| {"event": "start", "run": "pilot_Apc_s0", "condition": "A", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.016, "vi": 1.323, "tr": 1.088, "fi": 1.04, "ru": 1.83, "el": 1.821, "hi": 2.569, "ta": 2.733, "th": 2.89, "zh": 2.791, "my": 2.85}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "global", "p3_objective": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_char", "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": 2.4} |
| {"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 2.4} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6384.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.366833, "cps": 49.875, "mean_p": 0.496725, "boundary_entropy": 0.686995, "bpb": 8.160238}, "de": {"chunks": 8628.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.420202, "cps": 67.40625, "mean_p": 0.498889, "boundary_entropy": 0.687821, "bpb": 8.156329}, "vi": {"chunks": 13115.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.711418, "cps": 102.460938, "mean_p": 0.505745, "boundary_entropy": 0.688385, "bpb": 8.087291}, "tr": {"chunks": 5290.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.292217, "cps": 41.328125, "mean_p": 0.489567, "boundary_entropy": 0.687288, "bpb": 8.171434}, "fi": {"chunks": 8895.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.472862, "cps": 69.492188, "mean_p": 0.500688, "boundary_entropy": 0.687512, "bpb": 8.214758}, "ru": {"chunks": 10744.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.566966, "cps": 83.9375, "mean_p": 0.493655, "boundary_entropy": 0.689893, "bpb": 8.099397}, "el": {"chunks": 15039.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.732788, "cps": 117.492188, "mean_p": 0.497482, "boundary_entropy": 0.690234, "bpb": 8.158044}, "hi": {"chunks": 2563.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.152333, "cps": 20.023438, "mean_p": 0.484638, "boundary_entropy": 0.690256, "bpb": 8.262162}, "ta": {"chunks": 38089.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.03739, "cps": 297.570312, "mean_p": 0.512976, "boundary_entropy": 0.690378, "bpb": 8.051803}, "th": {"chunks": 34006.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.114144, "cps": 265.671875, "mean_p": 0.511916, "boundary_entropy": 0.690552, "bpb": 8.248459}, "zh": {"chunks": 11821.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.001524, "cps": 92.351562, "mean_p": 0.52004, "boundary_entropy": 0.685674, "bpb": 8.123118}, "my": {"chunks": 48201.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.543857, "cps": 376.570312, "mean_p": 0.521563, "boundary_entropy": 0.690206, "bpb": 7.913994}}, "bpb": {"en": 8.16024, "de": 8.15633, "vi": 8.08729, "tr": 8.17143, "fi": 8.21476, "ru": 8.0994, "el": 8.15804, "hi": 8.26216, "ta": 8.0518, "th": 8.24846, "zh": 8.12312, "my": 7.91399}, "gini_cpc": 0.424496, "gini_cps": 0.425905, "bpic": 1.892645, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1454838515955792, "vi": 1.939350894526797, "tr": 0.7965927738475616, "fi": 1.2890367847879758, "ru": 1.5455676790921895, "el": 1.997603802482379, "hi": 0.41526447061189375, "ta": 5.553993192060118, "th": 5.7632270176529286, "zh": 5.456221801328583, "my": 6.934639901884223}, "t": 4.3} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.63916, "bpb": 8.13559, "chunk_loss": 0.05247, "F": 0.47223, "G": 0.4995, "chunks": 483.56, "bytes_per_chunk": 2.118, "boundary_entropy": 0.69158, "router_noise": 0.0, "grad_norm": 7.8059, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 4.6} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6329.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.363673, "cps": 49.445312, "mean_p": 0.496549, "boundary_entropy": 0.686989, "bpb": 8.137961}, "de": {"chunks": 8591.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.4184, "cps": 67.117188, "mean_p": 0.498781, "boundary_entropy": 0.687819, "bpb": 8.136731}, "vi": {"chunks": 13023.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.706428, "cps": 101.742188, "mean_p": 0.505542, "boundary_entropy": 0.688387, "bpb": 8.072237}, "tr": {"chunks": 5242.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.289565, "cps": 40.953125, "mean_p": 0.489423, "boundary_entropy": 0.68728, "bpb": 8.156699}, "fi": {"chunks": 8839.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.469885, "cps": 69.054688, "mean_p": 0.500506, "boundary_entropy": 0.687509, "bpb": 8.196338}, "ru": {"chunks": 10508.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.554512, "cps": 82.09375, "mean_p": 0.493307, "boundary_entropy": 0.689882, "bpb": 8.05945}, "el": {"chunks": 15046.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.733129, "cps": 117.546875, "mean_p": 0.497467, "boundary_entropy": 0.690234, "bpb": 8.140773}, "hi": {"chunks": 2499.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.148529, "cps": 19.523438, "mean_p": 0.484438, "boundary_entropy": 0.690242, "bpb": 8.249371}, "ta": {"chunks": 37851.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.024659, "cps": 295.710938, "mean_p": 0.512695, "boundary_entropy": 0.690391, "bpb": 8.028667}, "th": {"chunks": 33863.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.105253, "cps": 264.554688, "mean_p": 0.511769, "boundary_entropy": 0.690558, "bpb": 8.229689}, "zh": {"chunks": 11820.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.001355, "cps": 92.34375, "mean_p": 0.520021, "boundary_entropy": 0.685675, "bpb": 8.120588}, "my": {"chunks": 48156.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.541482, "cps": 376.21875, "mean_p": 0.521509, "boundary_entropy": 0.69021, "bpb": 7.893437}}, "bpb": {"en": 8.13796, "de": 8.13673, "vi": 8.07224, "tr": 8.1567, "fi": 8.19634, "ru": 8.05945, "el": 8.14077, "hi": 8.24937, "ta": 8.02867, "th": 8.22969, "zh": 8.12059, "my": 7.89344}, "gini_cpc": 0.425856, "gini_cps": 0.427302, "bpic": 1.9021, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.150483346113175, "vi": 1.942481642293646, "tr": 0.7962244136833727, "fi": 1.2920528399610602, "ru": 1.5247543264222025, "el": 2.01590116971762, "hi": 0.40841361153416067, "ta": 5.567252420974174, "th": 5.7888645764352376, "zh": 5.5031716409026075, "my": 6.988372631746004}, "t": 6.0} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6249.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.359076, "cps": 48.820312, "mean_p": 0.4962, "boundary_entropy": 0.686976, "bpb": 8.08858}, "de": {"chunks": 8514.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.41465, "cps": 66.515625, "mean_p": 0.498575, "boundary_entropy": 0.687814, "bpb": 8.092687}, "vi": {"chunks": 12837.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.696338, "cps": 100.289062, "mean_p": 0.505085, "boundary_entropy": 0.68839, "bpb": 8.038675}, "tr": {"chunks": 5169.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.285533, "cps": 40.382812, "mean_p": 0.489116, "boundary_entropy": 0.687262, "bpb": 8.122182}, "fi": {"chunks": 8749.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.4651, "cps": 68.351562, "mean_p": 0.500139, "boundary_entropy": 0.687503, "bpb": 8.154257}, "ru": {"chunks": 10060.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.530871, "cps": 78.59375, "mean_p": 0.492759, "boundary_entropy": 0.689865, "bpb": 7.993319}, "el": {"chunks": 15089.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.735224, "cps": 117.882812, "mean_p": 0.497513, "boundary_entropy": 0.690235, "bpb": 8.120863}, "hi": {"chunks": 2410.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.143239, "cps": 18.828125, "mean_p": 0.484163, "boundary_entropy": 0.690223, "bpb": 8.217292}, "ta": {"chunks": 37473.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.00444, "cps": 292.757812, "mean_p": 0.512239, "boundary_entropy": 0.690413, "bpb": 7.987868}, "th": {"chunks": 33649.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.091949, "cps": 262.882812, "mean_p": 0.511607, "boundary_entropy": 0.690564, "bpb": 8.202645}, "zh": {"chunks": 11810.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.999661, "cps": 92.265625, "mean_p": 0.519977, "boundary_entropy": 0.685678, "bpb": 8.114118}, "my": {"chunks": 48090.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.537999, "cps": 375.703125, "mean_p": 0.521422, "boundary_entropy": 0.690216, "bpb": 7.860091}}, "bpb": {"en": 8.08858, "de": 8.09269, "vi": 8.03868, "tr": 8.12218, "fi": 8.15426, "ru": 7.99332, "el": 8.12086, "hi": 8.21729, "ta": 7.98787, "th": 8.20264, "zh": 8.11412, "my": 7.86009}, "gini_cpc": 0.428205, "gini_cps": 0.429748, "bpic": 1.917956, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1547682490153959, "vi": 1.9392508692359043, "tr": 0.7951875581317117, "fi": 1.2952694684086565, "ru": 1.4784354309354406, "el": 2.0475438385592653, "hi": 0.39891058943873814, "ta": 5.582215381794599, "th": 5.825922357089102, "zh": 5.568908092683926, "my": 7.068137615127983}, "t": 7.5} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5920.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.340171, "cps": 46.25, "mean_p": 0.494974, "boundary_entropy": 0.686928, "bpb": 7.918707}, "de": {"chunks": 8281.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.403302, "cps": 64.695312, "mean_p": 0.497822, "boundary_entropy": 0.687798, "bpb": 7.938997}, "vi": {"chunks": 12273.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.665745, "cps": 95.882812, "mean_p": 0.503599, "boundary_entropy": 0.688396, "bpb": 7.926941}, "tr": {"chunks": 4898.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.270563, "cps": 38.265625, "mean_p": 0.48808, "boundary_entropy": 0.687201, "bpb": 8.003561}, "fi": {"chunks": 8332.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.442932, "cps": 65.09375, "mean_p": 0.498783, "boundary_entropy": 0.687477, "bpb": 8.007465}, "ru": {"chunks": 8691.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.458628, "cps": 67.898438, "mean_p": 0.490852, "boundary_entropy": 0.689793, "bpb": 7.771116}, "el": {"chunks": 15238.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.742484, "cps": 119.046875, "mean_p": 0.497598, "boundary_entropy": 0.690236, "bpb": 8.033983}, "hi": {"chunks": 2070.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.123031, "cps": 16.171875, "mean_p": 0.482795, "boundary_entropy": 0.690125, "bpb": 8.097104}, "ta": {"chunks": 35896.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.920086, "cps": 280.4375, "mean_p": 0.510438, "boundary_entropy": 0.690491, "bpb": 7.841697}, "th": {"chunks": 32619.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.027914, "cps": 254.835938, "mean_p": 0.510754, "boundary_entropy": 0.690595, "bpb": 8.088073}, "zh": {"chunks": 11778.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.994243, "cps": 92.015625, "mean_p": 0.51976, "boundary_entropy": 0.68569, "bpb": 8.087311}, "my": {"chunks": 47878.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.52681, "cps": 374.046875, "mean_p": 0.521125, "boundary_entropy": 0.690237, "bpb": 7.783234}}, "bpb": {"en": 7.91871, "de": 7.939, "vi": 7.92694, "tr": 8.00356, "fi": 8.00747, "ru": 7.77112, "el": 8.03398, "hi": 8.0971, "ta": 7.8417, "th": 8.08807, "zh": 8.08731, "my": 7.78323}, "gini_cpc": 0.437272, "gini_cps": 0.438196, "bpic": 1.979538, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1855852592596494, "vi": 1.9570864304092541, "tr": 0.7953726312347812, "fi": 1.3020863655106918, "ru": 1.3482267792198532, "el": 2.182677329061264, "hi": 0.3616743303481788, "ta": 5.644467808273639, "th": 5.9614511694629035, "zh": 5.862468481434364, "my": 7.428053752374921}, "t": 9.1} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5082.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.292019, "cps": 39.703125, "mean_p": 0.49168, "boundary_entropy": 0.686769, "bpb": 7.437863}, "de": {"chunks": 7675.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.373789, "cps": 59.960938, "mean_p": 0.495828, "boundary_entropy": 0.68775, "bpb": 7.496012}, "vi": {"chunks": 10393.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.563765, "cps": 81.195312, "mean_p": 0.498904, "boundary_entropy": 0.688358, "bpb": 7.585987}, "tr": {"chunks": 4096.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.226261, "cps": 32.0, "mean_p": 0.485081, "boundary_entropy": 0.687, "bpb": 7.643259}, "fi": {"chunks": 7126.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.378821, "cps": 55.671875, "mean_p": 0.494861, "boundary_entropy": 0.687364, "bpb": 7.577115}, "ru": {"chunks": 5212.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.27504, "cps": 40.71875, "mean_p": 0.485324, "boundary_entropy": 0.689504, "bpb": 7.189116}, "el": {"chunks": 14862.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.724163, "cps": 116.109375, "mean_p": 0.497134, "boundary_entropy": 0.69023, "bpb": 7.711217}, "hi": {"chunks": 1134.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.0674, "cps": 8.859375, "mean_p": 0.477101, "boundary_entropy": 0.689635, "bpb": 7.697115}, "ta": {"chunks": 28681.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.534154, "cps": 224.070312, "mean_p": 0.503571, "boundary_entropy": 0.690676, "bpb": 7.366251}, "th": {"chunks": 27401.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.703513, "cps": 214.070312, "mean_p": 0.50624, "boundary_entropy": 0.690718, "bpb": 7.547651}, "zh": {"chunks": 11598.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.963766, "cps": 90.609375, "mean_p": 0.518922, "boundary_entropy": 0.685739, "bpb": 7.998384}, "my": {"chunks": 47102.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.485856, "cps": 367.984375, "mean_p": 0.52, "boundary_entropy": 0.690311, "bpb": 7.472518}}, "bpb": {"en": 7.43786, "de": 7.49601, "vi": 7.58599, "tr": 7.64326, "fi": 7.57711, "ru": 7.18912, "el": 7.71122, "hi": 7.69712, "ta": 7.36625, "th": 7.54765, "zh": 7.99838, "my": 7.47252}, "gini_cpc": 0.469742, "gini_cps": 0.463489, "bpic": 2.252738, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2800161125141312, "vi": 1.9305775197261255, "tr": 0.7748164918297442, "fi": 1.2972491477863093, "ru": 0.941856311115126, "el": 2.4798526213656906, "hi": 0.23080618422506846, "ta": 5.2536154380278965, "th": 5.833575283901674, "zh": 6.72479610713456, "my": 8.512662817444324}, "t": 10.7} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3939.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.22634, "cps": 30.773438, "mean_p": 0.488095, "boundary_entropy": 0.686662, "bpb": 6.812586}, "de": {"chunks": 6923.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.337165, "cps": 54.085938, "mean_p": 0.493886, "boundary_entropy": 0.687798, "bpb": 6.904777}, "vi": {"chunks": 7190.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.390019, "cps": 56.171875, "mean_p": 0.490785, "boundary_entropy": 0.688162, "bpb": 7.158099}, "tr": {"chunks": 3499.