{"event": "start", "run": "spec_D_s0", "condition": "D", "params": 82470528, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "parity_a", "p2_objective": "parity_a", "p3_objective": "parity_a", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 2.2}
{"event": "probe_schedule", "n": 21, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 11000], "t": 2.2}
{"event": "probe", "step": 0, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 8739.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.502155, "cps": 68.273438, "mean_p": 0.50363, "boundary_entropy": 0.687415, "bpb": 8.248358}, "de": {"chunks": 12222.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.595237, "cps": 95.484375, "mean_p": 0.507356, "boundary_entropy": 0.688163, "bpb": 8.204506}, "vi": {"chunks": 6386.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.346406, "cps": 49.890625, "mean_p": 0.491233, "boundary_entropy": 0.688507, "bpb": 8.13873}, "tr": {"chunks": 11403.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.629896, "cps": 89.085938, "mean_p": 0.50659, "boundary_entropy": 0.687851, "bpb": 8.143105}, "fi": {"chunks": 7137.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.379406, "cps": 55.757812, "mean_p": 0.497164, "boundary_entropy": 0.68792, "bpb": 8.234202}, "ru": {"chunks": 10473.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.552665, "cps": 81.820312, "mean_p": 0.496692, "boundary_entropy": 0.690192, "bpb": 8.071624}, "el": {"chunks": 36555.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.781172, "cps": 285.585938, "mean_p": 0.530217, "boundary_entropy": 0.688772, "bpb": 8.142597}, "hi": {"chunks": 42584.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.530996, "cps": 332.6875, "mean_p": 0.521073, "boundary_entropy": 0.69018, "bpb": 7.880816}, "ta": {"chunks": 49758.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.661567, "cps": 388.734375, "mean_p": 0.521091, "boundary_entropy": 0.690409, "bpb": 7.814466}, "th": {"chunks": 45716.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.842151, "cps": 357.15625, "mean_p": 0.525713, "boundary_entropy": 0.689893, "bpb": 8.063302}, "zh": {"chunks": 6218.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.052828, "cps": 48.578125, "mean_p": 0.496822, "boundary_entropy": 0.686552, "bpb": 8.034403}, "my": {"chunks": 53266.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.811167, "cps": 416.140625, "mean_p": 0.521325, "boundary_entropy": 0.690575, "bpb": 8.24483}}, "bpb": {"en": 8.24836, "de": 8.20451, "vi": 8.13873, "tr": 8.1431, "fi": 8.2342, "ru": 8.07162, "el": 8.1426, "hi": 7.88082, "ta": 7.81447, "th": 8.0633, "zh": 8.0344, "my": 8.24483}, "gini_cpc": 0.391191, "gini_cps": 0.409188, "bpic": 1.321301, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1853654183968407, "vi": 0.6898396505624507, "tr": 1.2543852937189204, "fi": 0.755555191784985, "ru": 1.1005867247811874, "el": 3.54705827773647, "hi": 5.040269530092341, "ta": 5.300292378412309, "th": 5.659910185771798, "zh": 2.096619669942964, "my": 5.598208760550946}, "t": 9.0}
{"event": "train", "step": 0, "phase": 1, "objective": "parity_a", "lm_loss": 5.63905, "bpb": 8.13544, "chunk_loss": 0.096583, "F": 0.53552, "G": 0.50248, "chunks": 548.38, "bytes_per_chunk": 1.867, "boundary_entropy": 0.69172, "router_noise": 0.0, "grad_norm": 9.2882, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.005, "n_target_mean": 9.4236, "alpha": 0.03, "t": 9.8}
{"event": "probe", "step": 1, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 8726.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.501408, "cps": 68.171875, "mean_p": 0.503578, "boundary_entropy": 0.687414, "bpb": 8.209931}, "de": {"chunks": 12191.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.593727, "cps": 95.242188, "mean_p": 0.507273, "boundary_entropy": 0.688163, "bpb": 8.167571}, "vi": {"chunks": 6361.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.34505, "cps": 49.695312, "mean_p": 0.491179, "boundary_entropy": 0.688503, "bpb": 8.118798}, "tr": {"chunks": 11369.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.628017, "cps": 88.820312, "mean_p": 0.506506, "boundary_entropy": 0.687851, "bpb": 8.116438}, "fi": {"chunks": 7085.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.376641, "cps": 55.351562, "mean_p": 0.497035, "boundary_entropy": 0.687916, "bpb": 8.205174}, "ru": {"chunks": 10470.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.552507, "cps": 81.796875, "mean_p": 0.496692, "boundary_entropy": 0.690192, "bpb": 8.028164}, "el": {"chunks": 36538.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.780344, "cps": 285.453125, "mean_p": 0.530011, "boundary_entropy": 0.688796, "bpb": 8.124557}, "hi": {"chunks": 42554.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.529212, "cps": 332.453125, "mean_p": 0.520912, "boundary_entropy": 0.690193, "bpb": 7.857743}, "ta": {"chunks": 49672.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.656967, "cps": 388.0625, "mean_p": 0.520764, "boundary_entropy": 0.690435, "bpb": 7.780861}, "th": {"chunks": 45710.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.841778, "cps": 357.109375, "mean_p": 0.525504, "boundary_entropy": 0.689914, "bpb": 8.031021}, "zh": {"chunks": 6211.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.051642, "cps": 48.523438, "mean_p": 0.496792, "boundary_entropy": 0.686551, "bpb": 8.029703}, "my": {"chunks": 53257.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.810692, "cps": 416.070312, "mean_p": 0.521137, "boundary_entropy": 0.69059, "bpb": 8.21694}}, "bpb": {"en": 8.20993, "de": 8.16757, "vi": 8.1188, "tr": 8.11644, "fi": 8.20517, "ru": 8.02816, "el": 8.12456, "hi": 7.85774, "ta": 7.78086, "th": 8.03102, "zh": 8.0297, "my": 8.21694}, "gini_cpc": 0.391601, "gini_cps": 0.409523, "bpic": 1.322726, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1841203249022088, "vi": 0.688162757055176, "tr": 1.2525083407094622, "fi": 0.7511676612134363, "ru": 1.1019106458302212, "el": 3.550690660855127, "hi": 5.044222417403677, "ta": 5.299014267966231, "th": 5.667598381307304, "zh": 2.097379400042177, "my": 5.605601674191271}, "t": 16.1}
{"event": "probe", "step": 2, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 8665.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.497903, "cps": 67.695312, "mean_p": 0.503432, "boundary_entropy": 0.687412, "bpb": 8.133869}, "de": {"chunks": 12076.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.588126, "cps": 94.34375, "mean_p": 0.507067, "boundary_entropy": 0.688166, "bpb": 8.0957}, "vi": {"chunks": 6326.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.343152, "cps": 49.421875, "mean_p": 0.491072, "boundary_entropy": 0.688497, "bpb": 8.079311}, "tr": {"chunks": 11300.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.624206, "cps": 88.28125, "mean_p": 0.506315, "boundary_entropy": 0.687853, "bpb": 8.064883}, "fi": {"chunks": 6975.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.370794, "cps": 54.492188, "mean_p": 0.496753, "boundary_entropy": 0.687909, "bpb": 8.148101}, "ru": {"chunks": 10451.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.551504, "cps": 81.648438, "mean_p": 0.496651, "boundary_entropy": 0.690191, "bpb": 7.949707}, "el": {"chunks": 36504.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.778687, "cps": 285.1875, "mean_p": 0.529611, "boundary_entropy": 0.688841, "bpb": 8.089679}, "hi": {"chunks": 42500.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.526003, "cps": 332.03125, "mean_p": 0.520543, "boundary_entropy": 0.690221, "bpb": 7.805827}, "ta": {"chunks": 49497.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.647606, "cps": 386.695312, "mean_p": 0.520121, "boundary_entropy": 0.690485, "bpb": 7.714495}, "th": {"chunks": 45705.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.841467, "cps": 357.070312, "mean_p": 0.525049, "boundary_entropy": 0.689958, "bpb": 7.952343}, "zh": {"chunks": 6192.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.048425, "cps": 48.375, "mean_p": 0.496739, "boundary_entropy": 0.68655, "bpb": 8.018642}, "my": {"chunks": 53234.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.809479, "cps": 415.890625, "mean_p": 0.520815, "boundary_entropy": 0.690615, "bpb": 8.175984}}, "bpb": {"en": 8.13387, "de": 8.0957, "vi": 8.07931, "tr": 8.06488, "fi": 8.1481, "ru": 7.94971, "el": 8.08968, "hi": 7.80583, "ta": 7.71449, "th": 7.95234, "zh": 8.01864, "my": 8.17598}, "gini_cpc": 0.392656, "gini_cps": 0.41036, "bpic": 1.326012, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1812076482570428, "vi": 0.6891941759470237, "tr": 1.253670611321081, "fi": 0.7447111935555503, "ru": 1.1076541693374156, "el": 3.572359554712924, "hi": 5.073286753338952, "ta": 5.317517904800032, "th": 5.706872911345723, "zh": 2.1056833261391343, "my": 5.642626151738392}, "t": 23.3}
{"event": "probe", "step": 4, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 8495.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.488134, "cps": 66.367188, "mean_p": 0.502995, "boundary_entropy": 0.687406, "bpb": 7.880173}, "de": {"chunks": 11746.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.572055, "cps": 91.765625, "mean_p": 0.506434, "boundary_entropy": 0.688171, "bpb": 7.85734}, "vi": {"chunks": 6152.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.333713, "cps": 48.0625, "mean_p": 0.49065, "boundary_entropy": 0.688472, "bpb": 7.947697}, "tr": {"chunks": 11059.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.610893, "cps": 86.398438, "mean_p": 0.505711, "boundary_entropy": 0.687855, "bpb": 7.887255}, "fi": {"chunks": 6557.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.348573, "cps": 51.226562, "mean_p": 0.495849, "boundary_entropy": 0.687881, "bpb": 7.953878}, "ru": {"chunks": 10418.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.549763, "cps": 81.390625, "mean_p": 0.496589, "boundary_entropy": 0.69019, "bpb": 7.71512}, "el": {"chunks": 36286.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.768065, "cps": 283.484375, "mean_p": 0.527791, "boundary_entropy": 0.689037, "bpb": 7.944995}, "hi": {"chunks": 42261.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.511798, "cps": 330.164062, "mean_p": 0.519333, "boundary_entropy": 0.690309, "bpb": 7.626739}, "ta": {"chunks": 48663.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.602995, "cps": 380.179688, "mean_p": 0.517803, "boundary_entropy": 0.690652, "bpb": 7.464029}, "th": {"chunks": 45642.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.837551, "cps": 356.578125, "mean_p": 0.523456, "boundary_entropy": 0.690105, "bpb": 7.68482}, "zh": {"chunks": 6131.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.038097, "cps": 47.898438, "mean_p": 0.496559, "boundary_entropy": 0.686548, "bpb": 7.983357}, "my": {"chunks": 53136.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.804307, "cps": 415.125, "mean_p": 0.519618, "boundary_entropy": 0.690706, "bpb": 8.006562}}, "bpb": {"en": 7.88017, "de": 7.85734, "vi": 7.9477, "tr": 7.88726, "fi": 7.95388, "ru": 7.71512, "el": 7.94499, "hi": 7.62674, "ta": 7.46403, "th": 7.68482, "zh": 7.98336, "my": 8.00656}, "gini_cpc": 0.396137, "gini_cps": 0.413072, "bpic": 1.339335, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1719209723608621, "vi": 0.6836501632367865, "tr": 1.2514861382598568, "fi": 0.7140917854877405, "ru": 1.1262527794558648, "el": 3.6220879217411137, "hi": 5.145711500603012, "ta": 5.332540303036838, "th": 5.813053746439374, "zh": 2.126662682994937, "my": 5.744949549356791}, "t": 30.7}
{"event": "probe", "step": 8, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 7807.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.448601, "cps": 60.992188, "mean_p": 0.501251, "boundary_entropy": 0.687376, "bpb": 7.151968}, "de": {"chunks": 10559.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.514245, "cps": 82.492188, "mean_p": 0.504071, "boundary_entropy": 0.688183, "bpb": 7.185544}, "vi": {"chunks": 5601.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.303824, "cps": 43.757812, "mean_p": 0.489145, "boundary_entropy": 0.688381, "bpb": 7.547238}, "tr": {"chunks": 10180.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.562338, "cps": 79.53125, "mean_p": 0.503555, "boundary_entropy": 0.68785, "bpb": 7.361281}, "fi": {"chunks": 5206.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.276753, "cps": 40.671875, "mean_p": 0.492488, "boundary_entropy": 0.68775, "bpb": 7.358177}, "ru": {"chunks": 11038.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.58248, "cps": 86.234375, "mean_p": 0.497073, "boundary_entropy": 0.690205, "bpb": 6.978729}, "el": {"chunks": 34647.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.688203, "cps": 270.679688, "mean_p": 0.521137, "boundary_entropy": 0.689644, "bpb": 7.475779}, "hi": {"chunks": 40285.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.394354, "cps": 314.726562, "mean_p": 0.51451, "boundary_entropy": 0.690603, "bpb": 7.076423}, "ta": {"chunks": 40607.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.172078, "cps": 317.242188, "mean_p": 0.509871, "boundary_entropy": 0.691059, "bpb": 6.800516}, "th": {"chunks": 44990.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.797016, "cps": 351.484375, "mean_p": 0.517848, "boundary_entropy": 0.690543, "bpb": 6.953838}, "zh": {"chunks": 5967.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.010328, "cps": 46.617188, "mean_p": 0.496019, "boundary_entropy": 0.686542, "bpb": 7.877131}, "my": {"chunks": 52704.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.781507, "cps": 411.75, "mean_p": 0.516575, "boundary_entropy": 0.69091, "bpb": 7.572034}}, "bpb": {"en": 7.15197, "de": 7.18554, "vi": 7.54724, "tr": 7.36128, "fi": 7.35818, "ru": 6.97873, "el": 7.47578, "hi": 7.07642, "ta": 6.80052, "th": 6.95384, "zh": 7.87713, "my": 7.57203}, "gini_cpc": 0.406356, "gini_cps": 0.41907, "bpic": 1.423568, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.146331756074759, "vi": 0.677270830886262, "tr": 1.2535370290951637, "fi": 0.6169247889230727, "ru": 1.2984376986622856, "el": 3.763264396752475, "hi": 5.3373813762214, "ta": 4.84189510699298, "th": 6.2349771864509185, "zh": 2.252177088376279, "my": 6.200406205664798}, "t": 40.1}
