| {"event": "start", "run": "pilot_C_s2", "condition": "C", "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": "parity_a", "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": 1.7} |
| {"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 1.7} |
| {"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": 3.4} |
| {"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": 7.8723, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 3.8} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10920.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.627478, "cps": 85.3125, "mean_p": 0.513205, "boundary_entropy": 0.686301, "bpb": 8.100817}, "de": {"chunks": 16284.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.793065, "cps": 127.21875, "mean_p": 0.524587, "boundary_entropy": 0.686399, "bpb": 8.104455}, "vi": {"chunks": 5101.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276702, "cps": 39.851562, "mean_p": 0.481628, "boundary_entropy": 0.687279, "bpb": 7.854743}, "tr": {"chunks": 8206.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.453295, "cps": 64.109375, "mean_p": 0.498086, "boundary_entropy": 0.686828, "bpb": 8.081298}, "fi": {"chunks": 13520.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.718728, "cps": 105.625, "mean_p": 0.517086, "boundary_entropy": 0.686669, "bpb": 8.042881}, "ru": {"chunks": 33468.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.766121, "cps": 261.46875, "mean_p": 0.530333, "boundary_entropy": 0.688547, "bpb": 8.195805}, "el": {"chunks": 20676.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.007455, "cps": 161.53125, "mean_p": 0.503625, "boundary_entropy": 0.690112, "bpb": 8.279663}, "hi": {"chunks": 885.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.0526, "cps": 6.914062, "mean_p": 0.462026, "boundary_entropy": 0.687437, "bpb": 8.016398}, "ta": {"chunks": 1279.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.068414, "cps": 9.992188, "mean_p": 0.477597, "boundary_entropy": 0.690061, "bpb": 8.085491}, "th": {"chunks": 3638.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.226173, "cps": 28.421875, "mean_p": 0.484856, "boundary_entropy": 0.690398, "bpb": 7.718626}, "zh": {"chunks": 11708.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.982391, "cps": 91.46875, "mean_p": 0.520268, "boundary_entropy": 0.685623, "bpb": 8.126799}, "my": {"chunks": 1618.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085392, "cps": 12.640625, "mean_p": 0.474595, "boundary_entropy": 0.689681, "bpb": 7.800315}}, "bpb": {"en": 8.10082, "de": 8.10446, "vi": 7.85474, "tr": 8.0813, "fi": 8.04288, "ru": 8.19581, "el": 8.27966, "hi": 8.0164, "ta": 8.08549, "th": 7.71863, "zh": 8.1268, "my": 7.80031}, "gini_cpc": 0.489134, "gini_cps": 0.463195, "bpic": 3.014705, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.263892592091102, "vi": 0.44097468980707394, "tr": 0.7224078274926201, "fi": 1.1454240300128344, "ru": 2.8146346371306796, "el": 1.6055622932833866, "hi": 0.08382811096778409, "ta": 0.10903013675278987, "th": 0.3604484076694211, "zh": 3.1592991668000137, "my": 0.13608699459712761}, "t": 5.1} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10901.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.626386, "cps": 85.164062, "mean_p": 0.513134, "boundary_entropy": 0.686312, "bpb": 8.052614}, "de": {"chunks": 16259.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.791847, "cps": 127.023438, "mean_p": 0.524502, "boundary_entropy": 0.686414, "bpb": 8.060213}, "vi": {"chunks": 5090.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276105, "cps": 39.765625, "mean_p": 0.481573, "boundary_entropy": 0.687273, "bpb": 7.831312}, "tr": {"chunks": 8250.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.455726, "cps": 64.453125, "mean_p": 0.498253, "boundary_entropy": 0.686833, "bpb": 8.048602}, "fi": {"chunks": 13500.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.717665, "cps": 105.46875, "mean_p": 0.516943, "boundary_entropy": 0.686683, "bpb": 7.99809}, "ru": {"chunks": 33462.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.765805, "cps": 261.421875, "mean_p": 0.530286, "boundary_entropy": 0.688551, "bpb": 8.105242}, "el": {"chunks": 20634.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.005409, "cps": 161.203125, "mean_p": 0.503586, "boundary_entropy": 0.690113, "bpb": 8.273511}, "hi": {"chunks": 903.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.05367, "cps": 7.054688, "mean_p": 0.462269, "boundary_entropy": 0.687476, "bpb": 7.991116}, "ta": {"chunks": 1317.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.070447, "cps": 10.289062, "mean_p": 0.477817, "boundary_entropy": 0.690082, "bpb": 8.069114}, "th": {"chunks": 3695.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.229717, "cps": 28.867188, "mean_p": 0.484972, "boundary_entropy": 0.690405, "bpb": 7.703154}, "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": 8.118736}, "my": {"chunks": 1617.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085339, "cps": 12.632812, "mean_p": 0.47459, "boundary_entropy": 0.689681, "bpb": 7.800299}}, "bpb": {"en": 8.05261, "de": 8.06021, "vi": 7.83131, "tr": 8.0486, "fi": 7.99809, "ru": 8.10524, "el": 8.27351, "hi": 7.99112, "ta": 8.06911, "th": 7.70315, "zh": 8.11874, "my": 7.8003}, "gini_cpc": 0.487982, "gini_cps": 0.462088, "bpic": 3.014468, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2641517332468664, "vi": 0.4407906979070231, "tr": 0.7275472067477093, "fi": 1.14572309000811, "ru": 2.819034955762815, "el": 1.6050935945402014, "hi": 0.08568217019499962, "ta": 0.11246518086543142, "th": 0.36673398535786383, "zh": 3.158588535001138, "my": 0.13623993394636594}, "t": 6.6} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10855.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.623743, "cps": 84.804688, "mean_p": 0.512764, "boundary_entropy": 0.686355, "bpb": 7.898113}, "de": {"chunks": 16230.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.790435, "cps": 126.796875, "mean_p": 0.52411, "boundary_entropy": 0.686474, "bpb": 7.921951}, "vi": {"chunks": 5064.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.274695, "cps": 39.5625, "mean_p": 0.481477, "boundary_entropy": 0.687264, "bpb": 7.753067}, "tr": {"chunks": 8385.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.463183, "cps": 65.507812, "mean_p": 0.49873, "boundary_entropy": 0.686848, "bpb": 7.941147}, "fi": {"chunks": 13395.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.712083, "cps": 104.648438, "mean_p": 0.51636, "boundary_entropy": 0.686734, "bpb": 7.856487}, "ru": {"chunks": 33405.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.762797, "cps": 260.976562, "mean_p": 0.530057, "boundary_entropy": 0.68857, "bpb": 7.817019}, "el": {"chunks": 20291.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.988696, "cps": 158.523438, "mean_p": 0.503179, "boundary_entropy": 0.690117, "bpb": 8.21817}, "hi": {"chunks": 994.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.059079, "cps": 7.765625, "mean_p": 0.463112, "boundary_entropy": 0.687609, "bpb": 7.901692}, "ta": {"chunks": 1469.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.078577, "cps": 11.476562, "mean_p": 0.47867, "boundary_entropy": 0.69016, "bpb": 7.989557}, "th": {"chunks": 3943.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.245135, "cps": 30.804688, "mean_p": 0.485304, "boundary_entropy": 0.690422, "bpb": 7.620845}, "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": 8.087994}, "my": {"chunks": 1606.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.084758, "cps": 12.546875, "mean_p": 0.474499, "boundary_entropy": 0.689669, "bpb": 7.793743}}, "bpb": {"en": 7.89811, "de": 7.92195, "vi": 7.75307, "tr": 7.94115, "fi": 7.85649, "ru": 7.81702, "el": 8.21817, "hi": 7.90169, "ta": 7.98956, "th": 7.62084, "zh": 8.08799, "my": 7.79374}, "gini_cpc": 0.483061, "gini_cps": 0.456892, "bpic": 3.016056, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2672444710061388, "vi": 0.44039750254763677, "tr": 0.7425860867338956, "fi": 1.1416293537981796, "ru": 2.826158756163338, "el": 1.5851008472973431, "hi": 0.09471649180778915, "ta": 0.12597681342316705, "th": 0.39300674517157314, "zh": 3.1513428949753726, "my": 0.13588654522202245}, "t": 8.3} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10127.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.581911, "cps": 79.117188, "mean_p": 0.509405, "boundary_entropy": 0.686568, "bpb": 7.422846}, "de": {"chunks": 15618.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.760629, "cps": 122.015625, "mean_p": 0.520556, "boundary_entropy": 0.686863, "bpb": 7.489078}, "vi": {"chunks": 4867.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.264009, "cps": 38.023438, "mean_p": 0.480771, "boundary_entropy": 0.6872, "bpb": 7.511513}, "tr": {"chunks": 8656.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.478153, "cps": 67.625, "mean_p": 0.499751, "boundary_entropy": 0.686899, "bpb": 7.58961}, "fi": {"chunks": 12504.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.664717, "cps": 97.6875, "mean_p": 0.512524, "boundary_entropy": 0.686991, "bpb": 7.426468}, "ru": {"chunks": 33185.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.751187, "cps": 259.257812, "mean_p": 0.529042, "boundary_entropy": 0.688661, "bpb": 7.19965}, "el": {"chunks": 18816.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.916825, "cps": 147.0, "mean_p": 0.501351, "boundary_entropy": 0.69013, "bpb": 7.947672}, "hi": {"chunks": 1283.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.076256, "cps": 10.023438, "mean_p": 0.465352, "boundary_entropy": 0.687949, "bpb": 7.609162}, "ta": {"chunks": 2106.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.11265, "cps": 16.453125, "mean_p": 0.480882, "boundary_entropy": 0.690349, "bpb": 7.734878}, "th": {"chunks": 4618.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.2871, "cps": 36.078125, "mean_p": 0.486152, "boundary_entropy": 0.690464, "bpb": 7.340397}, "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": 7.997354}, "my": {"chunks": 1471.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.077634, "cps": 11.492188, "mean_p": 0.474003, "boundary_entropy": 0.689621, "bpb": 7.66289}}, "bpb": {"en": 7.42285, "de": 7.48908, "vi": 7.51151, "tr": 7.58961, "fi": 7.42647, "ru": 7.19965, "el": 7.94767, "hi": 7.60916, "ta": 7.73488, "th": 7.3404, "zh": 7.99735, "my": 7.66289}, "gini_cpc": 0.471401, "gini_cps": 0.442829, "bpic": 3.081811, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.3071225937581465, "vi": 0.45369241069660227, "tr": 0.8216937933594943, "fi": 1.1423005654985154, "ru": 3.009372291163438, "el": 1.5755412185432118, "hi": 0.1310433218775342, "ta": 0.19358700798336653, "th": 0.49337390275610404, "zh": 3.2821540964127403, "my": 0.1334112971718954}, "t": 10.0} |
| {"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": 0.488711, "boundary_entropy": 0.686416, "bpb": 6.85762}, "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": 6.931266}, "vi": {"chunks": 3521.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.190995, "cps": 27.507812, "mean_p": 0.475362, "boundary_entropy": 0.686684, "bpb": 7.167738}, "tr": {"chunks": 7547.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.416892, "cps": 58.960938, "mean_p": 0.49526, "boundary_entropy": 0.686953, "bpb": 7.151015}, "fi": {"chunks": 6359.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.338047, "cps": 49.679688, "mean_p": 0.492187, "boundary_entropy": 0.687147, "bpb": 6.91724}, "ru": {"chunks": 25801.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.36153, "cps": 201.570312, "mean_p": 0.511539, "boundary_entropy": 0.689918, "bpb": 6.369007}, "el": {"chunks": 14723.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.71739, "cps": 115.023438, "mean_p": 0.496065, "boundary_entropy": 0.690088, "bpb": 7.323387}, "hi": {"chunks": 3471.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.2063, "cps": 27.117188, "mean_p": 0.475522, "boundary_entropy": 0.689263, "bpb": 6.56889}, "ta": {"chunks": 6916.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.369938, "cps": 54.03125, "mean_p": 0.490171, "boundary_entropy": 0.690934, "bpb": 6.74763}, "th": {"chunks": 6862.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.426609, "cps": 53.609375, "mean_p": 0.489636, "boundary_entropy": 0.690642, "bpb": 6.490437}, "zh": {"chunks": 9451.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.600237, "cps": 73.835938, "mean_p": 0.509541, "boundary_entropy": 0.686084, "bpb": 7.747893}, "my": {"chunks": 1207.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.063701, "cps": 9.429688, "mean_p": 0.473247, "boundary_entropy": 0.689569, "bpb": 7.215522}}, "bpb": {"en": 6.85762, "de": 6.93127, "vi": 7.16774, "tr": 7.15102, "fi": 6.91724, "ru": 6.36901, "el": 7.32339, "hi": 6.56889, "ta": 6.74763, "th": 6.49044, "zh": 7.74789, "my": 7.21552}, "gini_cpc": 0.418289, "gini_cps": 0.365769, "bpic": 3.868643, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.497600515816016, "vi": 0.6891753593904445, "tr": 1.5042868423252576, "fi": 1.2197864363377717, "ru": 4.91285767351581, "el": 2.588584373899331, "hi": 0.744400059275161, "ta": 1.3348620103246251, "th": 1.5393469188508513, "zh": 5.774191443121025, "my": 0.22985330477234833}, "t": 12.0} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5785.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.332414, "cps": 45.195312, "mean_p": 0.496372, "boundary_entropy": 0.687223, "bpb": 6.37825}, "de": {"chunks": 7925.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.385964, "cps": 61.914062, "mean_p": 0.498148, "boundary_entropy": 0.688173, "bpb": 6.45943}, "vi": {"chunks": 11198.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.607432, "cps": 87.484375, "mean_p": 0.499863, "boundary_entropy": 0.688482, "bpb": 6.772098}, "tr": {"chunks": 11103.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.613324, "cps": 86.742188, "mean_p": 0.506407, "boundary_entropy": 0.687556, "bpb": 6.734203}, "fi": {"chunks": 5809.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.308809, "cps": 45.382812, "mean_p": 0.494583, "boundary_entropy": 0.687785, "bpb": 6.467692}, "ru": {"chunks": 225.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.011873, "cps": 1.757812, "mean_p": 0.462424, "boundary_entropy": 0.687125, "bpb": 6.195343}, "el": {"chunks": 12921.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.629586, "cps": 100.945312, "mean_p": 0.495571, "boundary_entropy": 0.690207, "bpb": 6.247496}, "hi": {"chunks": 14768.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.877741, "cps": 115.375, "mean_p": 0.498408, "boundary_entropy": 0.690705, "bpb": 5.729038}, "ta": {"chunks": 16487.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.881894, "cps": 128.804688, "mean_p": 0.497633, "boundary_entropy": 0.69114, "bpb": 5.925641}, "th": {"chunks": 3350.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.208269, "cps": 26.171875, "mean_p": 0.483381, "boundary_entropy": 0.690306, "bpb": 5.67074}, "zh": {"chunks": 3546.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.600406, "cps": 27.703125, "mean_p": 0.483372, "boundary_entropy": 0.685424, "bpb": 7.347951}, "my": {"chunks": 1951.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.102966, "cps": 15.242188, "mean_p": 0.482419, "boundary_entropy": 0.690581, "bpb": 5.912374}}, "bpb": {"en": 6.37825, "de": 6.45943, "vi": 6.7721, "tr": 6.7342, "fi": 6.46769, "ru": 6.19534, "el": 6.2475, "hi": 5.72904, "ta": 5.92564, "th": 5.67074, "zh": 7.34795, "my": 5.91237}, "gini_cpc": 0.330513, "gini_cps": 0.36602, "bpic": 4.03691, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.161094641124322, "vi": 1.8273346024798285, "tr": 1.8450602193698966, "fi": 0.9289883119289652, "ru": 0.03571856977033224, "el": 1.893982832929271, "hi": 2.6405072708148984, "ta": 2.6529980168852547, "th": 0.6265337907491774, "zh": 1.8062008633676372, "my": 0.30975237875245704}, "t": 14.4} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.26393, "bpb": 6.15155, "chunk_loss": 0.030043, "F": 0.16486, "G": 0.48523, "chunks": 168.81, "bytes_per_chunk": 6.066, "boundary_entropy": 0.69148, "router_noise": 0.0, "grad_norm": 1.5511, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.24, "alpha": 0.03, "t": 15.8} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2948.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.169396, "cps": 23.03125, "mean_p": 0.487043, "boundary_entropy": 0.68688, "bpb": 5.799354}, "de": {"chunks": 4602.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.224127, "cps": 35.953125, "mean_p": 0.48987, "boundary_entropy": 0.687951, "bpb": 5.925948}, "vi": {"chunks": 3802.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.206238, "cps": 29.703125, "mean_p": 0.481405, "boundary_entropy": 0.687725, "bpb": 6.240389}, "tr": {"chunks": 2982.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164724, "cps": 23.296875, "mean_p": 0.482699, "boundary_entropy": 0.686979, "bpb": 6.226374}, "fi": {"chunks": 1359.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.072245, "cps": 10.617188, "mean_p": 0.478432, "boundary_entropy": 0.686832, "bpb": 5.915662}, "ru": {"chunks": 7992.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.421741, "cps": 62.4375, "mean_p": 0.494357, "boundary_entropy": 0.690254, "bpb": 5.485427}, "el": {"chunks": 7342.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.357745, "cps": 57.359375, "mean_p": 0.491754, "boundary_entropy": 0.690322, "bpb": 5.657431}, "hi": {"chunks": 12813.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.761545, "cps": 100.101562, "mean_p": 0.497281, "boundary_entropy": 0.69075, "bpb": 5.198742}, "ta": {"chunks": 9936.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.531479, "cps": 77.625, "mean_p": 0.49235, "boundary_entropy": 0.691028, "bpb": 5.25123}, "th": {"chunks": 7127.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.443084, "cps": 55.679688, "mean_p": 0.483961, "boundary_entropy": 0.690194, "bpb": 5.043331}, "zh": {"chunks": 4182.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.708093, "cps": 32.671875, "mean_p": 0.49191, "boundary_entropy": 0.6867, "bpb": 6.955378}, "my": {"chunks": 7796.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.411442, "cps": 60.90625, "mean_p": 0.493885, "boundary_entropy": 0.691268, "bpb": 5.093501}}, "bpb": {"en": 5.79935, "de": 5.92595, "vi": 6.24039, "tr": 6.22637, "fi": 5.91566, "ru": 5.48543, "el": 5.65743, "hi": 5.19874, "ta": 5.25123, "th": 5.04333, "zh": 6.95538, "my": 5.0935}, "gini_cpc": 0.313777, "gini_cps": 0.292636, "bpic": 5.265858, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.3230944514575307, "vi": 1.217490585389496, "tr": 0.972419655625045, "fi": 0.4264854450544865, "ru": 2.489676396143533, "el": 2.1118845646321556, "hi": 4.49564898054641, "ta": 3.1374929190755116, "th": 2.615666285240933, "zh": 4.180105345574563, "my": 2.42887461191366}, "t": 17.6} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.408, "bpb": 4.9167, "chunk_loss": 0.031931, "F": 0.19489, "G": 0.4748, "chunks": 199.56, "bytes_per_chunk": 5.131, "boundary_entropy": 0.68916, "router_noise": 0.0, "grad_norm": 1.6563, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.074, "alpha": 0.03, "t": 19.5} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3217.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.184853, "cps": 25.132812, "mean_p": 0.430748, "boundary_entropy": 0.665017, "bpb": 4.464213}, "de": {"chunks": 3908.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.190328, "cps": 30.53125, "mean_p": 0.437771, "boundary_entropy": 0.670056, "bpb": 4.665459}, "vi": {"chunks": 6314.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.342501, "cps": 49.328125, "mean_p": 0.466845, "boundary_entropy": 0.678264, "bpb": 5.045669}, "tr": {"chunks": 3925.