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.193283, "cps": 27.335938, "mean_p": 0.483496, "boundary_entropy": 0.686946, "bpb": 7.13095}, "fi": {"chunks": 5529.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.293924, "cps": 43.195312, "mean_p": 0.490063, "boundary_entropy": 0.687262, "bpb": 7.002157}, "ru": {"chunks": 2623.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.138417, "cps": 20.492188, "mean_p": 0.478868, "boundary_entropy": 0.689014, "bpb": 6.360271}, "el": {"chunks": 5605.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.273108, "cps": 43.789062, "mean_p": 0.487, "boundary_entropy": 0.68987, "bpb": 7.04147}, "hi": {"chunks": 405.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.024071, "cps": 3.164062, "mean_p": 0.467004, "boundary_entropy": 0.688448, "bpb": 7.053019}, "ta": {"chunks": 14613.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.781653, "cps": 114.164062, "mean_p": 0.490182, "boundary_entropy": 0.690532, "bpb": 6.622489}, "th": {"chunks": 17461.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.085546, "cps": 136.414062, "mean_p": 0.496577, "boundary_entropy": 0.690724, "bpb": 6.370152}, "zh": {"chunks": 10870.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.840501, "cps": 84.921875, "mean_p": 0.515358, "boundary_entropy": 0.685939, "bpb": 7.770742}, "my": {"chunks": 42520.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.244036, "cps": 332.1875, "mean_p": 0.515701, "boundary_entropy": 0.690552, "bpb": 6.672825}}, "bpb": {"en": 6.81259, "de": 6.90478, "vi": 7.1581, "tr": 7.13095, "fi": 7.00216, "ru": 6.36027, "el": 7.04147, "hi": 7.05302, "ta": 6.62249, "th": 6.37015, "zh": 7.77074, "my": 6.67283}, "gini_cpc": 0.523275, "gini_cps": 0.492091, "bpic": 3.167111, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.4896356723924322, "vi": 1.723153187823007, "tr": 0.8539482059143753, "fi": 1.2985923672618997, "ru": 0.6115433045232674, "el": 1.2066266446958864, "hi": 0.106350145815816, "ta": 3.4534411068609074, "th": 4.7960774361073515, "zh": 8.131566927304103, "my": 9.914436126225766}, "t": 12.6} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4861.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.27932, "cps": 37.976562, "mean_p": 0.495225, "boundary_entropy": 0.687434, "bpb": 6.360952}, "de": {"chunks": 8523.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.415088, "cps": 66.585938, "mean_p": 0.498961, "boundary_entropy": 0.68833, "bpb": 6.444327}, "vi": {"chunks": 5661.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.307079, "cps": 44.226562, "mean_p": 0.489051, "boundary_entropy": 0.688362, "bpb": 6.904572}, "tr": {"chunks": 6801.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.375684, "cps": 53.132812, "mean_p": 0.496479, "boundary_entropy": 0.68798, "bpb": 6.718628}, "fi": {"chunks": 7153.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.380256, "cps": 55.882812, "mean_p": 0.497712, "boundary_entropy": 0.688012, "bpb": 6.521174}, "ru": {"chunks": 5374.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.283588, "cps": 41.984375, "mean_p": 0.488007, "boundary_entropy": 0.68983, "bpb": 5.791016}, "el": {"chunks": 474.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.023096, "cps": 3.703125, "mean_p": 0.469262, "boundary_entropy": 0.688348, "bpb": 6.573207}, "hi": {"chunks": 2698.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.160357, "cps": 21.078125, "mean_p": 0.485716, "boundary_entropy": 0.690388, "bpb": 5.948363}, "ta": {"chunks": 5527.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.295641, "cps": 43.179688, "mean_p": 0.482416, "boundary_entropy": 0.690237, "bpb": 5.810354}, "th": {"chunks": 3698.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.229904, "cps": 28.890625, "mean_p": 0.487851, "boundary_entropy": 0.690624, "bpb": 5.623417}, "zh": {"chunks": 7406.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.253979, "cps": 57.859375, "mean_p": 0.501028, "boundary_entropy": 0.686375, "bpb": 7.331329}, "my": {"chunks": 25474.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.344416, "cps": 199.015625, "mean_p": 0.49649, "boundary_entropy": 0.690915, "bpb": 5.630428}}, "bpb": {"en": 6.36095, "de": 6.44433, "vi": 6.90457, "tr": 6.71863, "fi": 6.52117, "ru": 5.79102, "el": 6.57321, "hi": 5.94836, "ta": 5.81035, "th": 5.62342, "zh": 7.33133, "my": 5.63043}, "gini_cpc": 0.412363, "gini_cps": 0.367858, "bpic": 4.587938, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.4860676507387525, "vi": 1.0993817215415234, "tr": 1.344995163166249, "fi": 1.3613658144123393, "ru": 1.0152826084950546, "el": 0.08268676395732365, "hi": 0.5740971244275862, "ta": 1.0584308609512953, "th": 0.823084343442127, "zh": 4.489404775481027, "my": 4.8131818187148}, "t": 14.9} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.2243, "bpb": 6.09437, "chunk_loss": 0.03099, "F": 0.17831, "G": 0.48752, "chunks": 182.59, "bytes_per_chunk": 5.608, "boundary_entropy": 0.69168, "router_noise": 0.0, "grad_norm": 1.591, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.23, "alpha": 0.03, "t": 16.1} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4715.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.27093, "cps": 36.835938, "mean_p": 0.495963, "boundary_entropy": 0.687588, "bpb": 5.820065}, "de": {"chunks": 6108.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.297472, "cps": 47.71875, "mean_p": 0.495757, "boundary_entropy": 0.688412, "bpb": 5.934687}, "vi": {"chunks": 2908.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.157743, "cps": 22.71875, "mean_p": 0.481045, "boundary_entropy": 0.687917, "bpb": 6.312724}, "tr": {"chunks": 5040.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.278407, "cps": 39.375, "mean_p": 0.493897, "boundary_entropy": 0.688058, "bpb": 6.226978}, "fi": {"chunks": 5402.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.287172, "cps": 42.203125, "mean_p": 0.495046, "boundary_entropy": 0.688101, "bpb": 5.973529}, "ru": {"chunks": 7578.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.399894, "cps": 59.203125, "mean_p": 0.493554, "boundary_entropy": 0.690201, "bpb": 5.415797}, "el": {"chunks": 4163.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.202846, "cps": 32.523438, "mean_p": 0.490295, "boundary_entropy": 0.690334, "bpb": 5.661558}, "hi": {"chunks": 417.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.024785, "cps": 3.257812, "mean_p": 0.473574, "boundary_entropy": 0.689278, "bpb": 5.441769}, "ta": {"chunks": 2682.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.143461, "cps": 20.953125, "mean_p": 0.48333, "boundary_entropy": 0.690369, "bpb": 5.207834}, "th": {"chunks": 1473.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.091576, "cps": 11.507812, "mean_p": 0.484067, "boundary_entropy": 0.690428, "bpb": 5.102121}, "zh": {"chunks": 4634.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.784626, "cps": 36.203125, "mean_p": 0.492954, "boundary_entropy": 0.68664, "bpb": 6.993676}, "my": {"chunks": 21959.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.158909, "cps": 171.554688, "mean_p": 0.494025, "boundary_entropy": 0.690996, "bpb": 4.99626}}, "bpb": {"en": 5.82007, "de": 5.93469, "vi": 6.31272, "tr": 6.22698, "fi": 5.97353, "ru": 5.4158, "el": 5.66156, "hi": 5.44177, "ta": 5.20783, "th": 5.10212, "zh": 6.99368, "my": 4.99626}, "gini_cpc": 0.436552, "gini_cps": 0.411295, "bpic": 5.721329, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0979663856423556, "vi": 0.5822287990736419, "tr": 1.027596007270521, "fi": 1.059949366283677, "ru": 1.4760049356877647, "el": 0.7487002684673829, "hi": 0.09147942058398555, "ta": 0.5295119029960489, "th": 0.33800576273119576, "zh": 2.8960430268623423, "my": 4.277515636364614}, "t": 17.9} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.38298, "bpb": 4.88061, "chunk_loss": 0.029717, "F": 0.16199, "G": 0.47936, "chunks": 165.88, "bytes_per_chunk": 6.173, "boundary_entropy": 0.69027, "router_noise": 0.0, "grad_norm": 2.022, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.077, "alpha": 0.03, "t": 19.9} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2890.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166063, "cps": 22.578125, "mean_p": 0.449675, "boundary_entropy": 0.676329, "bpb": 4.495247}, "de": {"chunks": 3365.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.163883, "cps": 26.289062, "mean_p": 0.452042, "boundary_entropy": 0.678564, "bpb": 4.713403}, "vi": {"chunks": 6578.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.356821, "cps": 51.390625, "mean_p": 0.468635, "boundary_entropy": 0.683273, "bpb": 5.044461}, "tr": {"chunks": 3305.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.182566, "cps": 25.820312, "mean_p": 0.457515, "boundary_entropy": 0.680283, "bpb": 5.034498}, "fi": {"chunks": 4110.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.218489, "cps": 32.109375, "mean_p": 0.455546, "boundary_entropy": 0.678594, "bpb": 4.735896}, "ru": {"chunks": 5574.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.294142, "cps": 43.546875, "mean_p": 0.464582, "boundary_entropy": 0.684827, "bpb": 3.668242}, "el": {"chunks": 7144.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.348097, "cps": 55.8125, "mean_p": 0.478036, "boundary_entropy": 0.688098, "bpb": 4.710107}, "hi": {"chunks": 10071.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.598574, "cps": 78.679688, "mean_p": 0.490459, "boundary_entropy": 0.689656, "bpb": 3.641513}, "ta": {"chunks": 5531.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.295855, "cps": 43.210938, "mean_p": 0.483913, "boundary_entropy": 0.689628, "bpb": 3.448318}, "th": {"chunks": 9383.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.583339, "cps": 73.304688, "mean_p": 0.466869, "boundary_entropy": 0.68666, "bpb": 3.168764}, "zh": {"chunks": 2635.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.446156, "cps": 20.585938, "mean_p": 0.47883, "boundary_entropy": 0.685028, "bpb": 6.322387}, "my": {"chunks": 16237.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.856924, "cps": 126.851562, "mean_p": 0.475365, "boundary_entropy": 0.68885, "bpb": 3.999223}}, "bpb": {"en": 4.49525, "de": 4.7134, "vi": 5.04446, "tr": 5.0345, "fi": 4.7359, "ru": 3.66824, "el": 4.71011, "hi": 3.64151, "ta": 3.44832, "th": 3.16876, "zh": 6.32239, "my": 3.99922}, "gini_cpc": 0.290039, "gini_cps": 0.307133, "bpic": 4.995652, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9868677091160706, "vi": 2.1487060400190328, "tr": 1.0993783744574253, "fi": 1.3156980041310586, "ru": 1.7712669837762827, "el": 2.0961718197171644, "hi": 3.604489796757791, "ta": 1.7815764628145676, "th": 3.51274746077553, "zh": 2.6866646082747354, "my": 5.160225636654095}, "t": 22.2} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.89553, "bpb": 4.17737, "chunk_loss": 0.030335, "F": 0.17212, "G": 0.4326, "chunks": 176.25, "bytes_per_chunk": 5.81, "boundary_entropy": 0.67091, "router_noise": 0.0, "grad_norm": 1.6172, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.069, "alpha": 0.03, "t": 23.4} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2649.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.152215, "cps": 20.695312, "mean_p": 0.373057, "boundary_entropy": 0.624591, "bpb": 3.895521}, "de": {"chunks": 2877.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.140116, "cps": 22.476562, "mean_p": 0.376994, "boundary_entropy": 0.632386, "bpb": 4.109573}, "vi": {"chunks": 4130.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22403, "cps": 32.265625, "mean_p": 0.385528, "boundary_entropy": 0.610522, "bpb": 3.653165}, "tr": {"chunks": 2816.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.155554, "cps": 22.0, "mean_p": 0.391975, "boundary_entropy": 0.641599, "bpb": 4.168206}, "fi": {"chunks": 3222.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171283, "cps": 25.171875, "mean_p": 0.370109, "boundary_entropy": 0.624712, "bpb": 4.131064}, "ru": {"chunks": 5244.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.276728, "cps": 40.96875, "mean_p": 0.463584, "boundary_entropy": 0.679169, "bpb": 2.826355}, "el": {"chunks": 5279.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.257224, "cps": 41.242188, "mean_p": 0.400243, "boundary_entropy": 0.654252, "bpb": 3.177652}, "hi": {"chunks": 9827.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.584071, "cps": 76.773438, "mean_p": 0.407584, "boundary_entropy": 0.647135, "bpb": 2.414812}, "ta": {"chunks": 8752.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.468147, "cps": 68.375, "mean_p": 0.453768, "boundary_entropy": 0.678726, "bpb": 2.301073}, "th": {"chunks": 6132.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.381225, "cps": 47.90625, "mean_p": 0.446102, "boundary_entropy": 0.678257, "bpb": 2.334173}, "zh": {"chunks": 2932.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.496444, "cps": 22.90625, "mean_p": 0.464249, "boundary_entropy": 0.680409, "bpb": 5.321702}, "my": {"chunks": 7470.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.394237, "cps": 58.359375, "mean_p": 0.442782, "boundary_entropy": 0.679807, "bpb": 2.442}}, "bpb": {"en": 3.89552, "de": 4.10957, "vi": 3.65317, "tr": 4.16821, "fi": 4.13106, "ru": 2.82636, "el": 3.17765, "hi": 2.41481, "ta": 2.30107, "th": 2.33417, "zh": 5.3217, "my": 2.442}, "gini_cpc": 0.265124, "gini_cps": 0.255201, "bpic": 6.257639, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9205123437734323, "vi": 1.4718009252305473, "tr": 1.0219373236661562, "fi": 1.1252676005274618, "ru": 1.8180073328811288, "el": 1.6898687653092468, "hi": 3.8371435350579457, "ta": 3.0755585656014155, "th": 2.5045127263113462, "zh": 3.2614647215588577, "my": 2.589997755076718}, "t": 26.6} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.89722, "de": 4.10936, "vi": 3.65782, "tr": 4.17794, "fi": 4.13354, "ru": 2.81021, "el": 3.1727, "hi": 2.41046, "ta": 2.31975, "th": 2.36734, "zh": 5.33073, "my": 2.45962}, "bytes_per_chunk": {"en": 5.881, "de": 6.923, "vi": 5.856, "tr": 7.633, "fi": 6.037, "ru": 5.94, "el": 6.808, "hi": 4.363, "ta": 5.613, "th": 6.299, "zh": 8.229, "my": 7.155}, "bpb_mean": 3.40389, "t": 27.9} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.50071, "bpb": 3.60777, "chunk_loss": 0.029458, "F": 0.15738, "G": 0.40399, "chunks": 161.16, "bytes_per_chunk": 6.354, "boundary_entropy": 0.64965, "router_noise": 0.0, "grad_norm": 1.4604, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 27.9} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.57797, "bpb": 3.71922, "chunk_loss": 0.032548, "F": 0.21552, "G": 0.41131, "chunks": 220.69, "bytes_per_chunk": 4.64, "boundary_entropy": 0.64389, "router_noise": 0.0, "grad_norm": 1.5307, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 30.1} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3113.