{"event": "probe", "step": 16, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 4768.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.273976, "cps": 37.25, "mean_p": 0.493032, "boundary_entropy": 0.687179, "bpb": 6.322329}, "de": {"chunks": 6435.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.313398, "cps": 50.273438, "mean_p": 0.495262, "boundary_entropy": 0.68812, "bpb": 6.401743}, "vi": {"chunks": 3348.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.181611, "cps": 26.15625, "mean_p": 0.481212, "boundary_entropy": 0.687794, "bpb": 7.026453}, "tr": {"chunks": 6577.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.36331, "cps": 51.382812, "mean_p": 0.49441, "boundary_entropy": 0.687704, "bpb": 6.707499}, "fi": {"chunks": 2560.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.136091, "cps": 20.0, "mean_p": 0.483639, "boundary_entropy": 0.687236, "bpb": 6.600391}, "ru": {"chunks": 10813.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.570607, "cps": 84.476562, "mean_p": 0.496722, "boundary_entropy": 0.69025, "bpb": 5.861531}, "el": {"chunks": 13364.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.651172, "cps": 104.40625, "mean_p": 0.497777, "boundary_entropy": 0.690365, "bpb": 6.549071}, "hi": {"chunks": 9635.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.57266, "cps": 75.273438, "mean_p": 0.49507, "boundary_entropy": 0.690848, "bpb": 5.875948}, "ta": {"chunks": 992.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.053062, "cps": 7.75, "mean_p": 0.481113, "boundary_entropy": 0.690413, "bpb": 6.218597}, "th": {"chunks": 5281.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.328318, "cps": 41.257812, "mean_p": 0.491517, "boundary_entropy": 0.690902, "bpb": 5.431572}, "zh": {"chunks": 5247.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.888419, "cps": 40.992188, "mean_p": 0.493542, "boundary_entropy": 0.686487, "bpb": 7.500069}, "my": {"chunks": 33813.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.784516, "cps": 264.164062, "mean_p": 0.504201, "boundary_entropy": 0.691354, "bpb": 6.23522}}, "bpb": {"en": 6.32233, "de": 6.40174, "vi": 7.02645, "tr": 6.7075, "fi": 6.60039, "ru": 5.86153, "el": 6.54907, "hi": 5.87595, "ta": 6.2186, "th": 5.43157, "zh": 7.50007, "my": 6.23522}, "gini_cpc": 0.427813, "gini_cps": 0.437604, "bpic": 3.73208, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.143889352530237, "vi": 0.6628727726475937, "tr": 1.3260660729783746, "fi": 0.49672492783210165, "ru": 2.0826911047263197, "el": 2.3767499600219364, "hi": 2.09018400281221, "ta": 0.19367522809637577, "th": 1.1983480560487176, "zh": 3.242690468628195, "my": 6.51340678010677}, "t": 48.2}
{"event": "probe", "step": 32, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 6447.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.370453, "cps": 50.367188, "mean_p": 0.499364, "boundary_entropy": 0.687587, "bpb": 5.770918}, "de": {"chunks": 8860.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.431501, "cps": 69.21875, "mean_p": 0.500963, "boundary_entropy": 0.688534, "bpb": 5.894567}, "vi": {"chunks": 2671.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.144887, "cps": 20.867188, "mean_p": 0.481003, "boundary_entropy": 0.687929, "bpb": 6.499267}, "tr": {"chunks": 6087.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.336243, "cps": 47.554688, "mean_p": 0.495127, "boundary_entropy": 0.687998, "bpb": 6.252336}, "fi": {"chunks": 4702.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.24996, "cps": 36.734375, "mean_p": 0.493656, "boundary_entropy": 0.688026, "bpb": 5.947849}, "ru": {"chunks": 737.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.038892, "cps": 5.757812, "mean_p": 0.482885, "boundary_entropy": 0.689709, "bpb": 5.353197}, "el": {"chunks": 300.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.014618, "cps": 2.34375, "mean_p": 0.456466, "boundary_entropy": 0.686366, "bpb": 6.34812}, "hi": {"chunks": 5811.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.345379, "cps": 45.398438, "mean_p": 0.492464, "boundary_entropy": 0.690784, "bpb": 4.948915}, "ta": {"chunks": 285.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.015245, "cps": 2.226562, "mean_p": 0.475386, "boundary_entropy": 0.689936, "bpb": 5.647165}, "th": {"chunks": 233.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.014486, "cps": 1.820312, "mean_p": 0.478626, "boundary_entropy": 0.690077, "bpb": 5.20762}, "zh": {"chunks": 3024.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.512022, "cps": 23.625, "mean_p": 0.486045, "boundary_entropy": 0.686232, "bpb": 7.109671}, "my": {"chunks": 409.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.021585, "cps": 3.195312, "mean_p": 0.47786, "boundary_entropy": 0.690285, "bpb": 5.547672}}, "bpb": {"en": 5.77092, "de": 5.89457, "vi": 6.49927, "tr": 6.25234, "fi": 5.94785, "ru": 5.3532, "el": 6.34812, "hi": 4.94892, "ta": 5.64717, "th": 5.20762, "zh": 7.10967, "my": 5.54767}, "gini_cpc": 0.478485, "gini_cps": 0.48468, "bpic": 9.699767, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1647903520095912, "vi": 0.39110844725907423, "tr": 0.9076516475561859, "fi": 0.6747411412077609, "ru": 0.10498438863998841, "el": 0.03945907129766472, "hi": 0.9323141002726311, "ta": 0.0411515161396028, "th": 0.03910221017850579, "zh": 1.3821487781982877, "my": 0.05826749960927573}, "t": 57.2}
{"event": "probe", "step": 64, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 5967.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.342872, "cps": 46.617188, "mean_p": 0.498624, "boundary_entropy": 0.687645, "bpb": 5.378728}, "de": {"chunks": 5554.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.270491, "cps": 43.390625, "mean_p": 0.495726, "boundary_entropy": 0.688477, "bpb": 5.526125}, "vi": {"chunks": 4209.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228316, "cps": 32.882812, "mean_p": 0.489239, "boundary_entropy": 0.688695, "bpb": 5.91743}, "tr": {"chunks": 4705.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.259902, "cps": 36.757812, "mean_p": 0.493231, "boundary_entropy": 0.688043, "bpb": 5.867975}, "fi": {"chunks": 5699.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.302961, "cps": 44.523438, "mean_p": 0.496537, "boundary_entropy": 0.688185, "bpb": 5.572196}, "ru": {"chunks": 5031.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.265488, "cps": 39.304688, "mean_p": 0.495571, "boundary_entropy": 0.69042, "bpb": 4.845102}, "el": {"chunks": 10113.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.492764, "cps": 79.007812, "mean_p": 0.497506, "boundary_entropy": 0.690624, "bpb": 5.069932}, "hi": {"chunks": 4327.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.257177, "cps": 33.804688, "mean_p": 0.491719, "boundary_entropy": 0.69077, "bpb": 4.566829}, "ta": {"chunks": 2684.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.143568, "cps": 20.96875, "mean_p": 0.489701, "boundary_entropy": 0.69103, "bpb": 4.497794}, "th": {"chunks": 1285.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.079888, "cps": 10.039062, "mean_p": 0.489003, "boundary_entropy": 0.690823, "bpb": 4.589578}, "zh": {"chunks": 2380.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.40298, "cps": 18.59375, "mean_p": 0.487634, "boundary_entropy": 0.686621, "bpb": 6.728442}, "my": {"chunks": 5363.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.283038, "cps": 41.898438, "mean_p": 0.493164, "boundary_entropy": 0.691292, "bpb": 4.417682}}, "bpb": {"en": 5.37873, "de": 5.52612, "vi": 5.91743, "tr": 5.86798, "fi": 5.5722, "ru": 4.8451, "el": 5.06993, "hi": 4.56683, "ta": 4.49779, "th": 4.58958, "zh": 6.72844, "my": 4.41768}, "gini_cpc": 0.199772, "gini_cps": 0.233577, "bpic": 6.695762, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 0.7888992634867997, "vi": 0.6658921055226037, "tr": 0.7580138810744073, "fi": 0.883598267960359, "ru": 0.7743069998447928, "el": 1.4371670264513587, "hi": 0.7500667340506516, "ta": 0.4187213681668003, "th": 0.2329969010471488, "zh": 1.175307741511602, "my": 0.8254913051973738}, "t": 73.2}
{"event": "train", "step": 100, "phase": 1, "objective": "parity_a", "lm_loss": 3.34134, "bpb": 4.82054, "chunk_loss": 0.033571, "F": 0.1559, "G": 0.47075, "chunks": 159.64, "bytes_per_chunk": 6.414, "boundary_entropy": 0.68874, "router_noise": 0.0, "grad_norm": 1.7149, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.731, "n_target_mean": 8.4807, "alpha": 0.03, "t": 82.8}
{"event": "probe", "step": 128, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 4099.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.235534, "cps": 32.023438, "mean_p": 0.471183, "boundary_entropy": 0.67248, "bpb": 4.350919}, "de": {"chunks": 4616.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.224809, "cps": 36.0625, "mean_p": 0.470109, "boundary_entropy": 0.675463, "bpb": 4.599024}, "vi": {"chunks": 6978.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.378519, "cps": 54.515625, "mean_p": 0.491543, "boundary_entropy": 0.68662, "bpb": 4.888916}, "tr": {"chunks": 4448.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.245705, "cps": 34.75, "mean_p": 0.482378, "boundary_entropy": 0.682018, "bpb": 4.926181}, "fi": {"chunks": 4838.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.25719, "cps": 37.796875, "mean_p": 0.478627, "boundary_entropy": 0.68149, "bpb": 4.615057}, "ru": {"chunks": 2168.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.114406, "cps": 16.9375, "mean_p": 0.472602, "boundary_entropy": 0.687769, "bpb": 3.283067}, "el": {"chunks": 3688.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.179701, "cps": 28.8125, "mean_p": 0.490368, "boundary_entropy": 0.690341, "bpb": 4.353848}, "hi": {"chunks": 6818.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.40523, "cps": 53.265625, "mean_p": 0.485233, "boundary_entropy": 0.690067, "bpb": 3.074676}, "ta": {"chunks": 5123.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.27403, "cps": 40.023438, "mean_p": 0.455804, "boundary_entropy": 0.685475, "bpb": 3.240709}, "th": {"chunks": 8840.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.54958, "cps": 69.0625, "mean_p": 0.488488, "boundary_entropy": 0.690443, "bpb": 2.89828}, "zh": {"chunks": 3955.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.669658, "cps": 30.898438, "mean_p": 0.493138, "boundary_entropy": 0.686965, "bpb": 6.253639}, "my": {"chunks": 3962.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.209099, "cps": 30.953125, "mean_p": 0.490433, "boundary_entropy": 0.691156, "bpb": 3.253163}}, "bpb": {"en": 4.35092, "de": 4.59902, "vi": 4.88892, "tr": 4.92618, "fi": 4.61506, "ru": 3.28307, "el": 4.35385, "hi": 3.07468, "ta": 3.24071, "th": 2.89828, "zh": 6.25364, "my": 3.25316}, "gini_cpc": 0.260568, "gini_cps": 0.183698, "bpic": 6.446525, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9544640930331335, "vi": 1.6070671546434852, "tr": 1.0431828269767545, "fi": 1.0919435378351563, "ru": 0.48573149638788277, "el": 0.7629503368589835, "hi": 1.7204740474156166, "ta": 1.1634429391678713, "th": 2.3333366448051387, "zh": 2.8431465571831342, "my": 0.8877635240897355}, "t": 92.7}
{"event": "train", "step": 200, "phase": 1, "objective": "parity_a", "lm_loss": 2.45426, "bpb": 3.54075, "chunk_loss": 0.031487, "F": 0.15192, "G": 0.39065, "chunks": 155.56, "bytes_per_chunk": 6.583, "boundary_entropy": 0.58371, "router_noise": 0.0, "grad_norm": 1.7171, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.264, "n_target_mean": 9.111, "alpha": 0.03, "t": 109.3}
