": 0.38459, "router_noise": 0.0, "grad_norm": 0.9199, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.0389, "alpha": 0.03, "t": 270.0} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "parity_a", "lm_loss": 1.48109, "bpb": 2.13677, "chunk_loss": 0.030086, "F": 0.15082, "G": 0.22793, "chunks": 154.44, "bytes_per_chunk": 6.631, "boundary_entropy": 0.30634, "router_noise": 0.0, "grad_norm": 0.8727, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.3322, "alpha": 0.03, "t": 272.2} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2926.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168132, "cps": 22.859375, "mean_p": 0.196693, "boundary_entropy": 0.19149, "bpb": 2.05663}, "de": {"chunks": 3073.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.149662, "cps": 24.007812, "mean_p": 0.181292, "boundary_entropy": 0.212886, "bpb": 2.272173}, "vi": {"chunks": 2962.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.160673, "cps": 23.140625, "mean_p": 0.194077, "boundary_entropy": 0.317319, "bpb": 1.950931}, "tr": {"chunks": 2946.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.162735, "cps": 23.015625, "mean_p": 0.195026, "boundary_entropy": 0.245555, "bpb": 2.244665}, "fi": {"chunks": 2840.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.150975, "cps": 22.1875, "mean_p": 0.188642, "boundary_entropy": 0.227115, "bpb": 2.422125}, "ru": {"chunks": 3324.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.175409, "cps": 25.96875, "mean_p": 0.290049, "boundary_entropy": 0.529045, "bpb": 1.438188}, "el": {"chunks": 2772.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.135068, "cps": 21.65625, "mean_p": 0.288078, "boundary_entropy": 0.543413, "bpb": 1.529184}, "hi": {"chunks": 3065.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.182169, "cps": 23.945312, "mean_p": 0.241245, "boundary_entropy": 0.485432, "bpb": 1.201758}, "ta": {"chunks": 2505.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.133993, "cps": 19.570312, "mean_p": 0.262776, "boundary_entropy": 0.522684, "bpb": 0.997495}, "th": {"chunks": 2797.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.173889, "cps": 21.851562, "mean_p": 0.239474, "boundary_entropy": 0.48218, "bpb": 1.116545}, "zh": {"chunks": 2967.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50237, "cps": 23.179688, "mean_p": 0.329964, "boundary_entropy": 0.529661, "bpb": 3.157771}, "my": {"chunks": 3094.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.163289, "cps": 24.171875, "mean_p": 0.341314, "boundary_entropy": 0.617991, "bpb": 1.135285}}, "bpb": {"en": 2.05663, "de": 2.27217, "vi": 1.95093, "tr": 2.24466, "fi": 2.42213, "ru": 1.43819, "el": 1.52918, "hi": 1.20176, "ta": 0.99749, "th": 1.11654, "zh": 3.15777, "my": 1.13529}, "gini_cpc": 0.17689, "gini_cps": 0.035683, "bpic": 10.880922, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8901433495898887, "vi": 0.9556342591073437, "tr": 0.9679033421914073, "fi": 0.8979584777060498, "ru": 1.0432817231372986, "el": 0.8033451557559297, "hi": 1.0834907409006103, "ta": 0.7969518059536499, "th": 1.0342396882929905, "zh": 2.987953965046841, "my": 0.9711956571144597}, "t": 275.2} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.06713, "de": 2.28064, "vi": 1.95267, "tr": 2.24919, "fi": 2.42948, "ru": 1.41503, "el": 1.52076, "hi": 1.19835, "ta": 1.00458, "th": 1.13435, "zh": 3.15581, "my": 1.1426}, "bytes_per_chunk": {"en": 5.777, "de": 6.651, "vi": 8.159, "tr": 6.234, "fi": 6.293, "ru": 10.741, "el": 13.278, "hi": 14.044, "ta": 20.035, "th": 16.893, "zh": 5.362, "my": 16.268}, "bpb_mean": 1.79588, "t": 276.5} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "parity_a", "lm_loss": 1.3643, "bpb": 1.96826, "chunk_loss": 0.030416, "F": 0.14182, "G": 0.24156, "chunks": 145.22, "bytes_per_chunk": 7.051, "boundary_entropy": 0.34979, "router_noise": 0.0, "grad_norm": 0.8128, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 8.7435, "alpha": 0.03, "t": 276.5} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "parity_a", "lm_loss": 1.29726, "bpb": 1.87155, "chunk_loss": 0.030003, "F": 0.13931, "G": 0.22269, "chunks": 142.66, "bytes_per_chunk": 7.178, "boundary_entropy": 0.30423, "router_noise": 0.0, "grad_norm": 0.8136, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.8411, "alpha": 0.03, "t": 278.7} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "parity_a", "lm_loss": 1.39378, "bpb": 2.0108, "chunk_loss": 0.030132, "F": 0.13199, "G": 0.23553, "chunks": 135.16, "bytes_per_chunk": 7.576, "boundary_entropy": 0.34373, "router_noise": 0.0, "grad_norm": 0.8445, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.3676, "alpha": 0.03, "t": 280.8} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "parity_a", "lm_loss": 1.27069, "bpb": 1.83321, "chunk_loss": 0.030264, "F": 0.13159, "G": 0.23943, "chunks": 134.75, "bytes_per_chunk": 7.599, "boundary_entropy": 0.36041, "router_noise": 0.0, "grad_norm": 0.8534, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.4148, "alpha": 0.03, "t": 282.9} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2917.