| {"event": "start", "run": "pilot_A_s2", "condition": "A", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "global", "p3_objective": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 2.3} |
| {"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 2.3} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "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": 4.3} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.64089, "bpb": 8.13808, "chunk_loss": 0.065103, "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": 0.006, "alpha": 0.03, "t": 4.7} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10919.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.627421, "cps": 85.304688, "mean_p": 0.513204, "boundary_entropy": 0.686301, "bpb": 8.100851}, "de": {"chunks": 16285.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.793114, "cps": 127.226562, "mean_p": 0.524587, "boundary_entropy": 0.686399, "bpb": 8.104452}, "vi": {"chunks": 5099.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276593, "cps": 39.835938, "mean_p": 0.481628, "boundary_entropy": 0.687278, "bpb": 7.854846}, "tr": {"chunks": 8203.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.453129, "cps": 64.085938, "mean_p": 0.498086, "boundary_entropy": 0.686828, "bpb": 8.081317}, "fi": {"chunks": 13521.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.718782, "cps": 105.632812, "mean_p": 0.517085, "boundary_entropy": 0.686669, "bpb": 8.042953}, "ru": {"chunks": 33468.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.766121, "cps": 261.46875, "mean_p": 0.530332, "boundary_entropy": 0.688547, "bpb": 8.195778}, "el": {"chunks": 20672.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.00726, "cps": 161.5, "mean_p": 0.503626, "boundary_entropy": 0.690112, "bpb": 8.27967}, "hi": {"chunks": 885.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.0526, "cps": 6.914062, "mean_p": 0.462025, "boundary_entropy": 0.687437, "bpb": 8.01629}, "ta": {"chunks": 1278.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.068361, "cps": 9.984375, "mean_p": 0.477596, "boundary_entropy": 0.690061, "bpb": 8.085706}, "th": {"chunks": 3633.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.225863, "cps": 28.382812, "mean_p": 0.484856, "boundary_entropy": 0.690398, "bpb": 7.718654}, "zh": {"chunks": 11709.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.98256, "cps": 91.476562, "mean_p": 0.520269, "boundary_entropy": 0.685623, "bpb": 8.126806}, "my": {"chunks": 1616.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085286, "cps": 12.625, "mean_p": 0.474596, "boundary_entropy": 0.689681, "bpb": 7.800398}}, "bpb": {"en": 8.10085, "de": 8.10445, "vi": 7.85485, "tr": 8.08132, "fi": 8.04295, "ru": 8.19578, "el": 8.27967, "hi": 8.01629, "ta": 8.08571, "th": 7.71865, "zh": 8.12681, "my": 7.8004}, "gini_cpc": 0.489221, "gini_cps": 0.463272, "bpic": 3.01506, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.264085966489171, "vi": 0.4408421626190222, "tr": 0.7222098616257773, "fi": 1.1456136604009743, "ru": 2.814892411161005, "el": 1.6053986941007266, "hi": 0.08383578823776922, "ta": 0.10895486791171372, "th": 0.3599859797850751, "zh": 3.1598583721829123, "my": 0.13593122620185336}, "t": 6.0} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10904.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.626559, "cps": 85.1875, "mean_p": 0.513135, "boundary_entropy": 0.686312, "bpb": 8.052666}, "de": {"chunks": 16256.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.791701, "cps": 127.0, "mean_p": 0.524504, "boundary_entropy": 0.686413, "bpb": 8.060203}, "vi": {"chunks": 5092.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276214, "cps": 39.78125, "mean_p": 0.481575, "boundary_entropy": 0.687273, "bpb": 7.831458}, "tr": {"chunks": 8250.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.455726, "cps": 64.453125, "mean_p": 0.498254, "boundary_entropy": 0.686833, "bpb": 8.048639}, "fi": {"chunks": 13498.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.717559, "cps": 105.453125, "mean_p": 0.516944, "boundary_entropy": 0.686683, "bpb": 7.998151}, "ru": {"chunks": 33462.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.765805, "cps": 261.421875, "mean_p": 0.530287, "boundary_entropy": 0.688551, "bpb": 8.105147}, "el": {"chunks": 20636.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.005506, "cps": 161.21875, "mean_p": 0.503585, "boundary_entropy": 0.690113, "bpb": 8.273385}, "hi": {"chunks": 907.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.053908, "cps": 7.085938, "mean_p": 0.462267, "boundary_entropy": 0.687476, "bpb": 7.990529}, "ta": {"chunks": 1318.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.0705, "cps": 10.296875, "mean_p": 0.477815, "boundary_entropy": 0.690082, "bpb": 8.070294}, "th": {"chunks": 3695.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.229717, "cps": 28.867188, "mean_p": 0.484975, "boundary_entropy": 0.690405, "bpb": 7.703168}, "zh": {"chunks": 11685.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.978496, "cps": 91.289062, "mean_p": 0.52018, "boundary_entropy": 0.685628, "bpb": 8.118748}, "my": {"chunks": 1616.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085286, "cps": 12.625, "mean_p": 0.474587, "boundary_entropy": 0.68968, "bpb": 7.800255}}, "bpb": {"en": 8.05267, "de": 8.0602, "vi": 7.83146, "tr": 8.04864, "fi": 7.99815, "ru": 8.10515, "el": 8.27338, "hi": 7.99053, "ta": 8.07029, "th": 7.70317, "zh": 8.11875, "my": 7.80026}, "gini_cpc": 0.487926, "gini_cps": 0.462025, "bpic": 3.014326, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2635707407137569, "vi": 0.4408425749211851, "tr": 0.7273470378537032, "fi": 1.1452381790005715, "ru": 2.8182593592049203, "el": 1.6048075224658023, "hi": 0.08603803667070388, "ta": 0.11251960998042289, "th": 0.3666330864257221, "zh": 3.157719517612565, "my": 0.13611821890114056}, "t": 7.6} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10859.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.623973, "cps": 84.835938, "mean_p": 0.512768, "boundary_entropy": 0.686354, "bpb": 7.898079}, "de": {"chunks": 16233.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.790581, "cps": 126.820312, "mean_p": 0.524113, "boundary_entropy": 0.686473, "bpb": 7.921928}, "vi": {"chunks": 5067.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.274858, "cps": 39.585938, "mean_p": 0.481479, "boundary_entropy": 0.687264, "bpb": 7.752715}, "tr": {"chunks": 8387.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.463293, "cps": 65.523438, "mean_p": 0.498732, "boundary_entropy": 0.686848, "bpb": 7.941211}, "fi": {"chunks": 13398.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.712243, "cps": 104.671875, "mean_p": 0.516363, "boundary_entropy": 0.686734, "bpb": 7.856488}, "ru": {"chunks": 33404.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.762744, "cps": 260.96875, "mean_p": 0.530055, "boundary_entropy": 0.68857, "bpb": 7.817018}, "el": {"chunks": 20292.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.988744, "cps": 158.53125, "mean_p": 0.503181, "boundary_entropy": 0.690117, "bpb": 8.218084}, "hi": {"chunks": 995.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.059138, "cps": 7.773438, "mean_p": 0.463112, "boundary_entropy": 0.687609, "bpb": 7.903695}, "ta": {"chunks": 1471.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.078684, "cps": 11.492188, "mean_p": 0.478673, "boundary_entropy": 0.69016, "bpb": 7.989197}, "th": {"chunks": 3952.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.245695, "cps": 30.875, "mean_p": 0.485315, "boundary_entropy": 0.690423, "bpb": 7.620491}, "zh": {"chunks": 11609.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.965628, "cps": 90.695312, "mean_p": 0.519763, "boundary_entropy": 0.685653, "bpb": 8.08799}, "my": {"chunks": 1603.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.0846, "cps": 12.523438, "mean_p": 0.474497, "boundary_entropy": 0.689669, "bpb": 7.79375}}, "bpb": {"en": 7.89808, "de": 7.92193, "vi": 7.75272, "tr": 7.94121, "fi": 7.85649, "ru": 7.81702, "el": 8.21808, "hi": 7.9037, "ta": 7.9892, "th": 7.62049, "zh": 8.08799, "my": 7.79375}, "gini_cpc": 0.482952, "gini_cps": 0.456789, "bpic": 3.015487, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.267011826149086, "vi": 0.4404960814751646, "tr": 0.7424896064322807, "fi": 1.1414644153249383, "ru": 2.825033146149456, "el": 1.5845950522934753, "hi": 0.09477685534708236, "ta": 0.12610185937318397, "th": 0.39375869565298294, "zh": 3.1501820724705474, "my": 0.13558274843274326}, "t": 9.2} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10131.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.582141, "cps": 79.148438, "mean_p": 0.509409, "boundary_entropy": 0.686568, "bpb": 7.422677}, "de": {"chunks": 15619.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.760678, "cps": 122.023438, "mean_p": 0.520562, "boundary_entropy": 0.686863, "bpb": 7.489029}, "vi": {"chunks": 4868.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.264063, "cps": 38.03125, "mean_p": 0.480766, "boundary_entropy": 0.687199, "bpb": 7.511628}, "tr": {"chunks": 8659.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.478319, "cps": 67.648438, "mean_p": 0.499751, "boundary_entropy": 0.686899, "bpb": 7.589619}, "fi": {"chunks": 12503.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.664664, "cps": 97.679688, "mean_p": 0.512527, "boundary_entropy": 0.686991, "bpb": 7.426325}, "ru": {"chunks": 33185.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.751187, "cps": 259.257812, "mean_p": 0.529041, "boundary_entropy": 0.688661, "bpb": 7.199554}, "el": {"chunks": 18820.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.91702, "cps": 147.03125, "mean_p": 0.501353, "boundary_entropy": 0.69013, "bpb": 7.947578}, "hi": {"chunks": 1277.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.075899, "cps": 9.976562, "mean_p": 0.465339, "boundary_entropy": 0.687947, "bpb": 7.608555}, "ta": {"chunks": 2103.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.11249, "cps": 16.429688, "mean_p": 0.480873, "boundary_entropy": 0.690348, "bpb": 7.734847}, "th": {"chunks": 4626.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.287597, "cps": 36.140625, "mean_p": 0.486161, "boundary_entropy": 0.690465, "bpb": 7.339957}, "zh": {"chunks": 11280.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.909922, "cps": 88.125, "mean_p": 0.518091, "boundary_entropy": 0.685748, "bpb": 7.997294}, "my": {"chunks": 1470.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.077581, "cps": 11.484375, "mean_p": 0.474004, "boundary_entropy": 0.689621, "bpb": 7.663406}}, "bpb": {"en": 7.42268, "de": 7.48903, "vi": 7.51163, "tr": 7.58962, "fi": 7.42632, "ru": 7.19955, "el": 7.94758, "hi": 7.60855, "ta": 7.73485, "th": 7.33996, "zh": 7.99729, "my": 7.66341}, "gini_cpc": 0.471422, "gini_cps": 0.442855, "bpic": 3.081564, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.306690165773288, "vi": 0.45360646161805973, "tr": 0.8216540363714487, "fi": 1.1417582347868305, "ru": 3.0081841074535722, "el": 1.575253955380106, "hi": 0.13037899504602632, "ta": 0.19323491837036455, "th": 0.4940334646899628, "zh": 3.2808582108747237, "my": 0.13326796420970155}, "t": 10.9} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4823.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.277136, "cps": 37.679688, "mean_p": 0.488718, "boundary_entropy": 0.686417, "bpb": 6.857524}, "de": {"chunks": 8522.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.415039, "cps": 66.578125, "mean_p": 0.497538, "boundary_entropy": 0.687717, "bpb": 6.930973}, "vi": {"chunks": 3520.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.190941, "cps": 27.5, "mean_p": 0.47537, "boundary_entropy": 0.686685, "bpb": 7.167606}, "tr": {"chunks": 7547.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.416892, "cps": 58.960938, "mean_p": 0.495271, "boundary_entropy": 0.686954, "bpb": 7.150823}, "fi": {"chunks": 6363.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.33826, "cps": 49.710938, "mean_p": 0.492194, "boundary_entropy": 0.687148, "bpb": 6.917036}, "ru": {"chunks": 25784.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.360633, "cps": 201.4375, "mean_p": 0.511532, "boundary_entropy": 0.689919, "bpb": 6.368381}, "el": {"chunks": 14731.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.71778, "cps": 115.085938, "mean_p": 0.496079, "boundary_entropy": 0.690089, "bpb": 7.323162}, "hi": {"chunks": 3463.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.205825, "cps": 27.054688, "mean_p": 0.475508, "boundary_entropy": 0.689261, "bpb": 6.57384}, "ta": {"chunks": 6919.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.370099, "cps": 54.054688, "mean_p": 0.49017, "boundary_entropy": 0.690934, "bpb": 6.749592}, "th": {"chunks": 6882.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.427852, "cps": 53.765625, "mean_p": 0.489658, "boundary_entropy": 0.690643, "bpb": 6.489365}, "zh": {"chunks": 9449.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.599898, "cps": 73.820312, "mean_p": 0.509533, "boundary_entropy": 0.686084, "bpb": 7.747757}, "my": {"chunks": 1201.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.063384, "cps": 9.382812, "mean_p": 0.473236, "boundary_entropy": 0.689568, "bpb": 7.216407}}, "bpb": {"en": 6.85752, "de": 6.93097, "vi": 7.16761, "tr": 7.15082, "fi": 6.91704, "ru": 6.36838, "el": 7.32316, "hi": 6.57384, "ta": 6.74959, "th": 6.48937, "zh": 7.74776, "my": 7.21641}, "gini_cpc": 0.418341, "gini_cps": 0.365755, "bpic": 3.868604, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.497600515816016, "vi": 0.6889796265419951, "tr": 1.5042868423252576, "fi": 1.220553718260299, "ru": 4.909620644701045, "el": 2.589990926571422, "hi": 0.7426843576116053, "ta": 1.3354410424285832, "th": 1.543833502700606, "zh": 5.772969521325845, "my": 0.22871070342302433}, "t": 12.8} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5787.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.332529, "cps": 45.210938, "mean_p": 0.496387, "boundary_entropy": 0.687223, "bpb": 6.378327}, "de": {"chunks": 7938.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.386597, "cps": 62.015625, "mean_p": 0.498178, "boundary_entropy": 0.688173, "bpb": 6.459468}, "vi": {"chunks": 11200.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.60754, "cps": 87.5, "mean_p": 0.499859, "boundary_entropy": 0.688482, "bpb": 6.771567}, "tr": {"chunks": 11104.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.613379, "cps": 86.75, "mean_p": 0.506409, "boundary_entropy": 0.687556, "bpb": 6.73411}, "fi": {"chunks": 5821.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.309447, "cps": 45.476562, "mean_p": 0.494609, "boundary_entropy": 0.687786, "bpb": 6.467686}, "ru": {"chunks": 223.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.011768, "cps": 1.742188, "mean_p": 0.462216, "boundary_entropy": 0.687092, "bpb": 6.191952}, "el": {"chunks": 12983.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.632607, "cps": 101.429688, "mean_p": 0.495631, "boundary_entropy": 0.690208, "bpb": 6.248331}, "hi": {"chunks": 14946.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.888321, "cps": 116.765625, "mean_p": 0.498552, "boundary_entropy": 0.690708, "bpb": 5.729842}, "ta": {"chunks": 16864.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.902059, "cps": 131.75, "mean_p": 0.49779, "boundary_entropy": 0.691142, "bpb": 5.925055}, "th": {"chunks": 3375.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.209823, "cps": 26.367188, "mean_p": 0.483385, "boundary_entropy": 0.690306, "bpb": 5.672496}, "zh": {"chunks": 3539.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.599221, "cps": 27.648438, "mean_p": 0.4833, "boundary_entropy": 0.685419, "bpb": 7.348311}, "my": {"chunks": 1936.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.102174, "cps": 15.125, "mean_p": 0.482371, "boundary_entropy": 0.690577, "bpb": 5.911753}}, "bpb": {"en": 6.37833, "de": 6.45947, "vi": 6.77157, "tr": 6.73411, "fi": 6.46769, "ru": 6.19195, "el": 6.24833, "hi": 5.72984, "ta": 5.92506, "th": 5.6725, "zh": 7.34831, "my": 5.91175}, "gini_cpc": 0.333385, "gini_cps": 0.368947, "bpic": 4.00958, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.162597340744736, "vi": 1.827029326836349, "tr": 1.8445886817747885, "fi": 0.9305856552006413, "ru": 0.035388836683805436, "el": 1.902413196113948, "hi": 2.6714099722927678, "ta": 2.7127249505062783, "th": 0.6309912686829445, "zh": 1.802012328273415, "my": 0.3072646613640037}, "t": 15.2} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.26407, "bpb": 6.15176, "chunk_loss": 0.030022, "F": 0.16455, "G": 0.48523, "chunks": 168.5, "bytes_per_chunk": 6.077, "boundary_entropy": 0.69148, "router_noise": 0.0, "grad_norm": 1.5508, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.238, "alpha": 0.03, "t": 16.6} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2947.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.169339, "cps": 23.023438, "mean_p": 0.487057, "boundary_entropy": 0.686882, "bpb": 5.799501}, "de": {"chunks": 4608.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.224419, "cps": 36.0, "mean_p": 0.489894, "boundary_entropy": 0.687953, "bpb": 5.926009}, "vi": {"chunks": 3769.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.204448, "cps": 29.445312, "mean_p": 0.481252, "boundary_entropy": 0.687712, "bpb": 6.240996}, "tr": {"chunks": 2973.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164227, "cps": 23.226562, "mean_p": 0.482687, "boundary_entropy": 0.68698, "bpb": 6.226608}, "fi": {"chunks": 1360.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.072298, "cps": 10.625, "mean_p": 0.478464, "boundary_entropy": 0.686836, "bpb": 5.915859}, "ru": {"chunks": 8037.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.424116, "cps": 62.789062, "mean_p": 0.494418, "boundary_entropy": 0.690257, "bpb": 5.485718}, "el": {"chunks": 7357.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.358476, "cps": 57.476562, "mean_p": 0.491769, "boundary_entropy": 0.690323, "bpb": 5.658346}, "hi": {"chunks": 12390.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.736404, "cps": 96.796875, "mean_p": 0.496862, "boundary_entropy": 0.690744, "bpb": 5.198835}, "ta": {"chunks": 9655.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.516448, "cps": 75.429688, "mean_p": 0.49226, "boundary_entropy": 0.691026, "bpb": 5.251772}, "th": {"chunks": 7065.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.439229, "cps": 55.195312, "mean_p": 0.483894, "boundary_entropy": 0.69019, "bpb": 5.044049}, "zh": {"chunks": 4176.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.707078, "cps": 32.625, "mean_p": 0.491884, "boundary_entropy": 0.6867, "bpb": 6.955484}, "my": {"chunks": 7411.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.391123, "cps": 57.898438, "mean_p": 0.49359, "boundary_entropy": 0.691259, "bpb": 5.095473}}, "bpb": {"en": 5.7995, "de": 5.92601, "vi": 6.241, "tr": 6.22661, "fi": 5.91586, "ru": 5.48572, "el": 5.65835, "hi": 5.19883, "ta": 5.25177, "th": 5.04405, "zh": 6.95548, "my": 5.09547}, "gini_cpc": 0.310566, "gini_cps": 0.287302, "bpic": 5.349013, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.325269025334027, "vi": 1.2073327456063356, "tr": 0.9698137609093956, "fi": 0.4269440930433368, "ru": 2.5045444184103864, "el": 2.1169173152929748, "hi": 4.348707704362005, "ta": 3.0497960174185414, "th": 2.5937916348074084, "zh": 4.175524456158588, "my": 2.3097098239469385}, "t": 18.3} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.40418, "bpb": 4.9112, "chunk_loss": 0.031925, "F": 0.19485, "G": 0.4746, "chunks": 199.53, "bytes_per_chunk": 5.132, "boundary_entropy": 0.68913, "router_noise": 0.0, "grad_norm": 1.5999, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.073, "alpha": 0.03, "t": 20.2} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3079.