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216815, "cps": 30.664062, "mean_p": 0.449115, "boundary_entropy": 0.674603, "bpb": 5.01376}, "fi": {"chunks": 2509.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.133379, "cps": 19.601562, "mean_p": 0.43844, "boundary_entropy": 0.672454, "bpb": 4.732007}, "ru": {"chunks": 6366.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.335937, "cps": 49.734375, "mean_p": 0.480935, "boundary_entropy": 0.68719, "bpb": 3.747827}, "el": {"chunks": 2879.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.140282, "cps": 22.492188, "mean_p": 0.46629, "boundary_entropy": 0.687111, "bpb": 4.609691}, "hi": {"chunks": 9865.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.58633, "cps": 77.070312, "mean_p": 0.490806, "boundary_entropy": 0.689897, "bpb": 3.604367}, "ta": {"chunks": 10522.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.562824, "cps": 82.203125, "mean_p": 0.484856, "boundary_entropy": 0.690214, "bpb": 3.664953}, "th": {"chunks": 7823.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.486354, "cps": 61.117188, "mean_p": 0.47044, "boundary_entropy": 0.687546, "bpb": 3.370465}, "zh": {"chunks": 3388.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.573654, "cps": 26.46875, "mean_p": 0.486906, "boundary_entropy": 0.686229, "bpb": 6.39687}, "my": {"chunks": 2426.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.128035, "cps": 18.953125, "mean_p": 0.48409, "boundary_entropy": 0.690667, "bpb": 4.316608}}, "bpb": {"en": 4.46421, "de": 4.66546, "vi": 5.04567, "tr": 5.01376, "fi": 4.73201, "ru": 3.74783, "el": 4.60969, "hi": 3.60437, "ta": 3.66495, "th": 3.37047, "zh": 6.39687, "my": 4.31661}, "gini_cpc": 0.301546, "gini_cps": 0.282971, "bpic": 6.078062, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0296158207518467, "vi": 1.8528253964078647, "tr": 1.172903194034358, "fi": 0.7215423676782744, "ru": 1.8173161215606735, "el": 0.7588813484786571, "hi": 3.1718677844899936, "ta": 3.044709674957432, "th": 2.631027117459324, "zh": 3.103294690067718, "my": 0.6926286945755834}, "t": 21.8} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.69513, "bpb": 3.88826, "chunk_loss": 0.029537, "F": 0.15942, "G": 0.43523, "chunks": 163.25, "bytes_per_chunk": 6.273, "boundary_entropy": 0.6682, "router_noise": 0.0, "grad_norm": 1.5762, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.068, "alpha": 0.03, "t": 23.0} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2716.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.156065, "cps": 21.21875, "mean_p": 0.331736, "boundary_entropy": 0.580124, "bpb": 3.885563}, "de": {"chunks": 3750.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182633, "cps": 29.296875, "mean_p": 0.354316, "boundary_entropy": 0.601268, "bpb": 4.117992}, "vi": {"chunks": 4813.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.261079, "cps": 37.601562, "mean_p": 0.382753, "boundary_entropy": 0.62321, "bpb": 3.666892}, "tr": {"chunks": 3848.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212561, "cps": 30.0625, "mean_p": 0.367887, "boundary_entropy": 0.610168, "bpb": 4.212517}, "fi": {"chunks": 3417.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181649, "cps": 26.695312, "mean_p": 0.355952, "boundary_entropy": 0.605834, "bpb": 4.087368}, "ru": {"chunks": 4733.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.249763, "cps": 36.976562, "mean_p": 0.477915, "boundary_entropy": 0.681161, "bpb": 2.859735}, "el": {"chunks": 7594.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.370024, "cps": 59.328125, "mean_p": 0.482584, "boundary_entropy": 0.68779, "bpb": 3.062328}, "hi": {"chunks": 9053.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.538068, "cps": 70.726562, "mean_p": 0.452061, "boundary_entropy": 0.669122, "bpb": 2.454638}, "ta": {"chunks": 9234.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.493929, "cps": 72.140625, "mean_p": 0.466696, "boundary_entropy": 0.68453, "bpb": 2.332994}, "th": {"chunks": 6559.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.407771, "cps": 51.242188, "mean_p": 0.457912, "boundary_entropy": 0.683602, "bpb": 2.362251}, "zh": {"chunks": 1757.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.297494, "cps": 13.726562, "mean_p": 0.448169, "boundary_entropy": 0.677163, "bpb": 5.401611}, "my": {"chunks": 6998.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.369327, "cps": 54.671875, "mean_p": 0.449873, "boundary_entropy": 0.682844, "bpb": 2.543078}}, "bpb": {"en": 3.88556, "de": 4.11799, "vi": 3.66689, "tr": 4.21252, "fi": 4.08737, "ru": 2.85974, "el": 3.06233, "hi": 2.45464, "ta": 2.33299, "th": 2.36225, "zh": 5.40161, "my": 2.54308}, "gini_cpc": 0.218862, "gini_cps": 0.249426, "bpic": 5.952677, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1702353558950005, "vi": 1.6728888028750457, "tr": 1.3620055162990177, "fi": 1.1639315753034858, "ru": 1.6003745808091208, "el": 2.3709593181026034, "hi": 3.447718522344064, "ta": 3.164890985082998, "th": 2.6128285942271887, "zh": 1.9062184708194705, "my": 2.3664913243687162}, "t": 26.0} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.88853, "de": 4.11952, "vi": 3.67408, "tr": 4.22212, "fi": 4.08893, "ru": 2.84427, "el": 3.05438, "hi": 2.45064, "ta": 2.35018, "th": 2.39457, "zh": 5.41075, "my": 2.55942}, "bytes_per_chunk": {"en": 6.346, "de": 5.836, "vi": 4.972, "tr": 5.128, "fi": 5.904, "ru": 7.043, "el": 5.075, "hi": 4.827, "ta": 5.381, "th": 8.313, "zh": 12.868, "my": 6.853}, "bpb_mean": 3.42145, "t": 27.4} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.52992, "bpb": 3.6499, "chunk_loss": 0.029189, "F": 0.15283, "G": 0.39629, "chunks": 156.5, "bytes_per_chunk": 6.543, "boundary_entropy": 0.63195, "router_noise": 0.0, "grad_norm": 1.8555, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 27.4} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.48535, "bpb": 3.5856, "chunk_loss": 0.031082, "F": 0.18793, "G": 0.40673, "chunks": 192.44, "bytes_per_chunk": 5.321, "boundary_entropy": 0.64395, "router_noise": 0.0, "grad_norm": 1.3956, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 29.7} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2755.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.158306, "cps": 21.523438, "mean_p": 0.345034, "boundary_entropy": 0.591365, "bpb": 3.639508}, "de": {"chunks": 4256.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.207276, "cps": 33.25, "mean_p": 0.353875, "boundary_entropy": 0.598, "bpb": 3.860988}, "vi": {"chunks": 3867.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.209764, "cps": 30.210938, "mean_p": 0.376484, "boundary_entropy": 0.618809, "bpb": 3.259978}, "tr": {"chunks": 4089.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.225874, "cps": 31.945312, "mean_p": 0.369023, "boundary_entropy": 0.608001, "bpb": 3.864708}, "fi": {"chunks": 3495.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185796, "cps": 27.304688, "mean_p": 0.343595, "boundary_entropy": 0.589884, "bpb": 3.906599}, "ru": {"chunks": 5018.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.264802, "cps": 39.203125, "mean_p": 0.462391, "boundary_entropy": 0.678233, "bpb": 2.660089}, "el": {"chunks": 6673.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325147, "cps": 52.132812, "mean_p": 0.468983, "boundary_entropy": 0.684834, "bpb": 2.76411}, "hi": {"chunks": 9493.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.56422, "cps": 74.164062, "mean_p": 0.391691, "boundary_entropy": 0.617787, "bpb": 2.129326}, "ta": {"chunks": 11224.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.600374, "cps": 87.6875, "mean_p": 0.407186, "boundary_entropy": 0.637951, "bpb": 1.93909}, "th": {"chunks": 8585.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.533727, "cps": 67.070312, "mean_p": 0.423589, "boundary_entropy": 0.666753, "bpb": 2.073643}, "zh": {"chunks": 4279.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.724517, "cps": 33.429688, "mean_p": 0.44502, "boundary_entropy": 0.663809, "bpb": 5.024269}, "my": {"chunks": 13618.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.718704, "cps": 106.390625, "mean_p": 0.431934, "boundary_entropy": 0.669975, "bpb": 2.271485}}, "bpb": {"en": 3.63951, "de": 3.86099, "vi": 3.25998, "tr": 3.86471, "fi": 3.9066, "ru": 2.66009, "el": 2.76411, "hi": 2.12933, "ta": 1.93909, "th": 2.07364, "zh": 5.02427, "my": 2.27148}, "gini_cpc": 0.291977, "gini_cps": 0.278926, "bpic": 4.961488, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.3093378699049638, "vi": 1.3250539074602137, "tr": 1.4268196315237078, "fi": 1.1736478665753451, "ru": 1.6727227348691993, "el": 2.0539165610572105, "hi": 3.564108569317437, "ta": 3.792492281048001, "th": 3.371489032572535, "zh": 4.576688568578634, "my": 4.539964644856665}, "t": 30.9} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.32618, "bpb": 3.35597, "chunk_loss": 0.031638, "F": 0.20593, "G": 0.37518, "chunks": 210.88, "bytes_per_chunk": 4.856, "boundary_entropy": 0.61178, "router_noise": 0.0, "grad_norm": 1.9475, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.066, "alpha": 0.03, "t": 33.0} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.0908, "bpb": 3.01639, "chunk_loss": 0.030096, "F": 0.16809, "G": 0.36023, "chunks": 172.12, "bytes_per_chunk": 5.949, "boundary_entropy": 0.59179, "router_noise": 0.0, "grad_norm": 0.6564, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 35.3} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3145.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.180716, "cps": 24.570312, "mean_p": 0.349222, "boundary_entropy": 0.578546, "bpb": 3.233}, "de": {"chunks": 3966.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.193152, "cps": 30.984375, "mean_p": 0.358868, "boundary_entropy": 0.589283, "bpb": 3.538529}, "vi": {"chunks": 4764.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.258421, "cps": 37.21875, "mean_p": 0.369801, "boundary_entropy": 0.603598, "bpb": 2.809583}, "tr": {"chunks": 3699.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.204331, "cps": 28.898438, "mean_p": 0.342377, "boundary_entropy": 0.560365, "bpb": 3.464448}, "fi": {"chunks": 3662.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194673, "cps": 28.609375, "mean_p": 0.339457, "boundary_entropy": 0.55934, "bpb": 3.589242}, "ru": {"chunks": 5360.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.28285, "cps": 41.875, "mean_p": 0.326611, "boundary_entropy": 0.531569, "bpb": 2.500958}, "el": {"chunks": 7593.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.369975, "cps": 59.320312, "mean_p": 0.323783, "boundary_entropy": 0.535178, "bpb": 2.563508}, "hi": {"chunks": 10399.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.618068, "cps": 81.242188, "mean_p": 0.357519, "boundary_entropy": 0.561047, "bpb": 1.822094}, "ta": {"chunks": 10525.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.562985, "cps": 82.226562, "mean_p": 0.291151, "boundary_entropy": 0.457465, "bpb": 1.625414}, "th": {"chunks": 9372.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.582655, "cps": 73.21875, "mean_p": 0.340079, "boundary_entropy": 0.566787, "bpb": 1.900172}, "zh": {"chunks": 3021.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.511514, "cps": 23.601562, "mean_p": 0.368875, "boundary_entropy": 0.593375, "bpb": 4.156333}, "my": {"chunks": 11019.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.581539, "cps": 86.085938, "mean_p": 0.37755, "boundary_entropy": 0.635331, "bpb": 2.012805}}, "bpb": {"en": 3.233, "de": 3.53853, "vi": 2.80958, "tr": 3.46445, "fi": 3.58924, "ru": 2.50096, "el": 2.56351, "hi": 1.82209, "ta": 1.62541, "th": 1.90017, "zh": 4.15633, "my": 2.0128}, "gini_cpc": 0.250761, "gini_cps": 0.262876, "bpic": 5.015106, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0688180343604465, "vi": 1.4299869735240431, "tr": 1.1306735956332545, "fi": 1.077233689079699, "ru": 1.5651611008804862, "el": 2.0472742551747167, "hi": 3.4201092231117576, "ta": 3.11530165220205, "th": 3.224145890299596, "zh": 2.8304843179068793, "my": 3.217972122179923}, "t": 38.4} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.23904, "de": 3.53698, "vi": 2.81586, "tr": 3.47266, "fi": 3.58209, "ru": 2.48713, "el": 2.5579, "hi": 1.81778, "ta": 1.64258, "th": 1.92904, "zh": 4.16406, "my": 2.02874}, "bytes_per_chunk": {"en": 5.471, "de": 5.391, "vi": 5.147, "tr": 5.302, "fi": 5.341, "ru": 7.113, "el": 4.853, "hi": 3.857, "ta": 4.753, "th": 4.393, "zh": 6.205, "my": 4.409}, "bpb_mean": 2.77282, "t": 39.7} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.23748, "bpb": 3.228, "chunk_loss": 0.030318, "F": 0.17261, "G": 0.34402, "chunks": 176.75, "bytes_per_chunk": 5.793, "boundary_entropy": 0.57195, "router_noise": 0.0, "grad_norm": 0.9219, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 39.8} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.17737, "bpb": 3.14128, "chunk_loss": 0.03185, "F": 0.21201, "G": 0.352, "chunks": 217.09, "bytes_per_chunk": 4.717, "boundary_entropy": 0.5641, "router_noise": 0.0, "grad_norm": 1.2398, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 41.8} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.0678, "bpb": 2.98321, "chunk_loss": 0.030399, "F": 0.17731, "G": 0.33116, "chunks": 181.56, "bytes_per_chunk": 5.64, "boundary_entropy": 0.55152, "router_noise": 0.0, "grad_norm": 0.9265, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 43.9} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.91722, "bpb": 2.76597, "chunk_loss": 0.030164, "F": 0.16986, "G": 0.32673, "chunks": 173.94, "bytes_per_chunk": 5.887, "boundary_entropy": 0.54295, "router_noise": 0.0, "grad_norm": 0.7951, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 45.9} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3569.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20508, "cps": 27.882812, "mean_p": 0.318337, "boundary_entropy": 0.494302, "bpb": 2.890321}, "de": {"chunks": 4038.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196659, "cps": 31.546875, "mean_p": 0.321566, "boundary_entropy": 0.512918, "bpb": 3.202046}, "vi": {"chunks": 4530.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.245728, "cps": 35.390625, "mean_p": 0.339514, "boundary_entropy": 0.541321, "bpb": 2.57149}, "tr": {"chunks": 3421.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.188974, "cps": 26.726562, "mean_p": 0.301103, "boundary_entropy": 0.489633, "bpb": 3.163541}, "fi": {"chunks": 3748.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.199245, "cps": 29.28125, "mean_p": 0.303986, "boundary_entropy": 0.485257, "bpb": 3.372848}, "ru": {"chunks": 6114.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.322639, "cps": 47.765625, "mean_p": 0.357737, "boundary_entropy": 0.58802, "bpb": 2.149046}, "el": {"chunks": 8059.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.392681, "cps": 62.960938, "mean_p": 0.379243, "boundary_entropy": 0.583451, "bpb": 2.226628}, "hi": {"chunks": 6657.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.395661, "cps": 52.007812, "mean_p": 0.34095, "boundary_entropy": 0.581266, "bpb": 1.644883}, "ta": {"chunks": 7117.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.38069, "cps": 55.601562, "mean_p": 0.287383, "boundary_entropy": 0.493619, "bpb": 1.374067}, "th": {"chunks": 6586.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.40945, "cps": 51.453125, "mean_p": 0.359404, "boundary_entropy": 0.605794, "bpb": 1.746737}, "zh": {"chunks": 3742.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.633593, "cps": 29.234375, "mean_p": 0.393158, "boundary_entropy": 0.615694, "bpb": 3.849127}, "my": {"chunks": 15708.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.829006, "cps": 122.71875, "mean_p": 0.399385, "boundary_entropy": 0.607326, "bpb": 1.651221}}, "bpb": {"en": 2.89032, "de": 3.20205, "vi": 2.57149, "tr": 3.16354, "fi": 3.37285, "ru": 2.14905, "el": 2.22663, "hi": 1.64488, "ta": 1.37407, "th": 1.74674, "zh": 3.84913, "my": 1.65122}, "gini_cpc": 0.263217, "gini_cps": 0.253523, "bpic": 5.236543, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9589400995277481, "vi": 1.1982091516291289, "tr": 0.9214676541833786, "fi": 0.9715502570667303, "ru": 1.5732356978999016, "el": 1.9147755922060916, "hi": 1.929305739492751, "ta": 1.8563038635194102, "th": 1.996541001315772, "zh": 3.0894979598656716, "my": 4.042360939701188}, "t": 48.9} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.89945, "de": 3.20245, "vi": 2.57553, "tr": 3.17359, "fi": 3.36693, "ru": 2.13314, "el": 2.21385, "hi": 1.63784, "ta": 1.38704, "th": 1.77419, "zh": 3.85979, "my": 1.66345}, "bytes_per_chunk": {"en": 5.068, "de": 5.204, "vi": 5.854, "tr": 5.813, "fi": 5.631, "ru": 6.189, "el": 5.14, "hi": 6.447, "ta": 7.185, "th": 6.63, "zh": 4.751, "my": 3.911}, "bpb_mean": 2.4906, "t": 50.2} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.90302, "bpb": 2.74548, "chunk_loss": 0.030384, "F": 0.17773, "G": 0.33138, "chunks": 182.0, "bytes_per_chunk": 5.626, "boundary_entropy": 0.53689, "router_noise": 0.0, "grad_norm": 0.9621, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 50.2} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.94201, "bpb": 2.80173, "chunk_loss": 0.030741, "F": 0.18851, "G": 0.33274, "chunks": 193.03, "bytes_per_chunk": 5.305, "boundary_entropy": 0.53271, "router_noise": 0.0, "grad_norm": 0.8909, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 52.2} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.77894, "bpb": 2.56647, "chunk_loss": 0.030071, "F": 0.17197, "G": 0.31159, "chunks": 176.09, "bytes_per_chunk": 5.815, "boundary_entropy": 0.50226, "router_noise": 0.0, "grad_norm": 0.8542, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 54.2} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.86404, "bpb": 2.68925, "chunk_loss": 0.030317, "F": 0.17419, "G": 0.30637, "chunks": 178.38, "bytes_per_chunk": 5.741, "boundary_entropy": 0.49282, "router_noise": 0.0, "grad_norm": 0.8586, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 56.2} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3130.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.179854, "cps": 24.453125, "mean_p": 0.260627, "boundary_entropy": 0.399002, "bpb": 2.691054}, "de": {"chunks": 3814.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18575, "cps": 29.796875, "mean_p": 0.274482, "boundary_entropy": 0.426358, "bpb": 3.037815}, "vi": {"chunks": 4123.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223651, "cps": 32.210938, "mean_p": 0.303578, "boundary_entropy": 0.489852, "bpb": 2.411852}, "tr": {"chunks": 3354.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185273, "cps": 26.203125, "mean_p": 0.272615, "boundary_entropy": 0.424859, "bpb": 3.036545}, "fi": {"chunks": 3299.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175376, "cps": 25.773438, "mean_p": 0.261494, "boundary_entropy": 0.414238, "bpb": 3.236337}, "ru": {"chunks": 5765.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.304222, "cps": 45.039062, "mean_p": 0.356846, "boundary_entropy": 0.596526, "bpb": 2.000471}, "el": {"chunks": 7225.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.352044, "cps": 56.445312, "mean_p": 0.33899, "boundary_entropy": 0.56841, "bpb": 2.098325}, "hi": {"chunks": 6744.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.400832, "cps": 52.6875, "mean_p": 0.32421, "boundary_entropy": 0.544569, "bpb": 1.548639}, "ta": {"chunks": 7180.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.38406, "cps": 56.09375, "mean_p": 0.321928, "boundary_entropy": 0.535769, "bpb": 1.309167}, "th": {"chunks": 6893.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.428536, "cps": 53.851562, "mean_p": 0.402534, "boundary_entropy": 0.637413, "bpb": 1.589679}, "zh": {"chunks": 3525.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.596851, "cps": 27.539062, "mean_p": 0.369056, "boundary_entropy": 0.592241, "bpb": 3.717345}, "my": {"chunks": 11024.