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167615, "cps": 22.789062, "mean_p": 0.195904, "boundary_entropy": 0.191273, "bpb": 2.052498}, "de": {"chunks": 3059.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.14898, "cps": 23.898438, "mean_p": 0.182287, "boundary_entropy": 0.215178, "bpb": 2.271318}, "vi": {"chunks": 2947.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.159859, "cps": 23.023438, "mean_p": 0.194732, "boundary_entropy": 0.314604, "bpb": 1.94155}, "tr": {"chunks": 2945.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16268, "cps": 23.007812, "mean_p": 0.194815, "boundary_entropy": 0.242322, "bpb": 2.241991}, "fi": {"chunks": 2911.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15475, "cps": 22.742188, "mean_p": 0.190453, "boundary_entropy": 0.226392, "bpb": 2.410024}, "ru": {"chunks": 2648.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.139736, "cps": 20.6875, "mean_p": 0.276544, "boundary_entropy": 0.51863, "bpb": 1.429982}, "el": {"chunks": 2635.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.128393, "cps": 20.585938, "mean_p": 0.27721, "boundary_entropy": 0.532302, "bpb": 1.524634}, "hi": {"chunks": 2820.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.167608, "cps": 22.03125, "mean_p": 0.234365, "boundary_entropy": 0.477685, "bpb": 1.194014}, "ta": {"chunks": 3003.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.160631, "cps": 23.460938, "mean_p": 0.276271, "boundary_entropy": 0.533864, "bpb": 0.993182}, "th": {"chunks": 2850.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.177184, "cps": 22.265625, "mean_p": 0.249502, "boundary_entropy": 0.489321, "bpb": 1.113904}, "zh": {"chunks": 2865.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.4851, "cps": 22.382812, "mean_p": 0.32669, "boundary_entropy": 0.528925, "bpb": 3.147637}, "my": {"chunks": 2391.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.126187, "cps": 18.679688, "mean_p": 0.330332, "boundary_entropy": 0.609694, "bpb": 1.128737}}, "bpb": {"en": 2.0525, "de": 2.27132, "vi": 1.94155, "tr": 2.24199, "fi": 2.41002, "ru": 1.42998, "el": 1.52463, "hi": 1.19401, "ta": 0.99318, "th": 1.1139, "zh": 3.14764, "my": 1.12874}, "gini_cpc": 0.179769, "gini_cps": 0.033403, "bpic": 11.290665, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8888219288495033, "vi": 0.9537283336397031, "tr": 0.9705601120590361, "fi": 0.9232472294864927, "ru": 0.833674384442758, "el": 0.7659976997470884, "hi": 0.9999579245708791, "ta": 0.9583354505406466, "th": 1.057088835035932, "zh": 2.8941355954752677, "my": 0.7528420890141281}, "t": 286.0} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.06248, "de": 2.27995, "vi": 1.94426, "tr": 2.2466, "fi": 2.41821, "ru": 1.40793, "el": 1.51737, "hi": 1.19083, "ta": 0.99966, "th": 1.13318, "zh": 3.14512, "my": 1.13699}, "bytes_per_chunk": {"en": 5.802, "de": 6.629, "vi": 8.515, "tr": 6.17, "fi": 6.158, "ru": 13.043, "el": 13.425, "hi": 15.402, "ta": 17.319, "th": 16.457, "zh": 5.625, "my": 19.952}, "bpb_mean": 1.79021, "t": 287.2} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "parity_a", "lm_loss": 1.3814, "bpb": 1.99294, "chunk_loss": 0.029954, "F": 0.12506, "G": 0.25599, "chunks": 128.06, "bytes_per_chunk": 7.996, "boundary_entropy": 0.40844, "router_noise": 0.0, "grad_norm": 0.8277, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 9.4625, "alpha": 0.03, "t": 287.3} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "parity_a", "lm_loss": 1.39834, "bpb": 2.01737, "chunk_loss": 0.030136, "F": 0.12872, "G": 0.24593, "chunks": 131.81, "bytes_per_chunk": 7.769, "boundary_entropy": 0.37577, "router_noise": 0.0, "grad_norm": 0.9665, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.5005, "alpha": 0.03, "t": 289.4} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "parity_a", "lm_loss": 