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.178877, "cps": 24.320312, "mean_p": 0.382042, "boundary_entropy": 0.622848, "bpb": 3.635103}, "de": {"chunks": 3931.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.191448, "cps": 30.710938, "mean_p": 0.391009, "boundary_entropy": 0.634267, "bpb": 3.906151}, "vi": {"chunks": 4665.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.253051, "cps": 36.445312, "mean_p": 0.426983, "boundary_entropy": 0.642481, "bpb": 3.238321}, "tr": {"chunks": 4605.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.254378, "cps": 35.976562, "mean_p": 0.405366, "boundary_entropy": 0.635724, "bpb": 3.843866}, "fi": {"chunks": 3130.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.166392, "cps": 24.453125, "mean_p": 0.376923, "boundary_entropy": 0.620762, "bpb": 3.887001}, "ru": {"chunks": 5849.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.308654, "cps": 45.695312, "mean_p": 0.452657, "boundary_entropy": 0.668708, "bpb": 2.649369}, "el": {"chunks": 6779.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.330312, "cps": 52.960938, "mean_p": 0.377005, "boundary_entropy": 0.631496, "bpb": 2.784232}, "hi": {"chunks": 8667.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.515126, "cps": 67.710938, "mean_p": 0.331654, "boundary_entropy": 0.570877, "bpb": 2.103737}, "ta": {"chunks": 8719.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.466381, "cps": 68.117188, "mean_p": 0.373919, "boundary_entropy": 0.623575, "bpb": 1.952427}, "th": {"chunks": 4201.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.261175, "cps": 32.820312, "mean_p": 0.41389, "boundary_entropy": 0.662787, "bpb": 2.076048}, "zh": {"chunks": 3523.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.596512, "cps": 27.523438, "mean_p": 0.454633, "boundary_entropy": 0.67257, "bpb": 4.984426}, "my": {"chunks": 11833.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.624499, "cps": 92.445312, "mean_p": 0.396647, "boundary_entropy": 0.650644, "bpb": 2.182718}}, "bpb": {"en": 3.6351, "de": 3.90615, "vi": 3.23832, "tr": 3.84387, "fi": 3.887, "ru": 2.64937, "el": 2.78423, "hi": 2.10374, "ta": 1.95243, "th": 2.07605, "zh": 4.98443, "my": 2.18272}, "gini_cpc": 0.2493, "gini_cps": 0.243272, "bpic": 5.560835, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0702756286375106, "vi": 1.4146646638143359, "tr": 1.4220801769309532, "fi": 0.9302023955712225, "ru": 1.7255097060840274, "el": 1.84658706154334, "hi": 2.879776100702179, "ta": 2.607271201434906, "th": 1.460079878009143, "zh": 3.334757055307756, "my": 3.4912141407043573}, "t": 31.3} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.35474, "bpb": 3.39717, "chunk_loss": 0.030818, "F": 0.18558, "G": 0.3677, "chunks": 190.03, "bytes_per_chunk": 5.389, "boundary_entropy": 0.60698, "router_noise": 0.0, "grad_norm": 1.2333, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.065, "alpha": 0.03, "t": 33.4} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.05853, "bpb": 2.96982, "chunk_loss": 0.031588, "F": 0.20145, "G": 0.35939, "chunks": 206.28, "bytes_per_chunk": 4.964, "boundary_entropy": 0.58704, "router_noise": 0.0, "grad_norm": 0.8366, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 35.5} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3468.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199276, "cps": 27.09375, "mean_p": 0.319056, "boundary_entropy": 0.528865, "bpb": 3.255988}, "de": {"chunks": 4822.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.234841, "cps": 37.671875, "mean_p": 0.337978, "boundary_entropy": 0.53406, "bpb": 3.534977}, "vi": {"chunks": 5665.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.307296, "cps": 44.257812, "mean_p": 0.348901, "boundary_entropy": 0.560343, "bpb": 2.799861}, "tr": {"chunks": 4570.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.252444, "cps": 35.703125, "mean_p": 0.348998, "boundary_entropy": 0.552438, "bpb": 3.445892}, "fi": {"chunks": 3917.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.208229, "cps": 30.601562, "mean_p": 0.331847, "boundary_entropy": 0.558249, "bpb": 3.559213}, "ru": {"chunks": 5320.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.280739, "cps": 41.5625, "mean_p": 0.2963, "boundary_entropy": 0.475013, "bpb": 2.507414}, "el": {"chunks": 6782.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.330459, "cps": 52.984375, "mean_p": 0.303199, "boundary_entropy": 0.541906, "bpb": 2.553436}, "hi": {"chunks": 7088.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.421278, "cps": 55.375, "mean_p": 0.385149, "boundary_entropy": 0.626898, "bpb": 1.810283}, "ta": {"chunks": 5147.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.275314, "cps": 40.210938, "mean_p": 0.387902, "boundary_entropy": 0.64044, "bpb": 1.586573}, "th": {"chunks": 5314.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.33037, "cps": 41.515625, "mean_p": 0.326892, "boundary_entropy": 0.59057, "bpb": 1.902233}, "zh": {"chunks": 2882.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.487978, "cps": 22.515625, "mean_p": 0.388086, "boundary_entropy": 0.620555, "bpb": 4.137508}, "my": {"chunks": 8785.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.463637, "cps": 68.632812, "mean_p": 0.363679, "boundary_entropy": 0.602934, "bpb": 1.929967}}, "bpb": {"en": 3.25599, "de": 3.53498, "vi": 2.79986, "tr": 3.44589, "fi": 3.55921, "ru": 2.50741, "el": 2.55344, "hi": 1.81028, "ta": 1.58657, "th": 1.90223, "zh": 4.13751, "my": 1.92997}, "gini_cpc": 0.161777, "gini_cps": 0.162986, "bpic": 6.01915, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1784735248533165, "vi": 1.5420618876948187, "tr": 1.266807657909842, "fi": 1.0449288487796748, "ru": 1.4087938574467502, "el": 1.6582956888439961, "hi": 2.1140422731456012, "ta": 1.38157265926996, "th": 1.6578510787737721, "zh": 2.448756293816505, "my": 2.3266090933262626}, "t": 38.5} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.26384, "de": 3.53399, "vi": 2.80629, "tr": 3.45276, "fi": 3.55162, "ru": 2.49056, "el": 2.54673, "hi": 1.80682, "ta": 1.60229, "th": 1.93111, "zh": 4.14503, "my": 1.94515}, "bytes_per_chunk": {"en": 4.941, "de": 4.359, "vi": 4.318, "tr": 4.241, "fi": 5.054, "ru": 6.494, "el": 5.723, "hi": 5.249, "ta": 7.934, "th": 7.282, "zh": 5.787, "my": 5.674}, "bpb_mean": 2.75635, "t": 39.9} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.05394, "bpb": 2.96321, "chunk_loss": 0.030959, "F": 0.19183, "G": 0.33043, "chunks": 196.44, "bytes_per_chunk": 5.213, "boundary_entropy": 0.54594, "router_noise": 0.0, "grad_norm": 0.9158, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 39.9} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.12412, "bpb": 3.06445, "chunk_loss": 0.0301, "F": 0.16913, "G": 0.32294, "chunks": 173.19, "bytes_per_chunk": 5.913, "boundary_entropy": 0.54112, "router_noise": 0.0, "grad_norm": 0.791, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 42.0} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.96149, "bpb": 2.82984, "chunk_loss": 0.030881, "F": 0.18799, "G": 0.32865, "chunks": 192.5, "bytes_per_chunk": 5.319, "boundary_entropy": 0.53915, "router_noise": 0.0, "grad_norm": 1.075, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 44.0} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.90946, "bpb": 2.75477, "chunk_loss": 0.030547, "F": 0.18332, "G": 0.31174, "chunks": 187.72, "bytes_per_chunk": 5.455, "boundary_entropy": 0.5166, "router_noise": 0.0, "grad_norm": 0.9845, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 46.1} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2946.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.169281, "cps": 23.015625, "mean_p": 0.266836, "boundary_entropy": 0.448734, "bpb": 2.944773}, "de": {"chunks": 3688.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.179613, "cps": 28.8125, "mean_p": 0.27822, "boundary_entropy": 0.452305, "bpb": 3.268797}, "vi": {"chunks": 4940.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.267969, "cps": 38.59375, "mean_p": 0.303319, "boundary_entropy": 0.497103, "bpb": 2.58392}, "tr": {"chunks": 4015.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.221786, "cps": 31.367188, "mean_p": 0.320369, "boundary_entropy": 0.508837, "bpb": 3.157863}, "fi": {"chunks": 3469.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184413, "cps": 27.101562, "mean_p": 0.29463, "boundary_entropy": 0.498989, "bpb": 3.3723}, "ru": {"chunks": 5532.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.291926, "cps": 43.21875, "mean_p": 0.310255, "boundary_entropy": 0.531038, "bpb": 2.198196}, "el": {"chunks": 6900.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.336208, "cps": 53.90625, "mean_p": 0.326959, "boundary_entropy": 0.563231, "bpb": 2.260604}, "hi": {"chunks": 8375.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.497771, "cps": 65.429688, "mean_p": 0.318886, "boundary_entropy": 0.525007, "bpb": 1.629261}, "ta": {"chunks": 8010.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.428457, "cps": 62.578125, "mean_p": 0.356868, "boundary_entropy": 0.595403, "bpb": 1.401254}, "th": {"chunks": 6663.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.414237, "cps": 52.054688, "mean_p": 0.297745, "boundary_entropy": 0.515713, "bpb": 1.725528}, "zh": {"chunks": 2848.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482221, "cps": 22.25, "mean_p": 0.361606, "boundary_entropy": 0.57286, "bpb": 3.849121}, "my": {"chunks": 6128.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.323411, "cps": 47.875, "mean_p": 0.378734, "boundary_entropy": 0.629468, "bpb": 1.683728}}, "bpb": {"en": 2.94477, "de": 3.2688, "vi": 2.58392, "tr": 3.15786, "fi": 3.3723, "ru": 2.1982, "el": 2.2606, "hi": 1.62926, "ta": 1.40125, "th": 1.72553, "zh": 3.84912, "my": 1.68373}, "gini_cpc": 0.202855, "gini_cps": 0.199688, "bpic": 6.042463, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.06103542228927, "vi": 1.5829791136027558, "tr": 1.31016615377969, "fi": 1.0893910277613057, "ru": 1.7245045112822361, "el": 1.9860931945312366, "hi": 2.940499571786965, "ta": 2.5310365939682407, "th": 2.4470346877933995, "zh": 2.8486423072121765, "my": 1.9104987516483103}, "t": 49.0} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.95427, "de": 3.26793, "vi": 2.59158, "tr": 3.16847, "fi": 3.36223, "ru": 2.18292, "el": 2.25214, "hi": 1.62288, "ta": 1.41507, "th": 1.75425, "zh": 3.85987, "my": 1.69578}, "bytes_per_chunk": {"en": 5.698, "de": 5.397, "vi": 5.125, "tr": 4.708, "fi": 5.672, "ru": 6.207, "el": 4.961, "hi": 4.469, "ta": 5.252, "th": 5.841, "zh": 5.918, "my": 9.831}, "bpb_mean": 2.51062, "t": 50.3} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.93925, "bpb": 2.79774, "chunk_loss": 0.030219, "F": 0.17514, "G": 0.3027, "chunks": 179.34, "bytes_per_chunk": 5.71, "boundary_entropy": 0.50086, "router_noise": 0.0, "grad_norm": 0.8349, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 50.4} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.73193, "bpb": 2.49865, "chunk_loss": 0.030403, "F": 0.17859, "G": 0.31055, "chunks": 182.88, "bytes_per_chunk": 5.599, "boundary_entropy": 0.51385, "router_noise": 0.0, "grad_norm": 0.9119, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 52.4} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.75743, "bpb": 2.53543, "chunk_loss": 0.03045, "F": 0.17947, "G": 0.31113, "chunks": 183.78, "bytes_per_chunk": 5.572, "boundary_entropy": 0.51197, "router_noise": 0.0, "grad_norm": 0.6821, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 54.4} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.6891, "bpb": 2.43685, "chunk_loss": 0.030204, "F": 0.17676, "G": 0.31737, "chunks": 181.0, "bytes_per_chunk": 5.657, "boundary_entropy": 0.51634, "router_noise": 0.0, "grad_norm": 0.689, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 56.5} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2932.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168477, "cps": 22.90625, "mean_p": 0.262214, "boundary_entropy": 0.419845, "bpb": 2.743982}, "de": {"chunks": 3645.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.177519, "cps": 28.476562, "mean_p": 0.261896, "boundary_entropy": 0.420631, "bpb": 3.071267}, "vi": {"chunks": 4276.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23195, "cps": 33.40625, "mean_p": 0.275079, "boundary_entropy": 0.469579, "bpb": 2.428148}, "tr": {"chunks": 3192.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176324, "cps": 24.9375, "mean_p": 0.269899, "boundary_entropy": 0.455432, "bpb": 3.010089}, "fi": {"chunks": 3522.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.187231, "cps": 27.515625, "mean_p": 0.280068, "boundary_entropy": 0.461817, "bpb": 3.238985}, "ru": {"chunks": 6406.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.338047, "cps": 50.046875, "mean_p": 0.361227, "boundary_entropy": 0.580539, "bpb": 2.040909}, "el": {"chunks": 6720.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.327438, "cps": 52.5, "mean_p": 0.321815, "boundary_entropy": 0.546721, "bpb": 2.124484}, "hi": {"chunks": 6307.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.374859, "cps": 49.273438, "mean_p": 0.310705, "boundary_entropy": 0.535747, "bpb": 1.545703}, "ta": {"chunks": 7721.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.412998, "cps": 60.320312, "mean_p": 0.404296, "boundary_entropy": 0.639701, "bpb": 1.281065}, "th": {"chunks": 4852.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.301647, "cps": 37.90625, "mean_p": 0.35368, "boundary_entropy": 0.608502, "bpb": 1.619785}, "zh": {"chunks": 3343.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.566035, "cps": 26.117188, "mean_p": 0.341133, "boundary_entropy": 0.525544, "bpb": 3.734169}, "my": {"chunks": 10348.