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.176924, "cps": 24.054688, "mean_p": 0.427558, "boundary_entropy": 0.664242, "bpb": 4.46712}, "de": {"chunks": 3600.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.175328, "cps": 28.125, "mean_p": 0.434013, "boundary_entropy": 0.669144, "bpb": 4.656584}, "vi": {"chunks": 6136.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.332845, "cps": 47.9375, "mean_p": 0.464709, "boundary_entropy": 0.678001, "bpb": 5.03806}, "tr": {"chunks": 3722.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205601, "cps": 29.078125, "mean_p": 0.446433, "boundary_entropy": 0.67401, "bpb": 5.008576}, "fi": {"chunks": 2366.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.125777, "cps": 18.484375, "mean_p": 0.434891, "boundary_entropy": 0.671547, "bpb": 4.716824}, "ru": {"chunks": 6742.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.355778, "cps": 52.671875, "mean_p": 0.483543, "boundary_entropy": 0.68761, "bpb": 3.765363}, "el": {"chunks": 2614.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.127369, "cps": 20.421875, "mean_p": 0.466482, "boundary_entropy": 0.687252, "bpb": 4.506696}, "hi": {"chunks": 9225.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.548291, "cps": 72.070312, "mean_p": 0.489657, "boundary_entropy": 0.689848, "bpb": 3.624262}, "ta": {"chunks": 12004.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.642097, "cps": 93.78125, "mean_p": 0.484194, "boundary_entropy": 0.69013, "bpb": 3.699512}, "th": {"chunks": 7760.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.482437, "cps": 60.625, "mean_p": 0.471391, "boundary_entropy": 0.687747, "bpb": 3.421206}, "zh": {"chunks": 3292.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.557399, "cps": 25.71875, "mean_p": 0.486609, "boundary_entropy": 0.686222, "bpb": 6.400177}, "my": {"chunks": 2985.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.157536, "cps": 23.320312, "mean_p": 0.486395, "boundary_entropy": 0.690836, "bpb": 4.286373}}, "bpb": {"en": 4.46712, "de": 4.65658, "vi": 5.03806, "tr": 5.00858, "fi": 4.71682, "ru": 3.76536, "el": 4.5067, "hi": 3.62426, "ta": 3.69951, "th": 3.42121, "zh": 6.40018, "my": 4.28637}, "gini_cpc": 0.310501, "gini_cps": 0.296354, "bpic": 6.041417, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9909791677643354, "vi": 1.8812938695184658, "tr": 1.16209129682076, "fi": 0.7109143622457054, "ru": 2.010916164570445, "el": 0.7199115965821156, "hi": 3.0990296631497367, "ta": 3.6292338135786157, "th": 2.7268113140819876, "zh": 3.1505096572512503, "my": 0.89042099322033}, "t": 22.5} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.69705, "bpb": 3.89102, "chunk_loss": 0.028962, "F": 0.1506, "G": 0.435, "chunks": 154.22, "bytes_per_chunk": 6.64, "boundary_entropy": 0.6687, "router_noise": 0.0, "grad_norm": 1.7681, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.068, "alpha": 0.03, "t": 23.6} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3500.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201115, "cps": 27.34375, "mean_p": 0.357699, "boundary_entropy": 0.596906, "bpb": 3.854473}, "de": {"chunks": 4752.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.231432, "cps": 37.125, "mean_p": 0.376985, "boundary_entropy": 0.615355, "bpb": 4.096763}, "vi": {"chunks": 4639.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.251641, "cps": 36.242188, "mean_p": 0.377982, "boundary_entropy": 0.621521, "bpb": 3.609655}, "tr": {"chunks": 4257.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.235154, "cps": 33.257812, "mean_p": 0.39396, "boundary_entropy": 0.628252, "bpb": 4.218777}, "fi": {"chunks": 3958.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.210409, "cps": 30.921875, "mean_p": 0.384058, "boundary_entropy": 0.623201, "bpb": 4.089803}, "ru": {"chunks": 4352.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.229657, "cps": 34.0, "mean_p": 0.479159, "boundary_entropy": 0.681656, "bpb": 2.848696}, "el": {"chunks": 8104.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.394874, "cps": 63.3125, "mean_p": 0.481167, "boundary_entropy": 0.688499, "bpb": 3.036885}, "hi": {"chunks": 8523.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.506568, "cps": 66.585938, "mean_p": 0.455743, "boundary_entropy": 0.671837, "bpb": 2.450633}, "ta": {"chunks": 9775.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.522867, "cps": 76.367188, "mean_p": 0.462403, "boundary_entropy": 0.683229, "bpb": 2.310735}, "th": {"chunks": 15022.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.933914, "cps": 117.359375, "mean_p": 0.462313, "boundary_entropy": 0.683636, "bpb": 2.357541}, "zh": {"chunks": 3245.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.549441, "cps": 25.351562, "mean_p": 0.46711, "boundary_entropy": 0.681292, "bpb": 5.357706}, "my": {"chunks": 10258.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.541376, "cps": 80.140625, "mean_p": 0.453281, "boundary_entropy": 0.683664, "bpb": 2.516293}}, "bpb": {"en": 3.85447, "de": 4.09676, "vi": 3.60965, "tr": 4.21878, "fi": 4.0898, "ru": 2.8487, "el": 3.03689, "hi": 2.45063, "ta": 2.31074, "th": 2.35754, "zh": 5.35771, "my": 2.51629}, "gini_cpc": 0.272242, "gini_cps": 0.273462, "bpic": 4.774286, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1507476605603524, "vi": 1.2512304544926187, "tr": 1.169254835425857, "fi": 1.046212687105569, "ru": 1.1419201809272521, "el": 1.9634265667091277, "hi": 2.5187988792188496, "ta": 2.5998444962365794, "th": 4.643685172520983, "zh": 2.7319788592714436, "my": 2.6918781326336743}, "t": 26.7} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.85689, "de": 4.09754, "vi": 3.61589, "tr": 4.22806, "fi": 4.09179, "ru": 2.83357, "el": 3.02939, "hi": 2.44595, "ta": 2.32786, "th": 2.39028, "zh": 5.36341, "my": 2.53202}, "bytes_per_chunk": {"en": 4.883, "de": 4.534, "vi": 5.162, "tr": 4.824, "fi": 5.01, "ru": 7.597, "el": 4.473, "hi": 5.16, "ta": 5.121, "th": 3.156, "zh": 5.853, "my": 4.855}, "bpb_mean": 3.40105, "t": 28.1} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.51503, "bpb": 3.62842, "chunk_loss": 0.031162, "F": 0.19229, "G": 0.41294, "chunks": 196.91, "bytes_per_chunk": 5.2, "boundary_entropy": 0.64191, "router_noise": 0.0, "grad_norm": 0.8813, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 28.1} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.46931, "bpb": 3.56247, "chunk_loss": 0.031137, "F": 0.18893, "G": 0.40568, "chunks": 193.47, "bytes_per_chunk": 5.293, "boundary_entropy": 0.64476, "router_noise": 0.0, "grad_norm": 1.6171, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 30.4} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2616.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.150319, "cps": 20.4375, "mean_p": 0.345926, "boundary_entropy": 0.592793, "bpb": 3.628446}, "de": {"chunks": 4114.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.20036, "cps": 32.140625, "mean_p": 0.355676, "boundary_entropy": 0.6006, "bpb": 3.846412}, "vi": {"chunks": 3057.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.165826, "cps": 23.882812, "mean_p": 0.381109, "boundary_entropy": 0.628708, "bpb": 3.197745}, "tr": {"chunks": 3152.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174115, "cps": 24.625, "mean_p": 0.360757, "boundary_entropy": 0.608053, "bpb": 3.846097}, "fi": {"chunks": 3049.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162086, "cps": 23.820312, "mean_p": 0.34158, "boundary_entropy": 0.593541, "bpb": 3.875066}, "ru": {"chunks": 4724.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.249288, "cps": 36.90625, "mean_p": 0.460451, "boundary_entropy": 0.678107, "bpb": 2.649669}, "el": {"chunks": 6029.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.293768, "cps": 47.101562, "mean_p": 0.459582, "boundary_entropy": 0.682127, "bpb": 2.76021}, "hi": {"chunks": 8924.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.530401, "cps": 69.71875, "mean_p": 0.369687, "boundary_entropy": 0.598481, "bpb": 2.112584}, "ta": {"chunks": 13043.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.697673, "cps": 101.898438, "mean_p": 0.378447, "boundary_entropy": 0.60746, "bpb": 1.944292}, "th": {"chunks": 4677.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.290768, "cps": 36.539062, "mean_p": 0.4126, "boundary_entropy": 0.663828, "bpb": 2.069386}, "zh": {"chunks": 4117.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.697088, "cps": 32.164062, "mean_p": 0.416703, "boundary_entropy": 0.64609, "bpb": 4.982389}, "my": {"chunks": 11172.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.589614, "cps": 87.28125, "mean_p": 0.420736, "boundary_entropy": 0.668034, "bpb": 2.221303}}, "bpb": {"en": 3.62845, "de": 3.84641, "vi": 3.19774, "tr": 3.8461, "fi": 3.87507, "ru": 2.64967, "el": 2.76021, "hi": 2.11258, "ta": 1.94429, "th": 2.06939, "zh": 4.98239, "my": 2.2213}, "gini_cpc": 0.315489, "gini_cps": 0.298085, "bpic": 5.588447, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.3329021262259857, "vi": 1.1031604347533985, "tr": 1.158302617995801, "fi": 1.07828092239608, "ru": 1.6583914695843722, "el": 1.9542981425628085, "hi": 3.5285060730400644, "ta": 4.641286796008353, "th": 1.9343394324303516, "zh": 4.637391961836355, "my": 3.9224185264160925}, "t": 31.6} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.31952, "bpb": 3.34636, "chunk_loss": 0.031111, "F": 0.19196, "G": 0.37392, "chunks": 196.56, "bytes_per_chunk": 5.21, "boundary_entropy": 0.61096, "router_noise": 0.0, "grad_norm": 2.0235, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.064, "alpha": 0.03, "t": 33.6} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.08517, "bpb": 3.00826, "chunk_loss": 0.031353, "F": 0.19373, "G": 0.36897, "chunks": 198.38, "bytes_per_chunk": 5.162, "boundary_entropy": 0.58961, "router_noise": 0.0, "grad_norm": 0.7716, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 35.7} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3102.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.178245, "cps": 24.234375, "mean_p": 0.33507, "boundary_entropy": 0.555608, "bpb": 3.220415}, "de": {"chunks": 3524.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.171626, "cps": 27.53125, "mean_p": 0.34689, "boundary_entropy": 0.575484, "bpb": 3.535287}, "vi": {"chunks": 4564.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.247573, "cps": 35.65625, "mean_p": 0.393707, "boundary_entropy": 0.62242, "bpb": 2.805246}, "tr": {"chunks": 3396.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187593, "cps": 26.53125, "mean_p": 0.336927, "boundary_entropy": 0.550492, "bpb": 3.444982}, "fi": {"chunks": 3097.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.164638, "cps": 24.195312, "mean_p": 0.327985, "boundary_entropy": 0.54409, "bpb": 3.590796}, "ru": {"chunks": 4009.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.211557, "cps": 31.320312, "mean_p": 0.301439, "boundary_entropy": 0.480442, "bpb": 2.533724}, "el": {"chunks": 6714.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.327145, "cps": 52.453125, "mean_p": 0.294952, "boundary_entropy": 0.50483, "bpb": 2.551994}, "hi": {"chunks": 7481.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.444636, "cps": 58.445312, "mean_p": 0.346678, "boundary_entropy": 0.565712, "bpb": 1.791175}, "ta": {"chunks": 10855.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.580637, "cps": 84.804688, "mean_p": 0.2867, "boundary_entropy": 0.446182, "bpb": 1.581237}, "th": {"chunks": 8234.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.511906, "cps": 64.328125, "mean_p": 0.291417, "boundary_entropy": 0.517772, "bpb": 1.891468}, "zh": {"chunks": 2932.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.496444, "cps": 22.90625, "mean_p": 0.338315, "boundary_entropy": 0.56796, "bpb": 4.130733}, "my": {"chunks": 2725.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.143815, "cps": 21.289062, "mean_p": 0.377552, "boundary_entropy": 0.649297, "bpb": 2.011843}}, "bpb": {"en": 3.22041, "de": 3.53529, "vi": 2.80525, "tr": 3.44498, "fi": 3.5908, "ru": 2.53372, "el": 2.55199, "hi": 1.79117, "ta": 1.58124, "th": 1.89147, "zh": 4.13073, "my": 2.01184}, "gini_cpc": 0.272582, "gini_cps": 0.260976, "bpic": 6.329573, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9628659286117804, "vi": 1.388944270186588, "tr": 1.0524451155332837, "fi": 0.9236589624552982, "ru": 1.18688644146512, "el": 1.8353665760291018, "hi": 2.4945195306100745, "ta": 3.2575169562502615, "th": 2.871918585320129, "zh": 2.785177320248038, "my": 0.806836352284926}, "t": 38.6} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.22638, "de": 3.5335, "vi": 2.81122, "tr": 3.45508, "fi": 3.58252, "ru": 2.5219, "el": 2.54575, "hi": 1.78685, "ta": 1.5988, "th": 1.92096, "zh": 4.13831, "my": 2.02737}, "bytes_per_chunk": {"en": 5.721, "de": 6.025, "vi": 5.267, "tr": 5.49, "fi": 6.483, "ru": 9.339, "el": 5.662, "hi": 4.894, "ta": 4.67, "th": 5.181, "zh": 6.005, "my": 16.004}, "bpb_mean": 2.76239, "t": 39.9} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.23107, "bpb": 3.21875, "chunk_loss": 0.029938, "F": 0.16101, "G": 0.32955, "chunks": 164.88, "bytes_per_chunk": 6.211, "boundary_entropy": 0.54946, "router_noise": 0.0, "grad_norm": 1.022, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 39.9} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.16919, "bpb": 3.12948, "chunk_loss": 0.030795, "F": 0.18521, "G": 0.33301, "chunks": 189.66, "bytes_per_chunk": 5.399, "boundary_entropy": 0.54497, "router_noise": 0.0, "grad_norm": 1.0044, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 42.0} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.05378, "bpb": 2.96299, "chunk_loss": 0.030249, "F": 0.17285, "G": 0.31574, "chunks": 177.0, "bytes_per_chunk": 5.785, "boundary_entropy": 0.53017, "router_noise": 0.0, "grad_norm": 0.9456, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 44.0} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.90684, "bpb": 2.75099, "chunk_loss": 0.030566, "F": 0.18137, "G": 0.3248, "chunks": 185.72, "bytes_per_chunk": 5.514, "boundary_entropy": 0.5249, "router_noise": 0.0, "grad_norm": 0.8375, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 46.0} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3298.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189508, "cps": 25.765625, "mean_p": 0.28652, "boundary_entropy": 0.445989, "bpb": 2.87824}, "de": {"chunks": 3854.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187698, "cps": 30.109375, "mean_p": 0.289571, "boundary_entropy": 0.461443, "bpb": 3.201619}, "vi": {"chunks": 4344.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.235639, "cps": 33.9375, "mean_p": 0.347118, "boundary_entropy": 0.555913, "bpb": 2.556503}, "tr": {"chunks": 3797.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.209744, "cps": 29.664062, "mean_p": 0.294172, "boundary_entropy": 0.460756, "bpb": 3.15543}, "fi": {"chunks": 3599.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.191324, "cps": 28.117188, "mean_p": 0.289858, "boundary_entropy": 0.449181, "bpb": 3.358134}, "ru": {"chunks": 5179.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.273298, "cps": 40.460938, "mean_p": 0.324937, "boundary_entropy": 0.552886, "bpb": 2.142862}, "el": {"chunks": 6649.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.323978, "cps": 51.945312, "mean_p": 0.39597, "boundary_entropy": 0.639926, "bpb": 2.223917}, "hi": {"chunks": 6665.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.396137, "cps": 52.070312, "mean_p": 0.340767, "boundary_entropy": 0.569233, "bpb": 1.617225}, "ta": {"chunks": 6829.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.365285, "cps": 53.351562, "mean_p": 0.268764, "boundary_entropy": 0.488396, "bpb": 1.358612}, "th": {"chunks": 6712.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.417283, "cps": 52.4375, "mean_p": 0.349985, "boundary_entropy": 0.60188, "bpb": 1.721768}, "zh": {"chunks": 3197.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.541314, "cps": 24.976562, "mean_p": 0.386638, "boundary_entropy": 0.598228, "bpb": 3.843487}, "my": {"chunks": 13380.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.706143, "cps": 104.53125, "mean_p": 0.350673, "boundary_entropy": 0.558198, "bpb": 1.656772}}, "bpb": {"en": 2.87824, "de": 3.20162, "vi": 2.5565, "tr": 3.15543, "fi": 3.35813, "ru": 2.14286, "el": 2.22392, "hi": 1.61722, "ta": 1.35861, "th": 1.72177, "zh": 3.84349, "my": 1.65677}, "gini_cpc": 0.244022, "gini_cps": 0.233499, "bpic": 5.685392, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9904504919634612, "vi": 1.2434265706316079, "tr": 1.1067856370830254, "fi": 1.009586261654383, "ru": 1.442149107718598, "el": 1.7095781428441879, "hi": 2.090347790592164, "ta": 1.927547602338195, "th": 2.2019342738970296, "zh": 2.8564239283852113, "my": 3.7262003839815407}, "t": 48.9} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.88721, "de": 3.2004, "vi": 2.55959, "tr": 3.16565, "fi": 3.35118, "ru": 2.1265, "el": 2.21408, "hi": 1.61021, "ta": 1.37139, "th": 1.75056, "zh": 3.85418, "my": 1.66883}, "bytes_per_chunk": {"en": 5.654, "de": 5.666, "vi": 6.038, "tr": 5.194, "fi": 5.686, "ru": 7.25, "el": 5.542, "hi": 6.228, "ta": 7.723, "th": 7.578, "zh": 5.349, "my": 4.37}, "bpb_mean": 2.47998, "t": 50.2} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.89656, "bpb": 2.73615, "chunk_loss": 0.030199, "F": 0.16943, "G": 0.31588, "chunks": 173.5, "bytes_per_chunk": 5.902, "boundary_entropy": 0.51425, "router_noise": 0.0, "grad_norm": 0.8843, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 50.2} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.9351, "bpb": 2.79176, "chunk_loss": 0.030195, "F": 0.17419, "G": 0.31022, "chunks": 178.38, "bytes_per_chunk": 5.741, "boundary_entropy": 0.50084, "router_noise": 0.0, "grad_norm": 0.8314, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 52.3} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.77902, "bpb": 2.56658, "chunk_loss": 0.030285, "F": 0.17618, "G": 0.29603, "chunks": 180.41, "bytes_per_chunk": 5.676, "boundary_entropy": 0.4694, "router_noise": 0.0, "grad_norm": 0.9695, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 54.3} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.8572, "bpb": 2.67938, "chunk_loss": 0.030413, "F": 0.18222, "G": 0.30017, "chunks": 186.59, "bytes_per_chunk": 5.488, "boundary_entropy": 0.4613, "router_noise": 0.0, "grad_norm": 0.8148, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 56.3} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3202.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183991, "cps": 25.015625, "mean_p": 0.256109, "boundary_entropy": 0.366806, "bpb": 2.682189}, "de": {"chunks": 4062.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.197828, "cps": 31.734375, "mean_p": 0.274229, "boundary_entropy": 0.398062, "bpb": 3.034644}, "vi": {"chunks": 4105.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222674, "cps": 32.070312, "mean_p": 0.302479, "boundary_entropy": 0.499444, "bpb": 2.399351}, "tr": {"chunks": 3346.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.184831, "cps": 26.140625, "mean_p": 0.261934, "boundary_entropy": 0.399903, "bpb": 3.03554}, "fi": {"chunks": 3478.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184892, "cps": 27.171875, "mean_p": 0.252024, "boundary_entropy": 0.377886, "bpb": 3.23269}, "ru": {"chunks": 6950.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.366755, "cps": 54.296875, "mean_p": 0.344911, "boundary_entropy": 0.569369, "bpb": 1.988317}, "el": {"chunks": 7867.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.383326, "cps": 61.460938, "mean_p": 0.382214, "boundary_entropy": 0.625907, "bpb": 2.096759}, "hi": {"chunks": 7108.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.422467, "cps": 55.53125, "mean_p": 0.284789, "boundary_entropy": 0.497054, "bpb": 1.537165}, "ta": {"chunks": 8145.