1.49494, "bpb": 2.15674, "chunk_loss": 0.030086, "F": 0.14474, "G": 0.23448, "chunks": 148.22, "bytes_per_chunk": 6.909, "boundary_entropy": 0.33968, "router_noise": 0.0, "grad_norm": 0.8676, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 7.8644, "alpha": 0.03, "t": 291.5} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "parity_a", "lm_loss": 1.3547, "bpb": 1.95443, "chunk_loss": 0.029923, "F": 0.13309, "G": 0.22487, "chunks": 136.28, "bytes_per_chunk": 7.514, "boundary_entropy": 0.32696, "router_noise": 0.0, "grad_norm": 0.8611, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.3816, "alpha": 0.03, "t": 293.7} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2902.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166753, "cps": 22.671875, "mean_p": 0.193374, "boundary_entropy": 0.183786, "bpb": 2.045722}, "de": {"chunks": 3087.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.150343, "cps": 24.117188, "mean_p": 0.18088, "boundary_entropy": 0.208004, "bpb": 2.260566}, "vi": {"chunks": 2866.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.155465, "cps": 22.390625, "mean_p": 0.190639, "boundary_entropy": 0.31026, "bpb": 1.939269}, "tr": {"chunks": 2931.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.161907, "cps": 22.898438, "mean_p": 0.191656, "boundary_entropy": 0.237552, "bpb": 2.228439}, "fi": {"chunks": 2858.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.151932, "cps": 22.328125, "mean_p": 0.188372, "boundary_entropy": 0.222461, "bpb": 2.398306}, "ru": {"chunks": 2840.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.149868, "cps": 22.1875, "mean_p": 0.28133, "boundary_entropy": 0.521136, "bpb": 1.427911}, "el": {"chunks": 2839.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.138333, "cps": 22.179688, "mean_p": 0.277908, "boundary_entropy": 0.530897, "bpb": 1.517109}, "hi": {"chunks": 2831.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.168262, "cps": 22.117188, "mean_p": 0.231301, "boundary_entropy": 0.471254, "bpb": 1.189781}, "ta": {"chunks": 2684.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.143568, "cps": 20.96875, "mean_p": 0.261081, "boundary_entropy": 0.519468, "bpb": 0.988797}, "th": {"chunks": 2643.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.164315, "cps": 20.648438, "mean_p": 0.242958, "boundary_entropy": 0.482008, "bpb": 1.106219}, "zh": {"chunks": 3068.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.519472, "cps": 23.96875, "mean_p": 0.33288, "boundary_entropy": 0.525973, "bpb": 3.140185}, "my": {"chunks": 2294.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.121068, "cps": 17.921875, "mean_p": 0.33307, "boundary_entropy": 0.613691, "bpb": 1.123878}}, "bpb": {"en": 2.04572, "de": 2.26057, "vi": 1.93927, "tr": 2.22844, "fi": 2.39831, "ru": 1.42791, "el": 1.51711, "hi": 1.18978, "ta": 0.9888, "th": 1.10622, "zh": 3.14018, "my": 1.12388}, "gini_cpc": 0.187278, "gini_cps": 0.036273, "bpic": 11.340041, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9015938371281972, "vi": 0.9323087413691296, "tr": 0.9709390742985246, "fi": 0.9111230906180128, "ru": 0.898743655999229, "el": 0.8295666024701472, "hi": 1.009047262583677, "ta": 0.8609615450900405, "th": 0.9853778824460938, "zh": 3.115219299463709, "my": 0.7260336497857607}, "t": 296.7} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.05527, "de": 2.26957, "vi": 1.94149, "tr": 2.23354, "fi": 2.4049, "ru": 1.40645, "el": 1.50884, "hi": 1.1861, "ta": 0.99615, "th": 1.12494, "zh": 3.13747, "my": 1.13156}, "bytes_per_chunk": {"en": 5.836, "de": 6.613, "vi": 8.709, "tr": 6.257, "fi": 6.256, "ru": 12.128, "el": 13.295, "hi": 15.143, "ta": 19.233, "th": 17.76, "zh": 5.295, "my": 20.645}, "bpb_mean": 1.78302, "t": 298.0} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "parity_a", "lm_loss": 1.26919, "bpb": 1.83105, "chunk_loss": 0.029965, "F": 0.12173, "G": 0.24441, "chunks": 124.66, "bytes_per_chunk": 8.215, "boundary_entropy": 0.38159, "router_noise": 0.0, "grad_norm": 0.8461, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 10.3687, "alpha": 0.03, "t": 298.0} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "parity_a", "lm_loss": 1.31215, "bpb": 1.89303, "chunk_loss": 0.030221, "F": 0.13104, "G": 0.24735, "chunks": 134.19, "bytes_per_chunk": 7.631, "boundary_entropy": 0.37891, "router_noise": 0.0, "grad_norm": 0.8006, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.3915, "alpha": 