{"event": "probe", "step": 256, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2898.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166523, "cps": 22.640625, "mean_p": 0.327077, "boundary_entropy": 0.450877, "bpb": 3.526471}, "de": {"chunks": 3264.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.158964, "cps": 25.5, "mean_p": 0.326955, "boundary_entropy": 0.481736, "bpb": 3.746667}, "vi": {"chunks": 4242.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230106, "cps": 33.140625, "mean_p": 0.324573, "boundary_entropy": 0.452579, "bpb": 3.062301}, "tr": {"chunks": 3430.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.189471, "cps": 26.796875, "mean_p": 0.344064, "boundary_entropy": 0.514493, "bpb": 3.706058}, "fi": {"chunks": 3005.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159747, "cps": 23.476562, "mean_p": 0.31072, "boundary_entropy": 0.494807, "bpb": 3.763909}, "ru": {"chunks": 3699.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.195198, "cps": 28.898438, "mean_p": 0.23219, "boundary_entropy": 0.478597, "bpb": 2.615859}, "el": {"chunks": 3619.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.176339, "cps": 28.273438, "mean_p": 0.348819, "boundary_entropy": 0.612655, "bpb": 2.694205}, "hi": {"chunks": 5463.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.324695, "cps": 42.679688, "mean_p": 0.444753, "boundary_entropy": 0.670884, "bpb": 2.082252}, "ta": {"chunks": 4668.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.249692, "cps": 36.46875, "mean_p": 0.395903, "boundary_entropy": 0.656393, "bpb": 1.956507}, "th": {"chunks": 4401.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.273609, "cps": 34.382812, "mean_p": 0.430752, "boundary_entropy": 0.676836, "bpb": 2.0736}, "zh": {"chunks": 10007.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.694379, "cps": 78.179688, "mean_p": 0.537047, "boundary_entropy": 0.614467, "bpb": 4.608815}, "my": {"chunks": 281.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.01483, "cps": 2.195312, "mean_p": 0.295721, "boundary_entropy": 0.578807, "bpb": 2.202979}}, "bpb": {"en": 3.52647, "de": 3.74667, "vi": 3.0623, "tr": 3.70606, "fi": 3.76391, "ru": 2.61586, "el": 2.69421, "hi": 2.08225, "ta": 1.95651, "th": 2.0736, "zh": 4.60881, "my": 2.20298}, "gini_cpc": 0.464028, "gini_cps": 0.255074, "bpic": 7.835943, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.954604510297467, "vi": 1.3818256860178537, "tr": 1.1378088506428994, "fi": 0.9593085870090768, "ru": 1.1721976761336632, "el": 1.0589451778033412, "hi": 1.9498529805358413, "ta": 1.4994469906757544, "th": 1.6430699100469748, "zh": 10.175041663774566, "my": 0.08905712747233248}, "t": 123.6}
{"event": "train", "step": 300, "phase": 1, "objective": "parity_a", "lm_loss": 2.14311, "bpb": 3.09185, "chunk_loss": 0.032264, "F": 0.15733, "G": 0.3348, "chunks": 161.11, "bytes_per_chunk": 6.356, "boundary_entropy": 0.50572, "router_noise": 0.0, "grad_norm": 0.9799, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.263, "n_target_mean": 9.2864, "alpha": 0.03, "t": 135.5}
{"event": "train", "step": 400, "phase": 1, "objective": "parity_a", "lm_loss": 2.0767, "bpb": 2.99604, "chunk_loss": 0.031542, "F": 0.15015, "G": 0.29291, "chunks": 153.75, "bytes_per_chunk": 6.66, "boundary_entropy": 0.45097, "router_noise": 0.0, "grad_norm": 1.1785, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.225, "n_target_mean": 9.1232, "alpha": 0.03, "t": 158.0}
{"event": "train", "step": 500, "phase": 1, "objective": "parity_a", "lm_loss": 1.95919, "bpb": 2.82651, "chunk_loss": 0.030461, "F": 0.14996, "G": 0.2395, "chunks": 153.56, "bytes_per_chunk": 6.668, "boundary_entropy": 0.35393, "router_noise": 0.0, "grad_norm": 0.9164, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.194, "n_target_mean": 8.7302, "alpha": 0.03, "t": 177.5}
{"event": "train", "step": 600, "phase": 1, "objective": "parity_a", "lm_loss": 1.83082, "bpb": 2.64132, "chunk_loss": 0.030298, "F": 0.14583, "G": 0.22485, "chunks": 149.33, "bytes_per_chunk": 6.857, "boundary_entropy": 0.30977, "router_noise": 0.0, "grad_norm": 0.7559, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.166, "n_target_mean": 8.7164, "alpha": 0.03, "t": 194.1}
{"event": "train", "step": 700, "phase": 1, "objective": "parity_a", "lm_loss": 1.71573, "bpb": 2.47527, "chunk_loss": 0.03025, "F": 0.13811, "G": 0.21707, "chunks": 141.42, "bytes_per_chunk": 7.241, "boundary_entropy": 0.30086, "router_noise": 0.0, "grad_norm": 0.7006, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.187, "n_target_mean": 9.1975, "alpha": 0.03, "t": 212.8}
{"event": "train", "step": 800, "phase": 1, "objective": "parity_a", "lm_loss": 1.60508, "bpb": 2.31565, "chunk_loss": 0.030265, "F": 0.13354, "G": 0.19421, "chunks": 136.75, "bytes_per_chunk": 7.488, "boundary_entropy": 0.26207, "router_noise": 0.0, "grad_norm": 0.6618, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.163, "n_target_mean": 9.5182, "alpha": 0.03, "t": 229.1}
{"event": "train", "step": 900, "phase": 1, "objective": "parity_a", "lm_loss": 1.5848, "bpb": 2.28638, "chunk_loss": 0.030407, "F": 0.14322, "G": 0.19636, "chunks": 146.66, "bytes_per_chunk": 6.982, "boundary_entropy": 0.24697, "router_noise": 0.0, "grad_norm": 0.6897, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.162, "n_target_mean": 9.5986, "alpha": 0.03, "t": 245.3}
{"event": "probe", "step": 1000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2880.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165489, "cps": 22.5, "mean_p": 0.191015, "boundary_entropy": 0.148588, "bpb": 2.328853}, "de": {"chunks": 3215.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.156577, "cps": 25.117188, "mean_p": 0.181553, "boundary_entropy": 0.159285, "bpb": 2.620215}, "vi": {"chunks": 4136.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.224356, "cps": 32.3125, "mean_p": 0.157182, "boundary_entropy": 0.157287, "bpb": 2.130938}, "tr": {"chunks": 3148.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.173894, "cps": 24.59375, "mean_p": 0.187689, "boundary_entropy": 0.187988, "bpb": 2.615206}, "fi": {"chunks": 2808.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.149274, "cps": 21.9375, "mean_p": 0.178788, "boundary_entropy": 0.199421, "bpb": 2.851249}, "ru": {"chunks": 2885.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.152243, "cps": 22.539062, "mean_p": 0.124495, "boundary_entropy": 0.199259, "bpb": 1.705945}, "el": {"chunks": 2903.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.141451, "cps": 22.679688, "mean_p": 0.136002, "boundary_entropy": 0.259361, "bpb": 1.79643}, "hi": {"chunks": 3371.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.200357, "cps": 26.335938, "mean_p": 0.101668, "boundary_entropy": 0.237154, "bpb": 1.466475}, "ta": {"chunks": 2494.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.133405, "cps": 19.484375, "mean_p": 0.222738, "boundary_entropy": 0.433289, "bpb": 1.168251}, "th": {"chunks": 1231.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.076531, "cps": 9.617188, "mean_p": 0.212508, "boundary_entropy": 0.427019, "bpb": 1.409196}, "zh": {"chunks": 3371.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.570775, "cps": 26.335938, "mean_p": 0.328353, "boundary_entropy": 0.483437, "bpb": 3.479001}, "my": {"chunks": 1586.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.083703, "cps": 12.390625, "mean_p": 0.22536, "boundary_entropy": 0.456706, "bpb": 1.268494}}, "bpb": {"en": 2.32885, "de": 2.62022, "vi": 2.13094, "tr": 2.61521, "fi": 2.85125, "ru": 1.70594, "el": 1.79643, "hi": 1.46647, "ta": 1.16825, "th": 1.4092, "zh": 3.479, "my": 1.26849}, "gini_cpc": 0.276266, "gini_cps": 0.13963, "bpic": 11.278388, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9461504549586843, "vi": 1.3557169333694963, "tr": 1.0507896941575061, "fi": 0.9020214236351071, "ru": 0.919958498240985, "el": 0.8547474682876123, "hi": 1.2106965700842, "ta": 0.8061254123205848, "th": 0.4624540850689048, "zh": 3.4490297649283215, "my": 0.5057913983885722}, "t": 263.9}
{"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.33837, "de": 2.61094, "vi": 2.13548, "tr": 2.60755, "fi": 2.8622, "ru": 1.70589, "el": 1.79106, "hi": 1.47457, "ta": 1.16712, "th": 1.40727, "zh": 3.48573, "my": 1.28094}, "bytes_per_chunk": {"en": 5.607, "de": 6.362, "vi": 5.807, "tr": 5.982, "fi": 6.603, "ru": 11.65, "el": 12.287, "hi": 11.848, "ta": 18.177, "th": 29.182, "zh": 5.242, "my": 29.685}, "bpb_mean": 2.07226, "t": 268.8}
{"event": "train", "step": 1000, "phase": 1, "objective": "parity_a", "lm_loss": 1.50446, "bpb": 2.17048, "chunk_loss": 0.030029, "F": 0.13657, "G": 0.19081, "chunks": 139.84, "bytes_per_chunk": 7.322, "boundary_entropy": 0.22987, "router_noise": 0.0, "grad_norm": 0.7157, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.237, "n_target_mean": 9.1453, "alpha": 0.03, "t": 269.0}
{"event": "train", "step": 1100, "phase": 1, "objective": "parity_a", "lm_loss": 1.43043, "bpb": 2.06368, "chunk_loss": 0.030458, "F": 0.13094, "G": 0.18246, "chunks": 134.08, "bytes_per_chunk": 7.637, "boundary_entropy": 0.23602, "router_noise": 0.0, "grad_norm": 0.6094, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 10.1506, "alpha": 0.03, "t": 286.9}
{"event": "train", "step": 1200, "phase": 1, "objective": "parity_a", "lm_loss": 1.46222, "bpb": 2.10954, "chunk_loss": 0.030195, "F": 0.13525, "G": 0.18244, "chunks": 138.5, "bytes_per_chunk": 7.394, "boundary_entropy": 0.21784, "router_noise": 0.0, "grad_norm": 0.5885, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 9.7949, "alpha": 0.03, "t": 305.5}
{"event": "train", "step": 1300, "phase": 1, "objective": "parity_a", "lm_loss": 1.47374, "bpb": 2.12616, "chunk_loss": 0.029992, "F": 0.13844, "G": 0.18966, "chunks": 141.77, "bytes_per_chunk": 7.223, "boundary_entropy": 0.20857, "router_noise": 0.0, "grad_norm": 0.7001, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 9.1751, "alpha": 0.03, "t": 323.2}
{"event": "train", "step": 1400, "phase": 1, "objective": "parity_a", "lm_loss": 1.43701, "bpb": 2.07317, "chunk_loss": 0.030036, "F": 0.13689, "G": 0.17864, "chunks": 140.17, "bytes_per_chunk": 7.305, "boundary_entropy": 0.19366, "router_noise": 0.0, "grad_norm": 0.5595, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 9.4428, "alpha": 0.03, "t": 341.7}
{"event": "train", "step": 1500, "phase": 1, "objective": "parity_a", "lm_loss": 1.44171, "bpb": 2.07995, "chunk_loss": 0.030169, "F": 0.13213, "G": 0.17569, "chunks": 135.3, "bytes_per_chunk": 7.569, "boundary_entropy": 0.20954, "router_noise": 0.0, "grad_norm": 0.5581, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 9.4288, "alpha": 0.03, "t": 359.7}
{"event": "train", "step": 1600, "phase": 1, "objective": "parity_a", "lm_loss": 1.41686, "bpb": 2.0441, "chunk_loss": 0.030113, "F": 0.1402, "G": 0.18078, "chunks": 143.56, "bytes_per_chunk": 7.133, "boundary_entropy": 0.20072, "router_noise": 0.0, "grad_norm": 0.5522, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.188, "n_target_mean": 8.9822, "alpha": 0.03, "t": 378.4}
{"event": "train", "step": 1700, "phase": 1, "objective": "parity_a", "lm_loss": 1.36442, "bpb": 1.96844, "chunk_loss": 0.03017, "F": 0.13774, "G": 0.167, "chunks": 141.05, "bytes_per_chunk": 7.26, "boundary_entropy": 0.16349, "router_noise": 0.0, "grad_norm": 0.4909, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 9.2321, "alpha": 0.03, "t": 396.3}
{"event": "train", "step": 1800, "phase": 1, "objective": "parity_a", "lm_loss": 1.37059, "bpb": 1.97734, "chunk_loss": 0.030195, "F": 0.13446, "G": 0.18346, "chunks": 137.69, "bytes_per_chunk": 7.437, "boundary_entropy": 0.21084, "router_noise": 0.0, "grad_norm": 0.4909, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.19, "n_target_mean": 9.9916, "alpha": 0.03, "t": 415.4}
{"event": "train", "step": 1900, "phase": 1, "objective": "parity_a", "lm_loss": 1.36717, "bpb": 1.97241, "chunk_loss": 0.030345, "F": 0.14377, "G": 0.17949, "chunks": 147.22, "bytes_per_chunk": 6.956, "boundary_entropy": 0.17485, "router_noise": 0.0, "grad_norm": 0.516, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.1896, "alpha": 0.03, "t": 432.5}
{"event": "probe", "step": 2000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.178755, "boundary_entropy": 0.091655, "bpb": 2.079254}, "de": {"chunks": 3219.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.156772, "cps": 25.148438, "mean_p": 0.169124, "boundary_entropy": 0.095607, "bpb": 2.29599}, "vi": {"chunks": 4142.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.224681, "cps": 32.359375, "mean_p": 0.153266, "boundary_entropy": 0.117615, "bpb": 1.966634}, "tr": {"chunks": 2991.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.165221, "cps": 23.367188, "mean_p": 0.169805, "boundary_entropy": 0.123817, "bpb": 2.289895}, "fi": {"chunks": 2811.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.149434, "cps": 21.960938, "mean_p": 0.165934, "boundary_entropy": 0.13527, "bpb": 2.495648}, "ru": {"chunks": 2785.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.146966, "cps": 21.757812, "mean_p": 0.120749, "boundary_entropy": 0.168263, "bpb": 1.423606}, "el": {"chunks": 3148.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153389, "cps": 24.59375, "mean_p": 0.100059, "boundary_entropy": 0.16963, "bpb": 1.49369}, "hi": {"chunks": 3190.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.189599, "cps": 24.921875, "mean_p": 0.09155, "boundary_entropy": 0.209624, "bpb": 1.242918}, "ta": {"chunks": 2350.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.125702, "cps": 18.359375, "mean_p": 0.167977, "boundary_entropy": 0.356098, "bpb": 1.062128}, "th": {"chunks": 2616.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.162636, "cps": 20.4375, "mean_p": 0.193245, "boundary_entropy": 0.367093, "bpb": 1.179905}, "zh": {"chunks": 3028.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.512699, "cps": 23.65625, "mean_p": 0.253762, "boundary_entropy": 0.332852, "bpb": 3.179278}, "my": {"chunks": 2201.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.11616, "cps": 17.195312, "mean_p": 0.159172, "boundary_entropy": 0.347555, "bpb": 1.153515}}, "bpb": {"en": 2.07925, "de": 2.29599, "vi": 1.96663, "tr": 2.28989, "fi": 2.49565, "ru": 1.42361, "el": 1.49369, "hi": 1.24292, "ta": 1.06213, "th": 1.17991, "zh": 3.17928, "my": 1.15352}, "gini_cpc": 0.208405, "gini_cps": 0.084257, "bpic": 10.85108, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.945030675484146, "vi": 1.3543917164089296, "tr": 0.9959629844701339, "fi": 0.9007956877739419, "ru": 0.885917583416882, "el": 0.9246368867350513, "hi": 1.1429124048933812, "ta": 0.7577391545855836, "th": 0.9803790242755865, "zh": 3.090578396494861, "my": 0.7002191804321221}, "t": 453.7}