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.435678, "cps": 63.632812, "mean_p": 0.305973, "boundary_entropy": 0.501504, "bpb": 1.273775}, "th": {"chunks": 8539.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.530867, "cps": 66.710938, "mean_p": 0.424504, "boundary_entropy": 0.644879, "bpb": 1.581832}, "zh": {"chunks": 3496.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.59194, "cps": 27.3125, "mean_p": 0.361175, "boundary_entropy": 0.531619, "bpb": 3.709311}, "my": {"chunks": 10442.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.551087, "cps": 81.578125, "mean_p": 0.321402, "boundary_entropy": 0.532465, "bpb": 1.484709}}, "bpb": {"en": 2.68219, "de": 3.03464, "vi": 2.39935, "tr": 3.03554, "fi": 3.23269, "ru": 1.98832, "el": 2.09676, "hi": 1.53717, "ta": 1.27378, "th": 1.58183, "zh": 3.70931, "my": 1.48471}, "gini_cpc": 0.235155, "gini_cps": 0.227222, "bpic": 5.425233, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0752025965848975, "vi": 1.2102436485739945, "tr": 1.0045653121773206, "fi": 1.0048945407206722, "ru": 1.993326235746458, "el": 2.083392505835001, "hi": 2.296122946932313, "ta": 2.3679275573119836, "th": 2.8852851526390966, "zh": 3.2172201043497832, "my": 2.99518122974914}, "t": 59.2} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.68994, "de": 3.03643, "vi": 2.40433, "tr": 3.04748, "fi": 3.23005, "ru": 1.97309, "el": 2.08599, "hi": 1.53283, "ta": 1.28667, "th": 1.60984, "zh": 3.71942, "my": 1.49493}, "bytes_per_chunk": {"en": 5.733, "de": 5.244, "vi": 6.014, "tr": 5.893, "fi": 5.64, "ru": 5.848, "el": 4.558, "hi": 5.905, "ta": 6.32, "th": 5.571, "zh": 4.828, "my": 5.435}, "bpb_mean": 2.34258, "t": 60.5} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.83295, "bpb": 2.64438, "chunk_loss": 0.030596, "F": 0.18292, "G": 0.29812, "chunks": 187.31, "bytes_per_chunk": 5.467, "boundary_entropy": 0.456, "router_noise": 0.0, "grad_norm": 0.8968, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 60.5} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.63047, "bpb": 2.35227, "chunk_loss": 0.029921, "F": 0.16379, "G": 0.29398, "chunks": 167.72, "bytes_per_chunk": 6.105, "boundary_entropy": 0.46112, "router_noise": 0.0, "grad_norm": 0.8138, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 62.5} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.7208, "bpb": 2.48259, "chunk_loss": 0.030363, "F": 0.18018, "G": 0.29443, "chunks": 184.5, "bytes_per_chunk": 5.55, "boundary_entropy": 0.44361, "router_noise": 0.0, "grad_norm": 0.9356, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 64.5} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.87094, "bpb": 2.6992, "chunk_loss": 0.030602, "F": 0.18835, "G": 0.28418, "chunks": 192.88, "bytes_per_chunk": 5.309, "boundary_entropy": 0.40961, "router_noise": 0.0, "grad_norm": 0.9271, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 66.5} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3314.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.190427, "cps": 25.890625, "mean_p": 0.246812, "boundary_entropy": 0.323909, "bpb": 2.549457}, "de": {"chunks": 4242.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.206594, "cps": 33.140625, "mean_p": 0.26836, "boundary_entropy": 0.353481, "bpb": 2.897853}, "vi": {"chunks": 4257.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230919, "cps": 33.257812, "mean_p": 0.298162, "boundary_entropy": 0.471663, "bpb": 2.290669}, "tr": {"chunks": 4276.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.236204, "cps": 33.40625, "mean_p": 0.274482, "boundary_entropy": 0.371257, "bpb": 2.908347}, "fi": {"chunks": 3726.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.198076, "cps": 29.109375, "mean_p": 0.243884, "boundary_entropy": 0.329126, "bpb": 3.150684}, "ru": {"chunks": 5628.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.296992, "cps": 43.96875, "mean_p": 0.330621, "boundary_entropy": 0.565113, "bpb": 1.902014}, "el": {"chunks": 7463.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.363641, "cps": 58.304688, "mean_p": 0.37472, "boundary_entropy": 0.618088, "bpb": 1.989964}, "hi": {"chunks": 7879.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.468291, "cps": 61.554688, "mean_p": 0.315273, "boundary_entropy": 0.527348, "bpb": 1.472325}, "ta": {"chunks": 8661.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.463279, "cps": 67.664062, "mean_p": 0.349476, "boundary_entropy": 0.553427, "bpb": 1.258394}, "th": {"chunks": 10096.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.627666, "cps": 78.875, "mean_p": 0.426075, "boundary_entropy": 0.638145, "bpb": 1.516522}, "zh": {"chunks": 3222.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545547, "cps": 25.171875, "mean_p": 0.298188, "boundary_entropy": 0.437943, "bpb": 3.614912}, "my": {"chunks": 10876.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.573992, "cps": 84.96875, "mean_p": 0.331511, "boundary_entropy": 0.547059, "bpb": 1.40963}}, "bpb": {"en": 2.54946, "de": 2.89785, "vi": 2.29067, "tr": 2.90835, "fi": 3.15068, "ru": 1.90201, "el": 1.98996, "hi": 1.47233, "ta": 1.25839, "th": 1.51652, "zh": 3.61491, "my": 1.40963}, "gini_cpc": 0.23666, "gini_cps": 0.236434, "bpic": 5.211583, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0849004095178518, "vi": 1.212640655099823, "tr": 1.2403914421036921, "fi": 1.0401658308851056, "ru": 1.559611721600056, "el": 1.9096079435848516, "hi": 2.4591648560163057, "ta": 2.432843571075436, "th": 3.2960963079175976, "zh": 2.8648620175139747, "my": 3.0142372944840683}, "t": 69.4} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.5592, "de": 2.89901, "vi": 2.29361, "tr": 2.91918, "fi": 3.14693, "ru": 1.8846, "el": 1.98012, "hi": 1.4686, "ta": 1.26692, "th": 1.54283, "zh": 3.62209, "my": 1.41759}, "bytes_per_chunk": {"en": 5.647, "de": 5.099, "vi": 6.111, "tr": 4.719, "fi": 5.291, "ru": 6.752, "el": 4.86, "hi": 5.179, "ta": 6.107, "th": 4.72, "zh": 5.095, "my": 5.44}, "bpb_mean": 2.25006, "t": 70.7} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.54416, "bpb": 2.22775, "chunk_loss": 0.030619, "F": 0.1828, "G": 0.30323, "chunks": 187.19, "bytes_per_chunk": 5.47, "boundary_entropy": 0.46317, "router_noise": 0.0, "grad_norm": 0.792, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 70.7} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.76558, "bpb": 2.5472, "chunk_loss": 0.030291, "F": 0.17789, "G": 0.27995, "chunks": 182.16, "bytes_per_chunk": 5.622, "boundary_entropy": 0.41569, "router_noise": 0.0, "grad_norm": 0.9236, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 72.7} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.59007, "bpb": 2.29399, "chunk_loss": 0.030241, "F": 0.17688, "G": 0.29422, "chunks": 181.12, "bytes_per_chunk": 5.654, "boundary_entropy": 0.43363, "router_noise": 0.0, "grad_norm": 0.8158, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 74.7} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.61332, "bpb": 2.32753, "chunk_loss": 0.030145, "F": 0.1745, "G": 0.28851, "chunks": 178.69, "bytes_per_chunk": 5.731, "boundary_entropy": 0.42135, "router_noise": 0.0, "grad_norm": 0.935, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 76.7} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3368.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19353, "cps": 26.3125, "mean_p": 0.251751, "boundary_entropy": 0.319364, "bpb": 2.460248}, "de": {"chunks": 4070.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.198218, "cps": 31.796875, "mean_p": 0.262224, "boundary_entropy": 0.341151, "bpb": 2.802987}, "vi": {"chunks": 4101.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222457, "cps": 32.039062, "mean_p": 0.287168, "boundary_entropy": 0.452091, "bpb": 2.242921}, "tr": {"chunks": 3623.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200133, "cps": 28.304688, "mean_p": 0.251393, "boundary_entropy": 0.350561, "bpb": 2.814326}, "fi": {"chunks": 3757.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.199724, "cps": 29.351562, "mean_p": 0.237565, "boundary_entropy": 0.31126, "bpb": 3.03168}, "ru": {"chunks": 7193.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.379578, "cps": 56.195312, "mean_p": 0.353344, "boundary_entropy": 0.569424, "bpb": 1.823759}, "el": {"chunks": 6947.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.338498, "cps": 54.273438, "mean_p": 0.405277, "boundary_entropy": 0.640912, "bpb": 1.938795}, "hi": {"chunks": 6136.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.364695, "cps": 47.9375, "mean_p": 0.289875, "boundary_entropy": 0.520683, "bpb": 1.441694}, "ta": {"chunks": 8008.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.42835, "cps": 62.5625, "mean_p": 0.424279, "boundary_entropy": 0.627503, "bpb": 1.187957}, "th": {"chunks": 6813.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.423562, "cps": 53.226562, "mean_p": 0.409615, "boundary_entropy": 0.642559, "bpb": 1.456839}, "zh": {"chunks": 3251.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.550457, "cps": 25.398438, "mean_p": 0.29095, "boundary_entropy": 0.416135, "bpb": 3.583155}, "my": {"chunks": 9207.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.485909, "cps": 71.929688, "mean_p": 0.337273, "boundary_entropy": 0.572911, "bpb": 1.372669}}, "bpb": {"en": 2.46025, "de": 2.80299, "vi": 2.24292, "tr": 2.81433, "fi": 3.03168, "ru": 1.82376, "el": 1.93879, "hi": 1.44169, "ta": 1.18796, "th": 1.45684, "zh": 3.58315, "my": 1.37267}, "gini_cpc": 0.203476, "gini_cps": 0.198674, "bpic": 5.7734, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0242218568730044, "vi": 1.1494727092107018, "tr": 1.0341173389851657, "fi": 1.0320039249031905, "ru": 1.9613399902230524, "el": 1.7490752366008915, "hi": 1.8844400052235355, "ta": 2.2133527388971226, "th": 2.188614947905168, "zh": 2.844301067073784, "my": 2.5107692693893577}, "t": 79.6} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.4672, "de": 2.80566, "vi": 2.24853, "tr": 2.82274, "fi": 3.03417, "ru": 1.80655, "el": 1.92799, "hi": 1.43927, "ta": 1.19773, "th": 1.48302, "zh": 3.58801, "my": 1.38045}, "bytes_per_chunk": {"en": 5.567, "de": 5.263, "vi": 6.224, "tr": 5.634, "fi": 5.237, "ru": 5.245, "el": 5.592, "hi": 6.704, "ta": 7.033, "th": 7.154, "zh": 5.116, "my": 5.769}, "bpb_mean": 2.18344, "t": 80.8} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.6975, "bpb": 2.44898, "chunk_loss": 0.030426, "F": 0.18185, "G": 0.28059, "chunks": 186.22, "bytes_per_chunk": 5.499, "boundary_entropy": 0.40858, "router_noise": 0.0, "grad_norm": 0.7832, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 80.9} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.60131, "bpb": 2.3102, "chunk_loss": 0.031022, "F": 0.1908, "G": 0.30762, "chunks": 195.38, "bytes_per_chunk": 5.241, "boundary_entropy": 0.44603, "router_noise": 0.0, "grad_norm": 0.9461, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 82.8} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.71307, "bpb": 2.47144, "chunk_loss": 0.029905, "F": 0.17023, "G": 0.28715, "chunks": 174.31, "bytes_per_chunk": 5.875, "boundary_entropy": 0.41187, "router_noise": 0.0, "grad_norm": 0.8779, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 84.8} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.50486, "bpb": 2.17106, "chunk_loss": 0.030121, "F": 0.16965, "G": 0.29323, "chunks": 173.72, "bytes_per_chunk": 5.895, "boundary_entropy": 0.4318, "router_noise": 0.0, "grad_norm": 0.8825, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 86.8} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3318.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.190657, "cps": 25.921875, "mean_p": 0.240049, "boundary_entropy": 0.285017, "bpb": 2.399191}, "de": {"chunks": 3994.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.194516, "cps": 31.203125, "mean_p": 0.242993, "boundary_entropy": 0.294922, "bpb": 2.719121}, "vi": {"chunks": 4087.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.221698, "cps": 31.929688, "mean_p": 0.271948, "boundary_entropy": 0.424871, "bpb": 2.186549}, "tr": {"chunks": 4357.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.240678, "cps": 34.039062, "mean_p": 0.270847, "boundary_entropy": 0.339161, "bpb": 2.726278}, "fi": {"chunks": 3469.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184413, "cps": 27.101562, "mean_p": 0.213705, "boundary_entropy": 0.270321, "bpb": 2.971017}, "ru": {"chunks": 6056.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319578, "cps": 47.3125, "mean_p": 0.343257, "boundary_entropy": 0.570114, "bpb": 1.791565}, "el": {"chunks": 8797.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.428641, "cps": 68.726562, "mean_p": 0.409765, "boundary_entropy": 0.629288, "bpb": 1.860386}, "hi": {"chunks": 6419.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.381516, "cps": 50.148438, "mean_p": 0.375467, "boundary_entropy": 0.6102, "bpb": 1.426791}, "ta": {"chunks": 11691.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.625354, "cps": 91.335938, "mean_p": 0.464714, "boundary_entropy": 0.64154, "bpb": 1.146417}, "th": {"chunks": 8976.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.558035, "cps": 70.125, "mean_p": 0.422359, "boundary_entropy": 0.637019, "bpb": 1.407335}, "zh": {"chunks": 3206.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.542838, "cps": 25.046875, "mean_p": 0.279009, "boundary_entropy": 0.374141, "bpb": 3.538012}, "my": {"chunks": 13068.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.689677, "cps": 102.09375, "mean_p": 0.376009, "boundary_entropy": 0.593709, "bpb": 1.344514}}, "bpb": {"en": 2.39919, "de": 2.71912, "vi": 2.18655, "tr": 2.72628, "fi": 2.97102, "ru": 1.79157, "el": 1.86039, "hi": 1.42679, "ta": 1.14642, "th": 1.40734, "zh": 3.53801, "my": 1.34451}, "gini_cpc": 0.259349, "gini_cps": 0.274519, "bpic": 4.955978, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0202424553814442, "vi": 1.1628112752346718, "tr": 1.2623644248268324, "fi": 0.9672531548477415, "ru": 1.676194420455467, "el": 2.2482338652553975, "hi": 2.001059679876615, "ta": 3.280000648073408, "th": 2.926911002298481, "zh": 2.8471989437902665, "my": 3.617374628802307}, "t": 89.7} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.40804, "de": 2.72475, "vi": 2.18929, "tr": 2.73567, "fi": 2.9743, "ru": 1.77356, "el": 1.8514, "hi": 1.42302, "ta": 1.15409, "th": 1.42978, "zh": 3.5396, "my": 1.35257}, "bytes_per_chunk": {"en": 5.613, "de": 5.424, "vi": 6.301, "tr": 4.653, "fi": 5.662, "ru": 6.143, "el": 4.464, "hi": 6.731, "ta": 5.17, "th": 5.227, "zh": 5.171, "my": 4.4}, "bpb_mean": 2.12967, "t": 91.0} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.69714, "bpb": 2.44846, "chunk_loss": 0.030737, "F": 0.18832, "G": 0.29502, "chunks": 192.84, "bytes_per_chunk": 5.31, "boundary_entropy": 0.41356, "router_noise": 0.0, "grad_norm": 0.7396, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 91.0} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.69951, "bpb": 2.45187, "chunk_loss": 0.02997, "F": 0.17093, "G": 0.28839, "chunks": 175.03, "bytes_per_chunk": 5.85, "boundary_entropy": 0.41799, "router_noise": 0.0, "grad_norm": 0.7971, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 93.0} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.67693, "bpb": 2.41929, "chunk_loss": 0.030207, "F": 0.17685, "G": 0.27481, "chunks": 181.09, "bytes_per_chunk": 5.655, "boundary_entropy": 0.38344, "router_noise": 0.0, "grad_norm": 0.9328, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 95.0} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.56917, "bpb": 2.26383, "chunk_loss": 0.030722, "F": 0.1893, "G": 0.28881, "chunks": 193.84, "bytes_per_chunk": 5.283, "boundary_entropy": 0.40719, "router_noise": 0.0, "grad_norm": 0.8834, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 97.0} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3225.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185313, "cps": 25.195312, "mean_p": 0.233899, "boundary_entropy": 0.278, "bpb": 2.352719}, "de": {"chunks": 3899.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189889, "cps": 30.460938, "mean_p": 0.248734, "boundary_entropy": 0.309417, "bpb": 2.65483}, "vi": {"chunks": 4196.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227611, "cps": 32.78125, "mean_p": 0.271422, "boundary_entropy": 0.420239, "bpb": 2.144364}, "tr": {"chunks": 3898.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215323, "cps": 30.453125, "mean_p": 0.25146, "boundary_entropy": 0.331035, "bpb": 2.669487}, "fi": {"chunks": 3318.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176386, "cps": 25.921875, "mean_p": 0.213175, "boundary_entropy": 0.27293, "bpb": 2.903827}, "ru": {"chunks": 6830.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.360422, "cps": 53.359375, "mean_p": 0.346057, "boundary_entropy": 0.54879, "bpb": 1.730027}, "el": {"chunks": 6557.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.319495, "cps": 51.226562, "mean_p": 0.377792, "boundary_entropy": 0.612953, "bpb": 1.830674}, "hi": {"chunks": 6976.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.414621, "cps": 54.5, "mean_p": 0.342451, "boundary_entropy": 0.573159, "bpb": 1.374051}, "ta": {"chunks": 7623.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.407756, "cps": 59.554688, "mean_p": 0.432795, "boundary_entropy": 0.628353, "bpb": 1.116788}, "th": {"chunks": 6406.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.398259, "cps": 50.046875, "mean_p": 0.390735, "boundary_entropy": 0.62903, "bpb": 1.364504}, "zh": {"chunks": 3177.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.537928, "cps": 24.820312, "mean_p": 0.273041, "boundary_entropy": 0.366178, "bpb": 3.481589}, "my": {"chunks": 7933.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.418672, "cps": 61.976562, "mean_p": 0.370615, "boundary_entropy": 0.610503, "bpb": 1.318955}}, "bpb": {"en": 2.35272, "de": 2.65483, "vi": 2.14436, "tr": 2.66949, "fi": 2.90383, "ru": 1.73003, "el": 1.83067, "hi": 1.37405, "ta": 1.11679, "th": 1.3645, "zh": 3.48159, "my": 1.31896}, "gini_cpc": 0.198298, "gini_cps": 0.185239, "bpic": 5.99302, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0246964444076976, "vi": 1.2282499038103707, "tr": 1.1619453026544488, "fi": 0.9518289274088764, "ru": 1.9449385776523287, "el": 1.724085263385719, "hi": 2.2374111614084797, "ta": 2.2003656244363308, "th": 2.1491180429549415, "zh": 2.9028070783357878, "my": 2.2592717891315766}, "t": 99.8} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.36285, "de": 2.66078, "vi": 2.14871, "tr": 2.6739, "fi": 2.9095, "ru": 1.71288, "el": 1.82101, "hi": 1.37068, "ta": 1.12515, "th": 1.38498, "zh": 3.48902, "my": 1.32685}, "bytes_per_chunk": {"en": 5.74, "de": 5.456, "vi": 5.904, "tr": 5.175, "fi": 5.739, "ru": 5.35, "el": 5.835, "hi": 5.762, "ta": 7.023, "th": 7.434, "zh": 5.11, "my": 7.029}, "bpb_mean": 2.08219, "t": 101.1} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.78489, "bpb": 2.57505, "chunk_loss": 0.03038, "F": 0.18246, "G": 0.25685, "chunks": 186.84, "bytes_per_chunk": 5.481, "boundary_entropy": 0.34211, "router_noise": 0.0, "grad_norm": 1.0089, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 101.1} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.57197, "bpb": 2.26787, "chunk_loss": 0.030035, "F": 0.17407, "G": 0.28129, "chunks": 178.25, "bytes_per_chunk": 5.745, "boundary_entropy": 0.39979, "router_noise": 0.0, "grad_norm": 0.823, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 103.1} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.56075, "bpb": 2.25168, "chunk_loss": 0.030153, "F": 0.17328, "G": 0.28839, "chunks": 177.44, "bytes_per_chunk": 5.771, "boundary_entropy": 0.40963, "router_noise": 0.0, "grad_norm": 0.8405, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 105.1} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.5078, "bpb": 2.1753, "chunk_loss": 0.03013, "F": 0.17728, "G": 0.29505, "chunks": 181.53, "bytes_per_chunk": 5.641, "boundary_entropy": 0.40791, "router_noise": 0.0, "grad_norm": 0.8002, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 107.0} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3423.