0.03, "t": 300.2} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "parity_a", "lm_loss": 1.23875, "bpb": 1.78713, "chunk_loss": 0.0301, "F": 0.1243, "G": 0.23738, "chunks": 127.28, "bytes_per_chunk": 8.045, "boundary_entropy": 0.36433, "router_noise": 0.0, "grad_norm": 0.7883, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 10.1505, "alpha": 0.03, "t": 302.3} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "parity_a", "lm_loss": 1.46835, "bpb": 2.11838, "chunk_loss": 0.029781, "F": 0.12958, "G": 0.24991, "chunks": 132.69, "bytes_per_chunk": 7.717, "boundary_entropy": 0.39754, "router_noise": 0.0, "grad_norm": 0.8477, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.0422, "alpha": 0.03, "t": 304.5} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2913.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167385, "cps": 22.757812, "mean_p": 0.194149, "boundary_entropy": 0.18092, "bpb": 2.043066}, "de": {"chunks": 3109.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151415, "cps": 24.289062, "mean_p": 0.180555, "boundary_entropy": 0.202403, "bpb": 2.251633}, "vi": {"chunks": 3065.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.16626, "cps": 23.945312, "mean_p": 0.193621, "boundary_entropy": 0.306266, "bpb": 1.935109}, "tr": {"chunks": 2972.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164172, "cps": 23.21875, "mean_p": 0.194155, "boundary_entropy": 0.239583, "bpb": 2.222676}, "fi": {"chunks": 2874.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.152783, "cps": 22.453125, "mean_p": 0.186864, "boundary_entropy": 0.21728, "bpb": 2.391872}, "ru": {"chunks": 3148.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.166121, "cps": 24.59375, "mean_p": 0.281114, "boundary_entropy": 0.518494, "bpb": 1.41959}, "el": {"chunks": 2845.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.138625, "cps": 22.226562, "mean_p": 0.276413, "boundary_entropy": 0.528498, "bpb": 1.509382}, "hi": {"chunks": 2812.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.167132, "cps": 21.96875, "mean_p": 0.229781, "boundary_entropy": 0.467479, "bpb": 1.183192}, "ta": {"chunks": 2901.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.155175, "cps": 22.664062, "mean_p": 0.267281, "boundary_entropy": 0.523794, "bpb": 0.98493}, "th": {"chunks": 3063.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.190426, "cps": 23.929688, "mean_p": 0.250776, "boundary_entropy": 0.488418, "bpb": 1.102381}, "zh": {"chunks": 3016.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.510667, "cps": 23.5625, "mean_p": 0.326403, "boundary_entropy": 0.519602, "bpb": 3.129359}, "my": {"chunks": 2842.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.149989, "cps": 22.203125, "mean_p": 0.340131, "boundary_entropy": 0.617954, "bpb": 1.121801}}, "bpb": {"en": 2.04307, "de": 2.25163, "vi": 1.93511, "tr": 2.22268, "fi": 2.39187, "ru": 1.41959, "el": 1.50938, "hi": 1.18319, "ta": 0.98493, "th": 1.10238, "zh": 3.12936, "my": 1.1218}, "gini_cpc": 0.170863, "gini_cps": 0.021063, "bpic": 10.792492, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9045903499744177, "vi": 0.9932783568621569, "tr": 0.9808032341311842, "fi": 0.9127640274022896, "ru": 0.9924511628791688, "el": 0.8281806155240261, "hi": 0.9984903662375303, "ta": 0.9270558416600815, "th": 1.1376523470351092, "zh": 3.0508547400520967, "my": 0.8960749425986063}, "t": 307.6} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.05236, "de": 2.26006, "vi": 1.93745, "tr": 2.2272, "fi": 2.39831, "ru": 1.39648, "el": 1.5021, "hi": 1.17881, "ta": 0.99221, "th": 1.12049, "zh": 3.12693, "my": 1.13067}, "bytes_per_chunk": {"en": 5.808, "de": 6.559, "vi": 8.189, "tr": 6.212, "fi": 6.242, "ru": 11.268, "el": 12.915, "hi": 15.182, "ta": 18.408, "th": 15.407, "zh": 5.381, "my": 17.359}, "bpb_mean": 1.77692, "t": 308.8} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "parity_a", "lm_loss": 1.36735, "bpb": 1.97267, "chunk_loss": 0.030317, "F": 0.1358, "G": 0.23321, "chunks": 139.06, "bytes_per_chunk": 7.364, "boundary_entropy": 0.34145, "router_noise": 0.0, "grad_norm": 0.9072, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 9.1908, "alpha": 0.03, "t": 308.8} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "parity_a", "lm_loss": 1.22549, "bpb": 1.768, "chunk_loss": 0.030039, "F": 0.11862, "G": 0.23397, "chunks": 121.47, "bytes_per_chunk": 8.43, "boundary_entropy": 0.37094, "router_noise": 