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.546126, "cps": 80.84375, "mean_p": 0.39162, "boundary_entropy": 0.628772, "bpb": 1.459803}}, "bpb": {"en": 2.74398, "de": 3.07127, "vi": 2.42815, "tr": 3.01009, "fi": 3.23898, "ru": 2.04091, "el": 2.12448, "hi": 1.5457, "ta": 1.28106, "th": 1.61979, "zh": 3.73417, "my": 1.4598}, "gini_cpc": 0.232627, "gini_cps": 0.222557, "bpic": 6.066341, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0536716126417867, "vi": 1.3767488069693636, "tr": 1.04658013087005, "fi": 1.1113161416178148, "ru": 2.006494040826905, "el": 1.9435180666764649, "hi": 2.224989225819922, "ta": 2.451366445712334, "th": 1.790440450861958, "zh": 3.359720027128053, "my": 3.2415535330268432}, "t": 59.4} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.75536, "de": 3.07434, "vi": 2.43554, "tr": 3.02108, "fi": 3.23844, "ru": 2.02482, "el": 2.11061, "hi": 1.5399, "ta": 1.29241, "th": 1.64605, "zh": 3.74564, "my": 1.471}, "bytes_per_chunk": {"en": 5.629, "de": 5.436, "vi": 5.678, "tr": 6.059, "fi": 5.308, "ru": 5.571, "el": 5.794, "hi": 6.481, "ta": 5.792, "th": 9.029, "zh": 4.936, "my": 5.501}, "bpb_mean": 2.36293, "t": 60.7} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.81552, "bpb": 2.61924, "chunk_loss": 0.030176, "F": 0.17236, "G": 0.30039, "chunks": 176.5, "bytes_per_chunk": 5.802, "boundary_entropy": 0.48907, "router_noise": 0.0, "grad_norm": 0.7945, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 60.7} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.86477, "bpb": 2.6903, "chunk_loss": 0.030211, "F": 0.17682, "G": 0.29885, "chunks": 181.06, "bytes_per_chunk": 5.656, "boundary_entropy": 0.48291, "router_noise": 0.0, "grad_norm": 0.8728, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 62.8} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.6813, "bpb": 2.42561, "chunk_loss": 0.030012, "F": 0.17389, "G": 0.29884, "chunks": 178.06, "bytes_per_chunk": 5.751, "boundary_entropy": 0.47568, "router_noise": 0.0, "grad_norm": 0.9095, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 64.8} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.81396, "bpb": 2.61699, "chunk_loss": 0.030691, "F": 0.18848, "G": 0.29656, "chunks": 193.0, "bytes_per_chunk": 5.306, "boundary_entropy": 0.46517, "router_noise": 0.0, "grad_norm": 0.8544, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 66.8} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3233.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.185773, "cps": 25.257812, "mean_p": 0.264605, "boundary_entropy": 0.396762, "bpb": 2.60452}, "de": {"chunks": 3845.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18726, "cps": 30.039062, "mean_p": 0.267159, "boundary_entropy": 0.407291, "bpb": 2.943751}, "vi": {"chunks": 4547.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.24665, "cps": 35.523438, "mean_p": 0.276564, "boundary_entropy": 0.468126, "bpb": 2.325228}, "tr": {"chunks": 3517.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.194277, "cps": 27.476562, "mean_p": 0.267803, "boundary_entropy": 0.428855, "bpb": 2.912621}, "fi": {"chunks": 3647.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.193876, "cps": 28.492188, "mean_p": 0.274978, "boundary_entropy": 0.441549, "bpb": 3.143412}, "ru": {"chunks": 5279.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.278575, "cps": 41.242188, "mean_p": 0.332919, "boundary_entropy": 0.565774, "bpb": 1.947282}, "el": {"chunks": 6958.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.339034, "cps": 54.359375, "mean_p": 0.320739, "boundary_entropy": 0.547014, "bpb": 2.039518}, "hi": {"chunks": 9904.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.588648, "cps": 77.375, "mean_p": 0.352569, "boundary_entropy": 0.578175, "bpb": 1.504156}, "ta": {"chunks": 8377.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.448088, "cps": 65.445312, "mean_p": 0.396277, "boundary_entropy": 0.634099, "bpb": 1.214221}, "th": {"chunks": 9382.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.583276, "cps": 73.296875, "mean_p": 0.411658, "boundary_entropy": 0.644614, "bpb": 1.530804}, "zh": {"chunks": 3085.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.52235, "cps": 24.101562, "mean_p": 0.312182, "boundary_entropy": 0.473681, "bpb": 3.61849}, "my": {"chunks": 11026.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.581908, "cps": 86.140625, "mean_p": 0.412932, "boundary_entropy": 0.651298, "bpb": 1.414535}}, "bpb": {"en": 2.60452, "de": 2.94375, "vi": 2.32523, "tr": 2.91262, "fi": 3.14341, "ru": 1.94728, "el": 2.03952, "hi": 1.50416, "ta": 1.21422, "th": 1.5308, "zh": 3.61849, "my": 1.41454}, "gini_cpc": 0.249387, "gini_cps": 0.254311, "bpic": 5.271717, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.008004225286616, "vi": 1.3277008333255733, "tr": 1.0457797605511945, "fi": 1.0436197647647214, "ru": 1.4995496965250341, "el": 1.82499632747784, "hi": 3.16864783627607, "ta": 2.412023093289341, "th": 3.1397334432248414, "zh": 2.8117722554896285, "my": 3.132369796354762}, "t": 69.8} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.61416, "de": 2.94845, "vi": 2.3297, "tr": 2.92372, "fi": 3.14545, "ru": 1.92993, "el": 2.02818, "hi": 1.49886, "ta": 1.2255, "th": 1.55709, "zh": 3.62767, "my": 1.4243}, "bytes_per_chunk": {"en": 5.181, "de": 5.148, "vi": 5.244, "tr": 5.304, "fi": 5.144, "ru": 6.608, "el": 5.465, "hi": 4.332, "ta": 5.553, "th": 4.92, "zh": 5.211, "my": 5.196}, "bpb_mean": 2.27108, "t": 71.1} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.74103, "bpb": 2.51177, "chunk_loss": 0.030854, "F": 0.19366, "G": 0.30967, "chunks": 198.31, "bytes_per_chunk": 5.164, "boundary_entropy": 0.48762, "router_noise": 0.0, "grad_norm": 0.8028, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 71.1} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.61116, "bpb": 2.32442, "chunk_loss": 0.030732, "F": 0.1879, "G": 0.32331, "chunks": 192.41, "bytes_per_chunk": 5.322, "boundary_entropy": 0.50123, "router_noise": 0.0, "grad_norm": 0.7948, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 73.1} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.80183, "bpb": 2.59949, "chunk_loss": 0.030789, "F": 0.19296, "G": 0.31073, "chunks": 197.59, "bytes_per_chunk": 5.182, "boundary_entropy": 0.47456, "router_noise": 0.0, "grad_norm": 0.7971, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 75.1} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.6595, "bpb": 2.39415, "chunk_loss": 0.03038, "F": 0.18158, "G": 0.29221, "chunks": 185.94, "bytes_per_chunk": 5.507, "boundary_entropy": 0.45098, "router_noise": 0.0, "grad_norm": 0.8227, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 77.1} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3255.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.187037, "cps": 25.429688, "mean_p": 0.249077, "boundary_entropy": 0.35594, "bpb": 2.503157}, "de": {"chunks": 3972.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.193445, "cps": 31.03125, "mean_p": 0.256915, "boundary_entropy": 0.372153, "bpb": 2.81795}, "vi": {"chunks": 3784.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.205262, "cps": 29.5625, "mean_p": 0.268859, "boundary_entropy": 0.459776, "bpb": 2.300088}, "tr": {"chunks": 3933.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.217257, "cps": 30.726562, "mean_p": 0.270912, "boundary_entropy": 0.409012, "bpb": 2.800165}, "fi": {"chunks": 4102.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.218064, "cps": 32.046875, "mean_p": 0.274174, "boundary_entropy": 0.404883, "bpb": 3.058013}, "ru": {"chunks": 5334.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.281478, "cps": 41.671875, "mean_p": 0.326111, "boundary_entropy": 0.554506, "bpb": 1.855027}, "el": {"chunks": 6442.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.313892, "cps": 50.328125, "mean_p": 0.307488, "boundary_entropy": 0.524375, "bpb": 1.953798}, "hi": {"chunks": 4715.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.280238, "cps": 36.835938, "mean_p": 0.295103, "boundary_entropy": 0.52192, "bpb": 1.48707}, "ta": {"chunks": 6027.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.322386, "cps": 47.085938, "mean_p": 0.328918, "boundary_entropy": 0.591194, "bpb": 1.24429}, "th": {"chunks": 7884.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.490146, "cps": 61.59375, "mean_p": 0.377471, "boundary_entropy": 0.616232, "bpb": 1.48252}, "zh": {"chunks": 3049.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.516255, "cps": 23.820312, "mean_p": 0.294454, "boundary_entropy": 0.439124, "bpb": 3.565465}, "my": {"chunks": 9767.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.515463, "cps": 76.304688, "mean_p": 0.418528, "boundary_entropy": 0.655789, "bpb": 1.354521}}, "bpb": {"en": 2.50316, "de": 2.81795, "vi": 2.30009, "tr": 2.80016, "fi": 3.05801, "ru": 1.85503, "el": 1.9538, "hi": 1.48707, "ta": 1.24429, "th": 1.48252, "zh": 3.56547, "my": 1.35452}, "gini_cpc": 0.207795, "gini_cps": 0.199867, "bpic": 6.16377, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.034260550791454, "vi": 1.097440827718263, "tr": 1.1615730366252914, "fi": 1.1658881814175304, "ru": 1.5049321644395266, "el": 1.6782358828164243, "hi": 1.4983033531679462, "ta": 1.7236490695286821, "th": 2.6205875755097128, "zh": 2.7601781208206604, "my": 2.7559474924824574}, "t": 80.1} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.51485, "de": 2.82265, "vi": 2.3046, "tr": 2.80966, "fi": 3.06247, "ru": 1.83633, "el": 1.94384, "hi": 1.48361, "ta": 1.2511, "th": 1.50877, "zh": 3.57406, "my": 1.36308}, "bytes_per_chunk": {"en": 5.22, "de": 5.091, "vi": 6.598, "tr": 4.786, "fi": 4.624, "ru": 6.684, "el": 5.915, "hi": 8.406, "ta": 8.037, "th": 5.611, "zh": 5.317, "my": 5.758}, "bpb_mean": 2.20625, "t": 81.4} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.71863, "bpb": 2.47946, "chunk_loss": 0.030392, "F": 0.18408, "G": 0.27939, "chunks": 188.5, "bytes_per_chunk": 5.432, "boundary_entropy": 0.42762, "router_noise": 0.0, "grad_norm": 0.8617, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 81.4} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.83298, "bpb": 2.64443, "chunk_loss": 0.030344, "F": 0.18277, "G": 0.28172, "chunks": 187.16, "bytes_per_chunk": 5.471, "boundary_entropy": 0.42743, "router_noise": 0.0, "grad_norm": 0.8226, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 83.5} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.51739, "bpb": 2.18913, "chunk_loss": 0.030039, "F": 0.17157, "G": 0.29202, "chunks": 175.69, "bytes_per_chunk": 5.829, "boundary_entropy": 0.46153, "router_noise": 0.0, "grad_norm": 0.8633, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 85.5} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.50995, "bpb": 2.17839, "chunk_loss": 0.030456, "F": 0.18481, "G": 0.29132, "chunks": 189.25, "bytes_per_chunk": 5.411, "boundary_entropy": 0.45303, "router_noise": 0.0, "grad_norm": 0.7664, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 87.5} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3360.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19307, "cps": 26.25, "mean_p": 0.24682, "boundary_entropy": 0.341187, "bpb": 2.428261}, "de": {"chunks": 3967.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.193201, "cps": 30.992188, "mean_p": 0.256517, "boundary_entropy": 0.360164, "bpb": 2.748835}, "vi": {"chunks": 4213.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228533, "cps": 32.914062, "mean_p": 0.276214, "boundary_entropy": 0.458955, "bpb": 2.202801}, "tr": {"chunks": 3380.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.186709, "cps": 26.40625, "mean_p": 0.242336, "boundary_entropy": 0.369991, "bpb": 2.725321}, "fi": {"chunks": 3528.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18755, "cps": 27.5625, "mean_p": 0.242881, "boundary_entropy": 0.36442, "bpb": 3.009162}, "ru": {"chunks": 6881.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.363113, "cps": 53.757812, "mean_p": 0.372377, "boundary_entropy": 0.574739, "bpb": 1.79832}, "el": {"chunks": 6352.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.309506, "cps": 49.625, "mean_p": 0.303995, "boundary_entropy": 0.526823, "bpb": 1.871224}, "hi": {"chunks": 7006.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.416404, "cps": 54.734375, "mean_p": 0.301141, "boundary_entropy": 0.51671, "bpb": 1.429113}, "ta": {"chunks": 9523.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.509388, "cps": 74.398438, "mean_p": 0.351989, "boundary_entropy": 0.592197, "bpb": 1.139393}, "th": {"chunks": 6785.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.421822, "cps": 53.007812, "mean_p": 0.385004, "boundary_entropy": 0.629219, "bpb": 1.433002}, "zh": {"chunks": 3110.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.526583, "cps": 24.296875, "mean_p": 0.302231, "boundary_entropy": 0.436356, "bpb": 3.54317}, "my": {"chunks": 7987.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.421522, "cps": 62.398438, "mean_p": 0.408304, "boundary_entropy": 0.656549, "bpb": 1.310191}}, "bpb": {"en": 2.42826, "de": 2.74884, "vi": 2.2028, "tr": 2.72532, "fi": 3.00916, "ru": 1.79832, "el": 1.87122, "hi": 1.42911, "ta": 1.13939, "th": 1.433, "zh": 3.54317, "my": 1.31019}, "gini_cpc": 0.210476, "gini_cps": 0.20795, "bpic": 5.806769, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0006786549178674, "vi": 1.1836768665969235, "tr": 0.9670545923722194, "fi": 0.9714076869916538, "ru": 1.880733179419525, "el": 1.6030773835626928, "hi": 2.156750477605604, "ta": 2.6383545543116953, "th": 2.1848097717483017, "zh": 2.727418545305017, "my": 2.183258457356977}, "t": 90.4} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.43677, "de": 2.75159, "vi": 2.2056, "tr": 2.73823, "fi": 3.00847, "ru": 1.7805, "el": 1.86087, "hi": 1.42477, "ta": 1.14798, "th": 1.45476, "zh": 3.54833, "my": 1.31983}, "bytes_per_chunk": {"en": 5.221, "de": 5.068, "vi": 5.945, "tr": 5.545, "fi": 5.236, "ru": 5.269, "el": 6.038, "hi": 5.487, "ta": 5.043, "th": 6.656, "zh": 5.261, "my": 6.814}, "bpb_mean": 2.13981, "t": 91.7} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.57063, "bpb": 2.26594, "chunk_loss": 0.030434, "F": 0.18628, "G": 0.30174, "chunks": 190.75, "bytes_per_chunk": 5.368, "boundary_entropy": 0.44951, "router_noise": 0.0, "grad_norm": 0.7444, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 91.8} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.6662, "bpb": 2.40383, "chunk_loss": 0.030302, "F": 0.181, "G": 0.27566, "chunks": 185.34, "bytes_per_chunk": 5.525, "boundary_entropy": 0.42035, "router_noise": 0.0, "grad_norm": 0.7231, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 93.8} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.73738, "bpb": 2.50651, "chunk_loss": 0.030233, "F": 0.17636, "G": 0.29589, "chunks": 180.59, "bytes_per_chunk": 5.67, "boundary_entropy": 0.44982, "router_noise": 0.0, "grad_norm": 0.737, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 95.8} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.60291, "bpb": 2.3125, "chunk_loss": 0.030191, "F": 0.17807, "G": 0.28947, "chunks": 182.34, "bytes_per_chunk": 5.616, "boundary_entropy": 0.43566, "router_noise": 0.0, "grad_norm": 0.7412, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 97.8} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3371.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193702, "cps": 26.335938, "mean_p": 0.244827, "boundary_entropy": 0.328638, "bpb": 2.370436}, "de": {"chunks": 4151.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.202162, "cps": 32.429688, "mean_p": 0.258124, "boundary_entropy": 0.347606, "bpb": 2.650875}, "vi": {"chunks": 3842.