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19669, "cps": 26.742188, "mean_p": 0.243675, "boundary_entropy": 0.267951, "bpb": 2.314157}, "de": {"chunks": 4087.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.199045, "cps": 31.929688, "mean_p": 0.251026, "boundary_entropy": 0.287161, "bpb": 2.599411}, "vi": {"chunks": 4151.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22517, "cps": 32.429688, "mean_p": 0.282249, "boundary_entropy": 0.419598, "bpb": 2.131532}, "tr": {"chunks": 4116.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.227366, "cps": 32.15625, "mean_p": 0.260348, "boundary_entropy": 0.322186, "bpb": 2.592503}, "fi": {"chunks": 3652.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194142, "cps": 28.53125, "mean_p": 0.22701, "boundary_entropy": 0.262864, "bpb": 2.82731}, "ru": {"chunks": 6714.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.354301, "cps": 52.453125, "mean_p": 0.341683, "boundary_entropy": 0.548257, "bpb": 1.681029}, "el": {"chunks": 7434.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.362228, "cps": 58.078125, "mean_p": 0.391438, "boundary_entropy": 0.613442, "bpb": 1.780345}, "hi": {"chunks": 6366.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.378366, "cps": 49.734375, "mean_p": 0.337975, "boundary_entropy": 0.566647, "bpb": 1.349461}, "ta": {"chunks": 7442.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.398074, "cps": 58.140625, "mean_p": 0.416103, "boundary_entropy": 0.617738, "bpb": 1.106834}, "th": {"chunks": 8195.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.509481, "cps": 64.023438, "mean_p": 0.404842, "boundary_entropy": 0.626801, "bpb": 1.330273}, "zh": {"chunks": 3125.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.529123, "cps": 24.414062, "mean_p": 0.272621, "boundary_entropy": 0.362418, "bpb": 3.459748}, "my": {"chunks": 11026.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.581908, "cps": 86.140625, "mean_p": 0.394501, "boundary_entropy": 0.611769, "bpb": 1.284419}}, "bpb": {"en": 2.31416, "de": 2.59941, "vi": 2.13153, "tr": 2.5925, "fi": 2.82731, "ru": 1.68103, "el": 1.78034, "hi": 1.34946, "ta": 1.10683, "th": 1.33027, "zh": 3.45975, "my": 1.28442}, "gini_cpc": 0.215674, "gini_cps": 0.218857, "bpic": 5.503736, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0119742260518787, "vi": 1.1447925974365245, "tr": 1.155957948874013, "fi": 0.9870430763583934, "ru": 1.801313665048712, "el": 1.8416153775100148, "hi": 1.923662072710651, "ta": 2.023864428265067, "th": 2.5902704655400743, "zh": 2.6901332977975425, "my": 2.958501767927241}, "t": 109.9} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.32378, "de": 2.60372, "vi": 2.13708, "tr": 2.59883, "fi": 2.83333, "ru": 1.66112, "el": 1.7717, "hi": 1.34593, "ta": 1.11429, "th": 1.34978, "zh": 3.46111, "my": 1.2931}, "bytes_per_chunk": {"en": 5.367, "de": 5.143, "vi": 6.131, "tr": 4.863, "fi": 5.167, "ru": 5.435, "el": 5.241, "hi": 6.431, "ta": 7.096, "th": 5.714, "zh": 5.206, "my": 5.047}, "bpb_mean": 2.04115, "t": 111.2} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.45065, "bpb": 2.09285, "chunk_loss": 0.030631, "F": 0.18826, "G": 0.30705, "chunks": 192.78, "bytes_per_chunk": 5.312, "boundary_entropy": 0.43232, "router_noise": 0.0, "grad_norm": 0.7182, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 111.2} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.43946, "bpb": 2.0767, "chunk_loss": 0.029949, "F": 0.16803, "G": 0.28685, "chunks": 172.06, "bytes_per_chunk": 5.951, "boundary_entropy": 0.41943, "router_noise": 0.0, "grad_norm": 0.7274, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 113.2} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.54426, "bpb": 2.22789, "chunk_loss": 0.030349, "F": 0.18427, "G": 0.27536, "chunks": 188.69, "bytes_per_chunk": 5.427, "boundary_entropy": 0.36214, "router_noise": 0.0, "grad_norm": 0.7067, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 115.2} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.60769, "bpb": 2.31941, "chunk_loss": 0.030072, "F": 0.17453, "G": 0.28425, "chunks": 178.72, "bytes_per_chunk": 5.73, "boundary_entropy": 0.3965, "router_noise": 0.0, "grad_norm": 0.7765, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 117.1} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3224.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185255, "cps": 25.1875, "mean_p": 0.221911, "boundary_entropy": 0.2401, "bpb": 2.272786}, "de": {"chunks": 3774.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.183802, "cps": 29.484375, "mean_p": 0.224163, "boundary_entropy": 0.253307, "bpb": 2.566062}, "vi": {"chunks": 4611.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.250122, "cps": 36.023438, "mean_p": 0.285792, "boundary_entropy": 0.408672, "bpb": 2.100357}, "tr": {"chunks": 3547.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195934, "cps": 27.710938, "mean_p": 0.219556, "boundary_entropy": 0.287096, "bpb": 2.555828}, "fi": {"chunks": 3457.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183775, "cps": 27.007812, "mean_p": 0.216804, "boundary_entropy": 0.250379, "bpb": 2.812388}, "ru": {"chunks": 5818.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.307018, "cps": 45.453125, "mean_p": 0.31277, "boundary_entropy": 0.518267, "bpb": 1.650431}, "el": {"chunks": 5228.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.254739, "cps": 40.84375, "mean_p": 0.339857, "boundary_entropy": 0.586557, "bpb": 1.764699}, "hi": {"chunks": 6600.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.392273, "cps": 51.5625, "mean_p": 0.381612, "boundary_entropy": 0.5952, "bpb": 1.322354}, "ta": {"chunks": 9966.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.533084, "cps": 77.859375, "mean_p": 0.434991, "boundary_entropy": 0.626094, "bpb": 1.104608}, "th": {"chunks": 7473.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.464594, "cps": 58.382812, "mean_p": 0.407673, "boundary_entropy": 0.633157, "bpb": 1.306652}, "zh": {"chunks": 3094.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.523874, "cps": 24.171875, "mean_p": 0.265906, "boundary_entropy": 0.353859, "bpb": 3.421236}, "my": {"chunks": 9346.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.493245, "cps": 73.015625, "mean_p": 0.390466, "boundary_entropy": 0.618708, "bpb": 1.264722}}, "bpb": {"en": 2.27279, "de": 2.56606, "vi": 2.10036, "tr": 2.55583, "fi": 2.81239, "ru": 1.65043, "el": 1.7647, "hi": 1.32235, "ta": 1.10461, "th": 1.30665, "zh": 3.42124, "my": 1.26472}, "gini_cpc": 0.228532, "gini_cps": 0.227827, "bpic": 5.802731, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9921528305400529, "vi": 1.350147035960968, "tr": 1.057644521444631, "fi": 0.9920112176326796, "ru": 1.6572712243922558, "el": 1.3750668770944998, "hi": 2.1174733333579137, "ta": 2.877560745598171, "th": 2.507858357161204, "zh": 2.827847295806342, "my": 2.662511471908753}, "t": 120.0} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.28018, "de": 2.57297, "vi": 2.1043, "tr": 2.56146, "fi": 2.81481, "ru": 1.63019, "el": 1.75795, "hi": 1.31764, "ta": 1.11272, "th": 1.32743, "zh": 3.42169, "my": 1.27148}, "bytes_per_chunk": {"en": 5.758, "de": 5.609, "vi": 5.529, "tr": 5.675, "fi": 5.468, "ru": 6.17, "el": 7.478, "hi": 6.404, "ta": 5.823, "th": 6.303, "zh": 5.213, "my": 5.963}, "bpb_mean": 2.0144, "t": 121.2} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.41224, "bpb": 2.03744, "chunk_loss": 0.030564, "F": 0.17697, "G": 0.30582, "chunks": 181.22, "bytes_per_chunk": 5.651, "boundary_entropy": 0.45226, "router_noise": 0.0, "grad_norm": 1.1051, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 121.3} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.58496, "bpb": 2.28661, "chunk_loss": 0.030146, "F": 0.17719, "G": 0.27226, "chunks": 181.44, "bytes_per_chunk": 5.644, "boundary_entropy": 0.3636, "router_noise": 0.0, "grad_norm": 0.8716, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 123.3} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.6375, "bpb": 2.36242, "chunk_loss": 0.030492, "F": 0.18448, "G": 0.28437, "chunks": 188.91, "bytes_per_chunk": 5.421, "boundary_entropy": 0.39087, "router_noise": 0.0, "grad_norm": 0.8542, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 125.2} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.37103, "bpb": 1.97798, "chunk_loss": 0.030173, "F": 0.17648, "G": 0.2896, "chunks": 180.72, "bytes_per_chunk": 5.666, "boundary_entropy": 0.41412, "router_noise": 0.0, "grad_norm": 0.7515, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 127.2} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3189.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183244, "cps": 24.914062, "mean_p": 0.224811, "boundary_entropy": 0.236383, "bpb": 2.242383}, "de": {"chunks": 3853.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187649, "cps": 30.101562, "mean_p": 0.233604, "boundary_entropy": 0.260552, "bpb": 2.514735}, "vi": {"chunks": 4300.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.233252, "cps": 33.59375, "mean_p": 0.278246, "boundary_entropy": 0.405333, "bpb": 2.072183}, "tr": {"chunks": 3598.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.198752, "cps": 28.109375, "mean_p": 0.228096, "boundary_entropy": 0.280392, "bpb": 2.513507}, "fi": {"chunks": 3508.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.186487, "cps": 27.40625, "mean_p": 0.216016, "boundary_entropy": 0.241719, "bpb": 2.726788}, "ru": {"chunks": 6608.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.348707, "cps": 51.625, "mean_p": 0.336646, "boundary_entropy": 0.529615, "bpb": 1.619082}, "el": {"chunks": 6617.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.322419, "cps": 51.695312, "mean_p": 0.363789, "boundary_entropy": 0.597834, "bpb": 1.730466}, "hi": {"chunks": 7351.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.436909, "cps": 57.429688, "mean_p": 0.400091, "boundary_entropy": 0.604912, "bpb": 1.327002}, "ta": {"chunks": 7952.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.425354, "cps": 62.125, "mean_p": 0.407019, "boundary_entropy": 0.614119, "bpb": 1.06726}, "th": {"chunks": 8433.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.524277, "cps": 65.882812, "mean_p": 0.41596, "boundary_entropy": 0.633877, "bpb": 1.276148}, "zh": {"chunks": 3062.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.518456, "cps": 23.921875, "mean_p": 0.262691, "boundary_entropy": 0.337166, "bpb": 3.405234}, "my": {"chunks": 8597.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.453715, "cps": 67.164062, "mean_p": 0.368909, "boundary_entropy": 0.602763, "bpb": 1.239614}}, "bpb": {"en": 2.24238, "de": 2.51474, "vi": 2.07218, "tr": 2.51351, "fi": 2.72679, "ru": 1.61908, "el": 1.73047, "hi": 1.327, "ta": 1.06726, "th": 1.27615, "zh": 3.40523, "my": 1.23961}, "gini_cpc": 0.215525, "gini_cps": 0.209432, "bpic": 5.722267, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0240383066902055, "vi": 1.272901840924394, "tr": 1.0846264936675465, "fi": 1.0176942065632522, "ru": 1.90296333620435, "el": 1.75950250892134, "hi": 2.384300285574903, "ta": 2.3212424428684755, "th": 2.8610841803202702, "zh": 2.8293152777012325, "my": 2.4760143173444336}, "t": 130.1} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.24841, "de": 2.52193, "vi": 2.0778, "tr": 2.51863, "fi": 2.73361, "ru": 1.59734, "el": 1.72359, "hi": 1.32143, "ta": 1.07492, "th": 1.29415, "zh": 3.40565, "my": 1.24711}, "bytes_per_chunk": {"en": 5.753, "de": 5.483, "vi": 6.04, "tr": 5.582, "fi": 5.413, "ru": 5.328, "el": 5.691, "hi": 5.828, "ta": 6.571, "th": 5.471, "zh": 5.251, "my": 6.467}, "bpb_mean": 1.98038, "t": 131.3} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.28166, "bpb": 1.84904, "chunk_loss": 0.03022, "F": 0.17508, "G": 0.31568, "chunks": 179.28, "bytes_per_chunk": 5.712, "boundary_entropy": 0.45531, "router_noise": 0.0, "grad_norm": 0.739, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 131.4} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.5199, "bpb": 2.19276, "chunk_loss": 0.030264, "F": 0.17688, "G": 0.26565, "chunks": 181.12, "bytes_per_chunk": 5.654, "boundary_entropy": 0.36122, "router_noise": 0.0, "grad_norm": 0.6926, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 133.3} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.47627, "bpb": 2.12981, "chunk_loss": 0.030318, "F": 0.18216, "G": 0.27549, "chunks": 186.53, "bytes_per_chunk": 5.49, "boundary_entropy": 0.36758, "router_noise": 0.0, "grad_norm": 0.7599, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 135.3} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.45575, "bpb": 2.1002, "chunk_loss": 0.030202, "F": 0.17606, "G": 0.29069, "chunks": 180.28, "bytes_per_chunk": 5.68, "boundary_entropy": 0.41464, "router_noise": 0.0, "grad_norm": 0.7304, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 137.3} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3300.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189622, "cps": 25.78125, "mean_p": 0.226467, "boundary_entropy": 0.225362, "bpb": 2.221886}, "de": {"chunks": 4024.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.195977, "cps": 31.4375, "mean_p": 0.235679, "boundary_entropy": 0.249525, "bpb": 2.471411}, "vi": {"chunks": 4074.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.220993, "cps": 31.828125, "mean_p": 0.262471, "boundary_entropy": 0.387091, "bpb": 2.056088}, "tr": {"chunks": 3650.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201624, "cps": 28.515625, "mean_p": 0.227169, "boundary_entropy": 0.272748, "bpb": 2.481402}, "fi": {"chunks": 3608.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.191803, "cps": 28.1875, "mean_p": 0.22353, "boundary_entropy": 0.242257, "bpb": 2.683271}, "ru": {"chunks": 5825.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.307388, "cps": 45.507812, "mean_p": 0.314306, "boundary_entropy": 0.511469, "bpb": 1.581715}, "el": {"chunks": 6215.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.302831, "cps": 48.554688, "mean_p": 0.348177, "boundary_entropy": 0.582397, "bpb": 1.700007}, "hi": {"chunks": 7292.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.433403, "cps": 56.96875, "mean_p": 0.38201, "boundary_entropy": 0.601375, "bpb": 1.291207}, "ta": {"chunks": 7745.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.414282, "cps": 60.507812, "mean_p": 0.405452, "boundary_entropy": 0.615937, "bpb": 1.056012}, "th": {"chunks": 8546.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.531302, "cps": 66.765625, "mean_p": 0.418108, "boundary_entropy": 0.63457, "bpb": 1.248741}, "zh": {"chunks": 3040.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.514731, "cps": 23.75, "mean_p": 0.268939, "boundary_entropy": 0.354658, "bpb": 3.373269}, "my": {"chunks": 7766.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.409859, "cps": 60.671875, "mean_p": 0.36951, "boundary_entropy": 0.604441, "bpb": 1.219528}}, "bpb": {"en": 2.22189, "de": 2.47141, "vi": 2.05609, "tr": 2.4814, "fi": 2.68327, "ru": 1.58172, "el": 1.70001, "hi": 1.29121, "ta": 1.05601, "th": 1.24874, "zh": 3.37327, "my": 1.21953}, "gini_cpc": 0.212096, "gini_cps": 0.200063, "bpic": 5.896612, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0335125275056116, "vi": 1.1654350173829424, "tr": 1.0632918702575664, "fi": 1.011497528042103, "ru": 1.6210518109858478, "el": 1.597020416118501, "hi": 2.2856081048223693, "ta": 2.184772058644752, "th": 2.8018959504902927, "zh": 2.714502970784718, "my": 2.1614450073886426}, "t": 140.2} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.2312, "de": 2.47906, "vi": 2.05947, "tr": 2.48863, "fi": 2.69406, "ru": 1.56064, "el": 1.69103, "hi": 1.28787, "ta": 1.06197, "th": 1.26898, "zh": 3.37263, "my": 1.2268}, "bytes_per_chunk": {"en": 5.594, "de": 5.293, "vi": 6.109, "tr": 5.532, "fi": 5.204, "ru": 5.963, "el": 6.165, "hi": 5.825, "ta": 6.943, "th": 5.405, "zh": 5.312, "my": 7.607}, "bpb_mean": 1.95186, "t": 141.4} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 141.4} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "global", "lm_loss": 1.31317, "bpb": 1.8945, "chunk_loss": 0.030031, "F": 0.17212, "G": 0.29045, "chunks": 176.25, "bytes_per_chunk": 5.81, "boundary_entropy": 0.40438, "router_noise": 0.0, "grad_norm": 0.7275, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 141.4} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "global", "lm_loss": 1.46809, "bpb": 2.11801, "chunk_loss": 0.030092, "F": 0.17487, "G": 0.28092, "chunks": 179.06, "bytes_per_chunk": 5.719, "boundary_entropy": 0.38244, "router_noise": 0.0, "grad_norm": 0.7529, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 143.4} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "global", "lm_loss": 1.46167, "bpb": 2.10874, "chunk_loss": 0.030555, "F": 0.18381, "G": 0.28114, "chunks": 188.22, "bytes_per_chunk": 5.44, "boundary_entropy": 0.38033, "router_noise": 0.0, "grad_norm": 0.7182, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 145.3} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "global", "lm_loss": 1.52057, "bpb": 2.19372, "chunk_loss": 0.030123, "F": 0.17798, "G": 0.24915, "chunks": 182.25, "bytes_per_chunk": 5.619, "boundary_entropy": 0.30218, "router_noise": 0.0, "grad_norm": 0.7744, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 147.3} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3279.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188416, "cps": 25.617188, "mean_p": 0.226646, "boundary_entropy": 0.220146, "bpb": 2.194776}, "de": {"chunks": 3931.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.191448, "cps": 30.710938, "mean_p": 0.234253, "boundary_entropy": 0.242423, "bpb": 2.444013}, "vi": {"chunks": 4158.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225549, "cps": 32.484375, "mean_p": 0.278953, "boundary_entropy": 0.406631, "bpb": 2.049832}, "tr": {"chunks": 3811.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210518, "cps": 29.773438, "mean_p": 0.234812, "boundary_entropy": 0.273689, "bpb": 2.427022}, "fi": {"chunks": 3614.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.192122, "cps": 28.234375, "mean_p": 0.221989, "boundary_entropy": 0.238917, "bpb": 2.643864}, "ru": {"chunks": 5778.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.304908, "cps": 45.140625, "mean_p": 0.307813, "boundary_entropy": 0.504031, "bpb": 1.55224}, "el": {"chunks": 6949.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.338596, "cps": 54.289062, "mean_p": 0.363685, "boundary_entropy": 0.594177, "bpb": 1.666624}, "hi": {"chunks": 7363.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.437623, "cps": 57.523438, "mean_p": 0.383472, "boundary_entropy": 0.588194, "bpb": 1.278352}, "ta": {"chunks": 8168.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.436908, "cps": 63.8125, "mean_p": 0.411109, "boundary_entropy": 0.610878, "bpb": 1.041853}, "th": {"chunks": 7817.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.485981, "cps": 61.070312, "mean_p": 0.39107, "boundary_entropy": 0.613934, "bpb": 1.232001}, "zh": {"chunks": 3090.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.523197, "cps": 24.140625, "mean_p": 0.262531, "boundary_entropy": 0.335888, "bpb": 3.348134}, "my": {"chunks": 10156.