{"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.08793, "de": 2.28724, "vi": 1.97649, "tr": 2.28324, "fi": 2.51254, "ru": 1.42428, "el": 1.49059, "hi": 1.25051, "ta": 1.06152, "th": 1.17292, "zh": 3.18532, "my": 1.16426}, "bytes_per_chunk": {"en": 5.598, "de": 6.372, "vi": 5.81, "tr": 6.336, "fi": 6.61, "ru": 11.948, "el": 11.392, "hi": 12.146, "ta": 20.56, "th": 14.728, "zh": 5.51, "my": 22.747}, "bpb_mean": 1.82474, "t": 458.7}
{"event": "train", "step": 2000, "phase": 1, "objective": "parity_a", "lm_loss": 1.36306, "bpb": 1.96648, "chunk_loss": 0.030214, "F": 0.14079, "G": 0.1673, "chunks": 144.17, "bytes_per_chunk": 7.103, "boundary_entropy": 0.15965, "router_noise": 0.0, "grad_norm": 0.5105, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.264, "n_target_mean": 9.1542, "alpha": 0.03, "t": 458.8}
{"event": "train", "step": 2100, "phase": 1, "objective": "parity_a", "lm_loss": 1.3012, "bpb": 1.87723, "chunk_loss": 0.030033, "F": 0.13391, "G": 0.16659, "chunks": 137.12, "bytes_per_chunk": 7.468, "boundary_entropy": 0.16465, "router_noise": 0.0, "grad_norm": 0.532, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.18, "n_target_mean": 9.4996, "alpha": 0.03, "t": 476.9}
{"event": "train", "step": 2200, "phase": 1, "objective": "parity_a", "lm_loss": 1.26877, "bpb": 1.83045, "chunk_loss": 0.030088, "F": 0.12825, "G": 0.16148, "chunks": 131.33, "bytes_per_chunk": 7.797, "boundary_entropy": 0.17992, "router_noise": 0.0, "grad_norm": 0.5219, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 10.2711, "alpha": 0.03, "t": 494.6}
{"event": "train", "step": 2300, "phase": 1, "objective": "parity_a", "lm_loss": 1.34008, "bpb": 1.93333, "chunk_loss": 0.030189, "F": 0.13708, "G": 0.1667, "chunks": 140.38, "bytes_per_chunk": 7.295, "boundary_entropy": 0.16339, "router_noise": 0.0, "grad_norm": 0.4803, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 9.4625, "alpha": 0.03, "t": 513.2}
{"event": "train", "step": 2400, "phase": 1, "objective": "parity_a", "lm_loss": 1.35497, "bpb": 1.95481, "chunk_loss": 0.030145, "F": 0.14265, "G": 0.16587, "chunks": 146.08, "bytes_per_chunk": 7.01, "boundary_entropy": 0.14185, "router_noise": 0.0, "grad_norm": 0.5047, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.6679, "alpha": 0.03, "t": 530.9}
{"event": "train", "step": 2500, "phase": 1, "objective": "parity_a", "lm_loss": 1.26729, "bpb": 1.82832, "chunk_loss": 0.03023, "F": 0.13159, "G": 0.16084, "chunks": 134.75, "bytes_per_chunk": 7.599, "boundary_entropy": 0.15935, "router_noise": 0.0, "grad_norm": 0.4453, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.7102, "alpha": 0.03, "t": 548.6}
{"event": "train", "step": 2600, "phase": 1, "objective": "parity_a", "lm_loss": 1.28537, "bpb": 1.8544, "chunk_loss": 0.030305, "F": 0.13713, "G": 0.16572, "chunks": 140.42, "bytes_per_chunk": 7.292, "boundary_entropy": 0.15399, "router_noise": 0.0, "grad_norm": 0.4592, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.3894, "alpha": 0.03, "t": 566.2}
{"event": "train", "step": 2700, "phase": 1, "objective": "parity_a", "lm_loss": 1.28331, "bpb": 1.85143, "chunk_loss": 0.030027, "F": 0.13121, "G": 0.15994, "chunks": 134.36, "bytes_per_chunk": 7.621, "boundary_entropy": 0.15092, "router_noise": 0.0, "grad_norm": 0.486, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 9.7529, "alpha": 0.03, "t": 583.9}
{"event": "train", "step": 2800, "phase": 1, "objective": "parity_a", "lm_loss": 1.36158, "bpb": 1.96435, "chunk_loss": 0.030033, "F": 0.14076, "G": 0.17412, "chunks": 144.14, "bytes_per_chunk": 7.104, "boundary_entropy": 0.15739, "router_noise": 0.0, "grad_norm": 0.4632, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.0065, "alpha": 0.03, "t": 601.7}
{"event": "train", "step": 2900, "phase": 1, "objective": "parity_a", "lm_loss": 1.28252, "bpb": 1.85028, "chunk_loss": 0.029973, "F": 0.13638, "G": 0.16275, "chunks": 139.66, "bytes_per_chunk": 7.332, "boundary_entropy": 0.14258, "router_noise": 0.0, "grad_norm": 0.4317, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 9.3755, "alpha": 0.03, "t": 618.5}
{"event": "probe", "step": 3000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2878.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165374, "cps": 22.484375, "mean_p": 0.174231, "boundary_entropy": 0.071512, "bpb": 1.971714}, "de": {"chunks": 3172.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154483, "cps": 24.78125, "mean_p": 0.162992, "boundary_entropy": 0.073286, "bpb": 2.159385}, "vi": {"chunks": 4105.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222674, "cps": 32.070312, "mean_p": 0.147486, "boundary_entropy": 0.107515, "bpb": 1.848415}, "tr": {"chunks": 2996.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.165497, "cps": 23.40625, "mean_p": 0.165685, "boundary_entropy": 0.101172, "bpb": 2.144843}, "fi": {"chunks": 2818.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.149806, "cps": 22.015625, "mean_p": 0.162757, "boundary_entropy": 0.119392, "bpb": 2.317682}, "ru": {"chunks": 2791.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.147282, "cps": 21.804688, "mean_p": 0.116228, "boundary_entropy": 0.150626, "bpb": 1.318815}, "el": {"chunks": 3165.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.154217, "cps": 24.726562, "mean_p": 0.088411, "boundary_entropy": 0.11881, "bpb": 1.379059}, "hi": {"chunks": 3079.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.183001, "cps": 24.054688, "mean_p": 0.079601, "boundary_entropy": 0.176266, "bpb": 1.153485}, "ta": {"chunks": 2854.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.152661, "cps": 22.296875, "mean_p": 0.16769, "boundary_entropy": 0.325256, "bpb": 0.969129}, "th": {"chunks": 3151.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.195897, "cps": 24.617188, "mean_p": 0.19075, "boundary_entropy": 0.348261, "bpb": 1.084765}, "zh": {"chunks": 3070.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.51981, "cps": 23.984375, "mean_p": 0.24647, "boundary_entropy": 0.268727, "bpb": 2.98333}, "my": {"chunks": 2066.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.109035, "cps": 16.140625, "mean_p": 0.165166, "boundary_entropy": 0.342624, "bpb": 1.083143}}, "bpb": {"en": 1.97171, "de": 2.15939, "vi": 1.84841, "tr": 2.14484, "fi": 2.31768, "ru": 1.31882, "el": 1.37906, "hi": 1.15349, "ta": 0.96913, "th": 1.08476, "zh": 2.98333, "my": 1.08314}, "gini_cpc": 0.205344, "gini_cps": 0.070201, "bpic": 10.617817, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9341445880759348, "vi": 1.3464906750291628, "tr": 1.0007476713149837, "fi": 0.9058628220465537, "ru": 0.890602587915604, "el": 0.9325372739823727, "hi": 1.1065930711747778, "ta": 0.9231277998506064, "th": 1.1845698332542351, "zh": 3.143245217464154, "my": 0.6593261056227474}, "t": 638.8}
{"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 1.97996, "de": 2.15312, "vi": 1.85314, "tr": 2.13528, "fi": 2.33278, "ru": 1.31931, "el": 1.37334, "hi": 1.16108, "ta": 0.96933, "th": 1.08201, "zh": 2.99102, "my": 1.0931}, "bytes_per_chunk": {"en": 5.613, "de": 6.447, "vi": 5.829, "tr": 6.276, "fi": 6.585, "ru": 11.934, "el": 11.375, "hi": 13.36, "ta": 17.264, "th": 13.503, "zh": 5.338, "my": 23.935}, "bpb_mean": 1.70362, "t": 643.7}
{"event": "train", "step": 3000, "phase": 1, "objective": "parity_a", "lm_loss": 1.33472, "bpb": 1.92559, "chunk_loss": 0.030192, "F": 0.15015, "G": 0.17298, "chunks": 153.75, "bytes_per_chunk": 6.66, "boundary_entropy": 0.13317, "router_noise": 0.0, "grad_norm": 0.4399, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.254, "n_target_mean": 8.6315, "alpha": 0.03, "t": 643.9}
{"event": "train", "step": 3100, "phase": 1, "objective": "parity_a", "lm_loss": 1.33201, "bpb": 1.92168, "chunk_loss": 0.030049, "F": 0.14268, "G": 0.16442, "chunks": 146.11, "bytes_per_chunk": 7.008, "boundary_entropy": 0.12887, "router_noise": 0.0, "grad_norm": 0.4496, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.7209, "alpha": 0.03, "t": 661.6}
{"event": "train", "step": 3200, "phase": 1, "objective": "parity_a", "lm_loss": 1.21256, "bpb": 1.74935, "chunk_loss": 0.030237, "F": 0.13486, "G": 0.16037, "chunks": 138.09, "bytes_per_chunk": 7.415, "boundary_entropy": 0.14268, "router_noise": 0.0, "grad_norm": 0.4423, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 10.0048, "alpha": 0.03, "t": 679.4}
{"event": "train", "step": 3300, "phase": 1, "objective": "parity_a", "lm_loss": 1.28701, "bpb": 1.85677, "chunk_loss": 0.030042, "F": 0.13541, "G": 0.16963, "chunks": 138.66, "bytes_per_chunk": 7.385, "boundary_entropy": 0.15511, "router_noise": 0.0, "grad_norm": 0.445, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 9.5165, "alpha": 0.03, "t": 696.3}
{"event": "train", "step": 3400, "phase": 1, "objective": "parity_a", "lm_loss": 1.24302, "bpb": 1.79329, "chunk_loss": 0.029999, "F": 0.13913, "G": 0.16748, "chunks": 142.47, "bytes_per_chunk": 7.188, "boundary_entropy": 0.1314, "router_noise": 0.0, "grad_norm": 0.4592, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.1341, "alpha": 0.03, "t": 713.9}
{"event": "train", "step": 3500, "phase": 1, "objective": "parity_a", "lm_loss": 1.3215, "bpb": 1.90652, "chunk_loss": 0.029978, "F": 0.14674, "G": 0.1693, "chunks": 150.27, "bytes_per_chunk": 6.815, "boundary_entropy": 0.12971, "router_noise": 0.0, "grad_norm": 0.4275, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.167, "n_target_mean": 8.1605, "alpha": 0.03, "t": 730.6}
{"event": "train", "step": 3600, "phase": 1, "objective": "parity_a", "lm_loss": 1.23801, "bpb": 1.78607, "chunk_loss": 0.030205, "F": 0.13861, "G": 0.16416, "chunks": 141.94, "bytes_per_chunk": 7.214, "boundary_entropy": 0.13049, "router_noise": 0.0, "grad_norm": 0.4635, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 9.0714, "alpha": 0.03, "t": 748.4}
{"event": "train", "step": 3700, "phase": 1, "objective": "parity_a", "lm_loss": 1.14838, "bpb": 1.65677, "chunk_loss": 0.030089, "F": 0.11867, "G": 0.14674, "chunks": 121.52, "bytes_per_chunk": 8.427, "boundary_entropy": 0.14372, "router_noise": 0.0, "grad_norm": 0.4249, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 10.6964, "alpha": 0.03, "t": 766.0}
{"event": "train", "step": 3800, "phase": 1, "objective": "parity_a", "lm_loss": 1.23989, "bpb": 1.78878, "chunk_loss": 0.030014, "F": 0.13557, "G": 0.16039, "chunks": 138.83, "bytes_per_chunk": 7.376, "boundary_entropy": 0.12589, "router_noise": 0.0, "grad_norm": 0.4389, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.1599, "alpha": 0.03, "t": 783.0}
{"event": "train", "step": 3900, "phase": 1, "objective": "parity_a", "lm_loss": 1.25534, "bpb": 1.81108, "chunk_loss": 0.030243, "F": 0.15033, "G": 0.16774, "chunks": 153.94, "bytes_per_chunk": 6.652, "boundary_entropy": 0.12007, "router_noise": 0.0, "grad_norm": 0.4289, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.5621, "alpha": 0.03, "t": 800.6}
{"event": "probe", "step": 4000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.1717, "boundary_entropy": 0.059733, "bpb": 1.896732}, "de": {"chunks": 3182.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15497, "cps": 24.859375, "mean_p": 0.161573, "boundary_entropy": 0.063609, "bpb": 2.076914}, "vi": {"chunks": 4024.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.21828, "cps": 31.4375, "mean_p": 0.143937, "boundary_entropy": 0.103971, "bpb": 1.760323}, "tr": {"chunks": 2861.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.15804, "cps": 22.351562, "mean_p": 0.158355, "boundary_entropy": 0.093207, "bpb": 2.065817}, "fi": {"chunks": 2805.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.149115, "cps": 21.914062, "mean_p": 0.158004, "boundary_entropy": 0.100872, "bpb": 2.213789}, "ru": {"chunks": 2736.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.14438, "cps": 21.375, "mean_p": 0.108715, "boundary_entropy": 0.128895, "bpb": 1.260431}, "el": {"chunks": 3143.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153145, "cps": 24.554688, "mean_p": 0.085519, "boundary_entropy": 0.112671, "bpb": 1.304026}, "hi": {"chunks": 2819.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.167548, "cps": 22.023438, "mean_p": 0.072544, "boundary_entropy": 0.157475, "bpb": 1.08968}, "ta": {"chunks": 2333.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.124793, "cps": 18.226562, "mean_p": 0.133626, "boundary_entropy": 0.281656, "bpb": 0.924178}, "th": {"chunks": 1720.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.106932, "cps": 13.4375, "mean_p": 0.181444, "boundary_entropy": 0.333601, "bpb": 1.036915}, "zh": {"chunks": 2974.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503556, "cps": 23.234375, "mean_p": 0.228801, "boundary_entropy": 0.225446, "bpb": 2.884718}, "my": {"chunks": 2920.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.154106, "cps": 22.8125, "mean_p": 0.177647, "boundary_entropy": 0.338158, "bpb": 1.045612}}, "bpb": {"en": 1.89673, "de": 2.07691, "vi": 1.76032, "tr": 2.06582, "fi": 2.21379, "ru": 1.26043, "el": 1.30403, "hi": 1.08968, "ta": 0.92418, "th": 1.03692, "zh": 2.88472, "my": 1.04561}, "gini_cpc": 0.20698, "gini_cps": 0.08964, "bpic": 11.154479, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9335215469320997, "vi": 1.314896012793399, "tr": 0.9520152050207222, "fi": 0.8982506921200577, "ru": 0.8697280102581806, "el": 0.9225291847674486, "hi": 1.0092914199836753, "ta": 0.7517368635779526, "th": 0.6441454745702914, "zh": 3.033361008165658, "my": 0.9283164658328165}, "t": 820.4}