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.208408, "cps": 30.015625, "mean_p": 0.271673, "boundary_entropy": 0.457509, "bpb": 2.17412}, "tr": {"chunks": 3723.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205657, "cps": 29.085938, "mean_p": 0.251411, "boundary_entropy": 0.36672, "bpb": 2.660913}, "fi": {"chunks": 3843.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.204295, "cps": 30.023438, "mean_p": 0.250253, "boundary_entropy": 0.357102, "bpb": 2.884408}, "ru": {"chunks": 5941.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.313509, "cps": 46.414062, "mean_p": 0.346823, "boundary_entropy": 0.552515, "bpb": 1.73958}, "el": {"chunks": 7504.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.365639, "cps": 58.625, "mean_p": 0.313742, "boundary_entropy": 0.505944, "bpb": 1.819939}, "hi": {"chunks": 7502.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.445884, "cps": 58.609375, "mean_p": 0.307171, "boundary_entropy": 0.51669, "bpb": 1.394407}, "ta": {"chunks": 10423.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.557529, "cps": 81.429688, "mean_p": 0.363712, "boundary_entropy": 0.592899, "bpb": 1.176198}, "th": {"chunks": 7372.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.458315, "cps": 57.59375, "mean_p": 0.382528, "boundary_entropy": 0.625711, "bpb": 1.379235}, "zh": {"chunks": 3127.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.529462, "cps": 24.429688, "mean_p": 0.304422, "boundary_entropy": 0.420913, "bpb": 3.522325}, "my": {"chunks": 8739.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.46121, "cps": 68.273438, "mean_p": 0.400035, "boundary_entropy": 0.647722, "bpb": 1.270079}}, "bpb": {"en": 2.37044, "de": 2.65087, "vi": 2.17412, "tr": 2.66091, "fi": 2.88441, "ru": 1.73958, "el": 1.81994, "hi": 1.39441, "ta": 1.1762, "th": 1.37924, "zh": 3.52232, "my": 1.27008}, "gini_cpc": 0.216567, "gini_cps": 0.222555, "bpic": 5.519011, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0436759932492057, "vi": 1.0759190230943632, "tr": 1.0617147538673892, "fi": 1.0546876696497933, "ru": 1.61851118769514, "el": 1.8876320730257663, "hi": 2.301904779116342, "ta": 2.8782773223684956, "th": 2.3660811136919366, "zh": 2.733378702811967, "my": 2.3810237695923337}, "t": 100.8} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.37827, "de": 2.65608, "vi": 2.18049, "tr": 2.67046, "fi": 2.8902, "ru": 1.71947, "el": 1.80923, "hi": 1.39097, "ta": 1.18559, "th": 1.39788, "zh": 3.52656, "my": 1.28004}, "bytes_per_chunk": {"en": 5.201, "de": 4.952, "vi": 6.264, "tr": 5.114, "fi": 4.945, "ru": 6.078, "el": 5.146, "hi": 5.419, "ta": 4.846, "th": 6.283, "zh": 5.28, "my": 6.166}, "bpb_mean": 2.09044, "t": 102.1} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.58891, "bpb": 2.29231, "chunk_loss": 0.030591, "F": 0.18857, "G": 0.29438, "chunks": 193.09, "bytes_per_chunk": 5.303, "boundary_entropy": 0.43686, "router_noise": 0.0, "grad_norm": 0.7247, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 102.1} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.57572, "bpb": 2.27328, "chunk_loss": 0.030169, "F": 0.17856, "G": 0.28235, "chunks": 182.84, "bytes_per_chunk": 5.6, "boundary_entropy": 0.42624, "router_noise": 0.0, "grad_norm": 0.7459, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 104.1} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.54472, "bpb": 2.22856, "chunk_loss": 0.030712, "F": 0.1875, "G": 0.29751, "chunks": 192.0, "bytes_per_chunk": 5.333, "boundary_entropy": 0.43416, "router_noise": 0.0, "grad_norm": 0.7305, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 106.1} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.55619, "bpb": 2.2451, "chunk_loss": 0.030518, "F": 0.18713, "G": 0.28684, "chunks": 191.62, "bytes_per_chunk": 5.344, "boundary_entropy": 0.41967, "router_noise": 0.0, "grad_norm": 0.6831, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 108.1} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3318.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.190657, "cps": 25.921875, "mean_p": 0.234959, "boundary_entropy": 0.312318, "bpb": 2.327394}, "de": {"chunks": 4056.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.197536, "cps": 31.6875, "mean_p": 0.248664, "boundary_entropy": 0.330567, "bpb": 2.618683}, "vi": {"chunks": 4678.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.253756, "cps": 36.546875, "mean_p": 0.286012, "boundary_entropy": 0.462534, "bpb": 2.137521}, "tr": {"chunks": 3771.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.208308, "cps": 29.460938, "mean_p": 0.249096, "boundary_entropy": 0.360221, "bpb": 2.606267}, "fi": {"chunks": 3716.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.197544, "cps": 29.03125, "mean_p": 0.238079, "boundary_entropy": 0.33029, "bpb": 2.838017}, "ru": {"chunks": 6058.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319683, "cps": 47.328125, "mean_p": 0.356298, "boundary_entropy": 0.553524, "bpb": 1.703586}, "el": {"chunks": 6493.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.316377, "cps": 50.726562, "mean_p": 0.298492, "boundary_entropy": 0.49407, "bpb": 1.794913}, "hi": {"chunks": 4656.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.276731, "cps": 36.375, "mean_p": 0.263667, "boundary_entropy": 0.48511, "bpb": 1.387147}, "ta": {"chunks": 7240.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.387269, "cps": 56.5625, "mean_p": 0.288654, "boundary_entropy": 0.524581, "bpb": 1.101405}, "th": {"chunks": 8518.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.529562, "cps": 66.546875, "mean_p": 0.383374, "boundary_entropy": 0.616455, "bpb": 1.331148}, "zh": {"chunks": 2920.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.494412, "cps": 22.8125, "mean_p": 0.2966, "boundary_entropy": 0.421658, "bpb": 3.467812}, "my": {"chunks": 9858.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.520266, "cps": 77.015625, "mean_p": 0.422702, "boundary_entropy": 0.663043, "bpb": 1.254964}}, "bpb": {"en": 2.32739, "de": 2.61868, "vi": 2.13752, "tr": 2.60627, "fi": 2.83802, "ru": 1.70359, "el": 1.79491, "hi": 1.38715, "ta": 1.10141, "th": 1.33115, "zh": 3.46781, "my": 1.25496}, "gini_cpc": 0.210744, "gini_cps": 0.213719, "bpic": 5.878818, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0360799697113516, "vi": 1.3309594190231944, "tr": 1.0925811902735791, "fi": 1.0361235870320575, "ru": 1.676747985323517, "el": 1.659404625111208, "hi": 1.451461889625412, "ta": 2.0312381055556816, "th": 2.777565498838955, "zh": 2.5932067735082907, "my": 2.7288092355932925}, "t": 111.1} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.33552, "de": 2.62004, "vi": 2.14304, "tr": 2.61505, "fi": 2.84097, "ru": 1.68288, "el": 1.78674, "hi": 1.38275, "ta": 1.10916, "th": 1.35142, "zh": 3.47161, "my": 1.26312}, "bytes_per_chunk": {"en": 5.269, "de": 5.077, "vi": 5.489, "tr": 5.078, "fi": 5.072, "ru": 6.033, "el": 5.972, "hi": 8.838, "ta": 6.69, "th": 5.407, "zh": 5.657, "my": 5.364}, "bpb_mean": 2.05019, "t": 112.4} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.53228, "bpb": 2.21061, "chunk_loss": 0.030304, "F": 0.17844, "G": 0.29083, "chunks": 182.72, "bytes_per_chunk": 5.604, "boundary_entropy": 0.44176, "router_noise": 0.0, "grad_norm": 0.7122, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 112.5} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.57506, "bpb": 2.27233, "chunk_loss": 0.030656, "F": 0.1922, "G": 0.28108, "chunks": 196.81, "bytes_per_chunk": 5.203, "boundary_entropy": 0.40116, "router_noise": 0.0, "grad_norm": 0.6738, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 114.5} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.63158, "bpb": 2.35387, "chunk_loss": 0.030997, "F": 0.19394, "G": 0.28288, "chunks": 198.59, "bytes_per_chunk": 5.156, "boundary_entropy": 0.40031, "router_noise": 0.0, "grad_norm": 0.763, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 116.6} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.47142, "bpb": 2.12281, "chunk_loss": 0.030441, "F": 0.18207, "G": 0.29934, "chunks": 186.44, "bytes_per_chunk": 5.492, "boundary_entropy": 0.44628, "router_noise": 0.0, "grad_norm": 0.7307, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 118.7} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3326.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.191116, "cps": 25.984375, "mean_p": 0.235442, "boundary_entropy": 0.303403, "bpb": 2.292107}, "de": {"chunks": 4213.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.205182, "cps": 32.914062, "mean_p": 0.254138, "boundary_entropy": 0.322535, "bpb": 2.547701}, "vi": {"chunks": 4012.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.21763, "cps": 31.34375, "mean_p": 0.270255, "boundary_entropy": 0.448489, "bpb": 2.102884}, "tr": {"chunks": 3668.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.202618, "cps": 28.65625, "mean_p": 0.241461, "boundary_entropy": 0.338425, "bpb": 2.557104}, "fi": {"chunks": 3857.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.20504, "cps": 30.132812, "mean_p": 0.242792, "boundary_entropy": 0.327539, "bpb": 2.766988}, "ru": {"chunks": 5895.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.311082, "cps": 46.054688, "mean_p": 0.354904, "boundary_entropy": 0.551084, "bpb": 1.653228}, "el": {"chunks": 6974.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.339814, "cps": 54.484375, "mean_p": 0.311835, "boundary_entropy": 0.494304, "bpb": 1.708824}, "hi": {"chunks": 10044.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.596969, "cps": 78.46875, "mean_p": 0.319128, "boundary_entropy": 0.534531, "bpb": 1.345506}, "ta": {"chunks": 8651.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.462744, "cps": 67.585938, "mean_p": 0.329609, "boundary_entropy": 0.563489, "bpb": 1.101713}, "th": {"chunks": 8511.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.529127, "cps": 66.492188, "mean_p": 0.388241, "boundary_entropy": 0.620873, "bpb": 1.311984}, "zh": {"chunks": 2928.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.495767, "cps": 22.875, "mean_p": 0.301055, "boundary_entropy": 0.426638, "bpb": 3.435728}, "my": {"chunks": 11455.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.604549, "cps": 89.492188, "mean_p": 0.415191, "boundary_entropy": 0.651932, "bpb": 1.228699}}, "bpb": {"en": 2.29211, "de": 2.5477, "vi": 2.10288, "tr": 2.5571, "fi": 2.76699, "ru": 1.65323, "el": 1.70882, "hi": 1.34551, "ta": 1.10171, "th": 1.31198, "zh": 3.43573, "my": 1.2287}, "gini_cpc": 0.240401, "gini_cps": 0.252693, "bpic": 5.219096, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0735961051682814, "vi": 1.138727103614934, "tr": 1.0601825479901554, "fi": 1.07285154446262, "ru": 1.627707896686695, "el": 1.7780459211698274, "hi": 3.1235862791156888, "ta": 2.421267194353386, "th": 2.7686075608178213, "zh": 2.594056942204784, "my": 3.1632505540053852}, "t": 121.6} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.29872, "de": 2.55272, "vi": 2.10906, "tr": 2.56477, "fi": 2.77491, "ru": 1.62962, "el": 1.704, "hi": 1.34234, "ta": 1.10885, "th": 1.33061, "zh": 3.4382, "my": 1.23785}, "bytes_per_chunk": {"en": 5.23, "de": 4.928, "vi": 6.102, "tr": 5.21, "fi": 4.942, "ru": 6.079, "el": 5.482, "hi": 4.5, "ta": 5.696, "th": 5.384, "zh": 5.652, "my": 4.561}, "bpb_mean": 2.00764, "t": 122.9} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.59324, "bpb": 2.29856, "chunk_loss": 0.030522, "F": 0.19037, "G": 0.28318, "chunks": 194.94, "bytes_per_chunk": 5.253, "boundary_entropy": 0.39982, "router_noise": 0.0, "grad_norm": 0.7471, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 123.0} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.60226, "bpb": 2.31157, "chunk_loss": 0.030432, "F": 0.18509, "G": 0.29511, "chunks": 189.53, "bytes_per_chunk": 5.403, "boundary_entropy": 0.42875, "router_noise": 0.0, "grad_norm": 0.7702, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 125.0} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.54429, "bpb": 2.22794, "chunk_loss": 0.030601, "F": 0.19153, "G": 0.2802, "chunks": 196.12, "bytes_per_chunk": 5.221, "boundary_entropy": 0.39893, "router_noise": 0.0, "grad_norm": 0.7413, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 127.0} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.49961, "bpb": 2.16348, "chunk_loss": 0.03056, "F": 0.1879, "G": 0.29109, "chunks": 192.41, "bytes_per_chunk": 5.322, "boundary_entropy": 0.41671, "router_noise": 0.0, "grad_norm": 0.7041, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 129.0} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3389.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194737, "cps": 26.476562, "mean_p": 0.234437, "boundary_entropy": 0.29459, "bpb": 2.257266}, "de": {"chunks": 4082.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.198802, "cps": 31.890625, "mean_p": 0.245614, "boundary_entropy": 0.314177, "bpb": 2.509477}, "vi": {"chunks": 4241.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230052, "cps": 33.132812, "mean_p": 0.272507, "boundary_entropy": 0.438299, "bpb": 2.080978}, "tr": {"chunks": 3496.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.193117, "cps": 27.3125, "mean_p": 0.232873, "boundary_entropy": 0.324152, "bpb": 2.526698}, "fi": {"chunks": 3881.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.206315, "cps": 30.320312, "mean_p": 0.240822, "boundary_entropy": 0.311871, "bpb": 2.71346}, "ru": {"chunks": 5980.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.315567, "cps": 46.71875, "mean_p": 0.351721, "boundary_entropy": 0.543961, "bpb": 1.612575}, "el": {"chunks": 6469.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.315207, "cps": 50.539062, "mean_p": 0.294685, "boundary_entropy": 0.479917, "bpb": 1.682325}, "hi": {"chunks": 7778.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.462288, "cps": 60.765625, "mean_p": 0.292504, "boundary_entropy": 0.511175, "bpb": 1.313158}, "ta": {"chunks": 8299.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.443915, "cps": 64.835938, "mean_p": 0.319684, "boundary_entropy": 0.551635, "bpb": 1.066894}, "th": {"chunks": 6989.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.434504, "cps": 54.601562, "mean_p": 0.370141, "boundary_entropy": 0.616753, "bpb": 1.270771}, "zh": {"chunks": 3180.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.538435, "cps": 24.84375, "mean_p": 0.324325, "boundary_entropy": 0.44452, "bpb": 3.449126}, "my": {"chunks": 10585.