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.535993, "cps": 79.34375, "mean_p": 0.394501, "boundary_entropy": 0.609119, "bpb": 1.202105}}, "bpb": {"en": 2.19478, "de": 2.44401, "vi": 2.04983, "tr": 2.42702, "fi": 2.64386, "ru": 1.55224, "el": 1.66662, "hi": 1.27835, "ta": 1.04185, "th": 1.232, "zh": 3.34813, "my": 1.20211}, "gini_cpc": 0.218389, "gini_cps": 0.219708, "bpic": 5.634392, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.016092720935825, "vi": 1.1970824024768907, "tr": 1.1173033499120792, "fi": 1.0196684183005378, "ru": 1.618270162487598, "el": 1.7970665897396083, "hi": 2.322642834517645, "ta": 2.3188516384872293, "th": 2.5792993591626825, "zh": 2.77682007442615, "my": 2.8447363333288913}, "t": 150.2} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.20137, "de": 2.45038, "vi": 2.05287, "tr": 2.43597, "fi": 2.65287, "ru": 1.5289, "el": 1.6581, "hi": 1.27572, "ta": 1.04917, "th": 1.2518, "zh": 3.34846, "my": 1.21073}, "bytes_per_chunk": {"en": 5.612, "de": 5.412, "vi": 6.174, "tr": 5.318, "fi": 5.241, "ru": 6.033, "el": 5.377, "hi": 5.714, "ta": 6.274, "th": 6.153, "zh": 5.223, "my": 6.015}, "bpb_mean": 1.92636, "t": 151.4} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "global", "lm_loss": 1.49304, "bpb": 2.154, "chunk_loss": 0.030176, "F": 0.17758, "G": 0.27251, "chunks": 181.84, "bytes_per_chunk": 5.631, "boundary_entropy": 0.35474, "router_noise": 0.0, "grad_norm": 0.7603, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 151.4} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "global", "lm_loss": 1.44012, "bpb": 2.07766, "chunk_loss": 0.030307, "F": 0.18045, "G": 0.25843, "chunks": 184.78, "bytes_per_chunk": 5.542, "boundary_entropy": 0.33451, "router_noise": 0.0, "grad_norm": 0.7756, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 153.4} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "global", "lm_loss": 1.49256, "bpb": 2.15331, "chunk_loss": 0.03023, "F": 0.17957, "G": 0.25798, "chunks": 183.88, "bytes_per_chunk": 5.569, "boundary_entropy": 0.31971, "router_noise": 0.0, "grad_norm": 0.8373, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 155.4} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "global", "lm_loss": 1.52189, "bpb": 2.19562, "chunk_loss": 0.029995, "F": 0.17041, "G": 0.24564, "chunks": 174.5, "bytes_per_chunk": 5.868, "boundary_entropy": 0.31506, "router_noise": 0.0, "grad_norm": 0.7542, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 157.3} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3229.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185543, "cps": 25.226562, "mean_p": 0.221736, "boundary_entropy": 0.211426, "bpb": 2.176629}, "de": {"chunks": 3957.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.192714, "cps": 30.914062, "mean_p": 0.233782, "boundary_entropy": 0.241844, "bpb": 2.409769}, "vi": {"chunks": 4447.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.241226, "cps": 34.742188, "mean_p": 0.281006, "boundary_entropy": 0.40439, "bpb": 2.035074}, "tr": {"chunks": 3406.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.188146, "cps": 26.609375, "mean_p": 0.218863, "boundary_entropy": 0.259734, "bpb": 2.408492}, "fi": {"chunks": 3440.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182872, "cps": 26.875, "mean_p": 0.206896, "boundary_entropy": 0.220142, "bpb": 2.62578}, "ru": {"chunks": 6486.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.342269, "cps": 50.671875, "mean_p": 0.316014, "boundary_entropy": 0.495727, "bpb": 1.528353}, "el": {"chunks": 6298.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.306875, "cps": 49.203125, "mean_p": 0.342845, "boundary_entropy": 0.572711, "bpb": 1.653492}, "hi": {"chunks": 6735.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.400297, "cps": 52.617188, "mean_p": 0.386608, "boundary_entropy": 0.597363, "bpb": 1.253315}, "ta": {"chunks": 6938.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.371115, "cps": 54.203125, "mean_p": 0.400148, "boundary_entropy": 0.61189, "bpb": 1.045543}, "th": {"chunks": 6787.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.421946, "cps": 53.023438, "mean_p": 0.399424, "boundary_entropy": 0.623755, "bpb": 1.215862}, "zh": {"chunks": 3026.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.51236, "cps": 23.640625, "mean_p": 0.269754, "boundary_entropy": 0.351193, "bpb": 3.311096}, "my": {"chunks": 9238.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.487545, "cps": 72.171875, "mean_p": 0.378786, "boundary_entropy": 0.603765, "bpb": 1.203813}}, "bpb": {"en": 2.17663, "de": 2.40977, "vi": 2.03507, "tr": 2.40849, "fi": 2.62578, "ru": 1.52835, "el": 1.65349, "hi": 1.25332, "ta": 1.04554, "th": 1.21586, "zh": 3.3111, "my": 1.20381}, "gini_cpc": 0.205283, "gini_cps": 0.197036, "bpic": 5.997796, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0386511786682375, "vi": 1.3001098920205694, "tr": 1.0140285141352552, "fi": 0.9856044015547463, "ru": 1.8446917488361985, "el": 1.6539329002454506, "hi": 2.1574393831533234, "ta": 2.000160752328851, "th": 2.2741172847522546, "zh": 2.7614142155040304, "my": 2.6276689971798093}, "t": 160.2} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.18611, "de": 2.41595, "vi": 2.03908, "tr": 2.41542, "fi": 2.63304, "ru": 1.50379, "el": 1.6451, "hi": 1.24978, "ta": 1.0518, "th": 1.23472, "zh": 3.30758, "my": 1.21185}, "bytes_per_chunk": {"en": 5.641, "de": 5.319, "vi": 5.745, "tr": 5.875, "fi": 5.517, "ru": 5.439, "el": 6.021, "hi": 6.382, "ta": 7.656, "th": 7.018, "zh": 5.352, "my": 5.994}, "bpb_mean": 1.90785, "t": 161.4} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "global", "lm_loss": 1.40113, "bpb": 2.02141, "chunk_loss": 0.03016, "F": 0.17734, "G": 0.26666, "chunks": 181.59, "bytes_per_chunk": 5.639, "boundary_entropy": 0.34993, "router_noise": 0.0, "grad_norm": 0.7449, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 161.4} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "global", "lm_loss": 1.32119, "bpb": 1.90607, "chunk_loss": 0.030062, "F": 0.17117, "G": 0.26929, "chunks": 175.28, "bytes_per_chunk": 5.842, "boundary_entropy": 0.36214, "router_noise": 0.0, "grad_norm": 0.7732, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 163.4} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "global", "lm_loss": 1.40032, "bpb": 2.02024, "chunk_loss": 0.030029, "F": 0.17319, "G": 0.25929, "chunks": 177.34, "bytes_per_chunk": 5.774, "boundary_entropy": 0.33105, "router_noise": 0.0, "grad_norm": 0.7459, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 165.3} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "global", "lm_loss": 1.29681, "bpb": 1.87091, "chunk_loss": 0.029899, "F": 0.16858, "G": 0.27678, "chunks": 172.62, "bytes_per_chunk": 5.932, "boundary_entropy": 0.38368, "router_noise": 0.0, "grad_norm": 0.6816, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 167.3} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3221.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185083, "cps": 25.164062, "mean_p": 0.223563, "boundary_entropy": 0.21445, "bpb": 2.158145}, "de": {"chunks": 3844.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187211, "cps": 30.03125, "mean_p": 0.224259, "boundary_entropy": 0.229269, "bpb": 2.39665}, "vi": {"chunks": 4266.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.231408, "cps": 33.328125, "mean_p": 0.272342, "boundary_entropy": 0.395723, "bpb": 2.013026}, "tr": {"chunks": 3542.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195658, "cps": 27.671875, "mean_p": 0.221224, "boundary_entropy": 0.263082, "bpb": 2.385189}, "fi": {"chunks": 3540.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188188, "cps": 27.65625, "mean_p": 0.215433, "boundary_entropy": 0.225691, "bpb": 2.596073}, "ru": {"chunks": 6034.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.318417, "cps": 47.140625, "mean_p": 0.307684, "boundary_entropy": 0.48597, "bpb": 1.503093}, "el": {"chunks": 5407.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.263461, "cps": 42.242188, "mean_p": 0.325725, "boundary_entropy": 0.565134, "bpb": 1.622183}, "hi": {"chunks": 7154.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.425201, "cps": 55.890625, "mean_p": 0.377347, "boundary_entropy": 0.589784, "bpb": 1.234217}, "ta": {"chunks": 7293.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.390104, "cps": 56.976562, "mean_p": 0.393742, "boundary_entropy": 0.601332, "bpb": 1.022035}, "th": {"chunks": 7778.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.483556, "cps": 60.765625, "mean_p": 0.412511, "boundary_entropy": 0.625983, "bpb": 1.190216}, "zh": {"chunks": 3054.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.517101, "cps": 23.859375, "mean_p": 0.271193, "boundary_entropy": 0.347442, "bpb": 3.293611}, "my": {"chunks": 9859.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.520319, "cps": 77.023438, "mean_p": 0.40738, "boundary_entropy": 0.619784, "bpb": 1.175877}}, "bpb": {"en": 2.15814, "de": 2.39665, "vi": 2.01303, "tr": 2.38519, "fi": 2.59607, "ru": 1.50309, "el": 1.62218, "hi": 1.23422, "ta": 1.02203, "th": 1.19022, "zh": 3.29361, "my": 1.17588}, "gini_cpc": 0.22196, "gini_cps": 0.215296, "bpic": 5.90505, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0114964600717786, "vi": 1.2502909986816075, "tr": 1.0571373099218393, "fi": 1.0167747998848198, "ru": 1.720400210033266, "el": 1.4234719701658203, "hi": 2.297350494733438, "ta": 2.107725935018411, "th": 2.6126441945281185, "zh": 2.7938879557021505, "my": 2.811272121934133}, "t": 170.2} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.16633, "de": 2.40405, "vi": 2.01561, "tr": 2.39306, "fi": 2.60645, "ru": 1.47755, "el": 1.61567, "hi": 1.23114, "ta": 1.02953, "th": 1.2091, "zh": 3.28981, "my": 1.18349}, "bytes_per_chunk": {"en": 5.648, "de": 5.503, "vi": 6.059, "tr": 5.655, "fi": 5.345, "ru": 5.752, "el": 6.707, "hi": 6.017, "ta": 7.315, "th": 6.026, "zh": 5.293, "my": 5.662}, "bpb_mean": 1.88515, "t": 171.4} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "global", "lm_loss": 1.50671, "bpb": 2.17372, "chunk_loss": 0.030264, "F": 0.17767, "G": 0.27352, "chunks": 181.94, "bytes_per_chunk": 5.628, "boundary_entropy": 0.36026, "router_noise": 0.0, "grad_norm": 0.6856, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 171.4} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "global", "lm_loss": 1.46702, "bpb": 2.11646, "chunk_loss": 0.030229, "F": 0.17737, "G": 0.26296, "chunks": 181.62, "bytes_per_chunk": 5.638, "boundary_entropy": 0.3488, "router_noise": 0.0, "grad_norm": 0.8023, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 173.4} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "global", "lm_loss": 1.40334, "bpb": 2.02459, "chunk_loss": 0.030076, "F": 0.17172, "G": 0.2713, "chunks": 175.84, "bytes_per_chunk": 5.823, "boundary_entropy": 0.38076, "router_noise": 0.0, "grad_norm": 0.7194, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 175.3} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "global", "lm_loss": 1.31547, "bpb": 1.89782, "chunk_loss": 0.030188, "F": 0.17661, "G": 0.27479, "chunks": 180.84, "bytes_per_chunk": 5.662, "boundary_entropy": 0.37878, "router_noise": 0.0, "grad_norm": 0.7732, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 177.3} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3192.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183417, "cps": 24.9375, "mean_p": 0.216787, "boundary_entropy": 0.198358, "bpb": 2.140283}, "de": {"chunks": 3864.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188185, "cps": 30.1875, "mean_p": 0.224563, "boundary_entropy": 0.224669, "bpb": 2.365182}, "vi": {"chunks": 4242.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230106, "cps": 33.140625, "mean_p": 0.265814, "boundary_entropy": 0.386781, "bpb": 2.005723}, "tr": {"chunks": 3545.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195824, "cps": 27.695312, "mean_p": 0.222087, "boundary_entropy": 0.255446, "bpb": 2.353313}, "fi": {"chunks": 3414.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18149, "cps": 26.671875, "mean_p": 0.204998, "boundary_entropy": 0.211182, "bpb": 2.5556}, "ru": {"chunks": 5947.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.313826, "cps": 46.460938, "mean_p": 0.297796, "boundary_entropy": 0.473692, "bpb": 1.487954}, "el": {"chunks": 5651.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.27535, "cps": 44.148438, "mean_p": 0.326528, "boundary_entropy": 0.559905, "bpb": 1.612189}, "hi": {"chunks": 7351.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.436909, "cps": 57.429688, "mean_p": 0.386415, "boundary_entropy": 0.590188, "bpb": 1.22955}, "ta": {"chunks": 7441.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.398021, "cps": 58.132812, "mean_p": 0.391129, "boundary_entropy": 0.599035, "bpb": 1.018493}, "th": {"chunks": 7576.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.470998, "cps": 59.1875, "mean_p": 0.402518, "boundary_entropy": 0.619871, "bpb": 1.177392}, "zh": {"chunks": 3080.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.521504, "cps": 24.0625, "mean_p": 0.261406, "boundary_entropy": 0.322373, "bpb": 3.263145}, "my": {"chunks": 8121.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.428594, "cps": 63.445312, "mean_p": 0.381988, "boundary_entropy": 0.609929, "bpb": 1.162873}}, "bpb": {"en": 2.14028, "de": 2.36518, "vi": 2.00572, "tr": 2.35331, "fi": 2.5556, "ru": 1.48795, "el": 1.61219, "hi": 1.22955, "ta": 1.01849, "th": 1.17739, "zh": 3.26314, "my": 1.16287}, "gini_cpc": 0.213018, "gini_cps": 0.196802, "bpic": 6.051037, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0259966626252477, "vi": 1.2545522675688408, "tr": 1.0676451354592082, "fi": 0.9894933295708038, "ru": 1.7109998115341123, "el": 1.5012246941722087, "hi": 2.382059401847859, "ta": 2.1700366689658086, "th": 2.5679120088065983, "zh": 2.843272675423744, "my": 2.3367237461132246}, "t": 180.1} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.14795, "de": 2.37511, "vi": 2.00791, "tr": 2.36088, "fi": 2.56617, "ru": 1.46482, "el": 1.60278, "hi": 1.22323, "ta": 1.02576, "th": 1.19631, "zh": 3.26098, "my": 1.16998}, "bytes_per_chunk": {"en": 5.681, "de": 5.447, "vi": 5.953, "tr": 5.564, "fi": 5.575, "ru": 5.85, "el": 6.748, "hi": 5.827, "ta": 6.763, "th": 6.174, "zh": 5.278, "my": 6.53}, "bpb_mean": 1.86682, "t": 181.4} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "global", "lm_loss": 1.39814, "bpb": 2.01708, "chunk_loss": 0.030025, "F": 0.17136, "G": 0.26392, "chunks": 175.47, "bytes_per_chunk": 5.836, "boundary_entropy": 0.35008, "router_noise": 0.0, "grad_norm": 0.7289, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 181.4} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "global", "lm_loss": 1.38049, "bpb": 1.99163, "chunk_loss": 0.030054, "F": 0.16943, "G": 0.26245, "chunks": 173.5, "bytes_per_chunk": 5.902, "boundary_entropy": 0.33792, "router_noise": 0.0, "grad_norm": 0.7755, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 183.4} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "global", "lm_loss": 1.26957, "bpb": 1.8316, "chunk_loss": 0.030013, "F": 0.16959, "G": 0.28753, "chunks": 173.66, "bytes_per_chunk": 5.897, "boundary_entropy": 0.40361, "router_noise": 0.0, "grad_norm": 0.7446, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 185.3} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "global", "lm_loss": 1.35474, "bpb": 1.95448, "chunk_loss": 0.030082, "F": 0.17319, "G": 0.27922, "chunks": 177.34, "bytes_per_chunk": 5.774, "boundary_entropy": 0.37228, "router_noise": 0.0, "grad_norm": 0.6652, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 187.2} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3220.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185026, "cps": 25.15625, "mean_p": 0.218821, "boundary_entropy": 0.199726, "bpb": 2.12183}, "de": {"chunks": 3860.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18799, "cps": 30.15625, "mean_p": 0.222521, "boundary_entropy": 0.220017, "bpb": 2.34941}, "vi": {"chunks": 4457.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.241768, "cps": 34.820312, "mean_p": 0.279105, "boundary_entropy": 0.396748, "bpb": 1.990733}, "tr": {"chunks": 3572.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.197315, "cps": 27.90625, "mean_p": 0.222101, "boundary_entropy": 0.253814, "bpb": 2.332579}, "fi": {"chunks": 3495.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185796, "cps": 27.304688, "mean_p": 0.210524, "boundary_entropy": 0.210789, "bpb": 2.548083}, "ru": {"chunks": 6060.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319789, "cps": 47.34375, "mean_p": 0.297448, "boundary_entropy": 0.469447, "bpb": 1.46726}, "el": {"chunks": 6574.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.320324, "cps": 51.359375, "mean_p": 0.329965, "boundary_entropy": 0.551489, "bpb": 1.585433}, "hi": {"chunks": 6826.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.405706, "cps": 53.328125, "mean_p": 0.374878, "boundary_entropy": 0.583849, "bpb": 1.223811}, "ta": {"chunks": 8787.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.470019, "cps": 68.648438, "mean_p": 0.414887, "boundary_entropy": 0.605544, "bpb": 1.001583}, "th": {"chunks": 7818.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.486043, "cps": 61.078125, "mean_p": 0.408627, "boundary_entropy": 0.620951, "bpb": 1.159988}, "zh": {"chunks": 3012.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50999, "cps": 23.53125, "mean_p": 0.26526, "boundary_entropy": 0.331193, "bpb": 3.233601}, "my": {"chunks": 9594.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.506333, "cps": 74.953125, "mean_p": 0.40314, "boundary_entropy": 0.617522, "bpb": 1.171742}}, "bpb": {"en": 2.12183, "de": 2.34941, "vi": 1.99073, "tr": 2.33258, "fi": 2.54808, "ru": 1.46726, "el": 1.58543, "hi": 1.22381, "ta": 1.00158, "th": 1.15999, "zh": 3.2336, "my": 1.17174}, "gini_cpc": 0.215833, "gini_cps": 0.221455, "bpic": 5.70466, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0160220798938113, "vi": 1.3066754772096691, "tr": 1.0664221262527094, "fi": 1.0041614510605827, "ru": 1.7283498582408758, "el": 1.7312393033963107, "hi": 2.1927012265465655, "ta": 2.5402906247560133, "th": 2.6268958193757155, "zh": 2.7563208696072845, "my": 2.7365575547067684}, "t": 190.1} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.13101, "de": 2.35768, "vi": 1.99418, "tr": 2.33929, "fi": 2.55629, "ru": 1.44314, "el": 1.578, "hi": 1.22009, "ta": 1.01026, "th": 1.17877, "zh": 3.23107, "my": 1.17948}, "bytes_per_chunk": {"en": 5.671, "de": 5.468, "vi": 5.803, "tr": 5.61, "fi": 5.389, "ru": 5.722, "el": 5.624, "hi": 6.252, "ta": 5.981, "th": 6.041, "zh": 5.338, "my": 5.922}, "bpb_mean": 1.85161, "t": 191.3} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "global", "lm_loss": 1.29929, "bpb": 1.87448, "chunk_loss": 0.030145, "F": 0.17557, "G": 0.28072, "chunks": 179.78, "bytes_per_chunk": 5.696, "boundary_entropy": 0.38079, "router_noise": 0.0, "grad_norm": 0.7374, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 191.4} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "global", "lm_loss": 1.20507, "bpb": 1.73855, "chunk_loss": 0.030094, "F": 0.17209, "G": 0.29104, "chunks": 176.22, "bytes_per_chunk": 5.811, "boundary_entropy": 0.39616, "router_noise": 0.0, "grad_norm": 0.6797, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 193.3} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "global", "lm_loss": 1.31993, "bpb": 1.90426, "chunk_loss": 0.029957, "F": 0.16769, "G": 0.26994, "chunks": 171.72, "bytes_per_chunk": 5.963, "boundary_entropy": 0.35905, "router_noise": 0.0, "grad_norm": 0.7463, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 195.3} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "global", "lm_loss": 1.35968, "bpb": 1.96161, "chunk_loss": 0.030166, "F": 0.17783, "G": 0.2709, "chunks": 182.09, "bytes_per_chunk": 5.623, "boundary_entropy": 0.35998, "router_noise": 0.0, "grad_norm": 0.7311, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 197.2} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3198.