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.581803, "cps": 86.125, "mean_p": 0.33832, "boundary_entropy": 0.552869, "bpb": 1.4944}}, "bpb": {"en": 2.69105, "de": 3.03782, "vi": 2.41185, "tr": 3.03655, "fi": 3.23634, "ru": 2.00047, "el": 2.09833, "hi": 1.54864, "ta": 1.30917, "th": 1.58968, "zh": 3.71735, "my": 1.4944}, "gini_cpc": 0.240961, "gini_cps": 0.222247, "bpic": 5.808175, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.032780582772624, "vi": 1.2435119853799421, "tr": 1.0301306033861262, "fi": 0.9751023762012676, "ru": 1.691491734596677, "el": 1.9573874024919287, "hi": 2.2286520610114455, "ta": 2.13539763497954, "th": 2.3826870032505036, "zh": 3.318527808942874, "my": 3.234860903997556}, "t": 59.2} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.69934, "de": 3.04185, "vi": 2.41774, "tr": 3.04683, "fi": 3.23355, "ru": 1.98625, "el": 2.08747, "hi": 1.54481, "ta": 1.32173, "th": 1.61524, "zh": 3.72937, "my": 1.50402}, "bytes_per_chunk": {"en": 5.705, "de": 5.598, "vi": 5.948, "tr": 5.829, "fi": 5.908, "ru": 6.517, "el": 5.388, "hi": 6.205, "ta": 7.1, "th": 6.772, "zh": 5.125, "my": 5.072}, "bpb_mean": 2.35235, "t": 60.5} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.84132, "bpb": 2.65647, "chunk_loss": 0.030182, "F": 0.172, "G": 0.30459, "chunks": 176.12, "bytes_per_chunk": 5.814, "boundary_entropy": 0.48958, "router_noise": 0.0, "grad_norm": 0.8429, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 60.5} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.64059, "bpb": 2.36688, "chunk_loss": 0.030301, "F": 0.17401, "G": 0.30018, "chunks": 178.19, "bytes_per_chunk": 5.747, "boundary_entropy": 0.48481, "router_noise": 0.0, "grad_norm": 0.8714, "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": 1.72023, "bpb": 2.48177, "chunk_loss": 0.030078, "F": 0.17212, "G": 0.29688, "chunks": 176.25, "bytes_per_chunk": 5.81, "boundary_entropy": 0.46547, "router_noise": 0.0, "grad_norm": 0.8385, "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": 1.87118, "bpb": 2.69954, "chunk_loss": 0.030422, "F": 0.18207, "G": 0.29294, "chunks": 186.44, "bytes_per_chunk": 5.492, "boundary_entropy": 0.44848, "router_noise": 0.0, "grad_norm": 0.8782, "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": 3471.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199448, "cps": 27.117188, "mean_p": 0.257179, "boundary_entropy": 0.346525, "bpb": 2.557864}, "de": {"chunks": 4132.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201237, "cps": 32.28125, "mean_p": 0.262317, "boundary_entropy": 0.375129, "bpb": 2.894572}, "vi": {"chunks": 4928.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.267318, "cps": 38.5, "mean_p": 0.306982, "boundary_entropy": 0.466621, "bpb": 2.305752}, "tr": {"chunks": 4044.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.223388, "cps": 31.59375, "mean_p": 0.28122, "boundary_entropy": 0.394997, "bpb": 2.914704}, "fi": {"chunks": 3676.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.195418, "cps": 28.71875, "mean_p": 0.258931, "boundary_entropy": 0.37371, "bpb": 3.147739}, "ru": {"chunks": 5960.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.314512, "cps": 46.5625, "mean_p": 0.357929, "boundary_entropy": 0.597364, "bpb": 1.904823}, "el": {"chunks": 6812.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.33192, "cps": 53.21875, "mean_p": 0.366361, "boundary_entropy": 0.605986, "bpb": 1.988705}, "hi": {"chunks": 7445.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.442496, "cps": 58.164062, "mean_p": 0.388228, "boundary_entropy": 0.592696, "bpb": 1.478492}, "ta": {"chunks": 8071.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.43172, "cps": 63.054688, "mean_p": 0.392575, "boundary_entropy": 0.604821, "bpb": 1.268105}, "th": {"chunks": 10204.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.63438, "cps": 79.71875, "mean_p": 0.42826, "boundary_entropy": 0.640622, "bpb": 1.528223}, "zh": {"chunks": 3207.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.543007, "cps": 25.054688, "mean_p": 0.346134, "boundary_entropy": 0.561654, "bpb": 3.619434}, "my": {"chunks": 9634.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.508444, "cps": 75.265625, "mean_p": 0.33434, "boundary_entropy": 0.547205, "bpb": 1.413702}}, "bpb": {"en": 2.55786, "de": 2.89457, "vi": 2.30575, "tr": 2.9147, "fi": 3.14774, "ru": 1.90482, "el": 1.98871, "hi": 1.47849, "ta": 1.2681, "th": 1.52822, "zh": 3.61943, "my": 1.4137}, "gini_cpc": 0.226656, "gini_cps": 0.220343, "bpic": 5.361268, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0089680456626076, "vi": 1.3402847117075367, "tr": 1.1200311518106012, "fi": 0.9797902779167028, "ru": 1.576908709860594, "el": 1.6641915045164455, "hi": 2.218600649062943, "ta": 2.1645687505591162, "th": 3.1806720402012965, "zh": 2.7225447305978623, "my": 2.549252021955554}, "t": 69.4} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.56684, "de": 2.89976, "vi": 2.31112, "tr": 2.92215, "fi": 3.14385, "ru": 1.88796, "el": 1.97725, "hi": 1.476, "ta": 1.27735, "th": 1.55564, "zh": 3.62797, "my": 1.42074}, "bytes_per_chunk": {"en": 5.217, "de": 5.139, "vi": 5.25, "tr": 4.973, "fi": 5.379, "ru": 6.023, "el": 5.618, "hi": 5.8, "ta": 6.477, "th": 5.127, "zh": 5.36, "my": 5.905}, "bpb_mean": 2.25555, "t": 70.8} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.54873, "bpb": 2.23434, "chunk_loss": 0.030422, "F": 0.18018, "G": 0.3237, "chunks": 184.5, "bytes_per_chunk": 5.55, "boundary_entropy": 0.49837, "router_noise": 0.0, "grad_norm": 0.7775, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 70.8} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.77429, "bpb": 2.55976, "chunk_loss": 0.030334, "F": 0.17871, "G": 0.29756, "chunks": 183.0, "bytes_per_chunk": 5.596, "boundary_entropy": 0.45583, "router_noise": 0.0, "grad_norm": 0.9179, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 72.8} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.59218, "bpb": 2.29703, "chunk_loss": 0.030249, "F": 0.17606, "G": 0.30311, "chunks": 180.28, "bytes_per_chunk": 5.68, "boundary_entropy": 0.46111, "router_noise": 0.0, "grad_norm": 0.8148, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 74.9} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.61533, "bpb": 2.33043, "chunk_loss": 0.029893, "F": 0.16565, "G": 0.28882, "chunks": 169.62, "bytes_per_chunk": 6.037, "boundary_entropy": 0.44173, "router_noise": 0.0, "grad_norm": 0.9229, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 76.9} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3261.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187381, "cps": 25.476562, "mean_p": 0.244414, "boundary_entropy": 0.323771, "bpb": 2.473114}, "de": {"chunks": 4052.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.197341, "cps": 31.65625, "mean_p": 0.257567, "boundary_entropy": 0.360257, "bpb": 2.795117}, "vi": {"chunks": 4729.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.256523, "cps": 36.945312, "mean_p": 0.299937, "boundary_entropy": 0.457641, "bpb": 2.254138}, "tr": {"chunks": 4010.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.22151, "cps": 31.328125, "mean_p": 0.271035, "boundary_entropy": 0.373601, "bpb": 2.822078}, "fi": {"chunks": 3684.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.195843, "cps": 28.78125, "mean_p": 0.246225, "boundary_entropy": 0.34159, "bpb": 3.034362}, "ru": {"chunks": 5309.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.280158, "cps": 41.476562, "mean_p": 0.357902, "boundary_entropy": 0.60067, "bpb": 1.834006}, "el": {"chunks": 6689.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325927, "cps": 52.257812, "mean_p": 0.405383, "boundary_entropy": 0.640244, "bpb": 1.938289}, "hi": {"chunks": 6600.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.392273, "cps": 51.5625, "mean_p": 0.338714, "boundary_entropy": 0.558163, "bpb": 1.449516}, "ta": {"chunks": 7585.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.405723, "cps": 59.257812, "mean_p": 0.402162, "boundary_entropy": 0.617941, "bpb": 1.175925}, "th": {"chunks": 7598.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.472366, "cps": 59.359375, "mean_p": 0.410927, "boundary_entropy": 0.635657, "bpb": 1.458715}, "zh": {"chunks": 3061.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.518286, "cps": 23.914062, "mean_p": 0.330341, "boundary_entropy": 0.537922, "bpb": 3.582433}, "my": {"chunks": 9183.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.484642, "cps": 71.742188, "mean_p": 0.339354, "boundary_entropy": 0.571876, "bpb": 1.391462}}, "bpb": {"en": 2.47311, "de": 2.79512, "vi": 2.25414, "tr": 2.82208, "fi": 3.03436, "ru": 1.83401, "el": 1.93829, "hi": 1.44952, "ta": 1.17592, "th": 1.45872, "zh": 3.58243, "my": 1.39146}, "gini_cpc": 0.201847, "gini_cps": 0.1968, "bpic": 5.835997, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0531502881510415, "vi": 1.3689875339621622, "tr": 1.1821351815871575, "fi": 1.0451558589615377, "ru": 1.4951226900792043, "el": 1.7393767963163462, "hi": 2.093448030280869, "ta": 2.1652270148830315, "th": 2.5208763586043443, "zh": 2.7659428875130807, "my": 2.5863930800589383}, "t": 79.9} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.47896, "de": 2.80027, "vi": 2.26043, "tr": 2.82829, "fi": 3.03564, "ru": 1.81641, "el": 1.9253, "hi": 1.4464, "ta": 1.18567, "th": 1.48295, "zh": 3.58786, "my": 1.39839}, "bytes_per_chunk": {"en": 5.57, "de": 5.196, "vi": 5.429, "tr": 5.045, "fi": 5.221, "ru": 6.863, "el": 5.977, "hi": 6.563, "ta": 7.178, "th": 6.025, "zh": 5.621, "my": 6.304}, "bpb_mean": 2.18721, "t": 81.2} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.69995, "bpb": 2.45251, "chunk_loss": 0.030127, "F": 0.17563, "G": 0.29, "chunks": 179.84, "bytes_per_chunk": 5.694, "boundary_entropy": 0.43959, "router_noise": 0.0, "grad_norm": 0.8233, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 81.2} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.60353, "bpb": 2.31341, "chunk_loss": 0.030221, "F": 0.17245, "G": 0.30362, "chunks": 176.59, "bytes_per_chunk": 5.799, "boundary_entropy": 0.4627, "router_noise": 0.0, "grad_norm": 0.9726, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 83.3} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.72337, "bpb": 2.48629, "chunk_loss": 0.030083, "F": 0.17517, "G": 0.30008, "chunks": 179.38, "bytes_per_chunk": 5.709, "boundary_entropy": 0.44661, "router_noise": 0.0, "grad_norm": 0.8973, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 85.3} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.51285, "bpb": 2.18258, "chunk_loss": 0.030588, "F": 0.18192, "G": 0.30334, "chunks": 186.28, "bytes_per_chunk": 5.497, "boundary_entropy": 0.45823, "router_noise": 0.0, "grad_norm": 0.8418, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 87.3} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3322.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.190887, "cps": 25.953125, "mean_p": 0.240539, "boundary_entropy": 0.295848, "bpb": 2.404207}, "de": {"chunks": 3882.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189062, "cps": 30.328125, "mean_p": 0.241123, "boundary_entropy": 0.323714, "bpb": 2.736496}, "vi": {"chunks": 4318.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234228, "cps": 33.734375, "mean_p": 0.278893, "boundary_entropy": 0.429605, "bpb": 2.208207}, "tr": {"chunks": 3682.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203392, "cps": 28.765625, "mean_p": 0.246348, "boundary_entropy": 0.338624, "bpb": 2.735957}, "fi": {"chunks": 3384.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179895, "cps": 26.4375, "mean_p": 0.221611, "boundary_entropy": 0.304556, "bpb": 2.976814}, "ru": {"chunks": 5820.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.307124, "cps": 45.46875, "mean_p": 0.35827, "boundary_entropy": 0.597756, "bpb": 1.797698}, "el": {"chunks": 6952.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.338742, "cps": 54.3125, "mean_p": 0.394258, "boundary_entropy": 0.624962, "bpb": 1.869368}, "hi": {"chunks": 7123.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.423358, "cps": 55.648438, "mean_p": 0.389664, "boundary_entropy": 0.598452, "bpb": 1.428202}, "ta": {"chunks": 8419.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.450334, "cps": 65.773438, "mean_p": 0.42736, "boundary_entropy": 0.627852, "bpb": 1.150402}, "th": {"chunks": 8672.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.539136, "cps": 67.75, "mean_p": 0.421505, "boundary_entropy": 0.637388, "bpb": 1.417764}, "zh": {"chunks": 3109.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.526414, "cps": 24.289062, "mean_p": 0.327769, "boundary_entropy": 0.518351, "bpb": 3.539235}, "my": {"chunks": 9943.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.524752, "cps": 77.679688, "mean_p": 0.358064, "boundary_entropy": 0.589862, "bpb": 1.366002}}, "bpb": {"en": 2.40421, "de": 2.7365, "vi": 2.20821, "tr": 2.73596, "fi": 2.97681, "ru": 1.7977, "el": 1.86937, "hi": 1.4282, "ta": 1.1504, "th": 1.41776, "zh": 3.53923, "my": 1.366}, "gini_cpc": 0.226396, "gini_cps": 0.226452, "bpic": 5.592356, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9904387331088268, "vi": 1.2270548832670625, "tr": 1.0655104779883044, "fi": 0.9424166774977083, "ru": 1.608934122719092, "el": 1.7745711240465791, "hi": 2.2178509875627657, "ta": 2.3591716033497576, "th": 2.824377193929675, "zh": 2.757730176946352, "my": 2.749024152034587}, "t": 90.3} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.41077, "de": 2.7403, "vi": 2.2123, "tr": 2.74343, "fi": 2.98046, "ru": 1.7789, "el": 1.85877, "hi": 1.42443, "ta": 1.15745, "th": 1.43839, "zh": 3.5427, "my": 1.37437}, "bytes_per_chunk": {"en": 5.52, "de": 5.437, "vi": 5.983, "tr": 5.386, "fi": 5.795, "ru": 6.062, "el": 5.845, "hi": 6.332, "ta": 6.645, "th": 5.481, "zh": 5.566, "my": 5.756}, "bpb_mean": 2.13852, "t": 91.6} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.69968, "bpb": 2.45212, "chunk_loss": 0.030287, "F": 0.1777, "G": 0.30531, "chunks": 181.97, "bytes_per_chunk": 5.627, "boundary_entropy": 0.45727, "router_noise": 0.0, "grad_norm": 0.7736, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 91.6} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.70669, "bpb": 2.46224, "chunk_loss": 0.030222, "F": 0.17233, "G": 0.30146, "chunks": 176.47, "bytes_per_chunk": 5.803, "boundary_entropy": 0.45933, "router_noise": 0.0, "grad_norm": 0.816, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 93.6} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.68098, "bpb": 2.42515, "chunk_loss": 0.030361, "F": 0.18066, "G": 0.27882, "chunks": 185.0, "bytes_per_chunk": 5.535, "boundary_entropy": 0.39962, "router_noise": 0.0, "grad_norm": 0.9288, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 95.6} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.57209, "bpb": 2.26805, "chunk_loss": 0.030037, "F": 0.17325, "G": 0.28129, "chunks": 177.41, "bytes_per_chunk": 5.772, "boundary_entropy": 0.41149, "router_noise": 0.0, "grad_norm": 0.8583, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 97.6} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3327.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.191174, "cps": 25.992188, "mean_p": 0.23417, "boundary_entropy": 0.279145, "bpb": 2.351363}, "de": {"chunks": 3752.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18273, "cps": 29.3125, "mean_p": 0.229489, "boundary_entropy": 0.303823, "bpb": 2.661804}, "vi": {"chunks": 3957.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.214646, "cps": 30.914062, "mean_p": 0.259665, "boundary_entropy": 0.40997, "bpb": 2.162972}, "tr": {"chunks": 4037.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.223002, "cps": 31.539062, "mean_p": 0.259256, "boundary_entropy": 0.340875, "bpb": 2.672032}, "fi": {"chunks": 3272.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.173941, "cps": 25.5625, "mean_p": 0.209994, "boundary_entropy": 0.291099, "bpb": 2.893104}, "ru": {"chunks": 6315.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.333245, "cps": 49.335938, "mean_p": 0.366765, "boundary_entropy": 0.600314, "bpb": 1.742877}, "el": {"chunks": 7206.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.351118, "cps": 56.296875, "mean_p": 0.393958, "boundary_entropy": 0.623498, "bpb": 1.825678}, "hi": {"chunks": 7290.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.433284, "cps": 56.953125, "mean_p": 0.399305, "boundary_entropy": 0.592411, "bpb": 1.370928}, "ta": {"chunks": 8503.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.454827, "cps": 66.429688, "mean_p": 0.429633, "boundary_entropy": 0.628023, "bpb": 1.12389}, "th": {"chunks": 8179.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.508486, "cps": 63.898438, "mean_p": 0.419473, "boundary_entropy": 0.63776, "bpb": 1.37625}, "zh": {"chunks": 2782.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.471046, "cps": 21.734375, "mean_p": 0.309428, "boundary_entropy": 0.494869, "bpb": 3.496255}, "my": {"chunks": 9986.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.527021, "cps": 78.015625, "mean_p": 0.384902, "boundary_entropy": 0.619544, "bpb": 1.345424}}, "bpb": {"en": 2.35136, "de": 2.6618, "vi": 2.16297, "tr": 2.67203, "fi": 2.8931, "ru": 1.74288, "el": 1.82568, "hi": 1.37093, "ta": 1.12389, "th": 1.37625, "zh": 3.49625, "my": 1.34542}, "gini_cpc": 0.217828, "gini_cps": 0.230789, "bpic": 5.593986, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9558323869947943, "vi": 1.1227788616108605, "tr": 1.1664859637221676, "fi": 0.9098561414792938, "ru": 1.7431528082776802, "el": 1.8366429324911895, "hi": 2.2664376431118085, "ta": 2.3791292088292173, "th": 2.659809067754067, "zh": 2.4639676543577056, "my": 2.756763468068748}, "t": 100.5} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.35809, "de": 2.67114, "vi": 2.16724, "tr": 2.67905, "fi": 2.9, "ru": 1.72302, "el": 1.81643, "hi": 1.36896, "ta": 1.13087, "th": 1.39616, "zh": 3.50534, "my": 1.35287}, "bytes_per_chunk": {"en": 5.484, "de": 5.649, "vi": 6.476, "tr": 4.849, "fi": 5.909, "ru": 5.537, "el": 5.58, "hi": 5.991, "ta": 6.396, "th": 5.498, "zh": 6.087, "my": 5.544}, "bpb_mean": 2.0891, "t": 101.8} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.79029, "bpb": 2.58284, "chunk_loss": 0.030288, "F": 0.1785, "G": 0.26313, "chunks": 182.78, "bytes_per_chunk": 5.602, "boundary_entropy": 0.3723, "router_noise": 0.0, "grad_norm": 0.9679, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 101.8} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.57479, "bpb": 2.27194, "chunk_loss": 0.02957, "F": 0.16116, "G": 0.28349, "chunks": 165.03, "bytes_per_chunk": 6.205, "boundary_entropy": 0.42118, "router_noise": 0.0, "grad_norm": 0.7695, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 103.8} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.56388, "bpb": 2.2562, "chunk_loss": 0.030058, "F": 0.16833, "G": 0.29421, "chunks": 172.38, "bytes_per_chunk": 5.941, "boundary_entropy": 0.4377, "router_noise": 0.0, "grad_norm": 0.803, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 105.8} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.51412, "bpb": 2.18441, "chunk_loss": 0.030133, "F": 0.17511, "G": 0.30094, "chunks": 179.31, "bytes_per_chunk": 5.711, "boundary_entropy": 0.43607, "router_noise": 0.0, "grad_norm": 0.7974, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 107.8} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3332.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.191461, "cps": 26.03125, "mean_p": 0.234343, "boundary_entropy": 0.265556, "bpb": 2.313039}, "de": {"chunks": 3671.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178785, "cps": 28.679688, "mean_p": 0.223941, "boundary_entropy": 0.290206, "bpb": 2.59813}, "vi": {"chunks": 4293.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.232872, "cps": 33.539062, "mean_p": 0.282576, "boundary_entropy": 0.42229, "bpb": 2.13057}, "tr": {"chunks": 3910.