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.558634, "cps": 82.695312, "mean_p": 0.413618, "boundary_entropy": 0.651029, "bpb": 1.202834}}, "bpb": {"en": 2.25727, "de": 2.50948, "vi": 2.08098, "tr": 2.5267, "fi": 2.71346, "ru": 1.61258, "el": 1.68233, "hi": 1.31316, "ta": 1.06689, "th": 1.27077, "zh": 3.44913, "my": 1.20283}, "gini_cpc": 0.219152, "gini_cps": 0.21947, "bpic": 5.613377, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0208763539313759, "vi": 1.181347541620439, "tr": 0.9916842685066912, "fi": 1.059459380689516, "ru": 1.6204831555762824, "el": 1.618634740485838, "hi": 2.373916380984366, "ta": 2.279569546424405, "th": 2.2312412307257117, "zh": 2.7649432932605906, "my": 2.8686663405111834}, "t": 132.0} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.26487, "de": 2.51434, "vi": 2.08362, "tr": 2.53181, "fi": 2.72571, "ru": 1.58944, "el": 1.67528, "hi": 1.31047, "ta": 1.07369, "th": 1.28805, "zh": 3.45156, "my": 1.21196}, "bytes_per_chunk": {"en": 5.099, "de": 4.983, "vi": 5.782, "tr": 5.347, "fi": 4.789, "ru": 5.951, "el": 5.823, "hi": 5.462, "ta": 5.822, "th": 6.588, "zh": 5.373, "my": 4.731}, "bpb_mean": 1.97673, "t": 133.3} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.44607, "bpb": 2.08624, "chunk_loss": 0.030543, "F": 0.19162, "G": 0.28527, "chunks": 196.22, "bytes_per_chunk": 5.219, "boundary_entropy": 0.40931, "router_noise": 0.0, "grad_norm": 0.6744, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 133.3} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.53114, "bpb": 2.20897, "chunk_loss": 0.030251, "F": 0.18015, "G": 0.27277, "chunks": 184.47, "bytes_per_chunk": 5.551, "boundary_entropy": 0.39248, "router_noise": 0.0, "grad_norm": 0.7392, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 135.3} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.46471, "bpb": 2.11313, "chunk_loss": 0.03041, "F": 0.18518, "G": 0.29163, "chunks": 189.62, "bytes_per_chunk": 5.4, "boundary_entropy": 0.42066, "router_noise": 0.0, "grad_norm": 0.6886, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 137.3} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.488, "bpb": 2.14673, "chunk_loss": 0.030177, "F": 0.18005, "G": 0.27896, "chunks": 184.38, "bytes_per_chunk": 5.554, "boundary_entropy": 0.40144, "router_noise": 0.0, "grad_norm": 0.6943, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 139.3} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3222.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.18514, "cps": 25.171875, "mean_p": 0.222012, "boundary_entropy": 0.280083, "bpb": 2.229879}, "de": {"chunks": 3960.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19286, "cps": 30.9375, "mean_p": 0.239114, "boundary_entropy": 0.298893, "bpb": 2.47464}, "vi": {"chunks": 4168.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226092, "cps": 32.5625, "mean_p": 0.264774, "boundary_entropy": 0.426296, "bpb": 2.070467}, "tr": {"chunks": 3660.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.202176, "cps": 28.59375, "mean_p": 0.240156, "boundary_entropy": 0.327739, "bpb": 2.483089}, "fi": {"chunks": 3904.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.207538, "cps": 30.5, "mean_p": 0.239283, "boundary_entropy": 0.304115, "bpb": 2.673876}, "ru": {"chunks": 5941.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.313509, "cps": 46.414062, "mean_p": 0.338631, "boundary_entropy": 0.522133, "bpb": 1.575052}, "el": {"chunks": 6174.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.300833, "cps": 48.234375, "mean_p": 0.281425, "boundary_entropy": 0.45879, "bpb": 1.663818}, "hi": {"chunks": 7785.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.462704, "cps": 60.820312, "mean_p": 0.308504, "boundary_entropy": 0.535904, "bpb": 1.293654}, "ta": {"chunks": 8753.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.4682, "cps": 68.382812, "mean_p": 0.35074, "boundary_entropy": 0.584383, "bpb": 1.048805}, "th": {"chunks": 7206.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.447995, "cps": 56.296875, "mean_p": 0.393414, "boundary_entropy": 0.629353, "bpb": 1.255286}, "zh": {"chunks": 2940.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.497799, "cps": 22.96875, "mean_p": 0.309335, "boundary_entropy": 0.433625, "bpb": 3.383078}, "my": {"chunks": 9240.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.48765, "cps": 72.1875, "mean_p": 0.413489, "boundary_entropy": 0.65389, "bpb": 1.182817}}, "bpb": {"en": 2.22988, "de": 2.47464, "vi": 2.07047, "tr": 2.48309, "fi": 2.67388, "ru": 1.57505, "el": 1.66382, "hi": 1.29365, "ta": 1.0488, "th": 1.25529, "zh": 3.38308, "my": 1.18282}, "gini_cpc": 0.20682, "gini_cps": 0.215175, "bpic": 5.732096, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0416968787402319, "vi": 1.2211897557425329, "tr": 1.0920162928642032, "fi": 1.1209765038028483, "ru": 1.693358539329707, "el": 1.624891489776969, "hi": 2.4992064216756456, "ta": 2.528890605490569, "th": 2.4197571212986873, "zh": 2.6887626823966535, "my": 2.6339478938535694}, "t": 142.3} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.23873, "de": 2.47862, "vi": 2.07378, "tr": 2.49036, "fi": 2.68572, "ru": 1.55136, "el": 1.65652, "hi": 1.29235, "ta": 1.05684, "th": 1.27163, "zh": 3.3888, "my": 1.19289}, "bytes_per_chunk": {"en": 5.348, "de": 5.132, "vi": 5.862, "tr": 5.178, "fi": 4.796, "ru": 5.888, "el": 6.202, "hi": 5.674, "ta": 5.753, "th": 6.39, "zh": 5.615, "my": 5.798}, "bpb_mean": 1.94813, "t": 143.5} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 143.5} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "global", "lm_loss": 1.36769, "bpb": 1.97316, "chunk_loss": 0.030335, "F": 0.18341, "G": 0.28111, "chunks": 187.81, "bytes_per_chunk": 5.452, "boundary_entropy": 0.4098, "router_noise": 0.0, "grad_norm": 0.7135, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 143.6} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "global", "lm_loss": 1.45893, "bpb": 2.10478, "chunk_loss": 0.030305, "F": 0.17941, "G": 0.29299, "chunks": 183.72, "bytes_per_chunk": 5.574, "boundary_entropy": 0.43451, "router_noise": 0.0, "grad_norm": 0.6672, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 145.6} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "global", "lm_loss": 1.49834, "bpb": 2.16165, "chunk_loss": 0.03027, "F": 0.18301, "G": 0.2746, "chunks": 187.41, "bytes_per_chunk": 5.464, "boundary_entropy": 0.38525, "router_noise": 0.0, "grad_norm": 0.7204, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 147.6} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "global", "lm_loss": 1.56036, "bpb": 2.25112, "chunk_loss": 0.030511, "F": 0.1918, "G": 0.27942, "chunks": 196.41, "bytes_per_chunk": 5.214, "boundary_entropy": 0.38494, "router_noise": 0.0, "grad_norm": 0.7048, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 149.6} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3365.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193357, "cps": 26.289062, "mean_p": 0.224677, "boundary_entropy": 0.269444, "bpb": 2.194311}, "de": {"chunks": 4076.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19851, "cps": 31.84375, "mean_p": 0.238936, "boundary_entropy": 0.287993, "bpb": 2.4352}, "vi": {"chunks": 4062.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.220342, "cps": 31.734375, "mean_p": 0.256813, "boundary_entropy": 0.415164, "bpb": 2.042616}, "tr": {"chunks": 3578.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.197647, "cps": 27.953125, "mean_p": 0.237946, "boundary_entropy": 0.32095, "bpb": 2.442414}, "fi": {"chunks": 3927.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.208761, "cps": 30.679688, "mean_p": 0.240786, "boundary_entropy": 0.297301, "bpb": 2.63619}, "ru": {"chunks": 5806.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306385, "cps": 45.359375, "mean_p": 0.340351, "boundary_entropy": 0.523151, "bpb": 1.562697}, "el": {"chunks": 6399.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.311797, "cps": 49.992188, "mean_p": 0.287911, "boundary_entropy": 0.458394, "bpb": 1.636257}, "hi": {"chunks": 6518.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.3874, "cps": 50.921875, "mean_p": 0.287782, "boundary_entropy": 0.519181, "bpb": 1.287936}, "ta": {"chunks": 8317.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.444878, "cps": 64.976562, "mean_p": 0.357021, "boundary_entropy": 0.586498, "bpb": 1.045684}, "th": {"chunks": 9048.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.562512, "cps": 70.6875, "mean_p": 0.391936, "boundary_entropy": 0.618671, "bpb": 1.226019}, "zh": {"chunks": 2909.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.49255, "cps": 22.726562, "mean_p": 0.29921, "boundary_entropy": 0.419134, "bpb": 3.353061}, "my": {"chunks": 8110.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.428014, "cps": 63.359375, "mean_p": 0.403978, "boundary_entropy": 0.651236, "bpb": 1.185776}}, "bpb": {"en": 2.19431, "de": 2.4352, "vi": 2.04262, "tr": 2.44241, "fi": 2.63619, "ru": 1.5627, "el": 1.63626, "hi": 1.28794, "ta": 1.04568, "th": 1.22602, "zh": 3.35306, "my": 1.18578}, "gini_cpc": 0.211747, "gini_cps": 0.208623, "bpic": 5.804749, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0266462373583525, "vi": 1.139556411003522, "tr": 1.0221834296549241, "fi": 1.0796626298213328, "ru": 1.584553360489597, "el": 1.6125393327053277, "hi": 2.003541464556255, "ta": 2.3008074959455116, "th": 2.9091799000573197, "zh": 2.5473541652305136, "my": 2.2135866644960274}, "t": 152.5} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.20235, "de": 2.4389, "vi": 2.04577, "tr": 2.44506, "fi": 2.64533, "ru": 1.53822, "el": 1.62964, "hi": 1.2837, "ta": 1.05281, "th": 1.24293, "zh": 3.35532, "my": 1.19451}, "bytes_per_chunk": {"en": 5.172, "de": 4.992, "vi": 6.116, "tr": 5.206, "fi": 4.773, "ru": 6.086, "el": 6.012, "hi": 6.077, "ta": 5.862, "th": 5.269, "zh": 5.642, "my": 6.252}, "bpb_mean": 1.92288, "t": 153.8} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "global", "lm_loss": 1.58858, "bpb": 2.29183, "chunk_loss": 0.03058, "F": 0.18799, "G": 0.27512, "chunks": 192.5, "bytes_per_chunk": 5.319, "boundary_entropy": 0.37893, "router_noise": 0.0, "grad_norm": 0.7427, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 153.9} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "global", "lm_loss": 1.49296, "bpb": 2.15388, "chunk_loss": 0.030762, "F": 0.19165, "G": 0.2838, "chunks": 196.25, "bytes_per_chunk": 5.218, "boundary_entropy": 0.39981, "router_noise": 0.0, "grad_norm": 0.6629, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 155.9} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "global", "lm_loss": 1.50345, "bpb": 2.16901, "chunk_loss": 0.030593, "F": 0.1922, "G": 0.28104, "chunks": 196.81, "bytes_per_chunk": 5.203, "boundary_entropy": 0.38117, "router_noise": 0.0, "grad_norm": 0.7032, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 157.9} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "global", "lm_loss": 1.39358, "bpb": 2.01052, "chunk_loss": 0.030207, "F": 0.18262, "G": 0.26909, "chunks": 187.0, "bytes_per_chunk": 5.476, "boundary_entropy": 0.36404, "router_noise": 0.0, "grad_norm": 0.7033, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 159.9} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3390.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194794, "cps": 26.484375, "mean_p": 0.228884, "boundary_entropy": 0.272761, "bpb": 2.183029}, "de": {"chunks": 4129.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201091, "cps": 32.257812, "mean_p": 0.242045, "boundary_entropy": 0.288519, "bpb": 2.406831}, "vi": {"chunks": 4214.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228587, "cps": 32.921875, "mean_p": 0.261334, "boundary_entropy": 0.424982, "bpb": 2.031827}, "tr": {"chunks": 3552.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.196211, "cps": 27.75, "mean_p": 0.235972, "boundary_entropy": 0.321486, "bpb": 2.42118}, "fi": {"chunks": 3714.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.197438, "cps": 29.015625, "mean_p": 0.234034, "boundary_entropy": 0.297698, "bpb": 2.623671}, "ru": {"chunks": 6089.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.321319, "cps": 47.570312, "mean_p": 0.347662, "boundary_entropy": 0.524979, "bpb": 1.525985}, "el": {"chunks": 6363.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.310042, "cps": 49.710938, "mean_p": 0.287391, "boundary_entropy": 0.455215, "bpb": 1.590364}, "hi": {"chunks": 6576.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.390847, "cps": 51.375, "mean_p": 0.292619, "boundary_entropy": 0.52236, "bpb": 1.277681}, "ta": {"chunks": 8232.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.440332, "cps": 64.3125, "mean_p": 0.341006, "boundary_entropy": 0.571581, "bpb": 1.052049}, "th": {"chunks": 7535.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.468449, "cps": 58.867188, "mean_p": 0.373451, "boundary_entropy": 0.611849, "bpb": 1.227258}, "zh": {"chunks": 2859.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.484084, "cps": 22.335938, "mean_p": 0.298479, "boundary_entropy": 0.416041, "bpb": 3.346542}, "my": {"chunks": 10707.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.565073, "cps": 83.648438, "mean_p": 0.443493, "boundary_entropy": 0.665597, "bpb": 1.166783}}, "bpb": {"en": 2.18303, "de": 2.40683, "vi": 2.03183, "tr": 2.42118, "fi": 2.62367, "ru": 1.52598, "el": 1.59036, "hi": 1.27768, "ta": 1.05205, "th": 1.22726, "zh": 3.34654, "my": 1.16678}, "gini_cpc": 0.213744, "gini_cps": 0.223518, "bpic": 5.697461, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0323260764346966, "vi": 1.1734803219155558, "tr": 1.0072721531022824, "fi": 1.013571611037225, "ru": 1.649533658673267, "el": 1.5916424007771983, "hi": 2.0064629909663507, "ta": 2.260498973981496, "th": 2.404842361239103, "zh": 2.4851072405743073, "my": 2.9008738867996433}, "t": 162.8} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.1912, "de": 2.41204, "vi": 2.03371, "tr": 2.42653, "fi": 2.63371, "ru": 1.50264, "el": 1.58366, "hi": 1.27209, "ta": 1.05965, "th": 1.24247, "zh": 3.34543, "my": 1.17513}, "bytes_per_chunk": {"en": 5.164, "de": 4.941, "vi": 5.604, "tr": 5.27, "fi": 4.972, "ru": 5.76, "el": 6.003, "hi": 6.236, "ta": 5.943, "th": 6.128, "zh": 5.709, "my": 4.978}, "bpb_mean": 1.90652, "t": 164.1} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "global", "lm_loss": 1.46181, "bpb": 2.10895, "chunk_loss": 0.030436, "F": 0.18665, "G": 0.28367, "chunks": 191.12, "bytes_per_chunk": 5.358, "boundary_entropy": 0.38641, "router_noise": 0.0, "grad_norm": 0.7857, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 164.2} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "global", "lm_loss": 1.4286, "bpb": 2.06103, "chunk_loss": 0.030376, "F": 0.18509, "G": 0.27246, "chunks": 189.53, "bytes_per_chunk": 5.403, "boundary_entropy": 0.38483, "router_noise": 0.0, "grad_norm": 0.6792, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 166.2} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "global", "lm_loss": 1.44444, "bpb": 2.08388, "chunk_loss": 0.030336, "F": 0.18637, "G": 0.27568, "chunks": 190.84, "bytes_per_chunk": 5.366, "boundary_entropy": 0.38495, "router_noise": 0.0, "grad_norm": 0.7347, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 168.2} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "global", "lm_loss": 1.34439, "bpb": 1.93954, "chunk_loss": 0.030145, "F": 0.1777, "G": 0.28183, "chunks": 181.97, "bytes_per_chunk": 5.627, "boundary_entropy": 0.39908, "router_noise": 0.0, "grad_norm": 0.6472, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 170.2} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3330.