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183761, "cps": 24.984375, "mean_p": 0.215376, "boundary_entropy": 0.188043, "bpb": 2.106739}, "de": {"chunks": 3850.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187503, "cps": 30.078125, "mean_p": 0.220268, "boundary_entropy": 0.20826, "bpb": 2.319786}, "vi": {"chunks": 4221.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228967, "cps": 32.976562, "mean_p": 0.270055, "boundary_entropy": 0.388888, "bpb": 1.975058}, "tr": {"chunks": 3686.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203613, "cps": 28.796875, "mean_p": 0.225994, "boundary_entropy": 0.251647, "bpb": 2.307811}, "fi": {"chunks": 3597.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.191218, "cps": 28.101562, "mean_p": 0.215141, "boundary_entropy": 0.206499, "bpb": 2.504207}, "ru": {"chunks": 5902.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.311451, "cps": 46.109375, "mean_p": 0.291254, "boundary_entropy": 0.4593, "bpb": 1.457446}, "el": {"chunks": 6464.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.314964, "cps": 50.5, "mean_p": 0.327914, "boundary_entropy": 0.550882, "bpb": 1.583657}, "hi": {"chunks": 6675.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.396731, "cps": 52.148438, "mean_p": 0.375514, "boundary_entropy": 0.586148, "bpb": 1.209485}, "ta": {"chunks": 8100.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.433271, "cps": 63.28125, "mean_p": 0.412129, "boundary_entropy": 0.605415, "bpb": 0.990025}, "th": {"chunks": 7577.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.47106, "cps": 59.195312, "mean_p": 0.405008, "boundary_entropy": 0.619256, "bpb": 1.145396}, "zh": {"chunks": 3034.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.513715, "cps": 23.703125, "mean_p": 0.258274, "boundary_entropy": 0.311587, "bpb": 3.222014}, "my": {"chunks": 9010.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.475512, "cps": 70.390625, "mean_p": 0.384786, "boundary_entropy": 0.608414, "bpb": 1.147847}}, "bpb": {"en": 2.10674, "de": 2.31979, "vi": 1.97506, "tr": 2.30781, "fi": 2.50421, "ru": 1.45745, "el": 1.58366, "hi": 1.20948, "ta": 0.99003, "th": 1.1454, "zh": 3.22201, "my": 1.14785}, "gini_cpc": 0.209529, "gini_cps": 0.206926, "bpic": 5.875938, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.020361311023695, "vi": 1.2459995084397235, "tr": 1.1080272474610835, "fi": 1.0405769826704125, "ru": 1.6948671085655447, "el": 1.7139816317372016, "hi": 2.1589463882980353, "ta": 2.3577904818486948, "th": 2.563432480239534, "zh": 2.7955534137383102, "my": 2.5876591845520927}, "t": 200.1} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.11508, "de": 2.33074, "vi": 1.97879, "tr": 2.31463, "fi": 2.515, "ru": 1.43609, "el": 1.57733, "hi": 1.20326, "ta": 0.99841, "th": 1.16393, "zh": 3.2198, "my": 1.15508}, "bytes_per_chunk": {"en": 5.681, "de": 5.481, "vi": 6.155, "tr": 5.395, "fi": 5.258, "ru": 5.932, "el": 5.706, "hi": 6.362, "ta": 6.517, "th": 6.2, "zh": 5.315, "my": 6.317}, "bpb_mean": 1.83401, "t": 201.3} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "global", "lm_loss": 1.31419, "bpb": 1.89598, "chunk_loss": 0.030004, "F": 0.17142, "G": 0.28426, "chunks": 175.53, "bytes_per_chunk": 5.834, "boundary_entropy": 0.38689, "router_noise": 0.0, "grad_norm": 0.7309, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 201.4} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "global", "lm_loss": 1.30077, "bpb": 1.87661, "chunk_loss": 0.030213, "F": 0.1788, "G": 0.26689, "chunks": 183.09, "bytes_per_chunk": 5.593, "boundary_entropy": 0.33269, "router_noise": 0.0, "grad_norm": 0.7057, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 203.3} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "global", "lm_loss": 1.52167, "bpb": 2.1953, "chunk_loss": 0.030346, "F": 0.18365, "G": 0.25835, "chunks": 188.06, "bytes_per_chunk": 5.445, "boundary_entropy": 0.31176, "router_noise": 0.0, "grad_norm": 0.7914, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 205.3} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "global", "lm_loss": 1.41811, "bpb": 2.04589, "chunk_loss": 0.030235, "F": 0.17981, "G": 0.24186, "chunks": 184.12, "bytes_per_chunk": 5.561, "boundary_entropy": 0.27733, "router_noise": 0.0, "grad_norm": 0.7293, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 207.2} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3243.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186347, "cps": 25.335938, "mean_p": 0.218982, "boundary_entropy": 0.19036, "bpb": 2.09661}, "de": {"chunks": 3812.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.185652, "cps": 29.78125, "mean_p": 0.217516, "boundary_entropy": 0.206535, "bpb": 2.307647}, "vi": {"chunks": 4421.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.239816, "cps": 34.539062, "mean_p": 0.280033, "boundary_entropy": 0.388506, "bpb": 1.966711}, "tr": {"chunks": 3639.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201016, "cps": 28.429688, "mean_p": 0.225509, "boundary_entropy": 0.252781, "bpb": 2.286115}, "fi": {"chunks": 3475.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184732, "cps": 27.148438, "mean_p": 0.208568, "boundary_entropy": 0.204512, "bpb": 2.484301}, "ru": {"chunks": 6063.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319947, "cps": 47.367188, "mean_p": 0.2937, "boundary_entropy": 0.455523, "bpb": 1.433427}, "el": {"chunks": 6250.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.304536, "cps": 48.828125, "mean_p": 0.32261, "boundary_entropy": 0.543643, "bpb": 1.546288}, "hi": {"chunks": 6900.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.410104, "cps": 53.90625, "mean_p": 0.371771, "boundary_entropy": 0.578422, "bpb": 1.19331}, "ta": {"chunks": 8458.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.45242, "cps": 66.078125, "mean_p": 0.399214, "boundary_entropy": 0.59134, "bpb": 0.985156}, "th": {"chunks": 7701.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.478769, "cps": 60.164062, "mean_p": 0.408436, "boundary_entropy": 0.620624, "bpb": 1.139231}, "zh": {"chunks": 3003.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.508466, "cps": 23.460938, "mean_p": 0.262554, "boundary_entropy": 0.322093, "bpb": 3.205171}, "my": {"chunks": 10499.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.554095, "cps": 82.023438, "mean_p": 0.410881, "boundary_entropy": 0.620568, "bpb": 1.138075}}, "bpb": {"en": 2.09661, "de": 2.30765, "vi": 1.96671, "tr": 2.28611, "fi": 2.4843, "ru": 1.43343, "el": 1.54629, "hi": 1.19331, "ta": 0.98516, "th": 1.13923, "zh": 3.20517, "my": 1.13808}, "gini_cpc": 0.223523, "gini_cps": 0.227571, "bpic": 5.688678, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9962713842607012, "vi": 1.2869288725268444, "tr": 1.0787198630020436, "fi": 0.9913342179627908, "ru": 1.716941608351497, "el": 1.6342419098574137, "hi": 2.2007524264171225, "ta": 2.4278362011143884, "th": 2.5692314505185654, "zh": 2.7285948875898165, "my": 2.9734574707310357}, "t": 210.1} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.10316, "de": 2.31872, "vi": 1.96997, "tr": 2.29443, "fi": 2.49334, "ru": 1.40928, "el": 1.54022, "hi": 1.1878, "ta": 0.9912, "th": 1.15723, "zh": 3.20095, "my": 1.14599}, "bytes_per_chunk": {"en": 5.626, "de": 5.529, "vi": 5.848, "tr": 5.501, "fi": 5.408, "ru": 5.682, "el": 5.849, "hi": 6.156, "ta": 6.16, "th": 6.073, "zh": 5.347, "my": 5.342}, "bpb_mean": 1.81769, "t": 211.3} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "global", "lm_loss": 1.36681, "bpb": 1.97189, "chunk_loss": 0.030157, "F": 0.17709, "G": 0.27219, "chunks": 181.34, "bytes_per_chunk": 5.647, "boundary_entropy": 0.34936, "router_noise": 0.0, "grad_norm": 0.7792, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 211.4} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "global", "lm_loss": 1.52112, "bpb": 2.19451, "chunk_loss": 0.030253, "F": 0.17975, "G": 0.25471, "chunks": 184.06, "bytes_per_chunk": 5.563, "boundary_entropy": 0.3106, "router_noise": 0.0, "grad_norm": 0.7172, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 213.3} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "global", "lm_loss": 1.46063, "bpb": 2.10725, "chunk_loss": 0.030128, "F": 0.17603, "G": 0.25147, "chunks": 180.25, "bytes_per_chunk": 5.681, "boundary_entropy": 0.29833, "router_noise": 0.0, "grad_norm": 0.7373, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 215.3} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "global", "lm_loss": 1.41108, "bpb": 2.03576, "chunk_loss": 0.030077, "F": 0.17258, "G": 0.25208, "chunks": 176.72, "bytes_per_chunk": 5.795, "boundary_entropy": 0.31586, "router_noise": 0.0, "grad_norm": 0.7589, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 217.2} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3251.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186807, "cps": 25.398438, "mean_p": 0.219319, "boundary_entropy": 0.18884, "bpb": 2.086491}, "de": {"chunks": 4029.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196221, "cps": 31.476562, "mean_p": 0.231985, "boundary_entropy": 0.21954, "bpb": 2.29263}, "vi": {"chunks": 4190.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227285, "cps": 32.734375, "mean_p": 0.27356, "boundary_entropy": 0.383501, "bpb": 1.957911}, "tr": {"chunks": 3705.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.204662, "cps": 28.945312, "mean_p": 0.226426, "boundary_entropy": 0.243743, "bpb": 2.266603}, "fi": {"chunks": 3621.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.192494, "cps": 28.289062, "mean_p": 0.216709, "boundary_entropy": 0.207331, "bpb": 2.460444}, "ru": {"chunks": 6213.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327863, "cps": 48.539062, "mean_p": 0.303206, "boundary_entropy": 0.46643, "bpb": 1.421023}, "el": {"chunks": 6664.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.324709, "cps": 52.0625, "mean_p": 0.334342, "boundary_entropy": 0.552268, "bpb": 1.534233}, "hi": {"chunks": 7896.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.469302, "cps": 61.6875, "mean_p": 0.372774, "boundary_entropy": 0.573266, "bpb": 1.172828}, "ta": {"chunks": 8392.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.44889, "cps": 65.5625, "mean_p": 0.404556, "boundary_entropy": 0.5927, "bpb": 0.963704}, "th": {"chunks": 8223.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.511222, "cps": 64.242188, "mean_p": 0.41541, "boundary_entropy": 0.619303, "bpb": 1.123806}, "zh": {"chunks": 3056.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.51744, "cps": 23.875, "mean_p": 0.258532, "boundary_entropy": 0.308362, "bpb": 3.174053}, "my": {"chunks": 9741.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.514091, "cps": 76.101562, "mean_p": 0.400345, "boundary_entropy": 0.615482, "bpb": 1.136125}}, "bpb": {"en": 2.08649, "de": 2.29263, "vi": 1.95791, "tr": 2.2666, "fi": 2.46044, "ru": 1.42102, "el": 1.53423, "hi": 1.17283, "ta": 0.9637, "th": 1.12381, "zh": 3.17405, "my": 1.13612}, "gini_cpc": 0.216929, "gini_cps": 0.220956, "bpic": 5.563575, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.050393464494207, "vi": 1.2166846578345174, "tr": 1.0955818063739409, "fi": 1.030442569933582, "ru": 1.7550895888336366, "el": 1.7382061970879525, "hi": 2.5122289586973583, "ta": 2.402963401965976, "th": 2.7366318409763113, "zh": 2.769918927756007, "my": 2.75199295622782}, "t": 220.1} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.09275, "de": 2.3021, "vi": 1.95898, "tr": 2.27414, "fi": 2.47101, "ru": 1.39623, "el": 1.52842, "hi": 1.16791, "ta": 0.97177, "th": 1.14231, "zh": 3.17042, "my": 1.1441}, "bytes_per_chunk": {"en": 5.616, "de": 5.271, "vi": 6.295, "tr": 5.436, "fi": 5.241, "ru": 5.589, "el": 5.643, "hi": 5.481, "ta": 6.286, "th": 5.781, "zh": 5.295, "my": 5.695}, "bpb_mean": 1.80168, "t": 221.3} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "global", "lm_loss": 1.29725, "bpb": 1.87154, "chunk_loss": 0.030311, "F": 0.18024, "G": 0.27715, "chunks": 184.56, "bytes_per_chunk": 5.548, "boundary_entropy": 0.35974, "router_noise": 0.0, "grad_norm": 0.7488, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 221.3} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "global", "lm_loss": 1.45572, "bpb": 2.10015, "chunk_loss": 0.030144, "F": 0.17505, "G": 0.25428, "chunks": 179.25, "bytes_per_chunk": 5.713, "boundary_entropy": 0.30852, "router_noise": 0.0, "grad_norm": 0.7332, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 223.3} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "global", "lm_loss": 1.32424, "bpb": 1.91047, "chunk_loss": 0.030067, "F": 0.17096, "G": 0.26952, "chunks": 175.06, "bytes_per_chunk": 5.849, "boundary_entropy": 0.35, "router_noise": 0.0, "grad_norm": 0.7496, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 225.2} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "global", "lm_loss": 1.47465, "bpb": 2.12748, "chunk_loss": 0.030214, "F": 0.17865, "G": 0.24892, "chunks": 182.94, "bytes_per_chunk": 5.598, "boundary_entropy": 0.29593, "router_noise": 0.0, "grad_norm": 0.7017, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 227.2} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3247.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186577, "cps": 25.367188, "mean_p": 0.216061, "boundary_entropy": 0.177134, "bpb": 2.074651}, "de": {"chunks": 3901.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189987, "cps": 30.476562, "mean_p": 0.219786, "boundary_entropy": 0.197081, "bpb": 2.273894}, "vi": {"chunks": 4203.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22799, "cps": 32.835938, "mean_p": 0.268366, "boundary_entropy": 0.375813, "bpb": 1.949981}, "tr": {"chunks": 3709.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.204883, "cps": 28.976562, "mean_p": 0.226841, "boundary_entropy": 0.239306, "bpb": 2.255601}, "fi": {"chunks": 3607.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19175, "cps": 28.179688, "mean_p": 0.21289, "boundary_entropy": 0.195909, "bpb": 2.443141}, "ru": {"chunks": 6134.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.323694, "cps": 47.921875, "mean_p": 0.295915, "boundary_entropy": 0.455272, "bpb": 1.404163}, "el": {"chunks": 6309.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.307411, "cps": 49.289062, "mean_p": 0.316665, "boundary_entropy": 0.533879, "bpb": 1.522493}, "hi": {"chunks": 7291.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.433343, "cps": 56.960938, "mean_p": 0.368565, "boundary_entropy": 0.56678, "bpb": 1.165432}, "ta": {"chunks": 7830.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.418829, "cps": 61.171875, "mean_p": 0.386193, "boundary_entropy": 0.581849, "bpb": 0.958978}, "th": {"chunks": 7782.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.483805, "cps": 60.796875, "mean_p": 0.401088, "boundary_entropy": 0.611312, "bpb": 1.117243}, "zh": {"chunks": 3082.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.521842, "cps": 24.078125, "mean_p": 0.261963, "boundary_entropy": 0.312438, "bpb": 3.156849}, "my": {"chunks": 9846.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.519633, "cps": 76.921875, "mean_p": 0.397824, "boundary_entropy": 0.613431, "bpb": 1.124324}}, "bpb": {"en": 2.07465, "de": 2.27389, "vi": 1.94998, "tr": 2.2556, "fi": 2.44314, "ru": 1.40416, "el": 1.52249, "hi": 1.16543, "ta": 0.95898, "th": 1.11724, "zh": 3.15685, "my": 1.12432}, "gini_cpc": 0.216768, "gini_cps": 0.214045, "bpic": 5.733123, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0182756877535213, "vi": 1.22196306636432, "tr": 1.098115732068087, "fi": 1.027723034296536, "ru": 1.734907757353449, "el": 1.6476369155394561, "hi": 2.3225969794497696, "ta": 2.244802431656007, "th": 2.5930565781818733, "zh": 2.796926303902292, "my": 2.785083927035098}, "t": 230.0} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.08192, "de": 2.28456, "vi": 1.95353, "tr": 2.26126, "fi": 2.45116, "ru": 1.38115, "el": 1.51534, "hi": 1.15902, "ta": 0.96819, "th": 1.13406, "zh": 3.15381, "my": 1.13199}, "bytes_per_chunk": {"en": 5.617, "de": 5.411, "vi": 6.029, "tr": 5.414, "fi": 5.252, "ru": 5.661, "el": 5.766, "hi": 5.912, "ta": 6.638, "th": 6.093, "zh": 5.236, "my": 5.499}, "bpb_mean": 1.78967, "t": 231.3} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 231.3} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.28795, "bpb": 1.85812, "chunk_loss": 0.030208, "F": 0.17719, "G": 0.27069, "chunks": 181.44, "bytes_per_chunk": 5.644, "boundary_entropy": 0.35624, "router_noise": 0.0, "grad_norm": 0.6922, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 231.3} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.30037, "bpb": 1.87603, "chunk_loss": 0.030105, "F": 0.17212, "G": 0.26683, "chunks": 176.25, "bytes_per_chunk": 5.81, "boundary_entropy": 0.34477, "router_noise": 0.0, "grad_norm": 0.7229, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 233.3} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.34379, "bpb": 1.93868, "chunk_loss": 0.030111, "F": 0.17554, "G": 0.24666, "chunks": 179.75, "bytes_per_chunk": 5.697, "boundary_entropy": 0.28735, "router_noise": 0.0, "grad_norm": 0.7352, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 235.2} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.47235, "bpb": 2.12416, "chunk_loss": 0.030419, "F": 0.18347, "G": 0.26184, "chunks": 187.88, "bytes_per_chunk": 5.45, "boundary_entropy": 0.32289, "router_noise": 0.0, "grad_norm": 0.7678, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 237.2} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3186.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183072, "cps": 24.890625, "mean_p": 0.211499, "boundary_entropy": 0.172307, "bpb": 2.062503}, "de": {"chunks": 3812.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.185652, "cps": 29.78125, "mean_p": 0.215205, "boundary_entropy": 0.192933, "bpb": 2.267117}, "vi": {"chunks": 4306.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.233577, "cps": 33.640625, "mean_p": 0.267486, "boundary_entropy": 0.36917, "bpb": 1.939951}, "tr": {"chunks": 3625.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200243, "cps": 28.320312, "mean_p": 0.21929, "boundary_entropy": 0.229352, "bpb": 2.237294}, "fi": {"chunks": 3577.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.190155, "cps": 27.945312, "mean_p": 0.210052, "boundary_entropy": 0.192006, "bpb": 2.422936}, "ru": {"chunks": 5939.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.313404, "cps": 46.398438, "mean_p": 0.288853, "boundary_entropy": 0.448543, "bpb": 1.392414}, "el": {"chunks": 6295.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.306729, "cps": 49.179688, "mean_p": 0.312656, "boundary_entropy": 0.527812, "bpb": 1.51067}, "hi": {"chunks": 7091.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.421456, "cps": 55.398438, "mean_p": 0.365429, "boundary_entropy": 0.562799, "bpb": 1.154785}, "ta": {"chunks": 8004.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.428136, "cps": 62.53125, "mean_p": 0.399477, "boundary_entropy": 0.583033, "bpb": 0.950083}, "th": {"chunks": 7744.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.481442, "cps": 60.5, "mean_p": 0.399771, "boundary_entropy": 0.61231, "bpb": 1.112876}, "zh": {"chunks": 3045.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.515577, "cps": 23.789062, "mean_p": 0.259699, "boundary_entropy": 0.308673, "bpb": 3.151784}, "my": {"chunks": 9470.