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215986, "cps": 30.546875, "mean_p": 0.246461, "boundary_entropy": 0.316907, "bpb": 2.604015}, "fi": {"chunks": 3608.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.191803, "cps": 28.1875, "mean_p": 0.225249, "boundary_entropy": 0.285838, "bpb": 2.831838}, "ru": {"chunks": 5931.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.312982, "cps": 46.335938, "mean_p": 0.359354, "boundary_entropy": 0.590649, "bpb": 1.700081}, "el": {"chunks": 8055.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.392486, "cps": 62.929688, "mean_p": 0.407146, "boundary_entropy": 0.627133, "bpb": 1.791082}, "hi": {"chunks": 7819.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.464725, "cps": 61.085938, "mean_p": 0.407301, "boundary_entropy": 0.592668, "bpb": 1.340056}, "ta": {"chunks": 12822.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.685852, "cps": 100.171875, "mean_p": 0.437264, "boundary_entropy": 0.626245, "bpb": 1.112843}, "th": {"chunks": 7689.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.478023, "cps": 60.070312, "mean_p": 0.418959, "boundary_entropy": 0.639135, "bpb": 1.347403}, "zh": {"chunks": 3198.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.541483, "cps": 24.984375, "mean_p": 0.318327, "boundary_entropy": 0.491661, "bpb": 3.474294}, "my": {"chunks": 9384.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.49525, "cps": 73.3125, "mean_p": 0.370946, "boundary_entropy": 0.612503, "bpb": 1.308714}}, "bpb": {"en": 2.31304, "de": 2.59813, "vi": 2.13057, "tr": 2.60401, "fi": 2.83184, "ru": 1.70008, "el": 1.79108, "hi": 1.34006, "ta": 1.11284, "th": 1.3474, "zh": 3.47429, "my": 1.30871}, "gini_cpc": 0.246048, "gini_cps": 0.254567, "bpic": 5.206493, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9337940552931833, "vi": 1.2162892659097813, "tr": 1.1280941145170436, "fi": 1.0017832660681092, "ru": 1.6346991514283813, "el": 2.0499526370992776, "hi": 2.42725423601833, "ta": 3.5821967085844766, "th": 2.496708978897745, "zh": 2.828161098506453, "my": 2.5866862261053885}, "t": 110.7} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.32257, "de": 2.60713, "vi": 2.13642, "tr": 2.60933, "fi": 2.8391, "ru": 1.68034, "el": 1.78072, "hi": 1.33685, "ta": 1.11971, "th": 1.36594, "zh": 3.47732, "my": 1.31598}, "bytes_per_chunk": {"en": 5.419, "de": 5.628, "vi": 5.914, "tr": 5.085, "fi": 5.313, "ru": 5.824, "el": 4.891, "hi": 5.52, "ta": 4.981, "th": 5.967, "zh": 5.426, "my": 5.858}, "bpb_mean": 2.04928, "t": 112.0} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.45563, "bpb": 2.10004, "chunk_loss": 0.030533, "F": 0.18436, "G": 0.31034, "chunks": 188.78, "bytes_per_chunk": 5.424, "boundary_entropy": 0.45548, "router_noise": 0.0, "grad_norm": 0.7141, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 112.0} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.44257, "bpb": 2.08118, "chunk_loss": 0.03013, "F": 0.17303, "G": 0.30712, "chunks": 177.19, "bytes_per_chunk": 5.779, "boundary_entropy": 0.44976, "router_noise": 0.0, "grad_norm": 0.7669, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 114.0} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.55168, "bpb": 2.2386, "chunk_loss": 0.030479, "F": 0.18832, "G": 0.28659, "chunks": 192.84, "bytes_per_chunk": 5.31, "boundary_entropy": 0.38988, "router_noise": 0.0, "grad_norm": 0.8003, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 116.0} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.61616, "bpb": 2.33162, "chunk_loss": 0.030215, "F": 0.17792, "G": 0.29513, "chunks": 182.19, "bytes_per_chunk": 5.621, "boundary_entropy": 0.42927, "router_noise": 0.0, "grad_norm": 0.8115, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 117.9} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3400.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195369, "cps": 26.5625, "mean_p": 0.233903, "boundary_entropy": 0.248063, "bpb": 2.280229}, "de": {"chunks": 3796.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.184873, "cps": 29.65625, "mean_p": 0.224162, "boundary_entropy": 0.27118, "bpb": 2.558416}, "vi": {"chunks": 4321.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234391, "cps": 33.757812, "mean_p": 0.27653, "boundary_entropy": 0.404335, "bpb": 2.105584}, "tr": {"chunks": 3714.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205159, "cps": 29.015625, "mean_p": 0.234037, "boundary_entropy": 0.293822, "bpb": 2.555953}, "fi": {"chunks": 3588.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.190739, "cps": 28.03125, "mean_p": 0.222004, "boundary_entropy": 0.269797, "bpb": 2.814152}, "ru": {"chunks": 5812.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306702, "cps": 45.40625, "mean_p": 0.342327, "boundary_entropy": 0.56472, "bpb": 1.657729}, "el": {"chunks": 6519.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.317644, "cps": 50.929688, "mean_p": 0.369335, "boundary_entropy": 0.606399, "bpb": 1.782572}, "hi": {"chunks": 6791.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.403626, "cps": 53.054688, "mean_p": 0.391227, "boundary_entropy": 0.596341, "bpb": 1.319356}, "ta": {"chunks": 8344.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.446323, "cps": 65.1875, "mean_p": 0.426822, "boundary_entropy": 0.626138, "bpb": 1.107064}, "th": {"chunks": 6654.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.413677, "cps": 51.984375, "mean_p": 0.409147, "boundary_entropy": 0.63643, "bpb": 1.327369}, "zh": {"chunks": 3237.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.548087, "cps": 25.289062, "mean_p": 0.312247, "boundary_entropy": 0.477363, "bpb": 3.428008}, "my": {"chunks": 10523.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.555362, "cps": 82.210938, "mean_p": 0.397961, "boundary_entropy": 0.618931, "bpb": 1.279429}}, "bpb": {"en": 2.28023, "de": 2.55842, "vi": 2.10558, "tr": 2.55595, "fi": 2.81415, "ru": 1.65773, "el": 1.78257, "hi": 1.31936, "ta": 1.10706, "th": 1.32737, "zh": 3.42801, "my": 1.27943}, "gini_cpc": 0.220922, "gini_cps": 0.214175, "bpic": 5.753924, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9462785587619357, "vi": 1.1997377590580576, "tr": 1.0501142482071544, "fi": 0.9763055408756454, "ru": 1.5698624243364891, "el": 1.6258682081223075, "hi": 2.065969285901582, "ta": 2.284515079527398, "th": 2.117419627347364, "zh": 2.8053978506404254, "my": 2.8426369537682077}, "t": 120.8} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.28631, "de": 2.56558, "vi": 2.10865, "tr": 2.56083, "fi": 2.81755, "ru": 1.63891, "el": 1.77417, "hi": 1.31665, "ta": 1.11407, "th": 1.34607, "zh": 3.43152, "my": 1.28527}, "bytes_per_chunk": {"en": 5.366, "de": 5.614, "vi": 5.835, "tr": 5.343, "fi": 5.308, "ru": 5.848, "el": 6.005, "hi": 6.527, "ta": 6.733, "th": 6.543, "zh": 5.242, "my": 5.239}, "bpb_mean": 2.02047, "t": 122.1} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.41029, "bpb": 2.03462, "chunk_loss": 0.030734, "F": 0.18332, "G": 0.3219, "chunks": 187.72, "bytes_per_chunk": 5.455, "boundary_entropy": 0.47605, "router_noise": 0.0, "grad_norm": 1.0219, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 122.2} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.58669, "bpb": 2.28911, "chunk_loss": 0.03012, "F": 0.17322, "G": 0.27761, "chunks": 177.38, "bytes_per_chunk": 5.773, "boundary_entropy": 0.39506, "router_noise": 0.0, "grad_norm": 0.8738, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 124.1} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.64502, "bpb": 2.37326, "chunk_loss": 0.030266, "F": 0.17859, "G": 0.2886, "chunks": 182.88, "bytes_per_chunk": 5.599, "boundary_entropy": 0.4097, "router_noise": 0.0, "grad_norm": 0.8962, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 126.1} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.37049, "bpb": 1.97719, "chunk_loss": 0.030311, "F": 0.17816, "G": 0.30174, "chunks": 182.44, "bytes_per_chunk": 5.613, "boundary_entropy": 0.43838, "router_noise": 0.0, "grad_norm": 0.8342, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 128.1} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3237.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186002, "cps": 25.289062, "mean_p": 0.218321, "boundary_entropy": 0.225047, "bpb": 2.241799}, "de": {"chunks": 3647.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.177617, "cps": 28.492188, "mean_p": 0.212192, "boundary_entropy": 0.253871, "bpb": 2.513522}, "vi": {"chunks": 3843.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.208462, "cps": 30.023438, "mean_p": 0.254087, "boundary_entropy": 0.388508, "bpb": 2.083419}, "tr": {"chunks": 3725.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205767, "cps": 29.101562, "mean_p": 0.231679, "boundary_entropy": 0.281042, "bpb": 2.513982}, "fi": {"chunks": 3434.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182553, "cps": 26.828125, "mean_p": 0.211411, "boundary_entropy": 0.254226, "bpb": 2.729237}, "ru": {"chunks": 5421.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.286069, "cps": 42.351562, "mean_p": 0.32338, "boundary_entropy": 0.545276, "bpb": 1.626051}, "el": {"chunks": 6127.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.298543, "cps": 47.867188, "mean_p": 0.353757, "boundary_entropy": 0.592893, "bpb": 1.752688}, "hi": {"chunks": 7391.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.439287, "cps": 57.742188, "mean_p": 0.402583, "boundary_entropy": 0.592959, "bpb": 1.310307}, "ta": {"chunks": 8880.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.474993, "cps": 69.375, "mean_p": 0.409902, "boundary_entropy": 0.613652, "bpb": 1.076997}, "th": {"chunks": 6751.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.419708, "cps": 52.742188, "mean_p": 0.409073, "boundary_entropy": 0.637222, "bpb": 1.294051}, "zh": {"chunks": 2936.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.497122, "cps": 22.9375, "mean_p": 0.302682, "boundary_entropy": 0.460736, "bpb": 3.425902}, "my": {"chunks": 8236.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.434663, "cps": 64.34375, "mean_p": 0.374113, "boundary_entropy": 0.612463, "bpb": 1.252405}}, "bpb": {"en": 2.2418, "de": 2.51352, "vi": 2.08342, "tr": 2.51398, "fi": 2.72924, "ru": 1.62605, "el": 1.75269, "hi": 1.31031, "ta": 1.077, "th": 1.29405, "zh": 3.4259, "my": 1.25241}, "gini_cpc": 0.21234, "gini_cps": 0.212666, "bpic": 6.031637, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9549151353681118, "vi": 1.1207497826318105, "tr": 1.1062598393693752, "fi": 0.981453726151298, "ru": 1.537983278761484, "el": 1.6050495935367977, "hi": 2.3617262144933604, "ta": 2.553694358564548, "th": 2.2564642828025234, "zh": 2.6726619416267274, "my": 2.3368690820181395}, "t": 131.0} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.24958, "de": 2.52357, "vi": 2.08893, "tr": 2.51818, "fi": 2.7348, "ru": 1.60663, "el": 1.74403, "hi": 1.30661, "ta": 1.08278, "th": 1.31077, "zh": 3.42823, "my": 1.2606}, "bytes_per_chunk": {"en": 5.653, "de": 5.869, "vi": 6.524, "tr": 5.275, "fi": 5.593, "ru": 6.225, "el": 6.448, "hi": 6.088, "ta": 6.537, "th": 6.639, "zh": 5.803, "my": 7.148}, "bpb_mean": 1.98789, "t": 132.2} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.2858, "bpb": 1.85502, "chunk_loss": 0.029769, "F": 0.16373, "G": 0.3102, "chunks": 167.66, "bytes_per_chunk": 6.108, "boundary_entropy": 0.45592, "router_noise": 0.0, "grad_norm": 0.72, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 132.3} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.52765, "bpb": 2.20393, "chunk_loss": 0.030346, "F": 0.1785, "G": 0.27395, "chunks": 182.78, "bytes_per_chunk": 5.602, "boundary_entropy": 0.37935, "router_noise": 0.0, "grad_norm": 0.7415, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 134.2} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.47885, "bpb": 2.13353, "chunk_loss": 0.030129, "F": 0.17609, "G": 0.27786, "chunks": 180.31, "bytes_per_chunk": 5.679, "boundary_entropy": 0.38271, "router_noise": 0.0, "grad_norm": 0.8322, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 136.2} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.46302, "bpb": 2.11069, "chunk_loss": 0.029929, "F": 0.16815, "G": 0.29376, "chunks": 172.19, "bytes_per_chunk": 5.947, "boundary_entropy": 0.43324, "router_noise": 0.0, "grad_norm": 0.753, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 138.2} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3363.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193243, "cps": 26.273438, "mean_p": 0.224744, "boundary_entropy": 0.225477, "bpb": 2.217388}, "de": {"chunks": 3838.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186919, "cps": 29.984375, "mean_p": 0.220854, "boundary_entropy": 0.248552, "bpb": 2.465389}, "vi": {"chunks": 4067.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.220613, "cps": 31.773438, "mean_p": 0.251372, "boundary_entropy": 0.374403, "bpb": 2.065581}, "tr": {"chunks": 3681.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203336, "cps": 28.757812, "mean_p": 0.22954, "boundary_entropy": 0.27643, "bpb": 2.482201}, "fi": {"chunks": 3586.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.190633, "cps": 28.015625, "mean_p": 0.220009, "boundary_entropy": 0.252169, "bpb": 2.678627}, "ru": {"chunks": 5931.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.312982, "cps": 46.335938, "mean_p": 0.334895, "boundary_entropy": 0.54902, "bpb": 1.599231}, "el": {"chunks": 6713.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.327096, "cps": 52.445312, "mean_p": 0.3583, "boundary_entropy": 0.589111, "bpb": 1.723055}, "hi": {"chunks": 7873.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.467935, "cps": 61.507812, "mean_p": 0.40185, "boundary_entropy": 0.596742, "bpb": 1.275828}, "ta": {"chunks": 8284.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.443113, "cps": 64.71875, "mean_p": 0.416529, "boundary_entropy": 0.618049, "bpb": 1.058544}, "th": {"chunks": 6547.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.407025, "cps": 51.148438, "mean_p": 0.399009, "boundary_entropy": 0.634292, "bpb": 1.262184}, "zh": {"chunks": 2919.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.494243, "cps": 22.804688, "mean_p": 0.300381, "boundary_entropy": 0.462391, "bpb": 3.38338}, "my": {"chunks": 6531.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.34468, "cps": 51.023438, "mean_p": 0.345086, "boundary_entropy": 0.590832, "bpb": 1.237815}}, "bpb": {"en": 2.21739, "de": 2.46539, "vi": 2.06558, "tr": 2.4822, "fi": 2.67863, "ru": 1.59923, "el": 1.72306, "hi": 1.27583, "ta": 1.05854, "th": 1.26218, "zh": 3.38338, "my": 1.23782}, "gini_cpc": 0.198457, "gini_cps": 0.1918, "bpic": 6.059732, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9672746723546304, "vi": 1.1416376511053254, "tr": 1.0522344559920602, "fi": 0.9864967405950104, "ru": 1.6196305597857172, "el": 1.6926729537318133, "hi": 2.4214886146284695, "ta": 2.293041282677776, "th": 2.1062917590177297, "zh": 2.557631106182406, "my": 1.7836661094347703}, "t": 141.0} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.22607, "de": 2.47542, "vi": 2.07032, "tr": 2.48626, "fi": 2.68773, "ru": 1.57655, "el": 1.71398, "hi": 1.27394, "ta": 1.06593, "th": 1.28112, "zh": 3.3862, "my": 1.24359}, "bytes_per_chunk": {"en": 5.448, "de": 5.47, "vi": 6.114, "tr": 5.341, "fi": 5.339, "ru": 5.695, "el": 5.712, "hi": 5.572, "ta": 6.617, "th": 6.759, "zh": 5.715, "my": 8.393}, "bpb_mean": 1.95726, "t": 142.3} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 142.3} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "parity_a", "lm_loss": 1.319, "bpb": 1.90291, "chunk_loss": 0.036854, "F": 0.17331, "G": 0.29291, "chunks": 177.47, "bytes_per_chunk": 5.77, "boundary_entropy": 0.41691, "router_noise": 0.0, "grad_norm": 0.752, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 10.8309, "alpha": 0.03, "t": 142.4} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "parity_a", "lm_loss": 1.48143, "bpb": 2.13726, "chunk_loss": 0.030164, "F": 0.13327, "G": 0.21494, "chunks": 136.47, "bytes_per_chunk": 7.504, "boundary_entropy": 0.32602, "router_noise": 0.0, "grad_norm": 0.8477, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.4427, "alpha": 0.03, "t": 144.3} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "parity_a", "lm_loss": 1.47361, "bpb": 2.12596, "chunk_loss": 0.03088, "F": 0.14279, "G": 0.222, "chunks": 146.22, "bytes_per_chunk": 7.003, "boundary_entropy": 0.33113, "router_noise": 0.0, "grad_norm": 0.8023, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.5095, "alpha": 0.03, "t": 146.3} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "parity_a", "lm_loss": 1.53286, "bpb": 2.21145, "chunk_loss": 0.030171, "F": 0.1554, "G": 0.20396, "chunks": 159.12, "bytes_per_chunk": 6.435, "boundary_entropy": 0.24826, "router_noise": 0.0, "grad_norm": 0.8355, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 7.5354, "alpha": 0.03, "t": 148.2} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3270.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187899, "cps": 25.546875, "mean_p": 0.211397, "boundary_entropy": 0.182148, "bpb": 2.192632}, "de": {"chunks": 3750.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182633, "cps": 29.296875, "mean_p": 0.206784, "boundary_entropy": 0.20915, "bpb": 2.44641}, "vi": {"chunks": 3373.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.182967, "cps": 26.351562, "mean_p": 0.209906, "boundary_entropy": 0.319468, "bpb": 2.058675}, "tr": {"chunks": 3638.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200961, "cps": 28.421875, "mean_p": 0.216293, "boundary_entropy": 0.232325, "bpb": 2.430413}, "fi": {"chunks": 3409.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181224, "cps": 26.632812, "mean_p": 0.197439, "boundary_entropy": 0.206569, "bpb": 2.646783}, "ru": {"chunks": 3037.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.160264, "cps": 23.726562, "mean_p": 0.17758, "boundary_entropy": 0.351793, "bpb": 1.587419}, "el": {"chunks": 2687.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.130926, "cps": 20.992188, "mean_p": 0.232681, "boundary_entropy": 0.472532, "bpb": 1.695698}, "hi": {"chunks": 3247.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.192987, "cps": 25.367188, "mean_p": 0.157503, "boundary_entropy": 0.334075, "bpb": 1.338998}, "ta": {"chunks": 2741.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.146617, "cps": 21.414062, "mean_p": 0.166765, "boundary_entropy": 0.38348, "bpb": 1.088161}, "th": {"chunks": 2884.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.179297, "cps": 22.53125, "mean_p": 0.235383, "boundary_entropy": 0.490062, "bpb": 1.271823}, "zh": {"chunks": 3239.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.548425, "cps": 25.304688, "mean_p": 0.306029, "boundary_entropy": 0.453261, "bpb": 3.359397}, "my": {"chunks": 2632.