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.191346, "cps": 26.015625, "mean_p": 0.222122, "boundary_entropy": 0.25728, "bpb": 2.154272}, "de": {"chunks": 4038.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196659, "cps": 31.546875, "mean_p": 0.233263, "boundary_entropy": 0.271915, "bpb": 2.382059}, "vi": {"chunks": 3914.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.212314, "cps": 30.578125, "mean_p": 0.250747, "boundary_entropy": 0.407426, "bpb": 2.013508}, "tr": {"chunks": 3523.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.194609, "cps": 27.523438, "mean_p": 0.231815, "boundary_entropy": 0.308579, "bpb": 2.384514}, "fi": {"chunks": 3943.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.209611, "cps": 30.804688, "mean_p": 0.238899, "boundary_entropy": 0.288246, "bpb": 2.573663}, "ru": {"chunks": 5250.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.277045, "cps": 41.015625, "mean_p": 0.304482, "boundary_entropy": 0.483584, "bpb": 1.509596}, "el": {"chunks": 6526.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.317985, "cps": 50.984375, "mean_p": 0.27983, "boundary_entropy": 0.435734, "bpb": 1.557912}, "hi": {"chunks": 7028.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.417712, "cps": 54.90625, "mean_p": 0.305672, "boundary_entropy": 0.536293, "bpb": 1.246369}, "ta": {"chunks": 8122.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.434448, "cps": 63.453125, "mean_p": 0.336506, "boundary_entropy": 0.562874, "bpb": 1.02533}, "th": {"chunks": 7724.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.480199, "cps": 60.34375, "mean_p": 0.383589, "boundary_entropy": 0.616604, "bpb": 1.207739}, "zh": {"chunks": 2856.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.483576, "cps": 22.3125, "mean_p": 0.295521, "boundary_entropy": 0.408635, "bpb": 3.323538}, "my": {"chunks": 7205.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.380251, "cps": 56.289062, "mean_p": 0.411045, "boundary_entropy": 0.653347, "bpb": 1.165392}}, "bpb": {"en": 2.15427, "de": 2.38206, "vi": 2.01351, "tr": 2.38451, "fi": 2.57366, "ru": 1.5096, "el": 1.55791, "hi": 1.24637, "ta": 1.02533, "th": 1.20774, "zh": 3.32354, "my": 1.16539}, "gini_cpc": 0.198349, "gini_cps": 0.194058, "bpic": 6.0477, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0277649294938538, "vi": 1.109577307168845, "tr": 1.0170492372724047, "fi": 1.0954556012607153, "ru": 1.4478713542038082, "el": 1.6618281488622082, "hi": 2.1830142296828776, "ta": 2.270478544872768, "th": 2.5095802423499034, "zh": 2.5272292949908937, "my": 1.9872408031844382}, "t": 173.2} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.16312, "de": 2.38534, "vi": 2.01275, "tr": 2.39294, "fi": 2.58111, "ru": 1.48696, "el": 1.54971, "hi": 1.24214, "ta": 1.03272, "th": 1.22676, "zh": 3.32261, "my": 1.17468}, "bytes_per_chunk": {"en": 5.25, "de": 5.025, "vi": 6.068, "tr": 5.321, "fi": 4.727, "ru": 6.502, "el": 5.8, "hi": 5.923, "ta": 6.089, "th": 5.83, "zh": 5.718, "my": 7.711}, "bpb_mean": 1.8809, "t": 174.5} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "global", "lm_loss": 1.42522, "bpb": 2.05616, "chunk_loss": 0.03011, "F": 0.17673, "G": 0.27655, "chunks": 180.97, "bytes_per_chunk": 5.658, "boundary_entropy": 0.39526, "router_noise": 0.0, "grad_norm": 0.6926, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 174.5} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "global", "lm_loss": 1.38377, "bpb": 1.99636, "chunk_loss": 0.030222, "F": 0.18173, "G": 0.28281, "chunks": 186.09, "bytes_per_chunk": 5.503, "boundary_entropy": 0.39178, "router_noise": 0.0, "grad_norm": 0.6309, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 176.5} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "global", "lm_loss": 1.39885, "bpb": 2.01812, "chunk_loss": 0.03025, "F": 0.18262, "G": 0.26699, "chunks": 187.0, "bytes_per_chunk": 5.476, "boundary_entropy": 0.36546, "router_noise": 0.0, "grad_norm": 0.7552, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 178.5} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "global", "lm_loss": 1.33777, "bpb": 1.93, "chunk_loss": 0.030201, "F": 0.17969, "G": 0.28301, "chunks": 184.0, "bytes_per_chunk": 5.565, "boundary_entropy": 0.39655, "router_noise": 0.0, "grad_norm": 0.6668, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 180.6} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3411.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196001, "cps": 26.648438, "mean_p": 0.227157, "boundary_entropy": 0.256101, "bpb": 2.13611}, "de": {"chunks": 4202.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.204646, "cps": 32.828125, "mean_p": 0.245418, "boundary_entropy": 0.278373, "bpb": 2.350394}, "vi": {"chunks": 4168.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226092, "cps": 32.5625, "mean_p": 0.26142, "boundary_entropy": 0.409642, "bpb": 1.999052}, "tr": {"chunks": 3637.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200906, "cps": 28.414062, "mean_p": 0.23657, "boundary_entropy": 0.30516, "bpb": 2.357854}, "fi": {"chunks": 4145.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.22035, "cps": 32.382812, "mean_p": 0.249702, "boundary_entropy": 0.295607, "bpb": 2.538244}, "ru": {"chunks": 5656.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.29847, "cps": 44.1875, "mean_p": 0.322349, "boundary_entropy": 0.494083, "bpb": 1.478835}, "el": {"chunks": 6968.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.339522, "cps": 54.4375, "mean_p": 0.29162, "boundary_entropy": 0.435137, "bpb": 1.529402}, "hi": {"chunks": 6937.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.412303, "cps": 54.195312, "mean_p": 0.310539, "boundary_entropy": 0.543821, "bpb": 1.243605}, "ta": {"chunks": 8524.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.455951, "cps": 66.59375, "mean_p": 0.350468, "boundary_entropy": 0.576381, "bpb": 1.007133}, "th": {"chunks": 7699.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.478645, "cps": 60.148438, "mean_p": 0.387217, "boundary_entropy": 0.621613, "bpb": 1.19956}, "zh": {"chunks": 2885.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.488486, "cps": 22.539062, "mean_p": 0.290925, "boundary_entropy": 0.397051, "bpb": 3.280197}, "my": {"chunks": 9668.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.510239, "cps": 75.53125, "mean_p": 0.443691, "boundary_entropy": 0.665735, "bpb": 1.145357}}, "bpb": {"en": 2.13611, "de": 2.35039, "vi": 1.99905, "tr": 2.35785, "fi": 2.53824, "ru": 1.47883, "el": 1.5294, "hi": 1.2436, "ta": 1.00713, "th": 1.19956, "zh": 3.2802, "my": 1.14536}, "gini_cpc": 0.201337, "gini_cps": 0.212248, "bpic": 5.65215, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0441094865801757, "vi": 1.153524888010097, "tr": 1.0250266347006534, "fi": 1.1242296945681074, "ru": 1.522799015137275, "el": 1.7322465204214967, "hi": 2.1035799482750286, "ta": 2.3262713516964335, "th": 2.442056204819985, "zh": 2.492268197063285, "my": 2.6032487375994924}, "t": 183.6} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.14497, "de": 2.35431, "vi": 2.00018, "tr": 2.36397, "fi": 2.54952, "ru": 1.45458, "el": 1.52411, "hi": 1.2379, "ta": 1.01542, "th": 1.21871, "zh": 3.28111, "my": 1.15354}, "bytes_per_chunk": {"en": 5.124, "de": 4.895, "vi": 6.012, "tr": 5.191, "fi": 4.521, "ru": 6.129, "el": 5.483, "hi": 5.919, "ta": 5.752, "th": 5.971, "zh": 5.664, "my": 5.71}, "bpb_mean": 1.85819, "t": 184.9} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "global", "lm_loss": 1.36486, "bpb": 1.96908, "chunk_loss": 0.03026, "F": 0.17899, "G": 0.28348, "chunks": 183.28, "bytes_per_chunk": 5.587, "boundary_entropy": 0.40567, "router_noise": 0.0, "grad_norm": 0.6357, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 184.9} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "global", "lm_loss": 1.43391, "bpb": 2.0687, "chunk_loss": 0.030278, "F": 0.18451, "G": 0.27358, "chunks": 188.94, "bytes_per_chunk": 5.42, "boundary_entropy": 0.3735, "router_noise": 0.0, "grad_norm": 0.6484, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 187.0} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "global", "lm_loss": 1.3964, "bpb": 2.01458, "chunk_loss": 0.030247, "F": 0.18338, "G": 0.26305, "chunks": 187.78, "bytes_per_chunk": 5.453, "boundary_entropy": 0.35425, "router_noise": 0.0, "grad_norm": 0.7033, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 189.1} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "global", "lm_loss": 1.46025, "bpb": 2.10669, "chunk_loss": 0.030287, "F": 0.18314, "G": 0.27182, "chunks": 187.53, "bytes_per_chunk": 5.46, "boundary_entropy": 0.36963, "router_noise": 0.0, "grad_norm": 0.6775, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 191.1} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3361.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193128, "cps": 26.257812, "mean_p": 0.223406, "boundary_entropy": 0.251075, "bpb": 2.12072}, "de": {"chunks": 4116.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.200458, "cps": 32.15625, "mean_p": 0.235909, "boundary_entropy": 0.265554, "bpb": 2.325319}, "vi": {"chunks": 4026.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.218389, "cps": 31.453125, "mean_p": 0.258426, "boundary_entropy": 0.406379, "bpb": 1.980449}, "tr": {"chunks": 3553.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.196266, "cps": 27.757812, "mean_p": 0.226083, "boundary_entropy": 0.292497, "bpb": 2.332034}, "fi": {"chunks": 3833.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.203764, "cps": 29.945312, "mean_p": 0.232752, "boundary_entropy": 0.279707, "bpb": 2.530809}, "ru": {"chunks": 5524.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.291504, "cps": 43.15625, "mean_p": 0.314082, "boundary_entropy": 0.483841, "bpb": 1.463182}, "el": {"chunks": 6385.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.311114, "cps": 49.882812, "mean_p": 0.273884, "boundary_entropy": 0.415945, "bpb": 1.529851}, "hi": {"chunks": 7622.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.453016, "cps": 59.546875, "mean_p": 0.303376, "boundary_entropy": 0.52128, "bpb": 1.225744}, "ta": {"chunks": 8587.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.459321, "cps": 67.085938, "mean_p": 0.331966, "boundary_entropy": 0.551085, "bpb": 1.00522}, "th": {"chunks": 8311.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.516693, "cps": 64.929688, "mean_p": 0.382683, "boundary_entropy": 0.613663, "bpb": 1.172761}, "zh": {"chunks": 2847.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482052, "cps": 22.242188, "mean_p": 0.284219, "boundary_entropy": 0.389285, "bpb": 3.254937}, "my": {"chunks": 7660.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.404264, "cps": 59.84375, "mean_p": 0.424877, "boundary_entropy": 0.662071, "bpb": 1.13616}}, "bpb": {"en": 2.12072, "de": 2.32532, "vi": 1.98045, "tr": 2.33203, "fi": 2.53081, "ru": 1.46318, "el": 1.52985, "hi": 1.22574, "ta": 1.00522, "th": 1.17276, "zh": 3.25494, "my": 1.13616}, "gini_cpc": 0.207102, "gini_cps": 0.207455, "bpic": 5.830323, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.03795509881822, "vi": 1.1308011365713397, "tr": 1.0162493101435843, "fi": 1.0550730831920798, "ru": 1.5093851167238046, "el": 1.6109262714557606, "hi": 2.345683858152667, "ta": 2.3783271910805466, "th": 2.6753944683744133, "zh": 2.496029030835464, "my": 2.093249524827403}, "t": 194.1} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.12888, "de": 2.33058, "vi": 1.98109, "tr": 2.33855, "fi": 2.54147, "ru": 1.43987, "el": 1.52292, "hi": 1.21982, "ta": 1.01439, "th": 1.19057, "zh": 3.25661, "my": 1.14545}, "bytes_per_chunk": {"en": 5.169, "de": 4.999, "vi": 6.042, "tr": 5.347, "fi": 4.816, "ru": 6.256, "el": 5.948, "hi": 5.313, "ta": 5.738, "th": 5.525, "zh": 5.717, "my": 6.664}, "bpb_mean": 1.84252, "t": 195.4} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "global", "lm_loss": 1.3856, "bpb": 1.999, "chunk_loss": 0.030245, "F": 0.17874, "G": 0.28138, "chunks": 183.03, "bytes_per_chunk": 5.595, "boundary_entropy": 0.3972, "router_noise": 0.0, "grad_norm": 0.6742, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 195.4} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "global", "lm_loss": 1.24492, "bpb": 1.79603, "chunk_loss": 0.02996, "F": 0.17441, "G": 0.26986, "chunks": 178.59, "bytes_per_chunk": 5.734, "boundary_entropy": 0.37822, "router_noise": 0.0, "grad_norm": 0.6403, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 197.5} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "global", "lm_loss": 1.43651, "bpb": 2.07244, "chunk_loss": 0.030292, "F": 0.18668, "G": 0.27243, "chunks": 191.16, "bytes_per_chunk": 5.357, "boundary_entropy": 0.36007, "router_noise": 0.0, "grad_norm": 0.7568, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 199.5} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "global", "lm_loss": 1.36976, "bpb": 1.97614, "chunk_loss": 0.030315, "F": 0.18176, "G": 0.27819, "chunks": 186.12, "bytes_per_chunk": 5.502, "boundary_entropy": 0.38724, "router_noise": 0.0, "grad_norm": 0.6765, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 201.5} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3358.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.192955, "cps": 26.234375, "mean_p": 0.221521, "boundary_entropy": 0.241873, "bpb": 2.10607}, "de": {"chunks": 4087.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.199045, "cps": 31.929688, "mean_p": 0.235288, "boundary_entropy": 0.258351, "bpb": 2.315456}, "vi": {"chunks": 3987.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.216273, "cps": 31.148438, "mean_p": 0.255351, "boundary_entropy": 0.397184, "bpb": 1.983309}, "tr": {"chunks": 3647.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201458, "cps": 28.492188, "mean_p": 0.231261, "boundary_entropy": 0.290043, "bpb": 2.310309}, "fi": {"chunks": 3855.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.204933, "cps": 30.117188, "mean_p": 0.234041, "boundary_entropy": 0.273309, "bpb": 2.48256}, "ru": {"chunks": 5339.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.281741, "cps": 41.710938, "mean_p": 0.304782, "boundary_entropy": 0.470333, "bpb": 1.450971}, "el": {"chunks": 6669.