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.499789, "cps": 73.984375, "mean_p": 0.393539, "boundary_entropy": 0.610048, "bpb": 1.120784}}, "bpb": {"en": 2.0625, "de": 2.26712, "vi": 1.93995, "tr": 2.23729, "fi": 2.42294, "ru": 1.39241, "el": 1.51067, "hi": 1.15479, "ta": 0.95008, "th": 1.11288, "zh": 3.15178, "my": 1.12078}, "gini_cpc": 0.215395, "gini_cps": 0.21386, "bpic": 5.806594, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0140954485742164, "vi": 1.2758782447822732, "tr": 1.093794683758235, "fi": 1.0386886964903435, "ru": 1.7119160343653883, "el": 1.675456813736974, "hi": 2.3021348600722513, "ta": 2.3386216139848264, "th": 2.6297994353209497, "zh": 2.816256473833944, "my": 2.730014487066794}, "t": 240.0} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.07022, "de": 2.2768, "vi": 1.94204, "tr": 2.24536, "fi": 2.43272, "ru": 1.36908, "el": 1.50398, "hi": 1.14814, "ta": 0.95782, "th": 1.13076, "zh": 3.14774, "my": 1.12788}, "bytes_per_chunk": {"en": 5.717, "de": 5.515, "vi": 6.196, "tr": 5.529, "fi": 5.303, "ru": 5.802, "el": 5.842, "hi": 6.111, "ta": 6.485, "th": 6.079, "zh": 5.309, "my": 5.894}, "bpb_mean": 1.77938, "t": 241.2} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.44597, "bpb": 2.08609, "chunk_loss": 0.030112, "F": 0.17471, "G": 0.23403, "chunks": 178.91, "bytes_per_chunk": 5.724, "boundary_entropy": 0.26093, "router_noise": 0.0, "grad_norm": 0.8091, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 241.3} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.37544, "bpb": 1.98434, "chunk_loss": 0.030141, "F": 0.17636, "G": 0.24916, "chunks": 180.59, "bytes_per_chunk": 5.67, "boundary_entropy": 0.30683, "router_noise": 0.0, "grad_norm": 0.7642, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 243.2} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.30386, "bpb": 1.88108, "chunk_loss": 0.030209, "F": 0.1792, "G": 0.25302, "chunks": 183.5, "bytes_per_chunk": 5.58, "boundary_entropy": 0.29251, "router_noise": 0.0, "grad_norm": 0.7465, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 245.2} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.35242, "bpb": 1.95113, "chunk_loss": 0.030286, "F": 0.18079, "G": 0.25986, "chunks": 185.12, "bytes_per_chunk": 5.531, "boundary_entropy": 0.31773, "router_noise": 0.0, "grad_norm": 0.7451, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 247.1} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3254.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186979, "cps": 25.421875, "mean_p": 0.217687, "boundary_entropy": 0.1759, "bpb": 2.057415}, "de": {"chunks": 3963.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.193006, "cps": 30.960938, "mean_p": 0.224265, "boundary_entropy": 0.198663, "bpb": 2.250141}, "vi": {"chunks": 4393.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.238297, "cps": 34.320312, "mean_p": 0.27333, "boundary_entropy": 0.374425, "bpb": 1.931641}, "tr": {"chunks": 3766.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.208032, "cps": 29.421875, "mean_p": 0.22575, "boundary_entropy": 0.232007, "bpb": 2.226495}, "fi": {"chunks": 3630.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.192972, "cps": 28.359375, "mean_p": 0.212635, "boundary_entropy": 0.193496, "bpb": 2.411087}, "ru": {"chunks": 5938.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.313351, "cps": 46.390625, "mean_p": 0.286263, "boundary_entropy": 0.441821, "bpb": 1.382093}, "el": {"chunks": 6642.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.323637, "cps": 51.890625, "mean_p": 0.313521, "boundary_entropy": 0.523381, "bpb": 1.494911}, "hi": {"chunks": 7101.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.422051, "cps": 55.476562, "mean_p": 0.358426, "boundary_entropy": 0.563612, "bpb": 1.153584}, "ta": {"chunks": 8226.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.440011, "cps": 64.265625, "mean_p": 0.401483, "boundary_entropy": 0.587759, "bpb": 0.945951}, "th": {"chunks": 7667.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.476655, "cps": 59.898438, "mean_p": 0.39518, "boundary_entropy": 0.607276, "bpb": 1.099289}, "zh": {"chunks": 2991.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506434, "cps": 23.367188, "mean_p": 0.259109, "boundary_entropy": 0.311681, "bpb": 3.136593}, "my": {"chunks": 9231.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.487175, "cps": 72.117188, "mean_p": 0.393146, "boundary_entropy": 0.609058, "bpb": 1.112227}}, "bpb": {"en": 2.05741, "de": 2.25014, "vi": 1.93164, "tr": 2.22649, "fi": 2.41109, "ru": 1.38209, "el": 1.49491, "hi": 1.15358, "ta": 0.94595, "th": 1.09929, "zh": 3.13659, "my": 1.11223}, "gini_cpc": 0.206188, "gini_cps": 0.209455, "bpic": 5.745053, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.032234182754246, "vi": 1.2744553739454676, "tr": 1.1125930324058049, "fi": 1.032051366458777, "ru": 1.6758592874529907, "el": 1.7308705860494236, "hi": 2.257205040133483, "ta": 2.353259427818836, "th": 2.5492414395787284, "zh": 2.7085043784710723, "my": 2.605505213440232}, "t": 250.0} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.06571, "de": 2.26097, "vi": 1.93311, "tr": 2.23387, "fi": 2.42024, "ru": 1.35892, "el": 1.48773, "hi": 1.14826, "ta": 0.95302, "th": 1.11743, "zh": 3.13201, "my": 1.11981}, "bytes_per_chunk": {"en": 5.601, "de": 5.343, "vi": 5.862, "tr": 5.341, "fi": 5.235, "ru": 5.794, "el": 5.509, "hi": 6.039, "ta": 6.341, "th": 6.072, "zh": 5.418, "my": 6.039}, "bpb_mean": 1.76926, "t": 251.2} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.39705, "bpb": 2.01551, "chunk_loss": 0.030233, "F": 0.17868, "G": 0.25968, "chunks": 182.97, "bytes_per_chunk": 5.597, "boundary_entropy": 0.32452, "router_noise": 0.0, "grad_norm": 0.7544, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 251.2} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.50547, "bpb": 2.17194, "chunk_loss": 0.030121, "F": 0.17615, "G": 0.25487, "chunks": 180.38, "bytes_per_chunk": 5.677, "boundary_entropy": 0.30901, "router_noise": 0.0, "grad_norm": 0.7656, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 253.2} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.44528, "bpb": 2.08509, "chunk_loss": 0.030334, "F": 0.18042, "G": 0.26683, "chunks": 184.75, "bytes_per_chunk": 5.543, "boundary_entropy": 0.33752, "router_noise": 0.0, "grad_norm": 0.7974, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 255.1} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.46396, "bpb": 2.11205, "chunk_loss": 0.030371, "F": 0.18073, "G": 0.25609, "chunks": 185.06, "bytes_per_chunk": 5.533, "boundary_entropy": 0.28446, "router_noise": 0.0, "grad_norm": 0.764, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 257.0} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3206.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.184221, "cps": 25.046875, "mean_p": 0.215255, "boundary_entropy": 0.173382, "bpb": 2.045118}, "de": {"chunks": 3861.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188039, "cps": 30.164062, "mean_p": 0.217965, "boundary_entropy": 0.192107, "bpb": 2.239097}, "vi": {"chunks": 4228.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229346, "cps": 33.03125, "mean_p": 0.270508, "boundary_entropy": 0.373616, "bpb": 1.927575}, "tr": {"chunks": 3650.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201624, "cps": 28.515625, "mean_p": 0.220656, "boundary_entropy": 0.228179, "bpb": 2.213379}, "fi": {"chunks": 3542.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188294, "cps": 27.671875, "mean_p": 0.208472, "boundary_entropy": 0.186203, "bpb": 2.389485}, "ru": {"chunks": 6337.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.334406, "cps": 49.507812, "mean_p": 0.292453, "boundary_entropy": 0.43938, "bpb": 1.374857}, "el": {"chunks": 6696.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.326268, "cps": 52.3125, "mean_p": 0.321481, "boundary_entropy": 0.530047, "bpb": 1.48443}, "hi": {"chunks": 7737.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.459851, "cps": 60.445312, "mean_p": 0.370621, "boundary_entropy": 0.565162, "bpb": 1.135911}, "ta": {"chunks": 8220.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.43969, "cps": 64.21875, "mean_p": 0.399424, "boundary_entropy": 0.584807, "bpb": 0.935769}, "th": {"chunks": 7647.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.475412, "cps": 59.742188, "mean_p": 0.394951, "boundary_entropy": 0.608414, "bpb": 1.092745}, "zh": {"chunks": 3019.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.511175, "cps": 23.585938, "mean_p": 0.261438, "boundary_entropy": 0.312457, "bpb": 3.122337}, "my": {"chunks": 9551.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.504064, "cps": 74.617188, "mean_p": 0.393196, "boundary_entropy": 0.608249, "bpb": 1.109407}}, "bpb": {"en": 2.04512, "de": 2.2391, "vi": 1.92758, "tr": 2.21338, "fi": 2.38949, "ru": 1.37486, "el": 1.48443, "hi": 1.13591, "ta": 0.93577, "th": 1.09275, "zh": 3.12234, "my": 1.10941}, "gini_cpc": 0.214, "gini_cps": 0.217821, "bpic": 5.66935, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0207232251149412, "vi": 1.2449515173695374, "tr": 1.0944676144260665, "fi": 1.0221091487165248, "ru": 1.8152443554878137, "el": 1.771067860537402, "hi": 2.4961927997189517, "ta": 2.3867501043193116, "th": 2.5806590300941403, "zh": 2.7747909703677425, "my": 2.73618886492513}, "t": 259.9} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.05414, "de": 2.25025, "vi": 1.92849, "tr": 2.22106, "fi": 2.398, "ru": 1.35113, "el": 1.4774, "hi": 1.13103, "ta": 0.94468, "th": 1.11084, "zh": 3.1184, "my": 1.11614}, "bytes_per_chunk": {"en": 5.642, "de": 5.459, "vi": 6.248, "tr": 5.463, "fi": 5.327, "ru": 5.491, "el": 5.453, "hi": 5.622, "ta": 6.297, "th": 6.214, "zh": 5.339, "my": 5.923}, "bpb_mean": 1.75846, "t": 261.1} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.33867, "bpb": 1.93129, "chunk_loss": 0.030283, "F": 0.17889, "G": 0.25479, "chunks": 183.19, "bytes_per_chunk": 5.59, "boundary_entropy": 0.31105, "router_noise": 0.0, "grad_norm": 0.7195, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 261.2} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.28366, "bpb": 1.85192, "chunk_loss": 0.030026, "F": 0.1727, "G": 0.24973, "chunks": 176.84, "bytes_per_chunk": 5.79, "boundary_entropy": 0.29065, "router_noise": 0.0, "grad_norm": 0.7025, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 263.1} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.37861, "bpb": 1.98891, "chunk_loss": 0.030092, "F": 0.17416, "G": 0.2553, "chunks": 178.34, "bytes_per_chunk": 5.742, "boundary_entropy": 0.30909, "router_noise": 0.0, "grad_norm": 0.7951, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 265.0} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.24427, "bpb": 1.7951, "chunk_loss": 0.030214, "F": 0.17529, "G": 0.26393, "chunks": 179.5, "bytes_per_chunk": 5.705, "boundary_entropy": 0.33414, "router_noise": 0.0, "grad_norm": 0.7406, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 267.0} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3199.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183819, "cps": 24.992188, "mean_p": 0.212767, "boundary_entropy": 0.16686, "bpb": 2.040869}, "de": {"chunks": 3862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188087, "cps": 30.171875, "mean_p": 0.216976, "boundary_entropy": 0.188603, "bpb": 2.233266}, "vi": {"chunks": 4235.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229726, "cps": 33.085938, "mean_p": 0.268473, "boundary_entropy": 0.369846, "bpb": 1.918078}, "tr": {"chunks": 3551.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.196155, "cps": 27.742188, "mean_p": 0.216068, "boundary_entropy": 0.222157, "bpb": 2.208331}, "fi": {"chunks": 3516.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.186912, "cps": 27.46875, "mean_p": 0.206191, "boundary_entropy": 0.18324, "bpb": 2.37862}, "ru": {"chunks": 6092.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.321478, "cps": 47.59375, "mean_p": 0.28497, "boundary_entropy": 0.432607, "bpb": 1.365979}, "el": {"chunks": 6489.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.316182, "cps": 50.695312, "mean_p": 0.315945, "boundary_entropy": 0.525971, "bpb": 1.476459}, "hi": {"chunks": 7178.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.426627, "cps": 56.078125, "mean_p": 0.367096, "boundary_entropy": 0.563546, "bpb": 1.128051}, "ta": {"chunks": 8177.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.43739, "cps": 63.882812, "mean_p": 0.397974, "boundary_entropy": 0.58181, "bpb": 0.934022}, "th": {"chunks": 7891.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.490581, "cps": 61.648438, "mean_p": 0.40122, "boundary_entropy": 0.608319, "bpb": 1.087147}, "zh": {"chunks": 3001.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.508127, "cps": 23.445312, "mean_p": 0.259154, "boundary_entropy": 0.305445, "bpb": 3.113838}, "my": {"chunks": 9280.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.489761, "cps": 72.5, "mean_p": 0.392901, "boundary_entropy": 0.608741, "bpb": 1.10387}}, "bpb": {"en": 2.04087, "de": 2.23327, "vi": 1.91808, "tr": 2.20833, "fi": 2.37862, "ru": 1.36598, "el": 1.47646, "hi": 1.12805, "ta": 0.93402, "th": 1.08715, "zh": 3.11384, "my": 1.10387}, "gini_cpc": 0.214145, "gini_cps": 0.216095, "bpic": 5.773661, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.023221701212545, "vi": 1.2497413883661264, "tr": 1.0671119933231912, "fi": 1.016826516459803, "ru": 1.7488822117069895, "el": 1.7200727240857272, "hi": 2.320909793968862, "ta": 2.379459999409751, "th": 2.668829676821411, "zh": 2.7642825544343936, "my": 2.6643696642851302}, "t": 269.8} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.04905, "de": 2.24521, "vi": 1.91982, "tr": 2.21494, "fi": 2.38874, "ru": 1.34357, "el": 1.46871, "hi": 1.12228, "ta": 0.9417, "th": 1.10475, "zh": 3.10929, "my": 1.11034}, "bytes_per_chunk": {"en": 5.672, "de": 5.476, "vi": 6.138, "tr": 5.578, "fi": 5.366, "ru": 5.712, "el": 5.645, "hi": 6.061, "ta": 6.378, "th": 6.011, "zh": 5.373, "my": 6.016}, "bpb_mean": 1.75153, "t": 271.1} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.3572, "bpb": 1.95802, "chunk_loss": 0.030122, "F": 0.17374, "G": 0.26809, "chunks": 177.91, "bytes_per_chunk": 5.756, "boundary_entropy": 0.35098, "router_noise": 0.0, "grad_norm": 0.7219, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 271.1} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.37823, "bpb": 1.98837, "chunk_loss": 0.030142, "F": 0.17401, "G": 0.26254, "chunks": 178.19, "bytes_per_chunk": 5.747, "boundary_entropy": 0.32963, "router_noise": 0.0, "grad_norm": 0.9046, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 273.1} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.47281, "bpb": 2.12482, "chunk_loss": 0.030209, "F": 0.17737, "G": 0.24644, "chunks": 181.62, "bytes_per_chunk": 5.638, "boundary_entropy": 0.293, "router_noise": 0.0, "grad_norm": 0.7437, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 275.0} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.32666, "bpb": 1.91397, "chunk_loss": 0.030193, "F": 0.17587, "G": 0.24798, "chunks": 180.09, "bytes_per_chunk": 5.686, "boundary_entropy": 0.29521, "router_noise": 0.0, "grad_norm": 0.7268, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 276.9} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3178.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.182612, "cps": 24.828125, "mean_p": 0.209437, "boundary_entropy": 0.161242, "bpb": 2.03431}, "de": {"chunks": 3830.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186529, "cps": 29.921875, "mean_p": 0.213915, "boundary_entropy": 0.183529, "bpb": 2.226099}, "vi": {"chunks": 4196.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227611, "cps": 32.78125, "mean_p": 0.261074, "boundary_entropy": 0.357794, "bpb": 1.915217}, "tr": {"chunks": 3532.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195106, "cps": 27.59375, "mean_p": 0.214373, "boundary_entropy": 0.219843, "bpb": 2.196929}, "fi": {"chunks": 3525.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18739, "cps": 27.539062, "mean_p": 0.205398, "boundary_entropy": 0.179115, "bpb": 2.36826}, "ru": {"chunks": 6034.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.318417, "cps": 47.140625, "mean_p": 0.28126, "boundary_entropy": 0.428114, "bpb": 1.360916}, "el": {"chunks": 5684.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.276958, "cps": 44.40625, "mean_p": 0.295984, "boundary_entropy": 0.512188, "bpb": 1.475328}, "hi": {"chunks": 7248.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.430788, "cps": 56.625, "mean_p": 0.358796, "boundary_entropy": 0.556145, "bpb": 1.121296}, "ta": {"chunks": 8344.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.446323, "cps": 65.1875, "mean_p": 0.392955, "boundary_entropy": 0.577253, "bpb": 0.929392}, "th": {"chunks": 7710.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.479329, "cps": 60.234375, "mean_p": 0.39566, "boundary_entropy": 0.606508, "bpb": 1.078898}, "zh": {"chunks": 2994.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506942, "cps": 23.390625, "mean_p": 0.260461, "boundary_entropy": 0.308705, "bpb": 3.107807}, "my": {"chunks": 8084.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.426641, "cps": 63.15625, "mean_p": 0.383992, "boundary_entropy": 0.604898, "bpb": 1.108206}}, "bpb": {"en": 2.03431, "de": 2.2261, "vi": 1.91522, "tr": 2.19693, "fi": 2.36826, "ru": 1.36092, "el": 1.47533, "hi": 1.1213, "ta": 0.92939, "th": 1.0789, "zh": 3.10781, "my": 1.10821}, "gini_cpc": 0.212014, "gini_cps": 0.205849, "bpic": 5.963129, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0214487801904297, "vi": 1.246414707296553, "tr": 1.0684159696349123, "fi": 1.0261656317676826, "ru": 1.743678123510746, "el": 1.5166433018847878, "hi": 2.359029322132005, "ta": 2.444100462678777, "th": 2.6248442578005102, "zh": 2.776058288474543, "my": 2.3363244615152223}, "t": 279.8} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.04349, "de": 2.2368, "vi": 1.9163, "tr": 2.20419, "fi": 2.37705, "ru": 1.33894, "el": 1.46652, "hi": 1.11572, "ta": 0.93806, "th": 1.0975, "zh": 3.10345, "my": 1.11537}, "bytes_per_chunk": {"en": 5.705, "de": 5.494, "vi": 6.229, "tr": 5.639, "fi": 5.379, "ru": 5.755, "el": 6.507, "hi": 5.953, "ta": 6.202, "th": 6.185, "zh": 5.373, "my": 6.808}, "bpb_mean": 1.74612, "t": 281.0} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.24051, "bpb": 1.78967, "chunk_loss": 0.030047, "F": 0.17322, "G": 0.26948, "chunks": 177.38, "bytes_per_chunk": 5.773, "boundary_entropy": 0.34444, "router_noise": 0.0, "grad_norm": 0.7765, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 281.1} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.28959, "bpb": 1.86049, "chunk_loss": 0.030169, "F": 0.17648, "G": 0.26261, "chunks": 180.72, "bytes_per_chunk": 5.666, "boundary_entropy": 0.32876, "router_noise": 0.0, "grad_norm": 0.7163, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 283.0} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.22138, "bpb": 1.76208, "chunk_loss": 0.030148, "F": 0.1741, "G": 0.27326, "chunks": 178.28, "bytes_per_chunk": 5.744, "boundary_entropy": 0.35249, "router_noise": 0.0, "grad_norm": 0.7234, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 284.9} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.44464, "bpb": 2.08418, "chunk_loss": 0.030292, "F": 0.18137, "G": 0.26362, "chunks": 185.72, "bytes_per_chunk": 5.514, "boundary_entropy": 0.32808, "router_noise": 0.0, "grad_norm": 0.7881, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 286.9} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3195.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183589, "cps": 24.960938, "mean_p": 0.210425, "boundary_entropy": 0.157595, "bpb": 2.031131}, "de": {"chunks": 3839.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186967, "cps": 29.992188, "mean_p": 0.21281, "boundary_entropy": 0.176713, "bpb": 2.221601}, "vi": {"chunks": 4209.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228316, "cps": 32.882812, "mean_p": 0.266061, "boundary_entropy": 0.364103, "bpb": 1.91029}, "tr": {"chunks": 3525.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.194719, "cps": 27.539062, "mean_p": 0.213991, "boundary_entropy": 0.214787, "bpb": 2.190239}, "fi": {"chunks": 3540.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188188, "cps": 27.65625, "mean_p": 0.206177, "boundary_entropy": 0.176012, "bpb": 2.358224}, "ru": {"chunks": 6054.