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.138906, "cps": 20.5625, "mean_p": 0.174997, "boundary_entropy": 0.375126, "bpb": 1.236436}}, "bpb": {"en": 2.19263, "de": 2.44641, "vi": 2.05867, "tr": 2.43041, "fi": 2.64678, "ru": 1.58742, "el": 1.6957, "hi": 1.339, "ta": 1.08816, "th": 1.27182, "zh": 3.3594, "my": 1.23644}, "gini_cpc": 0.192695, "gini_cps": 0.06306, "bpic": 10.124278, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9719752986577437, "vi": 0.9737546997509225, "tr": 1.0695190163118586, "fi": 0.9644761336695932, "ru": 0.852927162257026, "el": 0.6967920529528965, "hi": 1.0270783701623567, "ta": 0.7802969935051074, "th": 0.9542244897377179, "zh": 2.9187296700292347, "my": 0.7392628400663913}, "t": 151.1} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.19981, "de": 2.454, "vi": 2.06126, "tr": 2.43687, "fi": 2.65628, "ru": 1.56631, "el": 1.68499, "hi": 1.33573, "ta": 1.09481, "th": 1.29015, "zh": 3.36047, "my": 1.24408}, "bytes_per_chunk": {"en": 5.646, "de": 5.696, "vi": 7.35, "tr": 5.456, "fi": 5.654, "ru": 11.278, "el": 13.803, "hi": 12.627, "ta": 19.274, "th": 15.7, "zh": 5.361, "my": 20.883}, "bpb_mean": 1.94873, "t": 152.4} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "parity_a", "lm_loss": 1.50823, "bpb": 2.17591, "chunk_loss": 0.030179, "F": 0.14114, "G": 0.21065, "chunks": 144.53, "bytes_per_chunk": 7.085, "boundary_entropy": 0.29134, "router_noise": 0.0, "grad_norm": 0.8034, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.2882, "alpha": 0.03, "t": 152.4} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "parity_a", "lm_loss": 1.45067, "bpb": 2.09287, "chunk_loss": 0.030369, "F": 0.14926, "G": 0.20597, "chunks": 152.84, "bytes_per_chunk": 6.7, "boundary_entropy": 0.27051, "router_noise": 0.0, "grad_norm": 0.8234, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.0661, "alpha": 0.03, "t": 154.4} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "parity_a", "lm_loss": 1.50028, "bpb": 2.16444, "chunk_loss": 0.030216, "F": 0.14969, "G": 0.20045, "chunks": 153.28, "bytes_per_chunk": 6.681, "boundary_entropy": 0.24866, "router_noise": 0.0, "grad_norm": 0.8675, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.2992, "alpha": 0.03, "t": 156.3} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "parity_a", "lm_loss": 1.53301, "bpb": 2.21166, "chunk_loss": 0.030012, "F": 0.14624, "G": 0.20079, "chunks": 149.75, "bytes_per_chunk": 6.838, "boundary_entropy": 0.25894, "router_noise": 0.0, "grad_norm": 0.8105, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.0733, "alpha": 0.03, "t": 158.3} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3318.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.190657, "cps": 25.921875, "mean_p": 0.212932, "boundary_entropy": 0.172255, "bpb": 2.172307}, "de": {"chunks": 3746.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182438, "cps": 29.265625, "mean_p": 0.200291, "boundary_entropy": 0.184332, "bpb": 2.411186}, "vi": {"chunks": 3439.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.186547, "cps": 26.867188, "mean_p": 0.204418, "boundary_entropy": 0.300284, "bpb": 2.050559}, "tr": {"chunks": 3436.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.189803, "cps": 26.84375, "mean_p": 0.205766, "boundary_entropy": 0.214856, "bpb": 2.407928}, "fi": {"chunks": 3342.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177662, "cps": 26.109375, "mean_p": 0.191183, "boundary_entropy": 0.188827, "bpb": 2.625344}, "ru": {"chunks": 3034.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.160106, "cps": 23.703125, "mean_p": 0.162327, "boundary_entropy": 0.318843, "bpb": 1.570802}, "el": {"chunks": 2685.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.130829, "cps": 20.976562, "mean_p": 0.220821, "boundary_entropy": 0.452511, "bpb": 1.686567}, "hi": {"chunks": 2778.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.165111, "cps": 21.703125, "mean_p": 0.13438, "boundary_entropy": 0.298009, "bpb": 1.3044}, "ta": {"chunks": 2446.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.130837, "cps": 19.109375, "mean_p": 0.149714, "boundary_entropy": 0.349077, "bpb": 1.082368}, "th": {"chunks": 2373.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.147529, "cps": 18.539062, "mean_p": 0.202093, "boundary_entropy": 0.441872, "bpb": 1.236848}, "zh": {"chunks": 3033.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.513546, "cps": 23.695312, "mean_p": 0.28619, "boundary_entropy": 0.421555, "bpb": 3.320548}, "my": {"chunks": 2811.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.148353, "cps": 21.960938, "mean_p": 0.149495, "boundary_entropy": 0.326902, "bpb": 1.222912}}, "bpb": {"en": 2.17231, "de": 2.41119, "vi": 2.05056, "tr": 2.40793, "fi": 2.62534, "ru": 1.5708, "el": 1.68657, "hi": 1.3044, "ta": 1.08237, "th": 1.23685, "zh": 3.32055, "my": 1.22291}, "gini_cpc": 0.190621, "gini_cps": 0.076926, "bpic": 10.531572, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9568923980618153, "vi": 0.9784457977812666, "tr": 0.9955208087456955, "fi": 0.9318420419432553, "ru": 0.8397579048318837, "el": 0.6862007420951168, "hi": 0.8660139882687704, "ta": 0.6862442549985079, "th": 0.7737923137761025, "zh": 2.6935603233050154, "my": 0.7781175452680812}, "t": 161.1} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.18239, "de": 2.42084, "vi": 2.05338, "tr": 2.41167, "fi": 2.63306, "ru": 1.55009, "el": 1.67708, "hi": 1.30135, "ta": 1.08714, "th": 1.25574, "zh": 3.31978, "my": 1.23136}, "bytes_per_chunk": {"en": 5.533, "de": 5.672, "vi": 7.291, "tr": 5.775, "fi": 5.705, "ru": 11.327, "el": 13.649, "hi": 14.929, "ta": 21.986, "th": 18.847, "zh": 5.594, "my": 17.415}, "bpb_mean": 1.92699, "t": 162.4} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "parity_a", "lm_loss": 1.40935, "bpb": 2.03327, "chunk_loss": 0.03006, "F": 0.13928, "G": 0.19875, "chunks": 142.62, "bytes_per_chunk": 7.18, "boundary_entropy": 0.26717, "router_noise": 0.0, "grad_norm": 0.789, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 8.7824, "alpha": 0.03, "t": 162.4} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "parity_a", "lm_loss": 1.33187, "bpb": 1.92148, "chunk_loss": 0.030159, "F": 0.13049, "G": 0.19439, "chunks": 133.62, "bytes_per_chunk": 7.663, "boundary_entropy": 0.27104, "router_noise": 0.0, "grad_norm": 0.7906, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.986, "alpha": 0.03, "t": 164.4} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "parity_a", "lm_loss": 1.41097, "bpb": 2.0356, "chunk_loss": 0.030218, "F": 0.14859, "G": 0.20113, "chunks": 152.16, "bytes_per_chunk": 6.73, "boundary_entropy": 0.25249, "router_noise": 0.0, "grad_norm": 0.7608, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.5507, "alpha": 0.03, "t": 166.4} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "parity_a", "lm_loss": 1.30772, "bpb": 1.88664, "chunk_loss": 0.0301, "F": 0.12881, "G": 0.1959, "chunks": 131.91, "bytes_per_chunk": 7.763, "boundary_entropy": 0.28411, "router_noise": 0.0, "grad_norm": 0.7205, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.6732, "alpha": 0.03, "t": 168.3} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3244.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186405, "cps": 25.34375, "mean_p": 0.20815, "boundary_entropy": 0.162621, "bpb": 2.161181}, "de": {"chunks": 3628.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.176691, "cps": 28.34375, "mean_p": 0.196995, "boundary_entropy": 0.179296, "bpb": 2.400985}, "vi": {"chunks": 3231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.175264, "cps": 25.242188, "mean_p": 0.194412, "boundary_entropy": 0.287819, "bpb": 2.026867}, "tr": {"chunks": 3534.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195216, "cps": 27.609375, "mean_p": 0.208106, "boundary_entropy": 0.21525, "bpb": 2.390825}, "fi": {"chunks": 3416.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181596, "cps": 26.6875, "mean_p": 0.195581, "boundary_entropy": 0.190538, "bpb": 2.608835}, "ru": {"chunks": 3154.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.166438, "cps": 24.640625, "mean_p": 0.166578, "boundary_entropy": 0.31577, "bpb": 1.537768}, "el": {"chunks": 2594.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.126395, "cps": 20.265625, "mean_p": 0.20403, "boundary_entropy": 0.425124, "bpb": 1.652178}, "hi": {"chunks": 3275.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.194651, "cps": 25.585938, "mean_p": 0.151206, "boundary_entropy": 0.311254, "bpb": 1.27691}, "ta": {"chunks": 3445.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.184274, "cps": 26.914062, "mean_p": 0.178295, "boundary_entropy": 0.379129, "bpb": 1.057168}, "th": {"chunks": 3548.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.220578, "cps": 27.71875, "mean_p": 0.232686, "boundary_entropy": 0.466624, "bpb": 1.209352}, "zh": {"chunks": 3390.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.573993, "cps": 26.484375, "mean_p": 0.307224, "boundary_entropy": 0.431335, "bpb": 3.306878}, "my": {"chunks": 3235.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.17073, "cps": 25.273438, "mean_p": 0.151894, "boundary_entropy": 0.319583, "bpb": 1.198434}}, "bpb": {"en": 2.16118, "de": 2.40098, "vi": 2.02687, "tr": 2.39082, "fi": 2.60883, "ru": 1.53777, "el": 1.65218, "hi": 1.27691, "ta": 1.05717, "th": 1.20935, "zh": 3.30688, "my": 1.19843}, "gini_cpc": 0.18647, "gini_cps": 0.039305, "bpic": 9.668489, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9478904458811392, "vi": 0.9402364658444357, "tr": 1.0472714600035296, "fi": 0.9742025303163317, "ru": 0.8928854568938311, "el": 0.6780666760273112, "hi": 1.0442380309379025, "ta": 0.9885690998569087, "th": 1.1833298517777207, "zh": 3.0792824743421705, "my": 0.9159129164457361}, "t": 171.2} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.16969, "de": 2.40988, "vi": 2.02983, "tr": 2.39606, "fi": 2.61927, "ru": 1.51495, "el": 1.6437, "hi": 1.27405, "ta": 1.06182, "th": 1.2288, "zh": 3.30496, "my": 1.20619}, "bytes_per_chunk": {"en": 5.611, "de": 5.789, "vi": 7.941, "tr": 5.601, "fi": 5.568, "ru": 10.858, "el": 14.11, "hi": 12.834, "ta": 15.619, "th": 13.08, "zh": 5.096, "my": 16.078}, "bpb_mean": 1.90493, "t": 172.5} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "parity_a", "lm_loss": 1.51496, "bpb": 2.18563, "chunk_loss": 0.030692, "F": 0.14981, "G": 0.21456, "chunks": 153.41, "bytes_per_chunk": 6.675, "boundary_entropy": 0.29116, "router_noise": 0.0, "grad_norm": 0.733, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.0106, "alpha": 0.03, "t": 172.5} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "parity_a", "lm_loss": 1.47558, "bpb": 2.12881, "chunk_loss": 0.030231, "F": 0.14178, "G": 0.20089, "chunks": 145.19, "bytes_per_chunk": 7.053, "boundary_entropy": 0.27396, "router_noise": 0.0, "grad_norm": 0.8847, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.0983, "alpha": 0.03, "t": 174.5} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "parity_a", "lm_loss": 1.41988, "bpb": 2.04846, "chunk_loss": 0.030088, "F": 0.12695, "G": 0.19228, "chunks": 130.0, "bytes_per_chunk": 7.877, "boundary_entropy": 0.28243, "router_noise": 0.0, "grad_norm": 0.7867, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.8865, "alpha": 0.03, "t": 176.4} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "parity_a", "lm_loss": 1.32891, "bpb": 1.91721, "chunk_loss": 0.030123, "F": 0.12399, "G": 0.18908, "chunks": 126.97, "bytes_per_chunk": 8.065, "boundary_entropy": 0.276, "router_noise": 0.0, "grad_norm": 0.8235, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.6709, "alpha": 0.03, "t": 178.4} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3242.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.18629, "cps": 25.328125, "mean_p": 0.201451, "boundary_entropy": 0.146043, "bpb": 2.140057}, "de": {"chunks": 3554.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.173087, "cps": 27.765625, "mean_p": 0.185732, "boundary_entropy": 0.163099, "bpb": 2.37167}, "vi": {"chunks": 3286.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.178248, "cps": 25.671875, "mean_p": 0.191512, "boundary_entropy": 0.275205, "bpb": 2.011698}, "tr": {"chunks": 3300.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.18229, "cps": 25.78125, "mean_p": 0.192482, "boundary_entropy": 0.192069, "bpb": 2.353939}, "fi": {"chunks": 3298.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175323, "cps": 25.765625, "mean_p": 0.182722, "boundary_entropy": 0.167723, "bpb": 2.562608}, "ru": {"chunks": 3052.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.161055, "cps": 23.84375, "mean_p": 0.157463, "boundary_entropy": 0.302111, "bpb": 1.525291}, "el": {"chunks": 3028.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.147542, "cps": 23.65625, "mean_p": 0.205849, "boundary_entropy": 0.417561, "bpb": 1.637974}, "hi": {"chunks": 2834.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.16844, "cps": 22.140625, "mean_p": 0.135364, "boundary_entropy": 0.292638, "bpb": 1.277415}, "ta": {"chunks": 2905.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.155389, "cps": 22.695312, "mean_p": 0.157832, "boundary_entropy": 0.356262, "bpb": 1.056881}, "th": {"chunks": 2943.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.182965, "cps": 22.992188, "mean_p": 0.215838, "boundary_entropy": 0.444518, "bpb": 1.195867}, "zh": {"chunks": 3222.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545547, "cps": 25.171875, "mean_p": 0.296271, "boundary_entropy": 0.418038, "bpb": 3.271544}, "my": {"chunks": 2397.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.126504, "cps": 18.726562, "mean_p": 0.158989, "boundary_entropy": 0.348807, "bpb": 1.183051}}, "bpb": {"en": 2.14006, "de": 2.37167, "vi": 2.0117, "tr": 2.35394, "fi": 2.56261, "ru": 1.52529, "el": 1.63797, "hi": 1.27742, "ta": 1.05688, "th": 1.19587, "zh": 3.27154, "my": 1.18305}, "gini_cpc": 0.184983, "gini_cps": 0.049915, "bpic": 10.355387, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9291292363869924, "vi": 0.9568316376444439, "tr": 0.9785307958984306, "fi": 0.941130563643393, "ru": 0.8645426533997222, "el": 0.7920017393404303, "hi": 0.9041820533433308, "ta": 0.8341262273922486, "th": 0.9821556210577023, "zh": 2.9284863436277955, "my": 0.6790719123880324}, "t": 181.2} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.14805, "de": 2.37867, "vi": 2.01306, "tr": 2.36103, "fi": 2.57248, "ru": 1.50283, "el": 1.62882, "hi": 1.27268, "ta": 1.06251, "th": 1.21571, "zh": 3.26915, "my": 1.19074}, "bytes_per_chunk": {"en": 5.699, "de": 6.01, "vi": 7.636, "tr": 6.033, "fi": 5.795, "ru": 11.23, "el": 12.014, "hi": 14.592, "ta": 18.685, "th": 15.377, "zh": 5.387, "my": 22.018}, "bpb_mean": 1.88464, "t": 182.5} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "parity_a", "lm_loss": 1.40697, "bpb": 2.02984, "chunk_loss": 0.030152, "F": 0.13306, "G": 0.18726, "chunks": 136.25, "bytes_per_chunk": 7.516, "boundary_entropy": 0.25436, "router_noise": 0.0, "grad_norm": 0.753, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.2811, "alpha": 0.03, "t": 182.5} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "parity_a", "lm_loss": 1.38457, "bpb": 1.99751, "chunk_loss": 0.030203, "F": 0.13898, "G": 0.19449, "chunks": 142.31, "bytes_per_chunk": 7.195, "boundary_entropy": 0.24631, "router_noise": 0.0, "grad_norm": 0.796, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.2351, "alpha": 0.03, "t": 184.5} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "parity_a", "lm_loss": 1.28722, "bpb": 1.85706, "chunk_loss": 0.030056, "F": 0.11487, "G": 0.17921, "chunks": 117.62, "bytes_per_chunk": 8.706, "boundary_entropy": 0.27246, "router_noise": 0.0, "grad_norm": 0.8092, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.7221, "alpha": 0.03, "t": 186.4} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "parity_a", "lm_loss": 1.36508, "bpb": 1.9694, "chunk_loss": 0.030006, "F": 0.12823, "G": 0.19142, "chunks": 131.31, "bytes_per_chunk": 7.798, "boundary_entropy": 0.27182, "router_noise": 0.0, "grad_norm": 0.7224, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.0553, "alpha": 0.03, "t": 188.4} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3292.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189163, "cps": 25.71875, "mean_p": 0.205166, "boundary_entropy": 0.14699, "bpb": 2.121909}, "de": {"chunks": 3665.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178493, "cps": 28.632812, "mean_p": 0.193661, "boundary_entropy": 0.165079, "bpb": 2.351669}, "vi": {"chunks": 3139.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.170274, "cps": 24.523438, "mean_p": 0.188695, "boundary_entropy": 0.27382, "bpb": 2.003162}, "tr": {"chunks": 3317.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.183229, "cps": 25.914062, "mean_p": 0.194769, "boundary_entropy": 0.191499, "bpb": 2.335673}, "fi": {"chunks": 3405.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181011, "cps": 26.601562, "mean_p": 0.189537, "boundary_entropy": 0.168864, "bpb": 2.54682}, "ru": {"chunks": 3232.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.170554, "cps": 25.25, "mean_p": 0.160124, "boundary_entropy": 0.3003, "bpb": 1.509738}, "el": {"chunks": 3208.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.156312, "cps": 25.0625, "mean_p": 0.209091, "boundary_entropy": 0.419647, "bpb": 1.610984}, "hi": {"chunks": 3194.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.189837, "cps": 24.953125, "mean_p": 0.149492, "boundary_entropy": 0.308092, "bpb": 1.265051}, "ta": {"chunks": 3046.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.162931, "cps": 23.796875, "mean_p": 0.164925, "boundary_entropy": 0.361674, "bpb": 1.048259}, "th": {"chunks": 3282.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.204041, "cps": 25.640625, "mean_p": 0.211239, "boundary_entropy": 0.431941, "bpb": 1.17974}, "zh": {"chunks": 3299.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.558584, "cps": 25.773438, "mean_p": 0.293318, "boundary_entropy": 0.413152, "bpb": 3.248767}, "my": {"chunks": 2423.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.127876, "cps": 18.929688, "mean_p": 0.15622, "boundary_entropy": 0.343566, "bpb": 1.192374}}, "bpb": {"en": 2.12191, "de": 2.35167, "vi": 2.00316, "tr": 2.33567, "fi": 2.54682, "ru": 1.50974, "el": 1.61098, "hi": 1.26505, "ta": 1.04826, "th": 1.17974, "zh": 3.24877, "my": 1.19237}, "gini_cpc": 0.185073, "gini_cps": 0.04105, "bpic": 9.96782, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9435955095608475, "vi": 0.90014498742915, "tr": 0.9686328951069715, "fi": 0.9569065481127643, "ru": 0.9016260094832921, "el": 0.8263381587288281, "hi": 1.0035618248751081, "ta": 0.8613283127469576, "th": 1.078653122782182, "zh": 2.952930098888253, "my": 0.6760118855697925}, "t": 191.3} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.1311, "de": 2.35971, "vi": 2.00465, "tr": 2.34337, "fi": 2.55622, "ru": 1.48793, "el": 1.60244, "hi": 1.26154, "ta": 1.05272, "th": 1.19907, "zh": 3.24653, "my": 1.19993}, "bytes_per_chunk": {"en": 5.639, "de": 5.876, "vi": 7.864, "tr": 5.968, "fi": 5.732, "ru": 10.632, "el": 11.596, "hi": 13.007, "ta": 17.455, "th": 15.105, "zh": 5.139, "my": 22.558}, "bpb_mean": 1.87044, "t": 192.5} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "parity_a", "lm_loss": 1.3118, "bpb": 1.89252, "chunk_loss": 0.030177, "F": 0.1243, "G": 0.18777, "chunks": 127.28, "bytes_per_chunk": 8.045, "boundary_entropy": 0.27257, "router_noise": 0.0, "grad_norm": 0.7891, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 10.5619, "alpha": 0.03, "t": 192.6} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "parity_a", "lm_loss": 1.22105, "bpb": 1.7616, "chunk_loss": 0.030084, "F": 0.11829, "G": 0.17611, "chunks": 121.12, "bytes_per_chunk": 8.454, "boundary_entropy": 0.25035, "router_noise": 0.0, "grad_norm": 0.7183, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.9543, "alpha": 0.03, "t": 194.5} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "parity_a", "lm_loss": 1.33431, "bpb": 1.92501, "chunk_loss": 0.030142, "F": 0.13153, "G": 0.18601, "chunks": 134.69, "bytes_per_chunk": 7.603, "boundary_entropy": 0.24242, "router_noise": 0.0, "grad_norm": 0.8184, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 10.0046, "alpha": 0.03, "t": 196.5} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "parity_a", "lm_loss": 1.37664, "bpb": 1.98607, "chunk_loss": 0.030006, "F": 0.12955, "G": 0.18552, "chunks": 132.66, "bytes_per_chunk": 7.719, "boundary_entropy": 0.25257, "router_noise": 0.0, "grad_norm": 0.8342, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.7442, "alpha": 0.03, "t": 198.5} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3248.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186634, "cps": 25.375, "mean_p": 0.200111, "boundary_entropy": 0.134577, "bpb": 2.109067}, "de": {"chunks": 3628.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.176691, "cps": 28.34375, "mean_p": 0.186938, "boundary_entropy": 0.15136, "bpb": 2.325045}, "vi": {"chunks": 3182.