{"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.90305, "de": 2.07096, "vi": 1.76894, "tr": 2.05683, "fi": 2.23057, "ru": 1.26068, "el": 1.29995, "hi": 1.09874, "ta": 0.92642, "th": 1.03643, "zh": 2.8917, "my": 1.05385}, "bytes_per_chunk": {"en": 5.607, "de": 6.44, "vi": 5.936, "tr": 6.611, "fi": 6.638, "ru": 12.265, "el": 11.482, "hi": 14.884, "ta": 20.77, "th": 23.354, "zh": 5.606, "my": 17.462}, "bpb_mean": 1.63318, "t": 825.3}
{"event": "train", "step": 4000, "phase": 1, "objective": "parity_a", "lm_loss": 1.31998, "bpb": 1.90433, "chunk_loss": 0.030173, "F": 0.13893, "G": 0.16494, "chunks": 142.27, "bytes_per_chunk": 7.198, "boundary_entropy": 0.13533, "router_noise": 0.0, "grad_norm": 0.4397, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.249, "n_target_mean": 8.9871, "alpha": 0.03, "t": 825.5}
{"event": "train", "step": 4100, "phase": 1, "objective": "parity_a", "lm_loss": 1.30056, "bpb": 1.87632, "chunk_loss": 0.03016, "F": 0.14308, "G": 0.16648, "chunks": 146.52, "bytes_per_chunk": 6.989, "boundary_entropy": 0.12282, "router_noise": 0.0, "grad_norm": 0.4339, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.5308, "alpha": 0.03, "t": 842.9}
{"event": "train", "step": 4200, "phase": 1, "objective": "parity_a", "lm_loss": 1.24028, "bpb": 1.78935, "chunk_loss": 0.030055, "F": 0.14343, "G": 0.16298, "chunks": 146.88, "bytes_per_chunk": 6.972, "boundary_entropy": 0.11632, "router_noise": 0.0, "grad_norm": 0.4421, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.165, "n_target_mean": 8.547, "alpha": 0.03, "t": 859.4}
{"event": "train", "step": 4300, "phase": 1, "objective": "parity_a", "lm_loss": 1.2155, "bpb": 1.75359, "chunk_loss": 0.030204, "F": 0.13757, "G": 0.16553, "chunks": 140.88, "bytes_per_chunk": 7.269, "boundary_entropy": 0.14423, "router_noise": 0.0, "grad_norm": 0.418, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.337, "alpha": 0.03, "t": 876.5}
{"event": "phase_transition", "step": 4400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 893.5}
{"event": "train", "step": 4400, "phase": 2, "objective": "parity_a", "lm_loss": 1.18055, "bpb": 1.70317, "chunk_loss": 0.030173, "F": 0.13408, "G": 0.15651, "chunks": 137.3, "bytes_per_chunk": 7.458, "boundary_entropy": 0.12285, "router_noise": 0.0, "grad_norm": 0.4078, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.7474, "alpha": 0.03, "t": 893.6}
{"event": "train", "step": 4500, "phase": 2, "objective": "parity_a", "lm_loss": 1.25596, "bpb": 1.81197, "chunk_loss": 0.030079, "F": 0.14044, "G": 0.16772, "chunks": 143.81, "bytes_per_chunk": 7.12, "boundary_entropy": 0.13307, "router_noise": 0.0, "grad_norm": 0.428, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.995, "alpha": 0.03, "t": 910.8}
{"event": "train", "step": 4600, "phase": 2, "objective": "parity_a", "lm_loss": 1.20185, "bpb": 1.73391, "chunk_loss": 0.03017, "F": 0.13702, "G": 0.16031, "chunks": 140.31, "bytes_per_chunk": 7.298, "boundary_entropy": 0.12912, "router_noise": 0.0, "grad_norm": 0.4088, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.4029, "alpha": 0.03, "t": 927.9}
{"event": "train", "step": 4700, "phase": 2, "objective": "parity_a", "lm_loss": 1.13288, "bpb": 1.6344, "chunk_loss": 0.029949, "F": 0.12605, "G": 0.16006, "chunks": 129.08, "bytes_per_chunk": 7.933, "boundary_entropy": 0.14088, "router_noise": 0.0, "grad_norm": 0.4174, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 10.3158, "alpha": 0.03, "t": 944.9}
{"event": "train", "step": 4800, "phase": 2, "objective": "parity_a", "lm_loss": 1.25921, "bpb": 1.81666, "chunk_loss": 0.030054, "F": 0.14384, "G": 0.16507, "chunks": 147.3, "bytes_per_chunk": 6.952, "boundary_entropy": 0.11291, "router_noise": 0.0, "grad_norm": 0.4189, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.5167, "alpha": 0.03, "t": 962.0}
{"event": "train", "step": 4900, "phase": 2, "objective": "parity_a", "lm_loss": 1.25332, "bpb": 1.80816, "chunk_loss": 0.030076, "F": 0.15048, "G": 0.16705, "chunks": 154.09, "bytes_per_chunk": 6.645, "boundary_entropy": 0.10287, "router_noise": 0.0, "grad_norm": 0.4376, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 7.9172, "alpha": 0.03, "t": 979.0}
{"event": "probe", "step": 5000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.171129, "boundary_entropy": 0.049553, "bpb": 1.838963}, "de": {"chunks": 3159.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15385, "cps": 24.679688, "mean_p": 0.158752, "boundary_entropy": 0.052932, "bpb": 2.014893}, "vi": {"chunks": 4115.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223217, "cps": 32.148438, "mean_p": 0.15094, "boundary_entropy": 0.090363, "bpb": 1.682447}, "tr": {"chunks": 2795.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.154394, "cps": 21.835938, "mean_p": 0.15411, "boundary_entropy": 0.084968, "bpb": 1.99651}, "fi": {"chunks": 2766.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.147042, "cps": 21.609375, "mean_p": 0.155759, "boundary_entropy": 0.101168, "bpb": 2.13597}, "ru": {"chunks": 2779.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.146649, "cps": 21.710938, "mean_p": 0.11256, "boundary_entropy": 0.134025, "bpb": 1.21684}, "el": {"chunks": 3150.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153486, "cps": 24.609375, "mean_p": 0.088157, "boundary_entropy": 0.0913, "bpb": 1.248358}, "hi": {"chunks": 2953.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.175513, "cps": 23.070312, "mean_p": 0.069668, "boundary_entropy": 0.144127, "bpb": 1.049691}, "ta": {"chunks": 2816.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.150629, "cps": 22.0, "mean_p": 0.143512, "boundary_entropy": 0.282138, "bpb": 0.891872}, "th": {"chunks": 2467.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.153373, "cps": 19.273438, "mean_p": 0.182529, "boundary_entropy": 0.328226, "bpb": 1.00982}, "zh": {"chunks": 2968.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50254, "cps": 23.1875, "mean_p": 0.2281, "boundary_entropy": 0.209467, "bpb": 2.804779}, "my": {"chunks": 2944.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.155373, "cps": 23.0, "mean_p": 0.187343, "boundary_entropy": 0.351507, "bpb": 1.008096}}, "bpb": {"en": 1.83896, "de": 2.01489, "vi": 1.68245, "tr": 1.99651, "fi": 2.13597, "ru": 1.21684, "el": 1.24836, "hi": 1.04969, "ta": 0.89187, "th": 1.00982, "zh": 2.80478, "my": 1.0081}, "gini_cpc": 0.176884, "gini_cps": 0.059572, "bpic": 10.71984, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9267739053295105, "vi": 1.3446314842556752, "tr": 0.9300533023533445, "fi": 0.8857616450638429, "ru": 0.8833969811796359, "el": 0.9245838154684896, "hi": 1.057267670525645, "ta": 0.9073686274477131, "th": 0.9238993521889003, "zh": 3.0272412482298834, "my": 0.9359464641821273}, "t": 998.7}
{"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.84523, "de": 2.00968, "vi": 1.69311, "tr": 1.99046, "fi": 2.15209, "ru": 1.21782, "el": 1.24506, "hi": 1.05655, "ta": 0.89407, "th": 1.00769, "zh": 2.81378, "my": 1.01569}, "bytes_per_chunk": {"en": 5.605, "de": 6.47, "vi": 5.833, "tr": 6.735, "fi": 6.695, "ru": 12.04, "el": 11.442, "hi": 14.308, "ta": 17.984, "th": 16.714, "zh": 5.571, "my": 18.267}, "bpb_mean": 1.57844, "t": 1003.7}
{"event": "train", "step": 5000, "phase": 2, "objective": "parity_a", "lm_loss": 1.21751, "bpb": 1.7565, "chunk_loss": 0.030091, "F": 0.13478, "G": 0.15961, "chunks": 138.02, "bytes_per_chunk": 7.419, "boundary_entropy": 0.12486, "router_noise": 0.0, "grad_norm": 0.4162, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.25, "n_target_mean": 9.2185, "alpha": 0.03, "t": 1004.0}
{"event": "train", "step": 5100, "phase": 2, "objective": "parity_a", "lm_loss": 1.22604, "bpb": 1.7688, "chunk_loss": 0.030096, "F": 0.13902, "G": 0.16316, "chunks": 142.36, "bytes_per_chunk": 7.193, "boundary_entropy": 0.12243, "router_noise": 0.0, "grad_norm": 0.4145, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.0005, "alpha": 0.03, "t": 1021.4}
{"event": "train", "step": 5200, "phase": 2, "objective": "parity_a", "lm_loss": 1.15257, "bpb": 1.6628, "chunk_loss": 0.030067, "F": 0.13115, "G": 0.16007, "chunks": 134.3, "bytes_per_chunk": 7.625, "boundary_entropy": 0.13791, "router_noise": 0.0, "grad_norm": 0.4334, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.8194, "alpha": 0.03, "t": 1038.3}
{"event": "train", "step": 5300, "phase": 2, "objective": "parity_a", "lm_loss": 1.18188, "bpb": 1.70509, "chunk_loss": 0.030055, "F": 0.13326, "G": 0.15806, "chunks": 136.45, "bytes_per_chunk": 7.504, "boundary_entropy": 0.1188, "router_noise": 0.0, "grad_norm": 0.4697, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.3411, "alpha": 0.03, "t": 1055.3}
{"event": "train", "step": 5400, "phase": 2, "objective": "parity_a", "lm_loss": 1.23492, "bpb": 1.78161, "chunk_loss": 0.030164, "F": 0.14148, "G": 0.16376, "chunks": 144.88, "bytes_per_chunk": 7.068, "boundary_entropy": 0.11624, "router_noise": 0.0, "grad_norm": 0.4187, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.0989, "alpha": 0.03, "t": 1072.6}
{"event": "train", "step": 5500, "phase": 2, "objective": "parity_a", "lm_loss": 1.18864, "bpb": 1.71485, "chunk_loss": 0.030071, "F": 0.13483, "G": 0.16076, "chunks": 138.06, "bytes_per_chunk": 7.417, "boundary_entropy": 0.1236, "router_noise": 0.0, "grad_norm": 0.4537, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.3685, "alpha": 0.03, "t": 1089.8}
{"event": "train", "step": 5600, "phase": 2, "objective": "parity_a", "lm_loss": 1.14498, "bpb": 1.65186, "chunk_loss": 0.029992, "F": 0.1335, "G": 0.15417, "chunks": 136.7, "bytes_per_chunk": 7.491, "boundary_entropy": 0.11177, "router_noise": 0.0, "grad_norm": 0.4105, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.17, "alpha": 0.03, "t": 1106.9}
{"event": "train", "step": 5700, "phase": 2, "objective": "parity_a", "lm_loss": 1.13204, "bpb": 1.63318, "chunk_loss": 0.030053, "F": 0.13179, "G": 0.16046, "chunks": 134.95, "bytes_per_chunk": 7.588, "boundary_entropy": 0.12712, "router_noise": 0.0, "grad_norm": 0.4045, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 10.0188, "alpha": 0.03, "t": 1124.0}
{"event": "train", "step": 5800, "phase": 2, "objective": "parity_a", "lm_loss": 1.11185, "bpb": 1.60405, "chunk_loss": 0.02999, "F": 0.1319, "G": 0.15105, "chunks": 135.06, "bytes_per_chunk": 7.582, "boundary_entropy": 0.1099, "router_noise": 0.0, "grad_norm": 0.4008, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.5687, "alpha": 0.03, "t": 1141.1}
{"event": "train", "step": 5900, "phase": 2, "objective": "parity_a", "lm_loss": 1.16888, "bpb": 1.68634, "chunk_loss": 0.030125, "F": 0.1433, "G": 0.16066, "chunks": 146.73, "bytes_per_chunk": 6.979, "boundary_entropy": 0.10023, "router_noise": 0.0, "grad_norm": 0.4275, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.6873, "alpha": 0.03, "t": 1158.1}
{"event": "probe", "step": 6000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.169895, "boundary_entropy": 0.044439, "bpb": 1.78002}, "de": {"chunks": 3176.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154678, "cps": 24.8125, "mean_p": 0.158906, "boundary_entropy": 0.047836, "bpb": 1.942354}, "vi": {"chunks": 4100.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222403, "cps": 32.03125, "mean_p": 0.149673, "boundary_entropy": 0.089291, "bpb": 1.60905}, "tr": {"chunks": 2899.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.160139, "cps": 22.648438, "mean_p": 0.158628, "boundary_entropy": 0.084102, "bpb": 1.905798}, "fi": {"chunks": 2767.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.147095, "cps": 21.617188, "mean_p": 0.151913, "boundary_entropy": 0.083497, "bpb": 2.068299}, "ru": {"chunks": 2761.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.145699, "cps": 21.570312, "mean_p": 0.110828, "boundary_entropy": 0.12792, "bpb": 1.182864}, "el": {"chunks": 3146.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153291, "cps": 24.578125, "mean_p": 0.087928, "boundary_entropy": 0.081312, "bpb": 1.202921}, "hi": {"chunks": 2623.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.155899, "cps": 20.492188, "mean_p": 0.066865, "boundary_entropy": 0.140617, "bpb": 1.023088}, "ta": {"chunks": 2762.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.14774, "cps": 21.578125, "mean_p": 0.146938, "boundary_entropy": 0.288, "bpb": 0.868962}, "th": {"chunks": 2671.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.166055, "cps": 20.867188, "mean_p": 0.177059, "boundary_entropy": 0.314023, "bpb": 0.979335}, "zh": {"chunks": 2986.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.505588, "cps": 23.328125, "mean_p": 0.226516, "boundary_entropy": 0.193033, "bpb": 2.746893}, "my": {"chunks": 1924.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.101541, "cps": 15.03125, "mean_p": 0.158329, "boundary_entropy": 0.314768, "bpb": 0.983554}}, "bpb": {"en": 1.78002, "de": 1.94235, "vi": 1.60905, "tr": 1.9058, "fi": 2.0683, "ru": 1.18286, "el": 1.20292, "hi": 1.02309, "ta": 0.86896, "th": 0.97934, "zh": 2.74689, "my": 0.98355}, "gini_cpc": 0.20142, "gini_cps": 0.081393, "bpic": 11.058696, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9317612926009895, "vi": 1.3397300329157396, "tr": 0.9646599368595155, "fi": 0.8860818770396433, "ru": 0.8776750863753059, "el": 0.9234097407821804, "hi": 0.9391172027730332, "ta": 0.8899688029156901, "th": 1.0002980014984002, "zh": 3.045600528037207, "my": 0.6116715343364175}, "t": 1177.8}