0.0, "grad_norm": 0.8191, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.9327, "alpha": 0.03, "t": 311.0} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "parity_a", "lm_loss": 1.3323, "bpb": 1.9221, "chunk_loss": 0.030812, "F": 0.13629, "G": 0.23403, "chunks": 139.56, "bytes_per_chunk": 7.337, "boundary_entropy": 0.33834, "router_noise": 0.0, "grad_norm": 0.8573, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.3141, "alpha": 0.03, "t": 313.1} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "parity_a", "lm_loss": 1.40754, "bpb": 2.03065, "chunk_loss": 0.030027, "F": 0.13998, "G": 0.23046, "chunks": 143.34, "bytes_per_chunk": 7.144, "boundary_entropy": 0.32183, "router_noise": 0.0, "grad_norm": 0.8891, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 8.7726, "alpha": 0.03, "t": 315.3} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2923.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.16796, "cps": 22.835938, "mean_p": 0.194782, "boundary_entropy": 0.181282, "bpb": 2.0373}, "de": {"chunks": 3103.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151123, "cps": 24.242188, "mean_p": 0.180888, "boundary_entropy": 0.202738, "bpb": 2.249575}, "vi": {"chunks": 2961.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.160618, "cps": 23.132812, "mean_p": 0.1867, "boundary_entropy": 0.299928, "bpb": 1.930454}, "tr": {"chunks": 3003.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.165884, "cps": 23.460938, "mean_p": 0.194825, "boundary_entropy": 0.238561, "bpb": 2.213316}, "fi": {"chunks": 2880.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.153102, "cps": 22.5, "mean_p": 0.187705, "boundary_entropy": 0.218173, "bpb": 2.389391}, "ru": {"chunks": 3009.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.158786, "cps": 23.507812, "mean_p": 0.280701, "boundary_entropy": 0.518608, "bpb": 1.414568}, "el": {"chunks": 2882.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.140428, "cps": 22.515625, "mean_p": 0.275054, "boundary_entropy": 0.525627, "bpb": 1.506149}, "hi": {"chunks": 2889.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.171709, "cps": 22.570312, "mean_p": 0.225987, "boundary_entropy": 0.461133, "bpb": 1.1804}, "ta": {"chunks": 2756.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.147419, "cps": 21.53125, "mean_p": 0.258148, "boundary_entropy": 0.516059, "bpb": 0.983232}, "th": {"chunks": 2805.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.174386, "cps": 21.914062, "mean_p": 0.241336, "boundary_entropy": 0.480187, "bpb": 1.099506}, "zh": {"chunks": 3143.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.532171, "cps": 24.554688, "mean_p": 0.328, "boundary_entropy": 0.51461, "bpb": 3.123301}, "my": {"chunks": 2220.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.117163, "cps": 17.34375, "mean_p": 0.317006, "boundary_entropy": 0.600304, "bpb": 1.11939}}, "bpb": {"en": 2.0373, "de": 2.24957, "vi": 1.93045, "tr": 2.21332, "fi": 2.38939, "ru": 1.41457, "el": 1.50615, "hi": 1.1804, "ta": 0.98323, "th": 1.09951, "zh": 3.1233, "my": 1.11939}, "gini_cpc": 0.19209, "gini_cps": 0.037576, "bpic": 11.100278, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8997558380817314, "vi": 0.9562921048990817, "tr": 0.9876432174736463, "fi": 0.911540373893737, "ru": 0.9453840661408878, "el": 0.8360811578455013, "hi": 1.022322157770086, "ta": 0.8777059623498629, "th": 1.0382623487883125, "zh": 3.1684455141034378, "my": 0.6975653554164228}, "t": 318.3} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.04609, "de": 2.25662, "vi": 1.93265, "tr": 2.21926, "fi": 2.39537, "ru": 1.39204, "el": 1.49854, "hi": 1.17746, "ta": 0.99062, "th": 1.11754, "zh": 3.12013, "my": 1.12689}, "bytes_per_chunk": {"en": 5.798, "de": 6.57, "vi": 8.442, "tr": 6.149, "fi": 6.237, "ru": 11.872, "el": 12.935, "hi": 14.841, "ta": 18.61, "th": 16.586, "zh": 5.224, "my": 20.567}, "bpb_mean": 1.77277, "t": 319.6} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "parity_a", "lm_loss": 1.34975, "bpb": 1.94728, "chunk_loss": 0.030141, "F": 0.13376, "G": 0.23758, "chunks": 136.97, "bytes_per_chunk": 7.476, "boundary_entropy": 0.35555, "router_noise": 0.0, "grad_norm": 0.8535, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.087, "n_target_mean": 9.2803, "alpha": 0.03, "t": 319.6} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "parity_a", "lm_loss": 1.33064, "bpb": 1.9197, "chunk_loss": 0.030108, "F": 0.13116, "G": 