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.172606, "cps": 24.859375, "mean_p": 0.188336, "boundary_entropy": 0.272271, "bpb": 1.991025}, "tr": {"chunks": 3403.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.18798, "cps": 26.585938, "mean_p": 0.200458, "boundary_entropy": 0.189056, "bpb": 2.316257}, "fi": {"chunks": 3497.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185902, "cps": 27.320312, "mean_p": 0.192505, "boundary_entropy": 0.161696, "bpb": 2.507707}, "ru": {"chunks": 3083.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.162691, "cps": 24.085938, "mean_p": 0.147854, "boundary_entropy": 0.269319, "bpb": 1.494243}, "el": {"chunks": 2312.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.112654, "cps": 18.0625, "mean_p": 0.176871, "boundary_entropy": 0.380723, "bpb": 1.608072}, "hi": {"chunks": 3003.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.178484, "cps": 23.460938, "mean_p": 0.135837, "boundary_entropy": 0.286466, "bpb": 1.254439}, "ta": {"chunks": 2980.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.159401, "cps": 23.28125, "mean_p": 0.157757, "boundary_entropy": 0.347075, "bpb": 1.050253}, "th": {"chunks": 2922.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.18166, "cps": 22.828125, "mean_p": 0.197399, "boundary_entropy": 0.40929, "bpb": 1.169536}, "zh": {"chunks": 3093.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.523705, "cps": 24.164062, "mean_p": 0.284264, "boundary_entropy": 0.396873, "bpb": 3.231543}, "my": {"chunks": 2463.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.129987, "cps": 19.242188, "mean_p": 0.155283, "boundary_entropy": 0.347552, "bpb": 1.173074}}, "bpb": {"en": 2.10907, "de": 2.32504, "vi": 1.99103, "tr": 2.31626, "fi": 2.50771, "ru": 1.49424, "el": 1.60807, "hi": 1.25444, "ta": 1.05025, "th": 1.16954, "zh": 3.23154, "my": 1.17307}, "gini_cpc": 0.19007, "gini_cps": 0.065727, "bpic": 10.424866, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9467230931152759, "vi": 0.9248368652474878, "tr": 1.0072087669381746, "fi": 0.9960745605822816, "ru": 0.8717107666012451, "el": 0.6036081349556391, "hi": 0.9563312752984576, "ta": 0.8540806727649052, "th": 0.9733459944169751, "zh": 2.806044648174437, "my": 0.6964807787400518}, "t": 201.3} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.11775, "de": 2.33664, "vi": 1.99358, "tr": 2.32148, "fi": 2.5159, "ru": 1.47355, "el": 1.59986, "hi": 1.24987, "ta": 1.05639, "th": 1.18746, "zh": 3.22939, "my": 1.18025}, "bytes_per_chunk": {"en": 5.704, "de": 5.943, "vi": 7.903, "tr": 5.791, "fi": 5.541, "ru": 11.304, "el": 15.858, "hi": 13.902, "ta": 18.47, "th": 16.385, "zh": 5.499, "my": 21.407}, "bpb_mean": 1.85518, "t": 202.6} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "parity_a", "lm_loss": 1.32283, "bpb": 1.90843, "chunk_loss": 0.030011, "F": 0.11908, "G": 0.18, "chunks": 121.94, "bytes_per_chunk": 8.398, "boundary_entropy": 0.25986, "router_noise": 0.0, "grad_norm": 0.7883, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 10.3328, "alpha": 0.03, "t": 202.7} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "parity_a", "lm_loss": 1.31444, "bpb": 1.89633, "chunk_loss": 0.030071, "F": 0.13388, "G": 0.17855, "chunks": 137.09, "bytes_per_chunk": 7.469, "boundary_entropy": 0.21745, "router_noise": 0.0, "grad_norm": 0.7316, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.3376, "alpha": 0.03, "t": 204.6} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "parity_a", "lm_loss": 1.52998, "bpb": 2.20729, "chunk_loss": 0.030207, "F": 0.15115, "G": 0.20163, "chunks": 154.78, "bytes_per_chunk": 6.616, "boundary_entropy": 0.24225, "router_noise": 0.0, "grad_norm": 0.8202, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.3595, "alpha": 0.03, "t": 206.6} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "parity_a", "lm_loss": 1.42907, "bpb": 2.06172, "chunk_loss": 0.030346, "F": 0.1557, "G": 0.18867, "chunks": 159.44, "bytes_per_chunk": 6.423, "boundary_entropy": 0.18771, "router_noise": 0.0, "grad_norm": 0.7849, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.2469, "alpha": 0.03, "t": 208.5} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3283.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188646, "cps": 25.648438, "mean_p": 0.203556, "boundary_entropy": 0.134681, "bpb": 2.099728}, "de": {"chunks": 3691.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.179759, "cps": 28.835938, "mean_p": 0.190715, "boundary_entropy": 0.149012, "bpb": 2.307678}, "vi": {"chunks": 3195.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.173312, "cps": 24.960938, "mean_p": 0.184632, "boundary_entropy": 0.260791, "bpb": 1.980946}, "tr": {"chunks": 3233.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178589, "cps": 25.257812, "mean_p": 0.191128, "boundary_entropy": 0.177057, "bpb": 2.289518}, "fi": {"chunks": 3435.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182606, "cps": 26.835938, "mean_p": 0.188065, "boundary_entropy": 0.155225, "bpb": 2.489198}, "ru": {"chunks": 2990.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.157784, "cps": 23.359375, "mean_p": 0.143407, "boundary_entropy": 0.267006, "bpb": 1.475311}, "el": {"chunks": 3000.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.146177, "cps": 23.4375, "mean_p": 0.193489, "boundary_entropy": 0.39469, "bpb": 1.579683}, "hi": {"chunks": 3018.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.179376, "cps": 23.578125, "mean_p": 0.141383, "boundary_entropy": 0.291711, "bpb": 1.241476}, "ta": {"chunks": 3118.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.166783, "cps": 24.359375, "mean_p": 0.171487, "boundary_entropy": 0.366083, "bpb": 1.026139}, "th": {"chunks": 3459.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.215045, "cps": 27.023438, "mean_p": 0.210248, "boundary_entropy": 0.420153, "bpb": 1.153796}, "zh": {"chunks": 3151.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.533525, "cps": 24.617188, "mean_p": 0.280883, "boundary_entropy": 0.387924, "bpb": 3.211474}, "my": {"chunks": 3652.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.192738, "cps": 28.53125, "mean_p": 0.164834, "boundary_entropy": 0.337889, "bpb": 1.159177}}, "bpb": {"en": 2.09973, "de": 2.30768, "vi": 1.98095, "tr": 2.28952, "fi": 2.4892, "ru": 1.47531, "el": 1.57968, "hi": 1.24148, "ta": 1.02614, "th": 1.1538, "zh": 3.21147, "my": 1.15918}, "gini_cpc": 0.168614, "gini_cps": 0.039696, "bpic": 9.784092, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9528946212139745, "vi": 0.9187153189731255, "tr": 0.9466913101190402, "fi": 0.9679838141120435, "ru": 0.8364022866658576, "el": 0.7748785642888956, "hi": 0.9508617992006189, "ta": 0.8841050653854209, "th": 1.1399419452571253, "zh": 2.8281874971053766, "my": 1.0216934996470775}, "t": 211.4} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.10685, "de": 2.31877, "vi": 1.98308, "tr": 2.29819, "fi": 2.4991, "ru": 1.45318, "el": 1.57247, "hi": 1.23749, "ta": 1.03082, "th": 1.17287, "zh": 3.2084, "my": 1.16703}, "bytes_per_chunk": {"en": 5.636, "de": 5.787, "vi": 7.841, "tr": 6.005, "fi": 5.567, "ru": 11.504, "el": 11.916, "hi": 13.857, "ta": 17.503, "th": 13.268, "zh": 5.331, "my": 13.106}, "bpb_mean": 1.83735, "t": 212.7} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "parity_a", "lm_loss": 1.38101, "bpb": 1.99237, "chunk_loss": 0.030179, "F": 0.13486, "G": 0.18342, "chunks": 138.09, "bytes_per_chunk": 7.415, "boundary_entropy": 0.23619, "router_noise": 0.0, "grad_norm": 0.8228, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 9.6644, "alpha": 0.03, "t": 212.7} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "parity_a", "lm_loss": 1.53035, "bpb": 2.20782, "chunk_loss": 0.030466, "F": 0.15454, "G": 0.19675, "chunks": 158.25, "bytes_per_chunk": 6.471, "boundary_entropy": 0.22819, "router_noise": 0.0, "grad_norm": 0.7552, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.3671, "alpha": 0.03, "t": 214.7} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "parity_a", "lm_loss": 1.4703, "bpb": 2.12119, "chunk_loss": 0.030152, "F": 0.15131, "G": 0.1941, "chunks": 154.94, "bytes_per_chunk": 6.609, "boundary_entropy": 0.21228, "router_noise": 0.0, "grad_norm": 0.7519, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.3124, "alpha": 0.03, "t": 216.7} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "parity_a", "lm_loss": 1.42856, "bpb": 2.06097, "chunk_loss": 0.029999, "F": 0.13812, "G": 0.18586, "chunks": 141.44, "bytes_per_chunk": 7.24, "boundary_entropy": 0.22641, "router_noise": 0.0, "grad_norm": 0.8076, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.5064, "alpha": 0.03, "t": 218.6} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3277.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188301, "cps": 25.601562, "mean_p": 0.202318, "boundary_entropy": 0.129479, "bpb": 2.08783}, "de": {"chunks": 3690.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.179711, "cps": 28.828125, "mean_p": 0.190727, "boundary_entropy": 0.145367, "bpb": 2.297417}, "vi": {"chunks": 3184.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.172715, "cps": 24.875, "mean_p": 0.183711, "boundary_entropy": 0.253342, "bpb": 1.971144}, "tr": {"chunks": 3371.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.186212, "cps": 26.335938, "mean_p": 0.194724, "boundary_entropy": 0.17483, "bpb": 2.272429}, "fi": {"chunks": 3561.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.189304, "cps": 27.820312, "mean_p": 0.194498, "boundary_entropy": 0.153187, "bpb": 2.468295}, "ru": {"chunks": 3101.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.163641, "cps": 24.226562, "mean_p": 0.146664, "boundary_entropy": 0.26554, "bpb": 1.464915}, "el": {"chunks": 3229.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.157336, "cps": 25.226562, "mean_p": 0.198402, "boundary_entropy": 0.395929, "bpb": 1.570822}, "hi": {"chunks": 3042.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.180802, "cps": 23.765625, "mean_p": 0.132312, "boundary_entropy": 0.270985, "bpb": 1.236821}, "ta": {"chunks": 3409.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.182348, "cps": 26.632812, "mean_p": 0.168431, "boundary_entropy": 0.350157, "bpb": 1.01609}, "th": {"chunks": 2774.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.172459, "cps": 21.671875, "mean_p": 0.186237, "boundary_entropy": 0.391509, "bpb": 1.144395}, "zh": {"chunks": 3302.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.559092, "cps": 25.796875, "mean_p": 0.286626, "boundary_entropy": 0.386215, "bpb": 3.184837}, "my": {"chunks": 3048.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.160861, "cps": 23.8125, "mean_p": 0.157248, "boundary_entropy": 0.327004, "bpb": 1.161619}}, "bpb": {"en": 2.08783, "de": 2.29742, "vi": 1.97114, "tr": 2.27243, "fi": 2.4683, "ru": 1.46492, "el": 1.57082, "hi": 1.23682, "ta": 1.01609, "th": 1.14439, "zh": 3.18484, "my": 1.16162}, "gini_cpc": 0.165367, "gini_cps": 0.040666, "bpic": 9.843567, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9543806771391664, "vi": 0.9172286168385668, "tr": 0.988907976499241, "fi": 1.0053279780802067, "ru": 0.8690409289016033, "el": 0.8355546850232605, "hi": 0.960178142857985, "ta": 0.9683875794589978, "th": 0.9158683902812994, "zh": 2.969144302339748, "my": 0.8542780849426879}, "t": 221.5} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.0956, "de": 2.30774, "vi": 1.97269, "tr": 2.27964, "fi": 2.47912, "ru": 1.44243, "el": 1.56287, "hi": 1.2318, "ta": 1.02123, "th": 1.16444, "zh": 3.18076, "my": 1.16983}, "bytes_per_chunk": {"en": 5.644, "de": 5.83, "vi": 7.938, "tr": 5.864, "fi": 5.403, "ru": 11.012, "el": 11.217, "hi": 13.796, "ta": 15.884, "th": 16.906, "zh": 5.159, "my": 16.053}, "bpb_mean": 1.82568, "t": 222.8} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "parity_a", "lm_loss": 1.31027, "bpb": 1.89033, "chunk_loss": 0.030311, "F": 0.13266, "G": 0.17834, "chunks": 135.84, "bytes_per_chunk": 7.538, "boundary_entropy": 0.22551, "router_noise": 0.0, "grad_norm": 0.7844, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 10.021, "alpha": 0.03, "t": 222.8} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "parity_a", "lm_loss": 1.4642, "bpb": 2.1124, "chunk_loss": 0.030295, "F": 0.15186, "G": 0.19659, "chunks": 155.5, "bytes_per_chunk": 6.585, "boundary_entropy": 0.23078, "router_noise": 0.0, "grad_norm": 0.7679, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.3995, "alpha": 0.03, "t": 224.8} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "parity_a", "lm_loss": 1.33453, "bpb": 1.92532, "chunk_loss": 0.030188, "F": 0.1322, "G": 0.18167, "chunks": 135.38, "bytes_per_chunk": 7.564, "boundary_entropy": 0.23115, "router_noise": 0.0, "grad_norm": 0.7864, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 9.8635, "alpha": 0.03, "t": 226.7} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "parity_a", "lm_loss": 1.48724, "bpb": 2.14564, "chunk_loss": 0.03015, "F": 0.15002, "G": 0.18866, "chunks": 153.62, "bytes_per_chunk": 6.666, "boundary_entropy": 0.20951, "router_noise": 0.0, "grad_norm": 0.7804, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.039, "n_target_mean": 8.0388, "alpha": 0.03, "t": 228.7} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3260.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187324, "cps": 25.46875, "mean_p": 0.200085, "boundary_entropy": 0.120432, "bpb": 2.077635}, "de": {"chunks": 3673.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178883, "cps": 28.695312, "mean_p": 0.188462, "boundary_entropy": 0.135377, "bpb": 2.279945}, "vi": {"chunks": 3155.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.171142, "cps": 24.648438, "mean_p": 0.174258, "boundary_entropy": 0.239717, "bpb": 1.965935}, "tr": {"chunks": 3295.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.182014, "cps": 25.742188, "mean_p": 0.190755, "boundary_entropy": 0.167099, "bpb": 2.264134}, "fi": {"chunks": 3570.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.189783, "cps": 27.890625, "mean_p": 0.193733, "boundary_entropy": 0.145015, "bpb": 2.456891}, "ru": {"chunks": 3067.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.161847, "cps": 23.960938, "mean_p": 0.139428, "boundary_entropy": 0.256706, "bpb": 1.453781}, "el": {"chunks": 2720.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.132534, "cps": 21.25, "mean_p": 0.179863, "boundary_entropy": 0.373594, "bpb": 1.55391}, "hi": {"chunks": 2874.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.170817, "cps": 22.453125, "mean_p": 0.12782, "boundary_entropy": 0.266262, "bpb": 1.220642}, "ta": {"chunks": 3095.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.165552, "cps": 24.179688, "mean_p": 0.153569, "boundary_entropy": 0.331855, "bpb": 1.009028}, "th": {"chunks": 2556.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.158906, "cps": 19.96875, "mean_p": 0.171292, "boundary_entropy": 0.367834, "bpb": 1.1392}, "zh": {"chunks": 3153.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.533864, "cps": 24.632812, "mean_p": 0.276612, "boundary_entropy": 0.374488, "bpb": 3.17083}, "my": {"chunks": 2713.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.143181, "cps": 21.195312, "mean_p": 0.159208, "boundary_entropy": 0.334288, "bpb": 1.145905}}, "bpb": {"en": 2.07764, "de": 2.27994, "vi": 1.96594, "tr": 2.26413, "fi": 2.45689, "ru": 1.45378, "el": 1.55391, "hi": 1.22064, "ta": 1.00903, "th": 1.1392, "zh": 3.17083, "my": 1.1459}, "gini_cpc": 0.181483, "gini_cps": 0.058974, "bpic": 10.335865, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9549377046496154, "vi": 0.9136139911245115, "tr": 0.9716534289642532, "fi": 1.0131245815250378, "ru": 0.863994706767891, "el": 0.7075132523767289, "hi": 0.9118810928085033, "ta": 0.8837749833048107, "th": 0.8482938225035329, "zh": 2.849948736753099, "my": 0.7643511755129634}, "t": 231.6} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.08552, "de": 2.29139, "vi": 1.96806, "tr": 2.27026, "fi": 2.46564, "ru": 1.43135, "el": 1.54682, "hi": 1.21611, "ta": 1.01516, "th": 1.15828, "zh": 3.16649, "my": 1.15291}, "bytes_per_chunk": {"en": 5.674, "de": 5.869, "vi": 7.885, "tr": 5.926, "fi": 5.395, "ru": 11.21, "el": 13.278, "hi": 14.64, "ta": 17.241, "th": 17.336, "zh": 5.305, "my": 18.038}, "bpb_mean": 1.814, "t": 232.9} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 232.9} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.30542, "bpb": 1.88333, "chunk_loss": 0.030223, "F": 0.13068, "G": 0.17979, "chunks": 133.81, "bytes_per_chunk": 7.652, "boundary_entropy": 0.23827, "router_noise": 0.0, "grad_norm": 0.7524, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 232.9} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.31503, "bpb": 1.89719, "chunk_loss": 0.030055, "F": 0.15857, "G": 0.21492, "chunks": 162.38, "bytes_per_chunk": 6.306, "boundary_entropy": 0.26755, "router_noise": 0.0, "grad_norm": 0.7782, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 234.9} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.35336, "bpb": 1.95249, "chunk_loss": 0.030064, "F": 0.16528, "G": 0.20887, "chunks": 169.25, "bytes_per_chunk": 6.05, "boundary_entropy": 0.22695, "router_noise": 0.0, "grad_norm": 0.7749, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 236.8} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.48067, "bpb": 2.13615, "chunk_loss": 0.03017, "F": 0.17215, "G": 0.23305, "chunks": 176.28, "bytes_per_chunk": 5.809, "boundary_entropy": 0.2853, "router_noise": 0.0, "grad_norm": 0.8114, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 238.8} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3292.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189163, "cps": 25.71875, "mean_p": 0.207873, "boundary_entropy": 0.140294, "bpb": 2.064926}, "de": {"chunks": 3800.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.185068, "cps": 29.6875, "mean_p": 0.199455, "boundary_entropy": 0.15325, "bpb": 2.268331}, "vi": {"chunks": 3880.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.210469, "cps": 30.3125, "mean_p": 0.223994, "boundary_entropy": 0.296499, "bpb": 1.949864}, "tr": {"chunks": 3531.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195051, "cps": 27.585938, "mean_p": 0.205446, "boundary_entropy": 0.188885, "bpb": 2.245342}, "fi": {"chunks": 3651.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194089, "cps": 28.523438, "mean_p": 0.20268, "boundary_entropy": 0.164377, "bpb": 2.435891}, "ru": {"chunks": 4912.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.259208, "cps": 38.375, "mean_p": 0.266812, "boundary_entropy": 0.441234, "bpb": 1.440501}, "el": {"chunks": 5871.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.286069, "cps": 45.867188, "mean_p": 0.270679, "boundary_entropy": 0.4679, "bpb": 1.542576}, "hi": {"chunks": 5438.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.32321, "cps": 42.484375, "mean_p": 0.268983, "boundary_entropy": 0.460632, "bpb": 1.200416}, "ta": {"chunks": 8065.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.431399, "cps": 63.007812, "mean_p": 0.29511, "boundary_entropy": 0.481336, "bpb": 1.000903}, "th": {"chunks": 7944.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.493876, "cps": 62.0625, "mean_p": 0.326796, "boundary_entropy": 0.550045, "bpb": 1.135494}, "zh": {"chunks": 2994.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506942, "cps": 23.390625, "mean_p": 0.273604, "boundary_entropy": 0.379314, "bpb": 3.161697}, "my": {"chunks": 8788.