{"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.78158, "de": 1.94145, "vi": 1.62199, "tr": 1.90133, "fi": 2.08555, "ru": 1.18298, "el": 1.19904, "hi": 1.02927, "ta": 0.87158, "th": 0.97757, "zh": 2.75623, "my": 0.99209}, "bytes_per_chunk": {"en": 5.609, "de": 6.442, "vi": 5.858, "tr": 6.562, "fi": 6.707, "ru": 12.11, "el": 11.458, "hi": 16.613, "ta": 17.642, "th": 15.425, "zh": 5.544, "my": 25.983}, "bpb_mean": 1.52839, "t": 1182.6}
{"event": "train", "step": 6000, "phase": 2, "objective": "parity_a", "lm_loss": 1.15185, "bpb": 1.66176, "chunk_loss": 0.030137, "F": 0.14055, "G": 0.16336, "chunks": 143.92, "bytes_per_chunk": 7.115, "boundary_entropy": 0.11235, "router_noise": 0.0, "grad_norm": 0.4455, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.247, "n_target_mean": 9.0576, "alpha": 0.03, "t": 1182.7}
{"event": "train", "step": 6100, "phase": 2, "objective": "parity_a", "lm_loss": 1.08198, "bpb": 1.56097, "chunk_loss": 0.029976, "F": 0.13422, "G": 0.16537, "chunks": 137.44, "bytes_per_chunk": 7.451, "boundary_entropy": 0.1249, "router_noise": 0.0, "grad_norm": 0.41, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.3786, "alpha": 0.03, "t": 1200.0}
{"event": "train", "step": 6200, "phase": 2, "objective": "parity_a", "lm_loss": 1.19614, "bpb": 1.72567, "chunk_loss": 0.030041, "F": 0.14546, "G": 0.16838, "chunks": 148.95, "bytes_per_chunk": 6.875, "boundary_entropy": 0.10616, "router_noise": 0.0, "grad_norm": 0.4283, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.4082, "alpha": 0.03, "t": 1216.9}
{"event": "train", "step": 6300, "phase": 2, "objective": "parity_a", "lm_loss": 1.11533, "bpb": 1.60909, "chunk_loss": 0.030141, "F": 0.14449, "G": 0.15964, "chunks": 147.95, "bytes_per_chunk": 6.921, "boundary_entropy": 0.09138, "router_noise": 0.0, "grad_norm": 0.4328, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.5209, "alpha": 0.03, "t": 1233.9}
{"event": "train", "step": 6400, "phase": 2, "objective": "parity_a", "lm_loss": 1.02392, "bpb": 1.47721, "chunk_loss": 0.030258, "F": 0.12686, "G": 0.15777, "chunks": 129.91, "bytes_per_chunk": 7.883, "boundary_entropy": 0.1325, "router_noise": 0.0, "grad_norm": 0.4127, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 10.5328, "alpha": 0.03, "t": 1250.9}
{"event": "train", "step": 6500, "phase": 2, "objective": "parity_a", "lm_loss": 1.0882, "bpb": 1.56994, "chunk_loss": 0.030193, "F": 0.13487, "G": 0.15536, "chunks": 138.11, "bytes_per_chunk": 7.414, "boundary_entropy": 0.1113, "router_noise": 0.0, "grad_norm": 0.4297, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.6646, "alpha": 0.03, "t": 1268.5}
{"event": "train", "step": 6600, "phase": 2, "objective": "parity_a", "lm_loss": 1.13923, "bpb": 1.64357, "chunk_loss": 0.030068, "F": 0.14612, "G": 0.16689, "chunks": 149.62, "bytes_per_chunk": 6.844, "boundary_entropy": 0.10359, "router_noise": 0.0, "grad_norm": 0.4101, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.167, "n_target_mean": 8.5192, "alpha": 0.03, "t": 1285.2}
{"event": "train", "step": 6700, "phase": 2, "objective": "parity_a", "lm_loss": 1.12556, "bpb": 1.62385, "chunk_loss": 0.030046, "F": 0.14226, "G": 0.16657, "chunks": 145.67, "bytes_per_chunk": 7.029, "boundary_entropy": 0.1094, "router_noise": 0.0, "grad_norm": 0.4157, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 8.8379, "alpha": 0.03, "t": 1302.0}
{"event": "train", "step": 6800, "phase": 2, "objective": "parity_a", "lm_loss": 1.13798, "bpb": 1.64176, "chunk_loss": 0.030096, "F": 0.13182, "G": 0.16072, "chunks": 134.98, "bytes_per_chunk": 7.586, "boundary_entropy": 0.12486, "router_noise": 0.0, "grad_norm": 0.4415, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.7731, "alpha": 0.03, "t": 1319.8}
{"event": "train", "step": 6900, "phase": 2, "objective": "parity_a", "lm_loss": 1.11236, "bpb": 1.6048, "chunk_loss": 0.030052, "F": 0.13283, "G": 0.15599, "chunks": 136.02, "bytes_per_chunk": 7.529, "boundary_entropy": 0.10518, "router_noise": 0.0, "grad_norm": 0.4057, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.167, "n_target_mean": 9.4099, "alpha": 0.03, "t": 1336.5}
{"event": "probe", "step": 7000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.169359, "boundary_entropy": 0.037293, "bpb": 1.714323}, "de": {"chunks": 3173.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154532, "cps": 24.789062, "mean_p": 0.157483, "boundary_entropy": 0.041038, "bpb": 1.874112}, "vi": {"chunks": 4110.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222945, "cps": 32.109375, "mean_p": 0.152626, "boundary_entropy": 0.08072, "bpb": 1.550853}, "tr": {"chunks": 2800.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.15467, "cps": 21.875, "mean_p": 0.153955, "boundary_entropy": 0.076767, "bpb": 1.827172}, "fi": {"chunks": 2746.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.145978, "cps": 21.453125, "mean_p": 0.148833, "boundary_entropy": 0.073615, "bpb": 1.988812}, "ru": {"chunks": 2790.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.14723, "cps": 21.796875, "mean_p": 0.110478, "boundary_entropy": 0.124719, "bpb": 1.157402}, "el": {"chunks": 3156.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153779, "cps": 24.65625, "mean_p": 0.087132, "boundary_entropy": 0.074265, "bpb": 1.166439}, "hi": {"chunks": 2917.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.173373, "cps": 22.789062, "mean_p": 0.066077, "boundary_entropy": 0.12894, "bpb": 0.992859}, "ta": {"chunks": 2367.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.126611, "cps": 18.492188, "mean_p": 0.132705, "boundary_entropy": 0.268981, "bpb": 0.84605}, "th": {"chunks": 1943.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.120796, "cps": 15.179688, "mean_p": 0.171886, "boundary_entropy": 0.307926, "bpb": 0.952349}, "zh": {"chunks": 2969.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.502709, "cps": 23.195312, "mean_p": 0.221207, "boundary_entropy": 0.175921, "bpb": 2.704889}, "my": {"chunks": 2433.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.128404, "cps": 19.007812, "mean_p": 0.178205, "boundary_entropy": 0.34341, "bpb": 0.952932}}, "bpb": {"en": 1.71432, "de": 1.87411, "vi": 1.55085, "tr": 1.82717, "fi": 1.98881, "ru": 1.1574, "el": 1.16644, "hi": 0.99286, "ta": 0.84605, "th": 0.95235, "zh": 2.70489, "my": 0.95293}, "gini_cpc": 0.208844, "gini_cps": 0.091702, "bpic": 11.191234, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9308811654354344, "vi": 1.3429976671423633, "tr": 0.9317170828584489, "fi": 0.8793570055478355, "ru": 0.886893694671171, "el": 0.9263449274979534, "hi": 1.0443785285889964, "ta": 0.7626923086536707, "th": 0.7276596843546953, "zh": 3.028261208219179, "my": 0.7734910826613844}, "t": 1355.9}
{"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.71505, "de": 1.8752, "vi": 1.56474, "tr": 1.82134, "fi": 2.00534, "ru": 1.15859, "el": 1.16398, "hi": 0.99916, "ta": 0.85053, "th": 0.95138, "zh": 2.71441, "my": 0.96088}, "bytes_per_chunk": {"en": 5.608, "de": 6.452, "vi": 5.842, "tr": 6.719, "fi": 6.744, "ru": 11.969, "el": 11.414, "hi": 15.349, "ta": 20.62, "th": 20.977, "zh": 5.5, "my": 21.585}, "bpb_mean": 1.48172, "t": 1361.4}
{"event": "train", "step": 7000, "phase": 2, "objective": "parity_a", "lm_loss": 1.10767, "bpb": 1.59803, "chunk_loss": 0.030062, "F": 0.13156, "G": 0.15622, "chunks": 134.72, "bytes_per_chunk": 7.601, "boundary_entropy": 0.10738, "router_noise": 0.0, "grad_norm": 0.43, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.251, "n_target_mean": 9.5778, "alpha": 0.03, "t": 1361.6}
{"event": "train", "step": 7100, "phase": 2, "objective": "parity_a", "lm_loss": 1.08255, "bpb": 1.56179, "chunk_loss": 0.030208, "F": 0.13712, "G": 0.16035, "chunks": 140.41, "bytes_per_chunk": 7.293, "boundary_entropy": 0.11, "router_noise": 0.0, "grad_norm": 0.4209, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.4019, "alpha": 0.03, "t": 1378.5}
{"event": "train", "step": 7200, "phase": 2, "objective": "parity_a", "lm_loss": 1.1357, "bpb": 1.63847, "chunk_loss": 0.030107, "F": 0.14175, "G": 0.1584, "chunks": 145.16, "bytes_per_chunk": 7.054, "boundary_entropy": 0.09655, "router_noise": 0.0, "grad_norm": 0.411, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 8.9796, "alpha": 0.03, "t": 1395.3}
{"event": "train", "step": 7300, "phase": 2, "objective": "parity_a", "lm_loss": 1.14302, "bpb": 1.64904, "chunk_loss": 0.03017, "F": 0.14024, "G": 0.1628, "chunks": 143.61, "bytes_per_chunk": 7.13, "boundary_entropy": 0.10918, "router_noise": 0.0, "grad_norm": 0.4276, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.1966, "alpha": 0.03, "t": 1412.3}
{"event": "train", "step": 7400, "phase": 2, "objective": "parity_a", "lm_loss": 1.11728, "bpb": 1.61189, "chunk_loss": 0.030105, "F": 0.13615, "G": 0.16008, "chunks": 139.42, "bytes_per_chunk": 7.345, "boundary_entropy": 0.11471, "router_noise": 0.0, "grad_norm": 0.4269, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.204, "alpha": 0.03, "t": 1429.2}
{"event": "train", "step": 7500, "phase": 2, "objective": "parity_a", "lm_loss": 1.09729, "bpb": 1.58306, "chunk_loss": 0.029954, "F": 0.12875, "G": 0.1543, "chunks": 131.84, "bytes_per_chunk": 7.767, "boundary_entropy": 0.1085, "router_noise": 0.0, "grad_norm": 0.423, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.4226, "alpha": 0.03, "t": 1446.8}
{"event": "train", "step": 7600, "phase": 2, "objective": "parity_a", "lm_loss": 1.14923, "bpb": 1.65799, "chunk_loss": 0.029995, "F": 0.14235, "G": 0.16464, "chunks": 145.77, "bytes_per_chunk": 7.025, "boundary_entropy": 0.09634, "router_noise": 0.0, "grad_norm": 0.4337, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.167, "n_target_mean": 8.7726, "alpha": 0.03, "t": 1463.5}
{"event": "phase_transition", "step": 7700, "phase": 3, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 1480.4}
{"event": "train", "step": 7700, "phase": 3, "objective": "parity_a", "lm_loss": 1.10272, "bpb": 1.59089, "chunk_loss": 0.030113, "F": 0.14188, "G": 0.1594, "chunks": 145.28, "bytes_per_chunk": 7.048, "boundary_entropy": 0.0933, "router_noise": 0.0, "grad_norm": 0.4251, "lr": 9.06e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.7942, "alpha": 0.03, "t": 1480.6}
{"event": "train", "step": 7800, "phase": 3, "objective": "parity_a", "lm_loss": 1.13697, "bpb": 1.6403, "chunk_loss": 0.030065, "F": 0.14343, "G": 0.15847, "chunks": 146.88, "bytes_per_chunk": 6.972, "boundary_entropy": 0.08469, "router_noise": 0.0, "grad_norm": 0.4471, "lr": 8.73e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 8.5505, "alpha": 0.03, "t": 1497.5}
{"event": "train", "step": 7900, "phase": 3, "objective": "parity_a", "lm_loss": 1.1236, "bpb": 1.62101, "chunk_loss": 0.030113, "F": 0.14265, "G": 0.1624, "chunks": 146.08, "bytes_per_chunk": 7.01, "boundary_entropy": 0.09288, "router_noise": 0.0, "grad_norm": 0.4284, "lr": 8.4e-05, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.3601, "alpha": 0.03, "t": 1514.9}
{"event": "probe", "step": 8000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.168972, "boundary_entropy": 0.03457, "bpb": 1.676276}, "de": {"chunks": 3173.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154532, "cps": 24.789062, "mean_p": 0.156717, "boundary_entropy": 0.03765, "bpb": 1.827418}, "vi": {"chunks": 4114.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223162, "cps": 32.140625, "mean_p": 0.153097, "boundary_entropy": 0.077868, "bpb": 1.50137}, "tr": {"chunks": 2769.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.152958, "cps": 21.632812, "mean_p": 0.151379, "boundary_entropy": 0.074243, "bpb": 1.775812}, "fi": {"chunks": 2737.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.1455, "cps": 21.382812, "mean_p": 0.147726, "boundary_entropy": 0.071007, "bpb": 1.940808}, "ru": {"chunks": 2793.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.147388, "cps": 21.820312, "mean_p": 0.110699, "boundary_entropy": 0.123584, "bpb": 1.134036}, "el": {"chunks": 3148.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153389, "cps": 24.59375, "mean_p": 0.089121, "boundary_entropy": 0.068067, "bpb": 1.143216}, "hi": {"chunks": 2953.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.175513, "cps": 23.070312, "mean_p": 0.067612, "boundary_entropy": 0.131828, "bpb": 0.977183}, "ta": {"chunks": 2696.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.14421, "cps": 21.0625, "mean_p": 0.14984, "boundary_entropy": 0.284598, "bpb": 0.830048}, "th": {"chunks": 2752.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.171091, "cps": 21.5, "mean_p": 0.173634, "boundary_entropy": 0.306753, "bpb": 0.937201}, "zh": {"chunks": 2975.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503725, "cps": 23.242188, "mean_p": 0.22242, "boundary_entropy": 0.170181, "bpb": 2.668321}, "my": {"chunks": 2872.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.151573, "cps": 22.4375, "mean_p": 0.190359, "boundary_entropy": 0.353694, "bpb": 0.929673}}, "bpb": {"en": 1.67628, "de": 1.82742, "vi": 1.50137, "tr": 1.77581, "fi": 1.94081, "ru": 1.13404, "el": 1.14322, "hi": 0.97718, "ta": 0.83005, "th": 0.9372, "zh": 2.66832, "my": 0.92967}, "gini_cpc": 0.181603, "gini_cps": 0.05534, "bpic": 10.698921, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9308811654354344, "vi": 1.344304720833013, "tr": 0.9214016437268018, "fi": 0.876474917765632, "ru": 0.8878473438052261, "el": 0.9239967781253351, "hi": 1.057267670525645, "ta": 0.8687023507098843, "th": 1.030632759312466, "zh": 3.0343809681549536, "my": 0.9130564691341949}, "t": 1534.5}