0.23418, "chunks": 134.31, "bytes_per_chunk": 7.624, "boundary_entropy": 0.35727, "router_noise": 0.0, "grad_norm": 0.8904, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.0063, "alpha": 0.03, "t": 321.8} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "parity_a", "lm_loss": 1.30782, "bpb": 1.88678, "chunk_loss": 0.030144, "F": 0.1351, "G": 0.21803, "chunks": 138.34, "bytes_per_chunk": 7.402, "boundary_entropy": 0.30076, "router_noise": 0.0, "grad_norm": 0.8511, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 8.7491, "alpha": 0.03, "t": 323.9} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "parity_a", "lm_loss": 1.21935, "bpb": 1.75915, "chunk_loss": 0.030005, "F": 0.12552, "G": 0.23061, "chunks": 128.53, "bytes_per_chunk": 7.967, "boundary_entropy": 0.34413, "router_noise": 0.0, "grad_norm": 0.8224, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 9.9933, "alpha": 0.03, "t": 326.0} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2926.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168132, "cps": 22.859375, "mean_p": 0.194425, "boundary_entropy": 0.179335, "bpb": 2.033066}, "de": {"chunks": 3109.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151415, "cps": 24.289062, "mean_p": 0.180506, "boundary_entropy": 0.200533, "bpb": 2.241215}, "vi": {"chunks": 2883.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.156387, "cps": 22.523438, "mean_p": 0.186403, "boundary_entropy": 0.3001, "bpb": 1.927241}, "tr": {"chunks": 3014.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.166492, "cps": 23.546875, "mean_p": 0.194808, "boundary_entropy": 0.236122, "bpb": 2.2087}, "fi": {"chunks": 2887.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.153474, "cps": 22.554688, "mean_p": 0.187016, "boundary_entropy": 0.2155, "bpb": 2.378249}, "ru": {"chunks": 2853.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.150554, "cps": 22.289062, "mean_p": 0.275961, "boundary_entropy": 0.514491, "bpb": 1.412056}, "el": {"chunks": 2924.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.142474, "cps": 22.84375, "mean_p": 0.272781, "boundary_entropy": 0.522209, "bpb": 1.503431}, "hi": {"chunks": 2870.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.170579, "cps": 22.421875, "mean_p": 0.227579, "boundary_entropy": 0.464694, "bpb": 1.177102}, "ta": {"chunks": 2882.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.154159, "cps": 22.515625, "mean_p": 0.258732, "boundary_entropy": 0.512042, "bpb": 0.981279}, "th": {"chunks": 2688.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.167112, "cps": 21.0, "mean_p": 0.239099, "boundary_entropy": 0.477763, "bpb": 1.096011}, "zh": {"chunks": 3185.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.539282, "cps": 24.882812, "mean_p": 0.328972, "boundary_entropy": 0.513217, "bpb": 3.112467}, "my": {"chunks": 2285.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.120593, "cps": 17.851562, "mean_p": 0.306428, "boundary_entropy": 0.590846, "bpb": 1.117021}}, "bpb": {"en": 2.03307, "de": 2.24122, "vi": 1.92724, "tr": 2.2087, "fi": 2.37825, "ru": 1.41206, "el": 1.50343, "hi": 1.1771, "ta": 0.98128, "th": 1.09601, "zh": 3.11247, "my": 1.11702}, "gini_cpc": 0.189727, "gini_cps": 0.036771, "bpic": 11.122153, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9005713224454813, "vi": 0.9301463771122458, "tr": 0.990244627754549, "fi": 0.9128190581469597, "ru": 0.8954520926927536, "el": 0.8473958280773227, "hi": 1.0145573984942093, "ta": 0.9168922573885904, "th": 0.9939350311517908, "zh": 3.2074935553333965, "my": 0.7172534183925471}, "t": 329.0} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.04206, "de": 2.25082, "vi": 1.93025, "tr": 2.21429, "fi": 2.38541, "ru": 1.38951, "el": 1.49539, "hi": 1.17327, "ta": 0.98866, "th": 1.11339, "zh": 3.11057, "my": 1.12498}, "bytes_per_chunk": {"en": 5.82, "de": 6.539, "vi": 8.725, "tr": 6.136, "fi": 6.245, "ru": 12.322, "el": 12.97, "hi": 15.029, "ta": 18.647, "th": 17.111, "zh": 5.137, "my": 19.864}, "bpb_mean": 1.76822, "t": 330.3} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "parity_a", "lm_loss": 1.30651, "bpb": 1.88489, "chunk_loss": 0.030193, "F": 0.13239, "G": 0.22972, "chunks": 135.56, "bytes_per_chunk": 7.554, "boundary_entropy": 0.34345, "router_noise": 0.0, "grad_norm": 0.8886, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.086, "n_target_mean": 9.4562, "alpha": 0.03, "t": 330.3} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "parity_a", "lm_loss": 