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319472, "cps": 47.296875, "mean_p": 0.279245, "boundary_entropy": 0.42518, "bpb": 1.353099}, "el": {"chunks": 6044.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.294499, "cps": 47.21875, "mean_p": 0.299414, "boundary_entropy": 0.510104, "bpb": 1.465683}, "hi": {"chunks": 7188.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.427221, "cps": 56.15625, "mean_p": 0.353652, "boundary_entropy": 0.549676, "bpb": 1.118157}, "ta": {"chunks": 8484.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.453811, "cps": 66.28125, "mean_p": 0.396955, "boundary_entropy": 0.575656, "bpb": 0.922331}, "th": {"chunks": 7666.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.476593, "cps": 59.890625, "mean_p": 0.396381, "boundary_entropy": 0.605141, "bpb": 1.072792}, "zh": {"chunks": 2994.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506942, "cps": 23.390625, "mean_p": 0.256979, "boundary_entropy": 0.303666, "bpb": 3.095876}, "my": {"chunks": 8737.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.461104, "cps": 68.257812, "mean_p": 0.390328, "boundary_entropy": 0.607203, "bpb": 1.098197}}, "bpb": {"en": 2.03113, "de": 2.2216, "vi": 1.91029, "tr": 2.19024, "fi": 2.35822, "ru": 1.3531, "el": 1.46568, "hi": 1.11816, "ta": 0.92233, "th": 1.07279, "zh": 3.09588, "my": 1.0982}, "gini_cpc": 0.212318, "gini_cps": 0.211889, "bpic": 5.861489, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0184013413424262, "vi": 1.2436238477788346, "tr": 1.0606249217116763, "fi": 1.025049023608452, "ru": 1.7401490950982943, "el": 1.6041200540147444, "hi": 2.32705288491723, "ta": 2.471886058259156, "th": 2.595977998339222, "zh": 2.7612873993026907, "my": 2.5116100798726633}, "t": 289.7} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.03945, "de": 2.23163, "vi": 1.91067, "tr": 2.19656, "fi": 2.36771, "ru": 1.33059, "el": 1.45773, "hi": 1.11109, "ta": 0.93051, "th": 1.09126, "zh": 3.09208, "my": 1.10602}, "bytes_per_chunk": {"en": 5.675, "de": 5.505, "vi": 6.112, "tr": 5.661, "fi": 5.346, "ru": 5.711, "el": 6.123, "hi": 6.003, "ta": 6.135, "th": 6.227, "zh": 5.387, "my": 6.353}, "bpb_mean": 1.73878, "t": 291.0} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.34549, "bpb": 1.94113, "chunk_loss": 0.030302, "F": 0.18054, "G": 0.26217, "chunks": 184.88, "bytes_per_chunk": 5.539, "boundary_entropy": 0.31279, "router_noise": 0.0, "grad_norm": 0.8237, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 291.0} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.20081, "bpb": 1.73241, "chunk_loss": 0.030299, "F": 0.17883, "G": 0.26916, "chunks": 183.12, "bytes_per_chunk": 5.592, "boundary_entropy": 0.35552, "router_noise": 0.0, "grad_norm": 0.7106, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 293.0} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.30776, "bpb": 1.8867, "chunk_loss": 0.030204, "F": 0.17822, "G": 0.25564, "chunks": 182.5, "bytes_per_chunk": 5.611, "boundary_entropy": 0.30727, "router_noise": 0.0, "grad_norm": 0.7739, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 294.9} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.38308, "bpb": 1.99536, "chunk_loss": 0.03022, "F": 0.17969, "G": 0.25072, "chunks": 184.0, "bytes_per_chunk": 5.565, "boundary_entropy": 0.28986, "router_noise": 0.0, "grad_norm": 0.7967, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 296.9} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3230.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.1856, "cps": 25.234375, "mean_p": 0.214196, "boundary_entropy": 0.162484, "bpb": 2.026254}, "de": {"chunks": 3923.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.191058, "cps": 30.648438, "mean_p": 0.218763, "boundary_entropy": 0.182712, "bpb": 2.214038}, "vi": {"chunks": 4302.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23336, "cps": 33.609375, "mean_p": 0.270129, "boundary_entropy": 0.366366, "bpb": 1.90552}, "tr": {"chunks": 3671.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.202784, "cps": 28.679688, "mean_p": 0.21985, "boundary_entropy": 0.218494, "bpb": 2.181351}, "fi": {"chunks": 3587.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.190686, "cps": 28.023438, "mean_p": 0.209182, "boundary_entropy": 0.178809, "bpb": 2.353667}, "ru": {"chunks": 6294.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.332137, "cps": 49.171875, "mean_p": 0.283548, "boundary_entropy": 0.423571, "bpb": 1.347057}, "el": {"chunks": 6290.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.306485, "cps": 49.140625, "mean_p": 0.303848, "boundary_entropy": 0.511392, "bpb": 1.456793}, "hi": {"chunks": 7472.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.444101, "cps": 58.375, "mean_p": 0.353852, "boundary_entropy": 0.545941, "bpb": 1.106088}, "ta": {"chunks": 8458.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.45242, "cps": 66.078125, "mean_p": 0.39281, "boundary_entropy": 0.574362, "bpb": 0.918443}, "th": {"chunks": 8052.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.500591, "cps": 62.90625, "mean_p": 0.398442, "boundary_entropy": 0.60233, "bpb": 1.067558}, "zh": {"chunks": 3028.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.512699, "cps": 23.65625, "mean_p": 0.259762, "boundary_entropy": 0.301876, "bpb": 3.092458}, "my": {"chunks": 9281.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.489814, "cps": 72.507812, "mean_p": 0.385473, "boundary_entropy": 0.602542, "bpb": 1.092837}}, "bpb": {"en": 2.02625, "de": 2.21404, "vi": 1.90552, "tr": 2.18135, "fi": 2.35367, "ru": 1.34706, "el": 1.45679, "hi": 1.10609, "ta": 0.91844, "th": 1.06756, "zh": 3.09246, "my": 1.09284}, "gini_cpc": 0.214111, "gini_cps": 0.21601, "bpic": 5.678242, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0294079047260476, "vi": 1.2573287955925807, "tr": 1.0925854958956054, "fi": 1.027403618823253, "ru": 1.789530571734318, "el": 1.6513205979427446, "hi": 2.392783405940776, "ta": 2.4376076780848175, "th": 2.697145023871826, "zh": 2.7623838485079455, "my": 2.639082667832641}, "t": 299.7} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.03404, "de": 2.22344, "vi": 1.90664, "tr": 2.18919, "fi": 2.3624, "ru": 1.32527, "el": 1.44976, "hi": 1.10115, "ta": 0.92748, "th": 1.08553, "zh": 3.08735, "my": 1.09894}, "bytes_per_chunk": {"en": 5.622, "de": 5.396, "vi": 6.002, "tr": 5.428, "fi": 5.29, "ru": 5.507, "el": 5.78, "hi": 5.776, "ta": 6.125, "th": 5.881, "zh": 5.328, "my": 5.921}, "bpb_mean": 1.7326, "t": 300.9} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.32393, "bpb": 1.91002, "chunk_loss": 0.030195, "F": 0.17761, "G": 0.26287, "chunks": 181.88, "bytes_per_chunk": 5.63, "boundary_entropy": 0.31973, "router_noise": 0.0, "grad_norm": 0.7389, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 301.0} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.31021, "bpb": 1.89023, "chunk_loss": 0.030189, "F": 0.17422, "G": 0.25587, "chunks": 178.41, "bytes_per_chunk": 5.74, "boundary_entropy": 0.31903, "router_noise": 0.0, "grad_norm": 0.7931, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 302.9} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.2849, "bpb": 1.85373, "chunk_loss": 0.030094, "F": 0.17313, "G": 0.24317, "chunks": 177.28, "bytes_per_chunk": 5.776, "boundary_entropy": 0.28463, "router_noise": 0.0, "grad_norm": 0.7636, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 304.9} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.19192, "bpb": 1.71958, "chunk_loss": 0.030115, "F": 0.1749, "G": 0.26563, "chunks": 179.09, "bytes_per_chunk": 5.718, "boundary_entropy": 0.32705, "router_noise": 0.0, "grad_norm": 0.7138, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 306.8} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3209.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.184393, "cps": 25.070312, "mean_p": 0.212032, "boundary_entropy": 0.160564, "bpb": 2.0217}, "de": {"chunks": 3880.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188964, "cps": 30.3125, "mean_p": 0.215346, "boundary_entropy": 0.179626, "bpb": 2.209773}, "vi": {"chunks": 4326.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234662, "cps": 33.796875, "mean_p": 0.268778, "boundary_entropy": 0.36428, "bpb": 1.901711}, "tr": {"chunks": 3568.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.197094, "cps": 27.875, "mean_p": 0.215417, "boundary_entropy": 0.215431, "bpb": 2.178413}, "fi": {"chunks": 3553.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188879, "cps": 27.757812, "mean_p": 0.206611, "boundary_entropy": 0.174941, "bpb": 2.344191}, "ru": {"chunks": 6148.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.324433, "cps": 48.03125, "mean_p": 0.282346, "boundary_entropy": 0.423768, "bpb": 1.344004}, "el": {"chunks": 6375.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.310627, "cps": 49.804688, "mean_p": 0.306974, "boundary_entropy": 0.51323, "bpb": 1.451712}, "hi": {"chunks": 7520.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.446954, "cps": 58.75, "mean_p": 0.362367, "boundary_entropy": 0.554066, "bpb": 1.104391}, "ta": {"chunks": 8423.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.450548, "cps": 65.804688, "mean_p": 0.393575, "boundary_entropy": 0.575083, "bpb": 0.914421}, "th": {"chunks": 7870.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.489276, "cps": 61.484375, "mean_p": 0.394394, "boundary_entropy": 0.602381, "bpb": 1.064271}, "zh": {"chunks": 3027.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.51253, "cps": 23.648438, "mean_p": 0.259165, "boundary_entropy": 0.303582, "bpb": 3.082743}, "my": {"chunks": 8270.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.436458, "cps": 64.609375, "mean_p": 0.377319, "boundary_entropy": 0.599106, "bpb": 1.093983}}, "bpb": {"en": 2.0217, "de": 2.20977, "vi": 1.90171, "tr": 2.17841, "fi": 2.34419, "ru": 1.344, "el": 1.45171, "hi": 1.10439, "ta": 0.91442, "th": 1.06427, "zh": 3.08274, "my": 1.09398}, "gini_cpc": 0.210448, "gini_cps": 0.206782, "bpic": 5.800012, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0247872690765545, "vi": 1.2726171637387829, "tr": 1.0688793758274622, "fi": 1.0243248933450104, "ru": 1.7594585807890242, "el": 1.684588171997801, "hi": 2.423913797422461, "ta": 2.4434065530625575, "th": 2.6534326589432715, "zh": 2.779542900053471, "my": 2.366990618334951}, "t": 309.7} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.02986, "de": 2.22002, "vi": 1.90274, "tr": 2.18534, "fi": 2.35355, "ru": 1.32241, "el": 1.44364, "hi": 1.09835, "ta": 0.92381, "th": 1.08196, "zh": 3.07762, "my": 1.10027}, "bytes_per_chunk": {"en": 5.645, "de": 5.46, "vi": 5.96, "tr": 5.585, "fi": 5.33, "ru": 5.639, "el": 5.758, "hi": 5.802, "ta": 6.233, "th": 6.013, "zh": 5.333, "my": 6.427}, "bpb_mean": 1.7283, "t": 310.9} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.2755, "bpb": 1.84016, "chunk_loss": 0.030176, "F": 0.17764, "G": 0.26346, "chunks": 181.91, "bytes_per_chunk": 5.629, "boundary_entropy": 0.31812, "router_noise": 0.0, "grad_norm": 0.7988, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.083, "alpha": 0.03, "t": 311.0} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.26667, "bpb": 1.82743, "chunk_loss": 0.030231, "F": 0.17813, "G": 0.26432, "chunks": 182.41, "bytes_per_chunk": 5.614, "boundary_entropy": 0.33147, "router_noise": 0.0, "grad_norm": 0.7361, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 312.9} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.34609, "bpb": 1.942, "chunk_loss": 0.030212, "F": 0.17633, "G": 0.26285, "chunks": 180.56, "bytes_per_chunk": 5.671, "boundary_entropy": 0.33253, "router_noise": 0.0, "grad_norm": 0.8129, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 314.9} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.38565, "bpb": 1.99907, "chunk_loss": 0.030212, "F": 0.17984, "G": 0.24955, "chunks": 184.16, "bytes_per_chunk": 5.56, "boundary_entropy": 0.30299, "router_noise": 0.0, "grad_norm": 0.7847, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 316.8} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3196.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183646, "cps": 24.96875, "mean_p": 0.211454, "boundary_entropy": 0.158796, "bpb": 2.017849}, "de": {"chunks": 3870.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188477, "cps": 30.234375, "mean_p": 0.216321, "boundary_entropy": 0.181637, "bpb": 2.20711}, "vi": {"chunks": 4398.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.238568, "cps": 34.359375, "mean_p": 0.269422, "boundary_entropy": 0.363053, "bpb": 1.898096}, "tr": {"chunks": 3539.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195492, "cps": 27.648438, "mean_p": 0.214653, "boundary_entropy": 0.214337, "bpb": 2.174232}, "fi": {"chunks": 3542.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188294, "cps": 27.671875, "mean_p": 0.206026, "boundary_entropy": 0.174449, "bpb": 2.346948}, "ru": {"chunks": 6226.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328549, "cps": 48.640625, "mean_p": 0.282357, "boundary_entropy": 0.420071, "bpb": 1.340354}, "el": {"chunks": 6687.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.32583, "cps": 52.242188, "mean_p": 0.312556, "boundary_entropy": 0.514591, "bpb": 1.444363}, "hi": {"chunks": 7239.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.430253, "cps": 56.554688, "mean_p": 0.352122, "boundary_entropy": 0.548444, "bpb": 1.103937}, "ta": {"chunks": 8302.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.444076, "cps": 64.859375, "mean_p": 0.391645, "boundary_entropy": 0.576094, "bpb": 0.910949}, "th": {"chunks": 7822.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.486292, "cps": 61.109375, "mean_p": 0.395604, "boundary_entropy": 0.60396, "bpb": 1.0617}, "zh": {"chunks": 3032.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.513376, "cps": 23.6875, "mean_p": 0.256037, "boundary_entropy": 0.299034, "bpb": 3.079185}, "my": {"chunks": 10027.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.529185, "cps": 78.335938, "mean_p": 0.396598, "boundary_entropy": 0.606604, "bpb": 1.086671}}, "bpb": {"en": 2.01785, "de": 2.20711, "vi": 1.8981, "tr": 2.17423, "fi": 2.34695, "ru": 1.34035, "el": 1.44436, "hi": 1.10394, "ta": 0.91095, "th": 1.0617, "zh": 3.07919, "my": 1.08667}, "gini_cpc": 0.218657, "gini_cps": 0.222996, "bpic": 5.653816, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.026303730244796, "vi": 1.2990606647243146, "tr": 1.064504154609104, "fi": 1.02530723741714, "ru": 1.7890284689635132, "el": 1.7742214586680944, "hi": 2.3428304139461105, "ta": 2.41810196015948, "th": 2.647976313257596, "zh": 2.795458851163006, "my": 2.881542228212752}, "t": 319.7} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.02704, "de": 2.21819, "vi": 1.89906, "tr": 2.18004, "fi": 2.35648, "ru": 1.31853, "el": 1.43686, "hi": 1.09654, "ta": 0.92087, "th": 1.07883, "zh": 3.07361, "my": 1.09346}, "bytes_per_chunk": {"en": 5.648, "de": 5.446, "vi": 5.849, "tr": 5.632, "fi": 5.348, "ru": 5.566, "el": 5.549, "hi": 5.964, "ta": 6.321, "th": 6.07, "zh": 5.318, "my": 5.44}, "bpb_mean": 1.72496, "t": 320.9} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3196.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.183646, "cps": 24.96875, "mean_p": 0.211454, "boundary_entropy": 0.158796, "bpb": 2.017849}, "de": {"chunks": 3870.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.188477, "cps": 30.234375, "mean_p": 0.216321, "boundary_entropy": 0.181637, "bpb": 2.20711}, "vi": {"chunks": 4398.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.238568, "cps": 34.359375, "mean_p": 0.269422, "boundary_entropy": 0.363053, "bpb": 1.898096}, "tr": {"chunks": 3539.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195492, "cps": 27.648438, "mean_p": 0.214653, "boundary_entropy": 0.214337, "bpb": 2.174232}, "fi": {"chunks": 3542.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188294, "cps": 27.671875, "mean_p": 0.206026, "boundary_entropy": 0.174449, "bpb": 2.346948}, "ru": {"chunks": 6226.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328549, "cps": 48.640625, "mean_p": 0.282357, "boundary_entropy": 0.420071, "bpb": 1.340354}, "el": {"chunks": 6687.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.32583, "cps": 52.242188, "mean_p": 0.312556, "boundary_entropy": 0.514591, "bpb": 1.444363}, "hi": {"chunks": 7239.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.430253, "cps": 56.554688, "mean_p": 0.352122, "boundary_entropy": 0.548444, "bpb": 1.103937}, "ta": {"chunks": 8302.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.444076, "cps": 64.859375, "mean_p": 0.391645, "boundary_entropy": 0.576094, "bpb": 0.910949}, "th": {"chunks": 7822.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.486292, "cps": 61.109375, "mean_p": 0.395604, "boundary_entropy": 0.60396, "bpb": 1.0617}, "zh": {"chunks": 3032.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.513376, "cps": 23.6875, "mean_p": 0.256037, "boundary_entropy": 0.299034, "bpb": 3.079185}, "my": {"chunks": 10027.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.529185, "cps": 78.335938, "mean_p": 0.396598, "boundary_entropy": 0.606604, "bpb": 1.086671}}, "bpb": {"en": 2.01785, "de": 2.20711, "vi": 1.8981, "tr": 2.17423, "fi": 2.34695, "ru": 1.34035, "el": 1.44436, "hi": 1.10394, "ta": 0.91095, "th": 1.0617, "zh": 3.07919, "my": 1.08667}, "gini_cpc": 0.218657, "gini_cps": 0.222996, "bpic": 5.653816, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.026303730244796, "vi": 1.2990606647243146, "tr": 1.064504154609104, "fi": 1.02530723741714, "ru": 1.7890284689635132, "el": 1.7742214586680944, "hi": 2.3428304139461105, "ta": 2.41810196015948, "th": 2.647976313257596, "zh": 2.795458851163006, "my": 2.881542228212752}, "t": 321.9} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_A_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_A_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_A_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_A_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_A_s2/th0.70"], "t": 324.8} |
| {"event": "final", "run": "pilot_A_s2", "condition": "A", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.1836464977302764, "cps": 24.96875, "bpb": 2.0366536574759873, "bytes_per_chunk": 5.642116137919159}, "de": {"cpc": 0.1884770856669751, "cps": 30.234375, "bpb": 2.2085573468042186, "bytes_per_chunk": 5.45974090890157}, "vi": {"cpc": 0.2385679414157852, "cps": 34.359375, "bpb": 1.9038478483110506, "bytes_per_chunk": 5.867275453792609}, "tr": {"cpc": 0.1954924598132906, "cps": 27.6484375, "bpb": 2.1745089955083285, "bytes_per_chunk": 5.644971306136072}, "fi": {"cpc": 0.18829408324916272, "cps": 27.671875, "bpb": 2.3741402975011927, "bytes_per_chunk": 5.352663120603579}, "ru": {"cpc": 0.32854881266490765, "cps": 48.640625, "bpb": 1.3188029527823422, "bytes_per_chunk": 5.544969963617904}, "el": {"cpc": 0.3258295570822979, "cps": 52.2421875, "bpb": 1.4323273247556754, "bytes_per_chunk": 5.576997947005074}, "hi": {"cpc": 0.4302526002971768, "cps": 56.5546875, "bpb": 1.104362611684646, "bytes_per_chunk": 5.94869234696772}, "ta": {"cpc": 0.4440759561380048, "cps": 64.859375, "bpb": 0.9217435379199496, "bytes_per_chunk": 6.291934186998212}, "th": {"cpc": 0.4862915760024868, "cps": 61.109375, "bpb": 1.0760360586894675, "bytes_per_chunk": 6.057351480000924}, "zh": {"cpc": 0.513376227565188, "cps": 23.6875, "bpb": 3.080435111671317, "bytes_per_chunk": 5.348949671995674}, "my": {"cpc": 0.5291851382731687, "cps": 78.3359375, "bpb": 1.1029825906473758, "bytes_per_chunk": 5.396852193068237}}, "gini_cpc": 0.2186565677816319, "gini_cps": 0.2229964643488509, "bpic": 5.653815556865056, "seed": 2, "mean_boundary_entropy": 0.328884482383728, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 327.1} |
|
|