{"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.67721, "de": 1.83127, "vi": 1.51672, "tr": 1.76989, "fi": 1.95485, "ru": 1.13466, "el": 1.14172, "hi": 0.98344, "ta": 0.83429, "th": 0.93688, "zh": 2.67893, "my": 0.93866}, "bytes_per_chunk": {"en": 5.609, "de": 6.456, "vi": 5.84, "tr": 6.836, "fi": 6.77, "ru": 11.98, "el": 11.437, "hi": 15.162, "ta": 17.718, "th": 15.011, "zh": 5.548, "my": 18.439}, "bpb_mean": 1.44988, "t": 1539.3}
{"event": "train", "step": 8000, "phase": 3, "objective": "parity_a", "lm_loss": 1.11693, "bpb": 1.61138, "chunk_loss": 0.030152, "F": 0.14316, "G": 0.16631, "chunks": 146.59, "bytes_per_chunk": 6.985, "boundary_entropy": 0.10188, "router_noise": 0.0, "grad_norm": 0.4229, "lr": 8.08e-05, "router_trainable": true, "sec_per_step": 0.246, "n_target_mean": 8.7462, "alpha": 0.03, "t": 1539.5}
{"event": "train", "step": 8100, "phase": 3, "objective": "parity_a", "lm_loss": 1.1491, "bpb": 1.6578, "chunk_loss": 0.030092, "F": 0.14345, "G": 0.16415, "chunks": 146.89, "bytes_per_chunk": 6.971, "boundary_entropy": 0.10154, "router_noise": 0.0, "grad_norm": 0.4473, "lr": 7.77e-05, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 8.6065, "alpha": 0.03, "t": 1557.1}
{"event": "train", "step": 8200, "phase": 3, "objective": "parity_a", "lm_loss": 1.09611, "bpb": 1.58135, "chunk_loss": 0.030078, "F": 0.13992, "G": 0.15856, "chunks": 143.28, "bytes_per_chunk": 7.147, "boundary_entropy": 0.09453, "router_noise": 0.0, "grad_norm": 0.4438, "lr": 7.47e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.8145, "alpha": 0.03, "t": 1574.1}
{"event": "train", "step": 8300, "phase": 3, "objective": "parity_a", "lm_loss": 1.02948, "bpb": 1.48523, "chunk_loss": 0.029944, "F": 0.13126, "G": 0.15852, "chunks": 134.41, "bytes_per_chunk": 7.619, "boundary_entropy": 0.11078, "router_noise": 0.0, "grad_norm": 0.4146, "lr": 7.17e-05, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 9.4833, "alpha": 0.03, "t": 1590.9}
{"event": "train", "step": 8400, "phase": 3, "objective": "parity_a", "lm_loss": 1.08494, "bpb": 1.56524, "chunk_loss": 0.030065, "F": 0.13596, "G": 0.15967, "chunks": 139.22, "bytes_per_chunk": 7.355, "boundary_entropy": 0.10431, "router_noise": 0.0, "grad_norm": 0.4394, "lr": 6.88e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.0726, "alpha": 0.03, "t": 1607.8}
{"event": "train", "step": 8500, "phase": 3, "objective": "parity_a", "lm_loss": 1.09109, "bpb": 1.57412, "chunk_loss": 0.03014, "F": 0.14012, "G": 0.15843, "chunks": 143.48, "bytes_per_chunk": 7.137, "boundary_entropy": 0.10008, "router_noise": 0.0, "grad_norm": 0.4291, "lr": 6.6e-05, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 9.0161, "alpha": 0.03, "t": 1625.5}
{"event": "train", "step": 8600, "phase": 3, "objective": "parity_a", "lm_loss": 1.14712, "bpb": 1.65495, "chunk_loss": 0.030078, "F": 0.14061, "G": 0.16078, "chunks": 143.98, "bytes_per_chunk": 7.112, "boundary_entropy": 0.09976, "router_noise": 0.0, "grad_norm": 0.4557, "lr": 6.33e-05, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 8.8052, "alpha": 0.03, "t": 1642.3}
{"event": "train", "step": 8700, "phase": 3, "objective": "parity_a", "lm_loss": 1.0188, "bpb": 1.46982, "chunk_loss": 0.030135, "F": 0.13177, "G": 0.15946, "chunks": 134.94, "bytes_per_chunk": 7.589, "boundary_entropy": 0.11243, "router_noise": 0.0, "grad_norm": 0.4323, "lr": 6.07e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.6548, "alpha": 0.03, "t": 1659.2}
{"event": "train", "step": 8800, "phase": 3, "objective": "parity_a", "lm_loss": 1.04239, "bpb": 1.50385, "chunk_loss": 0.030096, "F": 0.13246, "G": 0.15626, "chunks": 135.64, "bytes_per_chunk": 7.549, "boundary_entropy": 0.10462, "router_noise": 0.0, "grad_norm": 0.4263, "lr": 5.82e-05, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.4642, "alpha": 0.03, "t": 1676.8}
{"event": "train", "step": 8900, "phase": 3, "objective": "parity_a", "lm_loss": 1.09379, "bpb": 1.578, "chunk_loss": 0.030056, "F": 0.14597, "G": 0.16114, "chunks": 149.47, "bytes_per_chunk": 6.851, "boundary_entropy": 0.08346, "router_noise": 0.0, "grad_norm": 0.4171, "lr": 5.58e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.2885, "alpha": 0.03, "t": 1693.8}
{"event": "probe", "step": 9000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.169078, "boundary_entropy": 0.034196, "bpb": 1.652154}, "de": {"chunks": 3175.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154629, "cps": 24.804688, "mean_p": 0.156801, "boundary_entropy": 0.037399, "bpb": 1.799188}, "vi": {"chunks": 4118.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223379, "cps": 32.171875, "mean_p": 0.153012, "boundary_entropy": 0.076531, "bpb": 1.466705}, "tr": {"chunks": 2757.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.152295, "cps": 21.539062, "mean_p": 0.151727, "boundary_entropy": 0.074801, "bpb": 1.742603}, "fi": {"chunks": 2743.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.145819, "cps": 21.429688, "mean_p": 0.147464, "boundary_entropy": 0.067118, "bpb": 1.903464}, "ru": {"chunks": 2782.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.146807, "cps": 21.734375, "mean_p": 0.10759, "boundary_entropy": 0.11473, "bpb": 1.118051}, "el": {"chunks": 3157.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153827, "cps": 24.664062, "mean_p": 0.088903, "boundary_entropy": 0.059925, "bpb": 1.119491}, "hi": {"chunks": 3265.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.194056, "cps": 25.507812, "mean_p": 0.070297, "boundary_entropy": 0.131443, "bpb": 0.959812}, "ta": {"chunks": 2405.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.128644, "cps": 18.789062, "mean_p": 0.143836, "boundary_entropy": 0.278189, "bpb": 0.816417}, "th": {"chunks": 2564.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.159403, "cps": 20.03125, "mean_p": 0.169125, "boundary_entropy": 0.296343, "bpb": 0.919891}, "zh": {"chunks": 2956.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.500508, "cps": 23.09375, "mean_p": 0.217733, "boundary_entropy": 0.157152, "bpb": 2.634349}, "my": {"chunks": 2173.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.114682, "cps": 16.976562, "mean_p": 0.174364, "boundary_entropy": 0.338217, "bpb": 0.9094}}, "bpb": {"en": 1.65215, "de": 1.79919, "vi": 1.4667, "tr": 1.7426, "fi": 1.90346, "ru": 1.11805, "el": 1.11949, "hi": 0.95981, "ta": 0.81642, "th": 0.91989, "zh": 2.63435, "my": 0.9094}, "gini_cpc": 0.20656, "gini_cps": 0.086197, "bpic": 10.970186, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9314679168791377, "vi": 1.3456117745236624, "tr": 0.9174085705145512, "fi": 0.8783963096204344, "ru": 0.8843506303136911, "el": 0.9266384461695307, "hi": 1.1689735673099326, "ta": 0.7749366296206498, "th": 0.9602261609292017, "zh": 3.015001728358334, "my": 0.6908327672105172}, "t": 1713.5}
{"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.65249, "de": 1.80293, "vi": 1.48243, "tr": 1.73693, "fi": 1.91882, "ru": 1.1188, "el": 1.11831, "hi": 0.96513, "ta": 0.82104, "th": 0.9198, "zh": 2.64612, "my": 0.91889}, "bytes_per_chunk": {"en": 5.609, "de": 6.457, "vi": 5.828, "tr": 6.83, "fi": 6.768, "ru": 12.059, "el": 11.414, "hi": 13.06, "ta": 20.125, "th": 15.985, "zh": 5.546, "my": 23.685}, "bpb_mean": 1.42514, "t": 1718.9}
{"event": "train", "step": 9000, "phase": 3, "objective": "parity_a", "lm_loss": 1.08758, "bpb": 1.56904, "chunk_loss": 0.030087, "F": 0.13383, "G": 0.15477, "chunks": 137.05, "bytes_per_chunk": 7.472, "boundary_entropy": 0.10275, "router_noise": 0.0, "grad_norm": 0.4397, "lr": 5.35e-05, "router_trainable": true, "sec_per_step": 0.253, "n_target_mean": 9.2376, "alpha": 0.03, "t": 1719.1}
{"event": "train", "step": 9100, "phase": 3, "objective": "parity_a", "lm_loss": 1.07375, "bpb": 1.5491, "chunk_loss": 0.030079, "F": 0.1384, "G": 0.16032, "chunks": 141.72, "bytes_per_chunk": 7.226, "boundary_entropy": 0.10308, "router_noise": 0.0, "grad_norm": 0.4413, "lr": 5.12e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.2403, "alpha": 0.03, "t": 1736.0}
{"event": "train", "step": 9200, "phase": 3, "objective": "parity_a", "lm_loss": 0.99264, "bpb": 1.43208, "chunk_loss": 0.030141, "F": 0.12897, "G": 0.14564, "chunks": 132.06, "bytes_per_chunk": 7.754, "boundary_entropy": 0.09644, "router_noise": 0.0, "grad_norm": 0.4244, "lr": 4.91e-05, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 9.9164, "alpha": 0.03, "t": 1753.7}
{"event": "train", "step": 9300, "phase": 3, "objective": "parity_a", "lm_loss": 1.06127, "bpb": 1.53108, "chunk_loss": 0.030128, "F": 0.14503, "G": 0.16323, "chunks": 148.52, "bytes_per_chunk": 6.895, "boundary_entropy": 0.0918, "router_noise": 0.0, "grad_norm": 0.4256, "lr": 4.71e-05, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 8.6352, "alpha": 0.03, "t": 1771.2}
{"event": "train", "step": 9400, "phase": 3, "objective": "parity_a", "lm_loss": 1.11203, "bpb": 1.60432, "chunk_loss": 0.030078, "F": 0.14801, "G": 0.16765, "chunks": 151.56, "bytes_per_chunk": 6.756, "boundary_entropy": 0.08537, "router_noise": 0.0, "grad_norm": 0.4329, "lr": 4.52e-05, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 8.2723, "alpha": 0.03, "t": 1788.8}
{"event": "train", "step": 9500, "phase": 3, "objective": "parity_a", "lm_loss": 1.12696, "bpb": 1.62586, "chunk_loss": 0.030108, "F": 0.14522, "G": 0.16063, "chunks": 148.7, "bytes_per_chunk": 6.886, "boundary_entropy": 0.09111, "router_noise": 0.0, "grad_norm": 0.457, "lr": 4.34e-05, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 8.3913, "alpha": 0.03, "t": 1806.7}
{"event": "train", "step": 9600, "phase": 3, "objective": "parity_a", "lm_loss": 0.98483, "bpb": 1.42081, "chunk_loss": 0.030043, "F": 0.12607, "G": 0.15218, "chunks": 129.09, "bytes_per_chunk": 7.932, "boundary_entropy": 0.10842, "router_noise": 0.0, "grad_norm": 0.4416, "lr": 4.17e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 10.0668, "alpha": 0.03, "t": 1823.9}
{"event": "train", "step": 9700, "phase": 3, "objective": "parity_a", "lm_loss": 0.99258, "bpb": 1.43199, "chunk_loss": 0.030073, "F": 0.13342, "G": 0.1543, "chunks": 136.62, "bytes_per_chunk": 7.495, "boundary_entropy": 0.09627, "router_noise": 0.0, "grad_norm": 0.438, "lr": 4.01e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.4399, "alpha": 0.03, "t": 1840.9}
{"event": "train", "step": 9800, "phase": 3, "objective": "parity_a", "lm_loss": 1.016, "bpb": 1.46577, "chunk_loss": 0.030106, "F": 0.13106, "G": 0.15732, "chunks": 134.2, "bytes_per_chunk": 7.63, "boundary_entropy": 0.11086, "router_noise": 0.0, "grad_norm": 0.4456, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 9.9304, "alpha": 0.03, "t": 1858.7}
{"event": "train", "step": 9900, "phase": 3, "objective": "parity_a", "lm_loss": 1.08137, "bpb": 1.56008, "chunk_loss": 0.030219, "F": 0.14902, "G": 0.17016, "chunks": 152.59, "bytes_per_chunk": 6.711, "boundary_entropy": 0.09452, "router_noise": 0.0, "grad_norm": 0.4285, "lr": 3.72e-05, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 8.63, "alpha": 0.03, "t": 1876.3}