1.29774, "bpb": 1.87224, "chunk_loss": 0.030059, "F": 0.12604, "G": 0.23813, "chunks": 129.06, "bytes_per_chunk": 7.934, "boundary_entropy": 0.36485, "router_noise": 0.0, "grad_norm": 0.8147, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.9248, "alpha": 0.03, "t": 332.5} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "parity_a", "lm_loss": 1.37292, "bpb": 1.98071, "chunk_loss": 0.030137, "F": 0.1315, "G": 0.24078, "chunks": 134.66, "bytes_per_chunk": 7.605, "boundary_entropy": 0.37204, "router_noise": 0.0, "grad_norm": 0.9348, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.043, "n_target_mean": 9.5896, "alpha": 0.03, "t": 334.6} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "parity_a", "lm_loss": 1.40998, "bpb": 2.03417, "chunk_loss": 0.030083, "F": 0.13834, "G": 0.2478, "chunks": 141.66, "bytes_per_chunk": 7.229, "boundary_entropy": 0.37094, "router_noise": 0.0, "grad_norm": 0.8503, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.042, "n_target_mean": 8.6325, "alpha": 0.03, "t": 336.7} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2912.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167327, "cps": 22.75, "mean_p": 0.193075, "boundary_entropy": 0.176609, "bpb": 2.03092}, "de": {"chunks": 3073.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.149662, "cps": 24.007812, "mean_p": 0.179112, "boundary_entropy": 0.197926, "bpb": 2.241083}, "vi": {"chunks": 2952.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.16013, "cps": 23.0625, "mean_p": 0.185796, "boundary_entropy": 0.296375, "bpb": 1.924045}, "tr": {"chunks": 2984.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164835, "cps": 23.3125, "mean_p": 0.192429, "boundary_entropy": 0.231674, "bpb": 2.2056}, "fi": {"chunks": 2855.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.151773, "cps": 22.304688, "mean_p": 0.184115, "boundary_entropy": 0.210057, "bpb": 2.382605}, "ru": {"chunks": 3099.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.163536, "cps": 24.210938, "mean_p": 0.28128, "boundary_entropy": 0.518638, "bpb": 1.409159}, "el": {"chunks": 2926.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.142572, "cps": 22.859375, "mean_p": 0.276043, "boundary_entropy": 0.526308, "bpb": 1.496338}, "hi": {"chunks": 3036.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.180446, "cps": 23.71875, "mean_p": 0.228852, "boundary_entropy": 0.463725, "bpb": 1.176335}, "ta": {"chunks": 3017.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.16138, "cps": 23.570312, "mean_p": 0.261727, "boundary_entropy": 0.513043, "bpb": 0.979115}, "th": {"chunks": 2962.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.184147, "cps": 23.140625, "mean_p": 0.245104, "boundary_entropy": 0.479221, "bpb": 1.094705}, "zh": {"chunks": 3133.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.530477, "cps": 24.476562, "mean_p": 0.324595, "boundary_entropy": 0.510064, "bpb": 3.111488}, "my": {"chunks": 2885.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.152259, "cps": 22.539062, "mean_p": 0.31423, "boundary_entropy": 0.595955, "bpb": 1.113407}}, "bpb": {"en": 2.03092, "de": 2.24108, "vi": 1.92405, "tr": 2.2056, "fi": 2.38261, "ru": 1.40916, "el": 1.49634, "hi": 1.17633, "ta": 0.97912, "th": 1.09471, "zh": 3.11149, "my": 1.11341}, "gini_cpc": 0.175784, "gini_cps": 0.015995, "bpic": 10.709968, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8944228849244555, "vi": 0.9569868292813777, "tr": 0.9851015829444819, "fi": 0.9070411258376411, "ru": 0.9773387348139986, "el": 0.8520522471973283, "hi": 1.0783989190601373, "ta": 0.9644563797395437, "th": 1.1005169566212463, "zh": 3.1702951901214265, "my": 0.9099451006166094}, "t": 339.7} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.04084, "de": 2.25053, "vi": 1.92663, "tr": 2.2104, "fi": 2.38887, "ru": 1.38658, "el": 1.48922, "hi": 1.17197, "ta": 0.9864, "th": 1.11174, "zh": 3.10856, "my": 1.12147}, "bytes_per_chunk": {"en": 5.827, "de": 6.587, "vi": 8.623, "tr": 6.165, "fi": 6.308, "ru": 11.386, "el": 12.836, "hi": 14.233, "ta": 18.382, "th": 15.817, "zh": 5.251, "my": 16.817}, "bpb_mean": 1.7661, "t": 341.0} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2912.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.167327, "cps": 22.75, "mean_p": 0.193075, "boundary_entropy": 0.176609, "bpb": 2.03092}, "de": {"chunks": 3073.