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.463796, "cps": 68.65625, "mean_p": 0.224565, "boundary_entropy": 0.375151, "bpb": 1.144158}}, "bpb": {"en": 2.06493, "de": 2.26833, "vi": 1.94986, "tr": 2.24534, "fi": 2.43589, "ru": 1.4405, "el": 1.54258, "hi": 1.20042, "ta": 1.0009, "th": 1.13549, "zh": 3.1617, "my": 1.14416}, "gini_cpc": 0.216625, "gini_cps": 0.207278, "bpic": 6.173487, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9783527793536754, "vi": 1.1126354097563242, "tr": 1.0311253399525826, "fi": 1.0260398846284002, "ru": 1.3702929946107456, "el": 1.5122915616885753, "hi": 1.7086315603227418, "ta": 2.2805688911044695, "th": 2.6108532624563234, "zh": 2.679925042761876, "my": 2.4518334504281207}, "t": 241.6} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.07324, "de": 2.27674, "vi": 1.95293, "tr": 2.25197, "fi": 2.44482, "ru": 1.4183, "el": 1.53433, "hi": 1.19579, "ta": 1.00745, "th": 1.15366, "zh": 3.15783, "my": 1.15142}, "bytes_per_chunk": {"en": 5.582, "de": 5.624, "vi": 6.832, "tr": 5.591, "fi": 5.292, "ru": 7.066, "el": 6.183, "hi": 7.934, "ta": 6.457, "th": 5.899, "zh": 5.589, "my": 6.118}, "bpb_mean": 1.80154, "t": 243.0} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.44966, "bpb": 2.09142, "chunk_loss": 0.030035, "F": 0.17017, "G": 0.22166, "chunks": 174.25, "bytes_per_chunk": 5.877, "boundary_entropy": 0.24073, "router_noise": 0.0, "grad_norm": 0.8486, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 243.0} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.38084, "bpb": 1.99212, "chunk_loss": 0.029995, "F": 0.16931, "G": 0.23677, "chunks": 173.38, "bytes_per_chunk": 5.906, "boundary_entropy": 0.29238, "router_noise": 0.0, "grad_norm": 0.8184, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 244.9} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.31423, "bpb": 1.89603, "chunk_loss": 0.02993, "F": 0.17032, "G": 0.23327, "chunks": 174.41, "bytes_per_chunk": 5.871, "boundary_entropy": 0.26347, "router_noise": 0.0, "grad_norm": 0.7623, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 246.9} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.3655, "bpb": 1.97, "chunk_loss": 0.02989, "F": 0.16602, "G": 0.24198, "chunks": 170.0, "bytes_per_chunk": 6.024, "boundary_entropy": 0.30534, "router_noise": 0.0, "grad_norm": 0.783, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 248.9} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3293.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.18922, "cps": 25.726562, "mean_p": 0.209977, "boundary_entropy": 0.142466, "bpb": 2.06065}, "de": {"chunks": 3790.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.184581, "cps": 29.609375, "mean_p": 0.200856, "boundary_entropy": 0.157824, "bpb": 2.251999}, "vi": {"chunks": 4099.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222349, "cps": 32.023438, "mean_p": 0.239464, "boundary_entropy": 0.317437, "bpb": 1.942295}, "tr": {"chunks": 3438.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.189913, "cps": 26.859375, "mean_p": 0.206913, "boundary_entropy": 0.195982, "bpb": 2.231365}, "fi": {"chunks": 3631.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.193025, "cps": 28.367188, "mean_p": 0.203992, "boundary_entropy": 0.172829, "bpb": 2.422578}, "ru": {"chunks": 4991.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.263377, "cps": 38.992188, "mean_p": 0.29505, "boundary_entropy": 0.481257, "bpb": 1.426611}, "el": {"chunks": 5291.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.257808, "cps": 41.335938, "mean_p": 0.269115, "boundary_entropy": 0.476495, "bpb": 1.527985}, "hi": {"chunks": 6447.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.38318, "cps": 50.367188, "mean_p": 0.351943, "boundary_entropy": 0.560381, "bpb": 1.197218}, "ta": {"chunks": 7593.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.406151, "cps": 59.320312, "mean_p": 0.30986, "boundary_entropy": 0.502449, "bpb": 0.994138}, "th": {"chunks": 6790.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.422132, "cps": 53.046875, "mean_p": 0.363894, "boundary_entropy": 0.593272, "bpb": 1.117633}, "zh": {"chunks": 2989.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506095, "cps": 23.351562, "mean_p": 0.273077, "boundary_entropy": 0.383316, "bpb": 3.146621}, "my": {"chunks": 9213.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.486225, "cps": 71.976562, "mean_p": 0.25573, "boundary_entropy": 0.438079, "bpb": 1.136492}}, "bpb": {"en": 2.06065, "de": 2.252, "vi": 1.94229, "tr": 2.23136, "fi": 2.42258, "ru": 1.42661, "el": 1.52798, "hi": 1.19722, "ta": 0.99414, "th": 1.11763, "zh": 3.14662, "my": 1.13649}, "gini_cpc": 0.211231, "gini_cps": 0.203177, "bpic": 6.233753, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9754818478675991, "vi": 1.175079272871394, "tr": 1.0036625290525105, "fi": 1.0201094128864099, "ru": 1.3919086861312113, "el": 1.3624773958731968, "hi": 2.025046437826377, "ta": 2.1464477438303287, "th": 2.2309051113122185, "zh": 2.674637084338512, "my": 2.5696269862165644}, "t": 251.7} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.06795, "de": 2.26208, "vi": 1.94418, "tr": 2.23747, "fi": 2.43106, "ru": 1.40521, "el": 1.51976, "hi": 1.19178, "ta": 0.99941, "th": 1.13703, "zh": 3.14187, "my": 1.14348}, "bytes_per_chunk": {"en": 5.57, "de": 5.608, "vi": 6.275, "tr": 5.682, "fi": 5.301, "ru": 6.9, "el": 6.793, "hi": 6.877, "ta": 7.148, "th": 6.753, "zh": 5.581, "my": 5.805}, "bpb_mean": 1.79011, "t": 253.0} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.41271, "bpb": 2.0381, "chunk_loss": 0.02989, "F": 0.16626, "G": 0.25144, "chunks": 170.25, "bytes_per_chunk": 6.015, "boundary_entropy": 0.32471, "router_noise": 0.0, "grad_norm": 0.7802, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 253.1} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.51451, "bpb": 2.18498, "chunk_loss": 0.029958, "F": 0.16876, "G": 0.24748, "chunks": 172.81, "bytes_per_chunk": 5.925, "boundary_entropy": 0.31062, "router_noise": 0.0, "grad_norm": 0.8216, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 255.0} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.45841, "bpb": 2.10404, "chunk_loss": 0.029938, "F": 0.16504, "G": 0.25222, "chunks": 169.0, "bytes_per_chunk": 6.059, "boundary_entropy": 0.33268, "router_noise": 0.0, "grad_norm": 0.8515, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 257.0} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.4658, "bpb": 2.11471, "chunk_loss": 0.030175, "F": 0.17426, "G": 0.23709, "chunks": 178.44, "bytes_per_chunk": 5.739, "boundary_entropy": 0.26239, "router_noise": 0.0, "grad_norm": 0.8082, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 258.9} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3286.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188818, "cps": 25.671875, "mean_p": 0.209613, "boundary_entropy": 0.14557, "bpb": 2.048714}, "de": {"chunks": 3797.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.184922, "cps": 29.664062, "mean_p": 0.201967, "boundary_entropy": 0.161249, "bpb": 2.24313}, "vi": {"chunks": 3854.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.209059, "cps": 30.109375, "mean_p": 0.238879, "boundary_entropy": 0.322546, "bpb": 1.938597}, "tr": {"chunks": 3464.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.19135, "cps": 27.0625, "mean_p": 0.206206, "boundary_entropy": 0.196898, "bpb": 2.221753}, "fi": {"chunks": 3581.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.190367, "cps": 27.976562, "mean_p": 0.202858, "boundary_entropy": 0.174822, "bpb": 2.401083}, "ru": {"chunks": 5020.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.264908, "cps": 39.21875, "mean_p": 0.308073, "boundary_entropy": 0.496875, "bpb": 1.420919}, "el": {"chunks": 5857.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.285387, "cps": 45.757812, "mean_p": 0.289991, "boundary_entropy": 0.500964, "bpb": 1.516982}, "hi": {"chunks": 6675.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.396731, "cps": 52.148438, "mean_p": 0.375368, "boundary_entropy": 0.580192, "bpb": 1.185656}, "ta": {"chunks": 8417.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.450227, "cps": 65.757812, "mean_p": 0.325744, "boundary_entropy": 0.517295, "bpb": 0.984253}, "th": {"chunks": 7834.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.487038, "cps": 61.203125, "mean_p": 0.391289, "boundary_entropy": 0.611411, "bpb": 1.114239}, "zh": {"chunks": 2934.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.496783, "cps": 22.921875, "mean_p": 0.27229, "boundary_entropy": 0.382373, "bpb": 3.135759}, "my": {"chunks": 8721.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.46026, "cps": 68.132812, "mean_p": 0.260947, "boundary_entropy": 0.453229, "bpb": 1.13239}}, "bpb": {"en": 2.04871, "de": 2.24313, "vi": 1.9386, "tr": 2.22175, "fi": 2.40108, "ru": 1.42092, "el": 1.51698, "hi": 1.18566, "ta": 0.98425, "th": 1.11424, "zh": 3.13576, "my": 1.13239}, "gini_cpc": 0.216368, "gini_cps": 0.214431, "bpic": 6.049511, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.979365387173162, "vi": 1.1071975840498958, "tr": 1.013406984157638, "fi": 1.0082053523221495, "ru": 1.4029786557507102, "el": 1.5114400988704122, "hi": 2.1011291995669863, "ta": 2.3844510910234233, "th": 2.579402183054077, "zh": 2.6310144175981636, "my": 2.4375833318236113}, "t": 261.8} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.05688, "de": 2.25308, "vi": 1.93995, "tr": 2.22733, "fi": 2.41049, "ru": 1.39767, "el": 1.50958, "hi": 1.18115, "ta": 0.99128, "th": 1.13301, "zh": 3.13102, "my": 1.13932}, "bytes_per_chunk": {"en": 5.621, "de": 5.642, "vi": 6.725, "tr": 5.702, "fi": 5.406, "ru": 6.865, "el": 6.115, "hi": 6.78, "ta": 6.417, "th": 5.955, "zh": 5.739, "my": 6.203}, "bpb_mean": 1.7809, "t": 263.0} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.35315, "bpb": 1.95218, "chunk_loss": 0.029886, "F": 0.16541, "G": 0.24783, "chunks": 169.38, "bytes_per_chunk": 6.046, "boundary_entropy": 0.30724, "router_noise": 0.0, "grad_norm": 0.7829, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 263.1} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.28935, "bpb": 1.86014, "chunk_loss": 0.029989, "F": 0.1713, "G": 0.24021, "chunks": 175.41, "bytes_per_chunk": 5.838, "boundary_entropy": 0.27303, "router_noise": 0.0, "grad_norm": 0.7654, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 265.0} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.38764, "bpb": 2.00194, "chunk_loss": 0.030096, "F": 0.17258, "G": 0.24673, "chunks": 176.72, "bytes_per_chunk": 5.795, "boundary_entropy": 0.29956, "router_noise": 0.0, "grad_norm": 0.8345, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 267.0} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.26064, "bpb": 1.81871, "chunk_loss": 0.029996, "F": 0.16934, "G": 0.27008, "chunks": 173.41, "bytes_per_chunk": 5.905, "boundary_entropy": 0.34352, "router_noise": 0.0, "grad_norm": 0.8059, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 268.9} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3294.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189278, "cps": 25.734375, "mean_p": 0.209549, "boundary_entropy": 0.140274, "bpb": 2.041801}, "de": {"chunks": 3836.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186821, "cps": 29.96875, "mean_p": 0.203144, "boundary_entropy": 0.156855, "bpb": 2.237904}, "vi": {"chunks": 4133.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.224193, "cps": 32.289062, "mean_p": 0.246869, "boundary_entropy": 0.328658, "bpb": 1.932182}, "tr": {"chunks": 3412.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.188477, "cps": 26.65625, "mean_p": 0.204517, "boundary_entropy": 0.19528, "bpb": 2.214658}, "fi": {"chunks": 3602.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.191484, "cps": 28.140625, "mean_p": 0.20302, "boundary_entropy": 0.168994, "bpb": 2.389997}, "ru": {"chunks": 5446.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.287388, "cps": 42.546875, "mean_p": 0.32746, "boundary_entropy": 0.51525, "bpb": 1.410072}, "el": {"chunks": 6345.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.309165, "cps": 49.570312, "mean_p": 0.297511, "boundary_entropy": 0.506957, "bpb": 1.510642}, "hi": {"chunks": 6742.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.400713, "cps": 52.671875, "mean_p": 0.389631, "boundary_entropy": 0.590935, "bpb": 1.17786}, "ta": {"chunks": 8785.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.469912, "cps": 68.632812, "mean_p": 0.332046, "boundary_entropy": 0.526644, "bpb": 0.980636}, "th": {"chunks": 7481.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.465092, "cps": 58.445312, "mean_p": 0.384906, "boundary_entropy": 0.608316, "bpb": 1.111575}, "zh": {"chunks": 2914.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.493397, "cps": 22.765625, "mean_p": 0.270837, "boundary_entropy": 0.376571, "bpb": 3.129593}, "my": {"chunks": 8480.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.447541, "cps": 66.25, "mean_p": 0.272401, "boundary_entropy": 0.48036, "bpb": 1.12415}}, "bpb": {"en": 2.0418, "de": 2.2379, "vi": 1.93218, "tr": 2.21466, "fi": 2.39, "ru": 1.41007, "el": 1.51064, "hi": 1.17786, "ta": 0.98064, "th": 1.11158, "zh": 3.12959, "my": 1.12415}, "gini_cpc": 0.209067, "gini_cps": 0.212344, "bpic": 5.952862, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9870217359464974, "vi": 1.1844665177656148, "tr": 0.9957699047959425, "fi": 1.0116548091895616, "ru": 1.5183397013519422, "el": 1.6333953189664054, "hi": 2.1170650405843934, "ta": 2.4826575684545946, "th": 2.45719212539486, "zh": 2.606733481495463, "my": 2.364465608685337}, "t": 271.8} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.05018, "de": 2.24831, "vi": 1.93392, "tr": 2.22058, "fi": 2.39956, "ru": 1.38939, "el": 1.50285, "hi": 1.17206, "ta": 0.98683, "th": 1.13029, "zh": 3.12382, "my": 1.13209}, "bytes_per_chunk": {"en": 5.579, "de": 5.564, "vi": 6.289, "tr": 5.724, "fi": 5.379, "ru": 6.354, "el": 5.678, "hi": 6.714, "ta": 6.114, "th": 6.188, "zh": 5.706, "my": 6.385}, "bpb_mean": 1.77416, "t": 273.1} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.37334, "bpb": 1.98131, "chunk_loss": 0.029926, "F": 0.16736, "G": 0.2692, "chunks": 171.38, "bytes_per_chunk": 5.975, "boundary_entropy": 0.36563, "router_noise": 0.0, "grad_norm": 0.7913, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 273.2} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.38575, "bpb": 1.99922, "chunk_loss": 0.030151, "F": 0.17453, "G": 0.27097, "chunks": 178.72, "bytes_per_chunk": 5.73, "boundary_entropy": 0.34846, "router_noise": 0.0, "grad_norm": 0.9445, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 275.1} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.4806, "bpb": 2.13605, "chunk_loss": 0.029994, "F": 0.16815, "G": 0.24134, "chunks": 172.19, "bytes_per_chunk": 5.947, "boundary_entropy": 0.29596, "router_noise": 0.0, "grad_norm": 0.8091, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 277.1} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.33925, "bpb": 1.93213, "chunk_loss": 0.029902, "F": 0.16479, "G": 0.24425, "chunks": 168.75, "bytes_per_chunk": 6.068, "boundary_entropy": 0.29583, "router_noise": 0.0, "grad_norm": 0.7907, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 279.0} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3268.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187784, "cps": 25.53125, "mean_p": 0.206559, "boundary_entropy": 0.137975, "bpb": 2.03559}, "de": {"chunks": 3783.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18424, "cps": 29.554688, "mean_p": 0.198956, "boundary_entropy": 0.153669, "bpb": 2.228944}, "vi": {"chunks": 4053.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.219854, "cps": 31.664062, "mean_p": 0.24604, "boundary_entropy": 0.327859, "bpb": 1.927042}, "tr": {"chunks": 3346.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.184831, "cps": 26.140625, "mean_p": 0.201652, "boundary_entropy": 0.191787, "bpb": 2.203863}, "fi": {"chunks": 3594.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.191058, "cps": 28.078125, "mean_p": 0.201608, "boundary_entropy": 0.166716, "bpb": 2.376964}, "ru": {"chunks": 5461.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.288179, "cps": 42.664062, "mean_p": 0.3307, "boundary_entropy": 0.51939, "bpb": 1.406146}, "el": {"chunks": 6026.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.293622, "cps": 47.078125, "mean_p": 0.290779, "boundary_entropy": 0.501534, "bpb": 1.50417}, "hi": {"chunks": 7040.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.418425, "cps": 55.0, "mean_p": 0.397244, "boundary_entropy": 0.594057, "bpb": 1.171272}, "ta": {"chunks": 8397.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449158, "cps": 65.601562, "mean_p": 0.33057, "boundary_entropy": 0.528616, "bpb": 0.97316}, "th": {"chunks": 7438.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.462418, "cps": 58.109375, "mean_p": 0.385639, "boundary_entropy": 0.608226, "bpb": 1.10379}, "zh": {"chunks": 2995.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.507111, "cps": 23.398438, "mean_p": 0.273418, "boundary_entropy": 0.377564, "bpb": 3.119503}, "my": {"chunks": 9722.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.513088, "cps": 75.953125, "mean_p": 0.297831, "boundary_entropy": 0.505678, "bpb": 1.118986}}, "bpb": {"en": 2.03559, "de": 2.22894, "vi": 1.92704, "tr": 2.20386, "fi": 2.37696, "ru": 1.40615, "el": 1.50417, "hi": 1.17127, "ta": 0.97316, "th": 1.10379, "zh": 3.1195, "my": 1.11899}, "gini_cpc": 0.220781, "gini_cps": 0.222872, "bpic": 5.893171, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9811287600382488, "vi": 1.1707806448207487, "tr": 0.9842772734368973, "fi": 1.0174387209442222, "ru": 1.5346347729482017, "el": 1.5636168955378658, "hi": 2.228228160504582, "ta": 2.391887537592933, "th": 2.46250534092712, "zh": 2.700507926922023, "my": 2.732337313562416}, "t": 281.9} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.04418, "de": 2.23834, "vi": 1.92719, "tr": 2.21107, "fi": 2.3854, "ru": 1.38551, "el": 1.49584, "hi": 1.16621, "ta": 0.98021, "th": 1.12259, "zh": 3.11442, "my": 1.12625}, "bytes_per_chunk": {"en": 5.641, "de": 5.665, "vi": 6.438, "tr": 5.832, "fi": 5.407, "ru": 6.299, "el": 5.994, "hi": 6.392, "ta": 6.348, "th": 6.259, "zh": 5.535, "my": 5.62}, "bpb_mean": 1.76643, "t": 283.2} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.25513, "bpb": 1.81077, "chunk_loss": 0.030013, "F": 0.17032, "G": 0.26694, "chunks": 174.41, "bytes_per_chunk": 5.871, "boundary_entropy": 0.35031, "router_noise": 0.0, "grad_norm": 0.7741, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 283.2} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.30608, "bpb": 1.88428, "chunk_loss": 0.029997, "F": 0.16956, "G": 0.26751, "chunks": 173.62, "bytes_per_chunk": 5.898, "boundary_entropy": 0.34538, "router_noise": 0.0, "grad_norm": 0.7778, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 285.2} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.23392, "bpb": 1.78017, "chunk_loss": 0.030049, "F": 0.1702, "G": 0.27664, "chunks": 174.28, "bytes_per_chunk": 5.876, "boundary_entropy": 0.35791, "router_noise": 0.0, "grad_norm": 0.7654, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 287.1} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.45431, "bpb": 2.09813, "chunk_loss": 0.029922, "F": 0.16794, "G": 0.26202, "chunks": 171.97, "bytes_per_chunk": 5.955, "boundary_entropy": 0.34652, "router_noise": 0.0, "grad_norm": 0.8122, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 289.0} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3285.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188761, "cps": 25.664062, "mean_p": 0.207915, "boundary_entropy": 0.134914, "bpb": 2.032966}, "de": {"chunks": 3799.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.185019, "cps": 29.679688, "mean_p": 0.200907, "boundary_entropy": 0.151117, "bpb": 2.223076}, "vi": {"chunks": 3906.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.21188, "cps": 30.515625, "mean_p": 0.246489, "boundary_entropy": 0.331196, "bpb": 1.923959}, "tr": {"chunks": 3383.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.186875, "cps": 26.429688, "mean_p": 0.205204, "boundary_entropy": 0.192562, "bpb": 2.196716}, "fi": {"chunks": 3592.