{"event": "probe", "step": 10000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.168759, "boundary_entropy": 0.031087, "bpb": 1.637603}, "de": {"chunks": 3174.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15458, "cps": 24.796875, "mean_p": 0.15613, "boundary_entropy": 0.033927, "bpb": 1.781058}, "vi": {"chunks": 4124.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223705, "cps": 32.21875, "mean_p": 0.152459, "boundary_entropy": 0.075477, "bpb": 1.447223}, "tr": {"chunks": 2846.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.157212, "cps": 22.234375, "mean_p": 0.155065, "boundary_entropy": 0.072862, "bpb": 1.722588}, "fi": {"chunks": 2778.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.14768, "cps": 21.703125, "mean_p": 0.147698, "boundary_entropy": 0.061697, "bpb": 1.878492}, "ru": {"chunks": 2768.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.146069, "cps": 21.625, "mean_p": 0.105723, "boundary_entropy": 0.109617, "bpb": 1.106163}, "el": {"chunks": 3163.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.15412, "cps": 24.710938, "mean_p": 0.089362, "boundary_entropy": 0.057627, "bpb": 1.105616}, "hi": {"chunks": 3135.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.18633, "cps": 24.492188, "mean_p": 0.067643, "boundary_entropy": 0.125504, "bpb": 0.947535}, "ta": {"chunks": 2402.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.128484, "cps": 18.765625, "mean_p": 0.138252, "boundary_entropy": 0.267331, "bpb": 0.805724}, "th": {"chunks": 2276.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.141498, "cps": 17.78125, "mean_p": 0.16741, "boundary_entropy": 0.292833, "bpb": 0.912872}, "zh": {"chunks": 2977.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.504064, "cps": 23.257812, "mean_p": 0.217891, "boundary_entropy": 0.154251, "bpb": 2.61342}, "my": {"chunks": 2048.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.108085, "cps": 16.0, "mean_p": 0.168727, "boundary_entropy": 0.331073, "bpb": 0.901464}}, "bpb": {"en": 1.6376, "de": 1.78106, "vi": 1.44722, "tr": 1.72259, "fi": 1.87849, "ru": 1.10616, "el": 1.10562, "hi": 0.94754, "ta": 0.80572, "th": 0.91287, "zh": 2.61342, "my": 0.90146}, "gini_cpc": 0.212494, "gini_cps": 0.09331, "bpic": 11.098352, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.931174541157286, "vi": 1.3475723550596366, "tr": 0.9470238635054091, "fi": 0.8896044287734475, "ru": 0.8799002676881009, "el": 0.9283995581989946, "hi": 1.1224294436498128, "ta": 0.773969972702204, "th": 0.8523692442569669, "zh": 3.036420888133545, "my": 0.6510931924745234}, "t": 1895.8}
{"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.63751, "de": 1.78586, "vi": 1.46287, "tr": 1.71687, "fi": 1.89392, "ru": 1.10748, "el": 1.10408, "hi": 0.95354, "ta": 0.81085, "th": 0.91226, "zh": 2.62593, "my": 0.9109}, "bytes_per_chunk": {"en": 5.608, "de": 6.459, "vi": 5.831, "tr": 6.684, "fi": 6.705, "ru": 12.114, "el": 11.409, "hi": 13.934, "ta": 20.174, "th": 17.545, "zh": 5.556, "my": 24.835}, "bpb_mean": 1.41017, "t": 1901.2}
{"event": "train", "step": 10000, "phase": 3, "objective": "parity_a", "lm_loss": 1.06898, "bpb": 1.54221, "chunk_loss": 0.03007, "F": 0.13745, "G": 0.16325, "chunks": 140.75, "bytes_per_chunk": 7.275, "boundary_entropy": 0.10814, "router_noise": 0.0, "grad_norm": 0.4646, "lr": 3.6e-05, "router_trainable": true, "sec_per_step": 0.251, "n_target_mean": 9.4679, "alpha": 0.03, "t": 1901.4}
{"event": "train", "step": 10100, "phase": 3, "objective": "parity_a", "lm_loss": 1.0144, "bpb": 1.46347, "chunk_loss": 0.03004, "F": 0.12244, "G": 0.15813, "chunks": 125.38, "bytes_per_chunk": 8.167, "boundary_entropy": 0.13079, "router_noise": 0.0, "grad_norm": 0.4371, "lr": 3.49e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 10.5631, "alpha": 0.03, "t": 1918.5}
{"event": "train", "step": 10200, "phase": 3, "objective": "parity_a", "lm_loss": 1.10047, "bpb": 1.58765, "chunk_loss": 0.030033, "F": 0.14343, "G": 0.1614, "chunks": 146.88, "bytes_per_chunk": 6.972, "boundary_entropy": 0.08319, "router_noise": 0.0, "grad_norm": 0.4552, "lr": 3.38e-05, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 8.6149, "alpha": 0.03, "t": 1936.1}
{"event": "train", "step": 10300, "phase": 3, "objective": "parity_a", "lm_loss": 1.06851, "bpb": 1.54154, "chunk_loss": 0.030167, "F": 0.13577, "G": 0.15836, "chunks": 139.03, "bytes_per_chunk": 7.365, "boundary_entropy": 0.11015, "router_noise": 0.0, "grad_norm": 0.4796, "lr": 3.3e-05, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.3106, "alpha": 0.03, "t": 1953.9}
{"event": "train", "step": 10400, "phase": 3, "objective": "parity_a", "lm_loss": 1.02668, "bpb": 1.48119, "chunk_loss": 0.030071, "F": 0.13705, "G": 0.15203, "chunks": 140.34, "bytes_per_chunk": 7.296, "boundary_entropy": 0.08238, "router_noise": 0.0, "grad_norm": 0.4453, "lr": 3.22e-05, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 9.0706, "alpha": 0.03, "t": 1970.7}
{"event": "train", "step": 10500, "phase": 3, "objective": "parity_a", "lm_loss": 1.05503, "bpb": 1.52209, "chunk_loss": 0.030112, "F": 0.13651, "G": 0.15943, "chunks": 139.78, "bytes_per_chunk": 7.326, "boundary_entropy": 0.09704, "router_noise": 0.0, "grad_norm": 0.4377, "lr": 3.15e-05, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.367, "alpha": 0.03, "t": 1988.5}
{"event": "train", "step": 10600, "phase": 3, "objective": "parity_a", "lm_loss": 1.07379, "bpb": 1.54916, "chunk_loss": 0.030054, "F": 0.14001, "G": 0.15831, "chunks": 143.38, "bytes_per_chunk": 7.142, "boundary_entropy": 0.09285, "router_noise": 0.0, "grad_norm": 0.4382, "lr": 3.1e-05, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 8.8699, "alpha": 0.03, "t": 2006.1}
{"event": "train", "step": 10700, "phase": 3, "objective": "parity_a", "lm_loss": 1.04074, "bpb": 1.50147, "chunk_loss": 0.030087, "F": 0.13539, "G": 0.16053, "chunks": 138.64, "bytes_per_chunk": 7.386, "boundary_entropy": 0.10684, "router_noise": 0.0, "grad_norm": 0.4337, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 9.4706, "alpha": 0.03, "t": 2022.9}
{"event": "train", "step": 10800, "phase": 3, "objective": "parity_a", "lm_loss": 1.0632, "bpb": 1.53387, "chunk_loss": 0.030093, "F": 0.13715, "G": 0.15794, "chunks": 140.44, "bytes_per_chunk": 7.291, "boundary_entropy": 0.09337, "router_noise": 0.0, "grad_norm": 0.4521, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.0856, "alpha": 0.03, "t": 2040.5}
{"event": "train", "step": 10900, "phase": 3, "objective": "parity_a", "lm_loss": 1.05386, "bpb": 1.52041, "chunk_loss": 0.030046, "F": 0.12935, "G": 0.15491, "chunks": 132.45, "bytes_per_chunk": 7.731, "boundary_entropy": 0.1073, "router_noise": 0.0, "grad_norm": 0.461, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.6233, "alpha": 0.03, "t": 2057.5}
{"event": "probe", "step": 11000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.168646, "boundary_entropy": 0.031336, "bpb": 1.627837}, "de": {"chunks": 3172.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154483, "cps": 24.78125, "mean_p": 0.156239, "boundary_entropy": 0.034314, "bpb": 1.769876}, "vi": {"chunks": 4117.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223325, "cps": 32.164062, "mean_p": 0.152611, "boundary_entropy": 0.075575, "bpb": 1.430141}, "tr": {"chunks": 2853.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.157598, "cps": 22.289062, "mean_p": 0.155767, "boundary_entropy": 0.07414, "bpb": 1.709167}, "fi": {"chunks": 2757.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.146563, "cps": 21.539062, "mean_p": 0.147391, "boundary_entropy": 0.06361, "bpb": 1.865928}, "ru": {"chunks": 2798.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.147652, "cps": 21.859375, "mean_p": 0.109425, "boundary_entropy": 0.117802, "bpb": 1.09836}, "el": {"chunks": 3159.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153925, "cps": 24.679688, "mean_p": 0.0885, "boundary_entropy": 0.058342, "bpb": 1.097656}, "hi": {"chunks": 2695.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.160178, "cps": 21.054688, "mean_p": 0.065286, "boundary_entropy": 0.126959, "bpb": 0.943099}, "ta": {"chunks": 2580.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.138005, "cps": 20.15625, "mean_p": 0.142946, "boundary_entropy": 0.275502, "bpb": 0.804084}, "th": {"chunks": 2407.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.149643, "cps": 18.804688, "mean_p": 0.168255, "boundary_entropy": 0.296443, "bpb": 0.906581}, "zh": {"chunks": 2975.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503725, "cps": 23.242188, "mean_p": 0.218214, "boundary_entropy": 0.152768, "bpb": 2.602753}, "my": {"chunks": 2735.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.144342, "cps": 21.367188, "mean_p": 0.179823, "boundary_entropy": 0.34218, "bpb": 0.894355}}, "bpb": {"en": 1.62784, "de": 1.76988, "vi": 1.43014, "tr": 1.70917, "fi": 1.86593, "ru": 1.09836, "el": 1.09766, "hi": 0.9431, "ta": 0.80408, "th": 0.90658, "zh": 2.60275, "my": 0.89435}, "gini_cpc": 0.183962, "gini_cps": 0.068866, "bpic": 10.922418, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9305877897135826, "vi": 1.3452850111010002, "tr": 0.9493531562125552, "fi": 0.8828795572816396, "ru": 0.889436759028651, "el": 0.9272254835126853, "hi": 0.9648954866463303, "ta": 0.8313249498633166, "th": 0.9014291612155182, "zh": 3.0343809681549536, "my": 0.8695018952235456}, "t": 2078.0}
{"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.628, "de": 1.77462, "vi": 1.44682, "tr": 1.70345, "fi": 1.88084, "ru": 1.09919, "el": 1.0962, "hi": 0.94954, "ta": 0.80903, "th": 0.90574, "zh": 2.61463, "my": 0.90364}, "bytes_per_chunk": {"en": 5.609, "de": 6.458, "vi": 5.839, "tr": 6.656, "fi": 6.752, "ru": 11.957, "el": 11.419, "hi": 16.782, "ta": 18.922, "th": 17.403, "zh": 5.523, "my": 19.45}, "bpb_mean": 1.40098, "t": 2082.8}
{"event": "probe", "step": 11000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.168646, "boundary_entropy": 0.031336, "bpb": 1.627837}, "de": {"chunks": 3172.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154483, "cps": 24.78125, "mean_p": 0.156239, "boundary_entropy": 0.034314, "bpb": 1.769876}, "vi": {"chunks": 4117.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.223325, "cps": 32.164062, "mean_p": 0.152611, "boundary_entropy": 0.075575, "bpb": 1.430141}, "tr": {"chunks": 2853.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.157598, "cps": 22.289062, "mean_p": 0.155767, "boundary_entropy": 0.07414, "bpb": 1.709167}, "fi": {"chunks": 2757.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.146563, "cps": 21.539062, "mean_p": 0.147391, "boundary_entropy": 0.06361, "bpb": 1.865928}, "ru": {"chunks": 2798.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.147652, "cps": 21.859375, "mean_p": 0.109425, "boundary_entropy": 0.117802, "bpb": 1.09836}, "el": {"chunks": 3159.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153925, "cps": 24.679688, "mean_p": 0.0885, "boundary_entropy": 0.058342, "bpb": 1.097656}, "hi": {"chunks": 2695.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.160178, "cps": 21.054688, "mean_p": 0.065286, "boundary_entropy": 0.126959, "bpb": 0.943099}, "ta": {"chunks": 2580.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.138005, "cps": 20.15625, "mean_p": 0.142946, "boundary_entropy": 0.275502, "bpb": 0.804084}, "th": {"chunks": 2407.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.149643, "cps": 18.804688, "mean_p": 0.168255, "boundary_entropy": 0.296443, "bpb": 0.906581}, "zh": {"chunks": 2975.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503725, "cps": 23.242188, "mean_p": 0.218214, "boundary_entropy": 0.152768, "bpb": 2.602753}, "my": {"chunks": 2735.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.144342, "cps": 21.367188, "mean_p": 0.179823, "boundary_entropy": 0.34218, "bpb": 0.894355}}, "bpb": {"en": 1.62784, "de": 1.76988, "vi": 1.43014, "tr": 1.70917, "fi": 1.86593, "ru": 1.09836, "el": 1.09766, "hi": 0.9431, "ta": 0.80408, "th": 0.90658, "zh": 2.60275, "my": 0.89435}, "gini_cpc": 0.183962, "gini_cps": 0.068866, "bpic": 10.922418, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9305877897135826, "vi": 1.3452850111010002, "tr": 0.9493531562125552, "fi": 0.8828795572816396, "ru": 0.889436759028651, "el": 0.9272254835126853, "hi": 0.9648954866463303, "ta": 0.8313249498633166, "th": 0.9014291612155182, "zh": 3.0343809681549536, "my": 0.8695018952235456}, "t": 2085.7}
{"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s0/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s0/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s0/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s0/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s0/th0.70"], "t": 2088.9}
{"event": "final", "run": "spec_D_s0", "condition": "D", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 0.1660058610584382, "cps": 22.5703125, "bpb": 1.6234815696407647, "bytes_per_chunk": 5.615165470707936}, "de": {"cpc": 0.1544830273218721, "cps": 24.78125, "bpb": 1.7722732199479778, "bytes_per_chunk": 6.419433832892546}, "vi": {"cpc": 0.22332519663683212, "cps": 32.1640625, "bpb": 1.448364846684536, "bytes_per_chunk": 5.843960563788462}, "tr": {"cpc": 0.15759818814561122, "cps": 22.2890625, "bpb": 1.7031053341642146, "bytes_per_chunk": 6.663802636094411}, "fi": {"cpc": 0.1465631811174313, "cps": 21.5390625, "bpb": 1.886313518376068, "bytes_per_chunk": 6.782465831398245}, "ru": {"cpc": 0.14765171503957783, "cps": 21.859375, "bpb": 1.0986238784672824, "bytes_per_chunk": 11.91888696917341}, "el": {"cpc": 0.15392486478585002, "cps": 24.6796875, "bpb": 1.0938254872658002, "bytes_per_chunk": 11.34024766127724}, "hi": {"cpc": 0.16017830609212483, "cps": 21.0546875, "bpb": 0.9506838701237682, "bytes_per_chunk": 16.6912228200312}, "ta": {"cpc": 0.13800481412142285, "cps": 20.15625, "bpb": 0.8079301813485886, "bytes_per_chunk": 18.86708530507224}, "th": {"cpc": 0.1496425240907678, "cps": 18.8046875, "bpb": 0.9074946086877729, "bytes_per_chunk": 17.244897623550695}, "zh": {"cpc": 0.5037250253979004, "cps": 23.2421875, "bpb": 2.618067429552451, "bytes_per_chunk": 5.4947309951633105}, "my": {"cpc": 0.1443424108085286, "cps": 21.3671875, "bpb": 0.9009023739442976, "bytes_per_chunk": 19.446162976150735}}, "gini_cpc": 0.18396219951299186, "gini_cps": 0.06886615248882944, "bpic": 10.922417963969604, "seed": 0, "mean_boundary_entropy": 0.09221627563238144, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 7700, "phase": 3, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 2098.7}