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.149662, "cps": 24.007812, "mean_p": 0.179112, "boundary_entropy": 0.197926, "bpb": 2.241083}, "vi": {"chunks": 2952.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.16013, "cps": 23.0625, "mean_p": 0.185796, "boundary_entropy": 0.296375, "bpb": 1.924045}, "tr": {"chunks": 2984.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164835, "cps": 23.3125, "mean_p": 0.192429, "boundary_entropy": 0.231674, "bpb": 2.2056}, "fi": {"chunks": 2855.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.151773, "cps": 22.304688, "mean_p": 0.184115, "boundary_entropy": 0.210057, "bpb": 2.382605}, "ru": {"chunks": 3099.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.163536, "cps": 24.210938, "mean_p": 0.28128, "boundary_entropy": 0.518638, "bpb": 1.409159}, "el": {"chunks": 2926.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.142572, "cps": 22.859375, "mean_p": 0.276043, "boundary_entropy": 0.526308, "bpb": 1.496338}, "hi": {"chunks": 3036.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.180446, "cps": 23.71875, "mean_p": 0.228852, "boundary_entropy": 0.463725, "bpb": 1.176335}, "ta": {"chunks": 3017.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.16138, "cps": 23.570312, "mean_p": 0.261727, "boundary_entropy": 0.513043, "bpb": 0.979115}, "th": {"chunks": 2962.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.184147, "cps": 23.140625, "mean_p": 0.245104, "boundary_entropy": 0.479221, "bpb": 1.094705}, "zh": {"chunks": 3133.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.530477, "cps": 24.476562, "mean_p": 0.324595, "boundary_entropy": 0.510064, "bpb": 3.111488}, "my": {"chunks": 2885.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.152259, "cps": 22.539062, "mean_p": 0.31423, "boundary_entropy": 0.595955, "bpb": 1.113407}}, "bpb": {"en": 2.03092, "de": 2.24108, "vi": 1.92405, "tr": 2.2056, "fi": 2.38261, "ru": 1.40916, "el": 1.49634, "hi": 1.17633, "ta": 0.97912, "th": 1.09471, "zh": 3.11149, "my": 1.11341}, "gini_cpc": 0.175784, "gini_cps": 0.015995, "bpic": 10.709968, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8944228849244555, "vi": 0.9569868292813777, "tr": 0.9851015829444819, "fi": 0.9070411258376411, "ru": 0.9773387348139986, "el": 0.8520522471973283, "hi": 1.0783989190601373, "ta": 0.9644563797395437, "th": 1.1005169566212463, "zh": 3.1702951901214265, "my": 0.9099451006166094}, "t": 341.9} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_D_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_D_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_D_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_D_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_D_s2/th0.70"], "t": 344.1} |
| {"event": "final", "run": "pilot_D_s2", "condition": "D", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.16732747227489514, "cps": 22.75, "bpb": 2.0501788783388752, "bytes_per_chunk": 5.78371520921356}, "de": {"cpc": 0.14966152047922857, "cps": 24.0078125, "bpb": 2.2412111552777563, "bytes_per_chunk": 6.562541462742987}, "vi": {"cpc": 0.16013018714401953, "cps": 23.0625, "bpb": 1.930560698842099, "bytes_per_chunk": 8.555473964197713}, "tr": {"cpc": 0.1648345578080981, "cps": 23.3125, "bpb": 2.203401177944363, "bytes_per_chunk": 6.183443607071672}, "fi": {"cpc": 0.15177289883578757, "cps": 22.3046875, "bpb": 2.403979631391192, "bytes_per_chunk": 6.310564388969801}, "ru": {"cpc": 0.16353562005277045, "cps": 24.2109375, "bpb": 1.387356672396501, "bytes_per_chunk": 11.404755171738705}, "el": {"cpc": 0.14257174876967305, "cps": 22.859375, "bpb": 1.4845651416482626, "bytes_per_chunk": 12.836981538612212}, "hi": {"cpc": 0.18044576523031203, "cps": 23.71875, "bpb": 1.1793226742945286, "bytes_per_chunk": 14.2659519468858}, "ta": {"cpc": 0.16138004814121423, "cps": 23.5703125, "bpb": 0.9860990857841374, "bytes_per_chunk": 18.334953663227836}, "th": {"cpc": 0.18414672054709358, "cps": 23.140625, "bpb": 1.1074859563602184, "bytes_per_chunk": 15.86636000484203}, "zh": {"cpc": 0.5304774805282764, "cps": 24.4765625, "bpb": 3.1147953965002095, "bytes_per_chunk": 5.2603971224176505}, "my": {"cpc": 0.15225881359510238, "cps": 22.5390625, "bpb": 1.1316485030339114, "bytes_per_chunk": 16.763804956035173}}, "gini_cpc": 0.1757837534840907, "gini_cps": 0.01599505125486056, "bpic": 10.709968186638388, "seed": 2, "mean_boundary_entropy": 0.3650463819503784, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 346.4} |
|
|