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.190952, "cps": 28.0625, "mean_p": 0.201935, "boundary_entropy": 0.166555, "bpb": 2.366995}, "ru": {"chunks": 5190.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.273879, "cps": 40.546875, "mean_p": 0.323624, "boundary_entropy": 0.511419, "bpb": 1.39517}, "el": {"chunks": 5951.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.289967, "cps": 46.492188, "mean_p": 0.291256, "boundary_entropy": 0.50277, "bpb": 1.496025}, "hi": {"chunks": 6786.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.403328, "cps": 53.015625, "mean_p": 0.39145, "boundary_entropy": 0.591604, "bpb": 1.163546}, "ta": {"chunks": 8658.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.463118, "cps": 67.640625, "mean_p": 0.335645, "boundary_entropy": 0.532178, "bpb": 0.969097}, "th": {"chunks": 8194.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.509419, "cps": 64.015625, "mean_p": 0.393805, "boundary_entropy": 0.610713, "bpb": 1.098111}, "zh": {"chunks": 2956.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.500508, "cps": 23.09375, "mean_p": 0.270586, "boundary_entropy": 0.374739, "bpb": 3.110728}, "my": {"chunks": 8956.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.472662, "cps": 69.96875, "mean_p": 0.309551, "boundary_entropy": 0.529951, "bpb": 1.117214}}, "bpb": {"en": 2.03297, "de": 2.22308, "vi": 1.92396, "tr": 2.19672, "fi": 2.36699, "ru": 1.39517, "el": 1.49602, "hi": 1.16355, "ta": 0.9691, "th": 1.09811, "zh": 3.11073, "my": 1.11721}, "gini_cpc": 0.222065, "gini_cps": 0.221356, "bpic": 5.935737, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9801795394739329, "vi": 1.1224780142002073, "tr": 0.9900113934545524, "fi": 1.0116101795097545, "ru": 1.4509314345610294, "el": 1.5361649486921567, "hi": 2.136719574996438, "ta": 2.4534706005195144, "th": 2.6987561229370796, "zh": 2.6515494652137677, "my": 2.504029581560878}, "t": 291.9} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.04014, "de": 2.23353, "vi": 1.92455, "tr": 2.20311, "fi": 2.37696, "ru": 1.37293, "el": 1.48827, "hi": 1.15772, "ta": 0.97672, "th": 1.11697, "zh": 3.10494, "my": 1.12501}, "bytes_per_chunk": {"en": 5.616, "de": 5.627, "vi": 6.717, "tr": 5.742, "fi": 5.391, "ru": 6.618, "el": 6.124, "hi": 6.681, "ta": 6.248, "th": 5.745, "zh": 5.616, "my": 6.059}, "bpb_mean": 1.76007, "t": 293.2} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.35061, "bpb": 1.94852, "chunk_loss": 0.030439, "F": 0.18277, "G": 0.26428, "chunks": 187.16, "bytes_per_chunk": 5.471, "boundary_entropy": 0.31771, "router_noise": 0.0, "grad_norm": 0.8519, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 293.3} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.21558, "bpb": 1.75371, "chunk_loss": 0.030026, "F": 0.17068, "G": 0.27021, "chunks": 174.78, "bytes_per_chunk": 5.859, "boundary_entropy": 0.35865, "router_noise": 0.0, "grad_norm": 0.7788, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 295.2} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.31747, "bpb": 1.90071, "chunk_loss": 0.030219, "F": 0.17676, "G": 0.26123, "chunks": 181.0, "bytes_per_chunk": 5.657, "boundary_entropy": 0.31769, "router_noise": 0.0, "grad_norm": 0.8174, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 297.2} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.39034, "bpb": 2.00584, "chunk_loss": 0.030077, "F": 0.1749, "G": 0.24338, "chunks": 179.09, "bytes_per_chunk": 5.718, "boundary_entropy": 0.28104, "router_noise": 0.0, "grad_norm": 0.8375, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 299.1} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3292.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189163, "cps": 25.71875, "mean_p": 0.207999, "boundary_entropy": 0.135025, "bpb": 2.028587}, "de": {"chunks": 3847.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187357, "cps": 30.054688, "mean_p": 0.203856, "boundary_entropy": 0.154117, "bpb": 2.216079}, "vi": {"chunks": 4113.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223108, "cps": 32.132812, "mean_p": 0.255634, "boundary_entropy": 0.338255, "bpb": 1.917222}, "tr": {"chunks": 3475.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.191957, "cps": 27.148438, "mean_p": 0.207345, "boundary_entropy": 0.19422, "bpb": 2.188136}, "fi": {"chunks": 3653.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194195, "cps": 28.539062, "mean_p": 0.204841, "boundary_entropy": 0.165741, "bpb": 2.362884}, "ru": {"chunks": 5633.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.297256, "cps": 44.007812, "mean_p": 0.335958, "boundary_entropy": 0.522469, "bpb": 1.391128}, "el": {"chunks": 5906.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.287775, "cps": 46.140625, "mean_p": 0.289199, "boundary_entropy": 0.500722, "bpb": 1.487078}, "hi": {"chunks": 6993.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.415632, "cps": 54.632812, "mean_p": 0.399434, "boundary_entropy": 0.595411, "bpb": 1.158248}, "ta": {"chunks": 8577.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.458786, "cps": 67.007812, "mean_p": 0.336446, "boundary_entropy": 0.532232, "bpb": 0.967371}, "th": {"chunks": 7418.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.461175, "cps": 57.953125, "mean_p": 0.388932, "boundary_entropy": 0.609057, "bpb": 1.094873}, "zh": {"chunks": 3014.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.510328, "cps": 23.546875, "mean_p": 0.272622, "boundary_entropy": 0.373755, "bpb": 3.106026}, "my": {"chunks": 9328.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.492295, "cps": 72.875, "mean_p": 0.315422, "boundary_entropy": 0.531461, "bpb": 1.113499}}, "bpb": {"en": 2.02859, "de": 2.21608, "vi": 1.91722, "tr": 2.18814, "fi": 2.36288, "ru": 1.39113, "el": 1.48708, "hi": 1.15825, "ta": 0.96737, "th": 1.09487, "zh": 3.10603, "my": 1.1135}, "gini_cpc": 0.214789, "gini_cps": 0.215862, "bpic": 5.881791, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.990453458466734, "vi": 1.1794508866824127, "tr": 1.0147721768154134, "fi": 1.0266019442748688, "ru": 1.5714292423946115, "el": 1.5213070964627364, "hi": 2.1972159803856077, "ta": 2.425348962058653, "th": 2.437979544423591, "zh": 2.6978270136554086, "my": 2.602492310604632}, "t": 302.0} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.03591, "de": 2.22591, "vi": 1.9182, "tr": 2.19515, "fi": 2.3716, "ru": 1.36955, "el": 1.48021, "hi": 1.15302, "ta": 0.9744, "th": 1.11328, "zh": 3.09897, "my": 1.12075}, "bytes_per_chunk": {"en": 5.62, "de": 5.578, "vi": 6.292, "tr": 5.693, "fi": 5.322, "ru": 6.168, "el": 6.113, "hi": 6.457, "ta": 6.259, "th": 6.266, "zh": 5.539, "my": 5.804}, "bpb_mean": 1.75475, "t": 303.3} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.33591, "bpb": 1.9273, "chunk_loss": 0.030025, "F": 0.17154, "G": 0.26932, "chunks": 175.66, "bytes_per_chunk": 5.83, "boundary_entropy": 0.33739, "router_noise": 0.0, "grad_norm": 0.7872, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 303.4} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.32193, "bpb": 1.90714, "chunk_loss": 0.029921, "F": 0.16458, "G": 0.26051, "chunks": 168.53, "bytes_per_chunk": 6.076, "boundary_entropy": 0.33557, "router_noise": 0.0, "grad_norm": 0.8577, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 305.3} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.29486, "bpb": 1.86809, "chunk_loss": 0.029969, "F": 0.17078, "G": 0.24883, "chunks": 174.88, "bytes_per_chunk": 5.856, "boundary_entropy": 0.28649, "router_noise": 0.0, "grad_norm": 0.7814, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 307.3} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.2086, "bpb": 1.74364, "chunk_loss": 0.029965, "F": 0.16946, "G": 0.26004, "chunks": 173.53, "bytes_per_chunk": 5.901, "boundary_entropy": 0.32698, "router_noise": 0.0, "grad_norm": 0.781, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 309.2} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3299.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189565, "cps": 25.773438, "mean_p": 0.209607, "boundary_entropy": 0.136011, "bpb": 2.023268}, "de": {"chunks": 3825.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186285, "cps": 29.882812, "mean_p": 0.202511, "boundary_entropy": 0.150107, "bpb": 2.210599}, "vi": {"chunks": 4078.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22121, "cps": 31.859375, "mean_p": 0.253574, "boundary_entropy": 0.337598, "bpb": 1.913044}, "tr": {"chunks": 3435.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.189748, "cps": 26.835938, "mean_p": 0.206999, "boundary_entropy": 0.191829, "bpb": 2.184599}, "fi": {"chunks": 3620.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.192441, "cps": 28.28125, "mean_p": 0.203815, "boundary_entropy": 0.165371, "bpb": 2.354209}, "ru": {"chunks": 5519.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.29124, "cps": 43.117188, "mean_p": 0.334517, "boundary_entropy": 0.521715, "bpb": 1.386446}, "el": {"chunks": 5971.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.290942, "cps": 46.648438, "mean_p": 0.291836, "boundary_entropy": 0.50301, "bpb": 1.483531}, "hi": {"chunks": 7396.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.439584, "cps": 57.78125, "mean_p": 0.40572, "boundary_entropy": 0.599417, "bpb": 1.153205}, "ta": {"chunks": 8376.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.448034, "cps": 65.4375, "mean_p": 0.338932, "boundary_entropy": 0.537784, "bpb": 0.963117}, "th": {"chunks": 7029.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.436991, "cps": 54.914062, "mean_p": 0.386131, "boundary_entropy": 0.607425, "bpb": 1.092306}, "zh": {"chunks": 2928.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.495767, "cps": 22.875, "mean_p": 0.265486, "boundary_entropy": 0.368259, "bpb": 3.098494}, "my": {"chunks": 9318.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.491767, "cps": 72.796875, "mean_p": 0.302532, "boundary_entropy": 0.517025, "bpb": 1.110762}}, "bpb": {"en": 2.02327, "de": 2.2106, "vi": 1.91304, "tr": 2.1846, "fi": 2.35421, "ru": 1.38645, "el": 1.48353, "hi": 1.15321, "ta": 0.96312, "th": 1.09231, "zh": 3.09849, "my": 1.11076}, "gini_cpc": 0.211163, "gini_cps": 0.216885, "bpic": 5.923095, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9826997305677594, "vi": 1.1669328995077286, "tr": 1.0009629315072353, "fi": 1.015169337579869, "ru": 1.536359974758482, "el": 1.534786712730364, "hi": 2.3189086161009667, "ta": 2.363485835099114, "th": 2.305230105712086, "zh": 2.615287477954869, "my": 2.5941861400416695}, "t": 312.1} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.03134, "de": 2.2217, "vi": 1.9149, "tr": 2.19041, "fi": 2.36457, "ru": 1.36444, "el": 1.47516, "hi": 1.14763, "ta": 0.97067, "th": 1.11074, "zh": 3.09164, "my": 1.11776}, "bytes_per_chunk": {"en": 5.59, "de": 5.592, "vi": 6.464, "tr": 5.659, "fi": 5.362, "ru": 6.32, "el": 6.036, "hi": 5.999, "ta": 6.278, "th": 6.538, "zh": 5.694, "my": 5.788}, "bpb_mean": 1.75008, "t": 313.4} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.29135, "bpb": 1.86303, "chunk_loss": 0.029798, "F": 0.16556, "G": 0.2672, "chunks": 169.53, "bytes_per_chunk": 6.04, "boundary_entropy": 0.33359, "router_noise": 0.0, "grad_norm": 0.8195, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 313.4} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.28382, "bpb": 1.85216, "chunk_loss": 0.030138, "F": 0.17072, "G": 0.26309, "chunks": 174.81, "bytes_per_chunk": 5.858, "boundary_entropy": 0.34209, "router_noise": 0.0, "grad_norm": 0.7784, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 315.4} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.35579, "bpb": 1.95599, "chunk_loss": 0.030022, "F": 0.17087, "G": 0.267, "chunks": 174.97, "bytes_per_chunk": 5.852, "boundary_entropy": 0.34586, "router_noise": 0.0, "grad_norm": 0.8518, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 317.3} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.3983, "bpb": 2.01732, "chunk_loss": 0.030117, "F": 0.17322, "G": 0.25802, "chunks": 177.38, "bytes_per_chunk": 5.773, "boundary_entropy": 0.32585, "router_noise": 0.0, "grad_norm": 0.8004, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.039, "alpha": 0.03, "t": 319.3} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3263.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187496, "cps": 25.492188, "mean_p": 0.204978, "boundary_entropy": 0.132001, "bpb": 2.020282}, "de": {"chunks": 3771.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.183656, "cps": 29.460938, "mean_p": 0.198301, "boundary_entropy": 0.148121, "bpb": 2.212045}, "vi": {"chunks": 4090.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.221861, "cps": 31.953125, "mean_p": 0.253828, "boundary_entropy": 0.337658, "bpb": 1.911062}, "tr": {"chunks": 3386.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187041, "cps": 26.453125, "mean_p": 0.202889, "boundary_entropy": 0.188182, "bpb": 2.18005}, "fi": {"chunks": 3561.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.189304, "cps": 27.820312, "mean_p": 0.19994, "boundary_entropy": 0.160717, "bpb": 2.354586}, "ru": {"chunks": 5384.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.284116, "cps": 42.0625, "mean_p": 0.333853, "boundary_entropy": 0.521813, "bpb": 1.384495}, "el": {"chunks": 5938.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.289334, "cps": 46.390625, "mean_p": 0.289116, "boundary_entropy": 0.499698, "bpb": 1.480265}, "hi": {"chunks": 8513.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.505973, "cps": 66.507812, "mean_p": 0.415607, "boundary_entropy": 0.599228, "bpb": 1.152837}, "ta": {"chunks": 9098.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.486654, "cps": 71.078125, "mean_p": 0.343427, "boundary_entropy": 0.538149, "bpb": 0.960092}, "th": {"chunks": 7981.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.496177, "cps": 62.351562, "mean_p": 0.399699, "boundary_entropy": 0.613797, "bpb": 1.089401}, "zh": {"chunks": 2944.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.498476, "cps": 23.0, "mean_p": 0.265952, "boundary_entropy": 0.366812, "bpb": 3.095445}, "my": {"chunks": 9333.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.492559, "cps": 72.914062, "mean_p": 0.308933, "boundary_entropy": 0.523373, "bpb": 1.10918}}, "bpb": {"en": 2.02028, "de": 2.21205, "vi": 1.91106, "tr": 2.18005, "fi": 2.35459, "ru": 1.3845, "el": 1.48026, "hi": 1.15284, "ta": 0.96009, "th": 1.0894, "zh": 3.09545, "my": 1.10918}, "gini_cpc": 0.224252, "gini_cps": 0.233636, "bpic": 5.705763, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9795151818203623, "vi": 1.1832791510306024, "tr": 0.9975701577155233, "fi": 1.0096413681179397, "ru": 1.5153148639458809, "el": 1.5431437866090427, "hi": 2.69857570973393, "ta": 2.5955387043182423, "th": 2.6463256845647707, "zh": 2.65859026059403, "my": 2.6270294182552685}, "t": 322.1} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.02893, "de": 2.2227, "vi": 1.91189, "tr": 2.18581, "fi": 2.36539, "ru": 1.36225, "el": 1.47257, "hi": 1.14651, "ta": 0.96716, "th": 1.10759, "zh": 3.08867, "my": 1.11652}, "bytes_per_chunk": {"en": 5.646, "de": 5.656, "vi": 6.392, "tr": 5.784, "fi": 5.436, "ru": 6.44, "el": 6.073, "hi": 5.328, "ta": 5.894, "th": 5.797, "zh": 5.622, "my": 5.832}, "bpb_mean": 1.748, "t": 323.4} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3263.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187496, "cps": 25.492188, "mean_p": 0.204978, "boundary_entropy": 0.132001, "bpb": 2.020282}, "de": {"chunks": 3771.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.183656, "cps": 29.460938, "mean_p": 0.198301, "boundary_entropy": 0.148121, "bpb": 2.212045}, "vi": {"chunks": 4090.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.221861, "cps": 31.953125, "mean_p": 0.253828, "boundary_entropy": 0.337658, "bpb": 1.911062}, "tr": {"chunks": 3386.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187041, "cps": 26.453125, "mean_p": 0.202889, "boundary_entropy": 0.188182, "bpb": 2.18005}, "fi": {"chunks": 3561.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.189304, "cps": 27.820312, "mean_p": 0.19994, "boundary_entropy": 0.160717, "bpb": 2.354586}, "ru": {"chunks": 5384.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.284116, "cps": 42.0625, "mean_p": 0.333853, "boundary_entropy": 0.521813, "bpb": 1.384495}, "el": {"chunks": 5938.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.289334, "cps": 46.390625, "mean_p": 0.289116, "boundary_entropy": 0.499698, "bpb": 1.480265}, "hi": {"chunks": 8513.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.505973, "cps": 66.507812, "mean_p": 0.415607, "boundary_entropy": 0.599228, "bpb": 1.152837}, "ta": {"chunks": 9098.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.486654, "cps": 71.078125, "mean_p": 0.343427, "boundary_entropy": 0.538149, "bpb": 0.960092}, "th": {"chunks": 7981.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.496177, "cps": 62.351562, "mean_p": 0.399699, "boundary_entropy": 0.613797, "bpb": 1.089401}, "zh": {"chunks": 2944.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.498476, "cps": 23.0, "mean_p": 0.265952, "boundary_entropy": 0.366812, "bpb": 3.095445}, "my": {"chunks": 9333.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.492559, "cps": 72.914062, "mean_p": 0.308933, "boundary_entropy": 0.523373, "bpb": 1.10918}}, "bpb": {"en": 2.02028, "de": 2.21205, "vi": 1.91106, "tr": 2.18005, "fi": 2.35459, "ru": 1.3845, "el": 1.48026, "hi": 1.15284, "ta": 0.96009, "th": 1.0894, "zh": 3.09545, "my": 1.10918}, "gini_cpc": 0.224252, "gini_cps": 0.233636, "bpic": 5.705763, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9795151818203623, "vi": 1.1832791510306024, "tr": 0.9975701577155233, "fi": 1.0096413681179397, "ru": 1.5153148639458809, "el": 1.5431437866090427, "hi": 2.69857570973393, "ta": 2.5955387043182423, "th": 2.6463256845647707, "zh": 2.65859026059403, "my": 2.6270294182552685}, "t": 324.4} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s2/th0.70"], "t": 327.8} |
| {"event": "final", "run": "pilot_C_s2", "condition": "C", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.18749640866517267, "cps": 25.4921875, "bpb": 2.0378489691462307, "bytes_per_chunk": 5.63847543663426}, "de": {"cpc": 0.18365557882433156, "cps": 29.4609375, "bpb": 2.2129623061940205, "bytes_per_chunk": 5.673130193905817}, "vi": {"cpc": 0.22186059126661242, "cps": 31.953125, "bpb": 1.9167235659440396, "bytes_per_chunk": 6.390093483003644}, "tr": {"cpc": 0.1870408219632105, "cps": 26.453125, "bpb": 2.1776285028210034, "bytes_per_chunk": 5.795350790896128}, "fi": {"cpc": 0.18930413056190526, "cps": 27.8203125, "bpb": 2.3810676081931272, "bytes_per_chunk": 5.465660314415579}, "ru": {"cpc": 0.2841160949868074, "cps": 42.0625, "bpb": 1.363596296402935, "bytes_per_chunk": 6.417312329404277}, "el": {"cpc": 0.2893339180431711, "cps": 46.390625, "bpb": 1.4675325979694616, "bytes_per_chunk": 6.087311907858072}, "hi": {"cpc": 0.5059732540861813, "cps": 66.5078125, "bpb": 1.15428091468842, "bytes_per_chunk": 5.313280972890803}, "ta": {"cpc": 0.48665418561112594, "cps": 71.078125, "bpb": 0.9669115941155675, "bytes_per_chunk": 5.884396956160632}, "th": {"cpc": 0.49617656201429905, "cps": 62.3515625, "bpb": 1.1033083956560619, "bytes_per_chunk": 5.784161867566912}, "zh": {"cpc": 0.4984761259735862, "cps": 23.0, "bpb": 3.0968593688373898, "bytes_per_chunk": 5.620763961105095}, "my": {"cpc": 0.49255858138062064, "cps": 72.9140625, "bpb": 1.124886870753794, "bytes_per_chunk": 5.808963492327295}}, "gini_cpc": 0.2242521343695385, "gini_cps": 0.23363612936477263, "bpic": 5.705762540513217, "seed": 2, "mean_boundary_entropy": 0.3473655879497528, "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": "parity_a", "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": 330.1} |
|
|