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.324952, "cps": 52.101562, "mean_p": 0.274524, "boundary_entropy": 0.410446, "bpb": 1.50103}, "hi": {"chunks": 5753.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.341932, "cps": 44.945312, "mean_p": 0.318834, "boundary_entropy": 0.559788, "bpb": 1.228823}, "ta": {"chunks": 7822.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.418401, "cps": 61.109375, "mean_p": 0.340463, "boundary_entropy": 0.562633, "bpb": 0.995682}, "th": {"chunks": 7023.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.436618, "cps": 54.867188, "mean_p": 0.372811, "boundary_entropy": 0.610961, "bpb": 1.168177}, "zh": {"chunks": 2835.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.48002, "cps": 22.148438, "mean_p": 0.291359, "boundary_entropy": 0.398952, "bpb": 3.257751}, "my": {"chunks": 7035.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.371279, "cps": 54.960938, "mean_p": 0.407415, "boundary_entropy": 0.654589, "bpb": 1.134187}}, "bpb": {"en": 2.10607, "de": 2.31546, "vi": 1.98331, "tr": 2.31031, "fi": 2.48256, "ru": 1.45097, "el": 1.50103, "hi": 1.22882, "ta": 0.99568, "th": 1.16818, "zh": 3.25775, "my": 1.13419}, "gini_cpc": 0.18318, "gini_cps": 0.179398, "bpic": 6.249487, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0315627682476416, "vi": 1.120847486129089, "tr": 1.0440676517934548, "fi": 1.0620768125451163, "ru": 1.4601387777936672, "el": 1.6840822584704789, "hi": 1.7720775739468497, "ta": 2.1683818852553154, "th": 2.262794068067886, "zh": 2.4877288861611344, "my": 1.9241731673268228}, "t": 204.5} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.11607, "de": 2.32123, "vi": 1.98489, "tr": 2.31687, "fi": 2.49556, "ru": 1.42615, "el": 1.49471, "hi": 1.22551, "ta": 1.00262, "th": 1.18617, "zh": 3.25765, "my": 1.14313}, "bytes_per_chunk": {"en": 5.178, "de": 4.996, "vi": 6.114, "tr": 5.118, "fi": 4.809, "ru": 6.447, "el": 5.681, "hi": 7.164, "ta": 6.298, "th": 6.478, "zh": 5.697, "my": 7.383}, "bpb_mean": 1.83088, "t": 205.8} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "global", "lm_loss": 1.62902, "bpb": 2.35018, "chunk_loss": 0.030457, "F": 0.18842, "G": 0.27489, "chunks": 192.94, "bytes_per_chunk": 5.307, "boundary_entropy": 0.35887, "router_noise": 0.0, "grad_norm": 0.7299, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 205.8} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "global", "lm_loss": 1.42799, "bpb": 2.06016, "chunk_loss": 0.030336, "F": 0.18588, "G": 0.27617, "chunks": 190.34, "bytes_per_chunk": 5.38, "boundary_entropy": 0.36757, "router_noise": 0.0, "grad_norm": 0.7465, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 207.8} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "global", "lm_loss": 1.33858, "bpb": 1.93116, "chunk_loss": 0.030365, "F": 0.18469, "G": 0.28412, "chunks": 189.12, "bytes_per_chunk": 5.414, "boundary_entropy": 0.39392, "router_noise": 0.0, "grad_norm": 0.7012, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 209.8} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "global", "lm_loss": 1.38742, "bpb": 2.00162, "chunk_loss": 0.030302, "F": 0.18304, "G": 0.27887, "chunks": 187.44, "bytes_per_chunk": 5.463, "boundary_entropy": 0.3742, "router_noise": 0.0, "grad_norm": 0.6791, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 211.8} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3358.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.192955, "cps": 26.234375, "mean_p": 0.220349, "boundary_entropy": 0.236909, "bpb": 2.096657}, "de": {"chunks": 4044.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196951, "cps": 31.59375, "mean_p": 0.233293, "boundary_entropy": 0.257361, "bpb": 2.293688}, "vi": {"chunks": 4246.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230323, "cps": 33.171875, "mean_p": 0.266438, "boundary_entropy": 0.405558, "bpb": 1.964766}, "tr": {"chunks": 3592.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.19842, "cps": 28.0625, "mean_p": 0.228456, "boundary_entropy": 0.286462, "bpb": 2.294733}, "fi": {"chunks": 3876.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.20605, "cps": 30.28125, "mean_p": 0.231796, "boundary_entropy": 0.269152, "bpb": 2.469718}, "ru": {"chunks": 5598.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.295409, "cps": 43.734375, "mean_p": 0.307123, "boundary_entropy": 0.467153, "bpb": 1.430049}, "el": {"chunks": 6372.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.310481, "cps": 49.78125, "mean_p": 0.270934, "boundary_entropy": 0.412818, "bpb": 1.482869}, "hi": {"chunks": 7135.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.424071, "cps": 55.742188, "mean_p": 0.352417, "boundary_entropy": 0.582929, "bpb": 1.202111}, "ta": {"chunks": 8952.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.478845, "cps": 69.9375, "mean_p": 0.357239, "boundary_entropy": 0.571996, "bpb": 0.979748}, "th": {"chunks": 8378.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.520858, "cps": 65.453125, "mean_p": 0.386603, "boundary_entropy": 0.616675, "bpb": 1.147022}, "zh": {"chunks": 2839.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.480698, "cps": 22.179688, "mean_p": 0.283774, "boundary_entropy": 0.385061, "bpb": 3.217691}, "my": {"chunks": 8028.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.423686, "cps": 62.71875, "mean_p": 0.439161, "boundary_entropy": 0.666552, "bpb": 1.129803}}, "bpb": {"en": 2.09666, "de": 2.29369, "vi": 1.96477, "tr": 2.29473, "fi": 2.46972, "ru": 1.43005, "el": 1.48287, "hi": 1.20211, "ta": 0.97975, "th": 1.14702, "zh": 3.21769, "my": 1.1298}, "gini_cpc": 0.206481, "gini_cps": 0.210919, "bpic": 5.778268, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0207095264970547, "vi": 1.1936589982704067, "tr": 1.028322184053219, "fi": 1.0678624449869962, "ru": 1.5309715074150494, "el": 1.6090826437207812, "hi": 2.197770465863162, "ta": 2.481635724470159, "th": 2.699371878438381, "zh": 2.491238909280938, "my": 2.1957728766595213}, "t": 214.8} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.10594, "de": 2.29757, "vi": 1.96566, "tr": 2.2997, "fi": 2.48035, "ru": 1.40739, "el": 1.47537, "hi": 1.19688, "ta": 0.98833, "th": 1.16502, "zh": 3.21729, "my": 1.13813}, "bytes_per_chunk": {"en": 5.174, "de": 5.03, "vi": 5.908, "tr": 5.229, "fi": 4.818, "ru": 6.148, "el": 5.947, "hi": 5.889, "ta": 5.563, "th": 5.676, "zh": 5.747, "my": 6.472}, "bpb_mean": 1.81147, "t": 216.1} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "global", "lm_loss": 1.4582, "bpb": 2.10373, "chunk_loss": 0.030384, "F": 0.18222, "G": 0.2792, "chunks": 186.59, "bytes_per_chunk": 5.488, "boundary_entropy": 0.38536, "router_noise": 0.0, "grad_norm": 0.6892, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 216.1} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "global", "lm_loss": 1.45863, "bpb": 2.10436, "chunk_loss": 0.030416, "F": 0.18478, "G": 0.28176, "chunks": 189.22, "bytes_per_chunk": 5.412, "boundary_entropy": 0.38354, "router_noise": 0.0, "grad_norm": 0.6884, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 218.1} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "global", "lm_loss": 1.46431, "bpb": 2.11255, "chunk_loss": 0.030228, "F": 0.18619, "G": 0.25792, "chunks": 190.66, "bytes_per_chunk": 5.371, "boundary_entropy": 0.32724, "router_noise": 0.0, "grad_norm": 0.7597, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 220.2} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "global", "lm_loss": 1.53458, "bpb": 2.21393, "chunk_loss": 0.030641, "F": 0.18909, "G": 0.2897, "chunks": 193.62, "bytes_per_chunk": 5.289, "boundary_entropy": 0.39077, "router_noise": 0.0, "grad_norm": 0.6988, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 222.2} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3372.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19376, "cps": 26.34375, "mean_p": 0.220959, "boundary_entropy": 0.235299, "bpb": 2.079328}, "de": {"chunks": 4123.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.200799, "cps": 32.210938, "mean_p": 0.23535, "boundary_entropy": 0.250983, "bpb": 2.281776}, "vi": {"chunks": 3826.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.20754, "cps": 29.890625, "mean_p": 0.25673, "boundary_entropy": 0.403417, "bpb": 1.96013}, "tr": {"chunks": 3536.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195327, "cps": 27.625, "mean_p": 0.226215, "boundary_entropy": 0.279695, "bpb": 2.280724}, "fi": {"chunks": 3959.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.210462, "cps": 30.929688, "mean_p": 0.237343, "boundary_entropy": 0.268032, "bpb": 2.442533}, "ru": {"chunks": 5648.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.298047, "cps": 44.125, "mean_p": 0.309705, "boundary_entropy": 0.470134, "bpb": 1.416315}, "el": {"chunks": 6831.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332846, "cps": 53.367188, "mean_p": 0.279653, "boundary_entropy": 0.414524, "bpb": 1.459672}, "hi": {"chunks": 6111.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.36321, "cps": 47.742188, "mean_p": 0.313995, "boundary_entropy": 0.551735, "bpb": 1.199126}, "ta": {"chunks": 8015.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.428724, "cps": 62.617188, "mean_p": 0.342476, "boundary_entropy": 0.558421, "bpb": 0.972469}, "th": {"chunks": 7252.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.450855, "cps": 56.65625, "mean_p": 0.382675, "boundary_entropy": 0.616644, "bpb": 1.138238}, "zh": {"chunks": 2905.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.491873, "cps": 22.695312, "mean_p": 0.29166, "boundary_entropy": 0.390044, "bpb": 3.199451}, "my": {"chunks": 7614.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.401837, "cps": 59.484375, "mean_p": 0.417071, "boundary_entropy": 0.656365, "bpb": 1.121171}}, "bpb": {"en": 2.07933, "de": 2.28178, "vi": 1.96013, "tr": 2.28072, "fi": 2.44253, "ru": 1.41632, "el": 1.45967, "hi": 1.19913, "ta": 0.97247, "th": 1.13824, "zh": 3.19945, "my": 1.12117}, "gini_cpc": 0.190692, "gini_cps": 0.189114, "bpic": 6.073253, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0363285955963952, "vi": 1.0711206151844463, "tr": 1.0080875685062718, "fi": 1.0862009356606106, "ru": 1.5382326594615912, "el": 1.7178292543232894, "hi": 1.8745359123910044, "ta": 2.2126596260061544, "th": 2.3268762368954077, "zh": 2.5385706133939427, "my": 2.0738915909208724}, "t": 225.1} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.08978, "de": 2.28668, "vi": 1.96006, "tr": 2.28889, "fi": 2.4547, "ru": 1.39229, "el": 1.45212, "hi": 1.19517, "ta": 0.98287, "th": 1.1556, "zh": 3.19763, "my": 1.12932}, "bytes_per_chunk": {"en": 5.195, "de": 4.969, "vi": 6.439, "tr": 5.346, "fi": 4.698, "ru": 6.064, "el": 5.544, "hi": 6.682, "ta": 6.174, "th": 6.314, "zh": 5.641, "my": 6.727}, "bpb_mean": 1.79876, "t": 226.4} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "global", "lm_loss": 1.33214, "bpb": 1.92187, "chunk_loss": 0.030181, "F": 0.18039, "G": 0.25765, "chunks": 184.72, "bytes_per_chunk": 5.544, "boundary_entropy": 0.3433, "router_noise": 0.0, "grad_norm": 0.705, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 226.4} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "global", "lm_loss": 1.52517, "bpb": 2.20036, "chunk_loss": 0.030363, "F": 0.18469, "G": 0.26902, "chunks": 189.12, "bytes_per_chunk": 5.414, "boundary_entropy": 0.35399, "router_noise": 0.0, "grad_norm": 0.713, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 228.5} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "global", "lm_loss": 1.39227, "bpb": 2.00861, "chunk_loss": 0.030351, "F": 0.18814, "G": 0.25857, "chunks": 192.66, "bytes_per_chunk": 5.315, "boundary_entropy": 0.33645, "router_noise": 0.0, "grad_norm": 0.7053, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 230.5} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "global", "lm_loss": 1.37874, "bpb": 1.9891, "chunk_loss": 0.030253, "F": 0.18585, "G": 0.25743, "chunks": 190.31, "bytes_per_chunk": 5.381, "boundary_entropy": 0.32125, "router_noise": 0.0, "grad_norm": 0.7407, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 232.5} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3395.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195081, "cps": 26.523438, "mean_p": 0.221612, "boundary_entropy": 0.230233, "bpb": 2.064015}, "de": {"chunks": 4132.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201237, "cps": 32.28125, "mean_p": 0.234638, "boundary_entropy": 0.246509, "bpb": 2.25931}, "vi": {"chunks": 3880.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.210469, "cps": 30.3125, "mean_p": 0.261146, "boundary_entropy": 0.404864, "bpb": 1.946541}, "tr": {"chunks": 3511.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.193946, "cps": 27.429688, "mean_p": 0.22442, "boundary_entropy": 0.281203, "bpb": 2.25771}, "fi": {"chunks": 3990.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.21211, "cps": 31.171875, "mean_p": 0.237524, "boundary_entropy": 0.270784, "bpb": 2.425848}, "ru": {"chunks": 6042.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.318839, "cps": 47.203125, "mean_p": 0.3172, "boundary_entropy": 0.467608, "bpb": 1.403721}, "el": {"chunks": 6956.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.338937, "cps": 54.34375, "mean_p": 0.282285, "boundary_entropy": 0.402375, "bpb": 1.44458}, "hi": {"chunks": 7156.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.425319, "cps": 55.90625, "mean_p": 0.323817, "boundary_entropy": 0.552583, "bpb": 1.179753}, "ta": {"chunks": 7312.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.391121, "cps": 57.125, "mean_p": 0.331185, "boundary_entropy": 0.550631, "bpb": 0.966477}, "th": {"chunks": 7591.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.47193, "cps": 59.304688, "mean_p": 0.381209, "boundary_entropy": 0.61476, "bpb": 1.122316}, "zh": {"chunks": 2870.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.485946, "cps": 22.421875, "mean_p": 0.287726, "boundary_entropy": 0.386043, "bpb": 3.196977}, "my": {"chunks": 8772.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.462951, "cps": 68.53125, "mean_p": 0.443034, "boundary_entropy": 0.6675, "bpb": 1.119157}}, "bpb": {"en": 2.06401, "de": 2.25931, "vi": 1.94654, "tr": 2.25771, "fi": 2.42585, "ru": 1.40372, "el": 1.44458, "hi": 1.17975, "ta": 0.96648, "th": 1.12232, "zh": 3.19698, "my": 1.11916}, "gini_cpc": 0.195481, "gini_cps": 0.198761, "bpic": 5.849696, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0315546646524039, "vi": 1.0788794606532606, "tr": 0.9941790785292831, "fi": 1.0872899000201133, "ru": 1.6343905712653637, "el": 1.7374130120328377, "hi": 2.180216391993295, "ta": 2.004910797741121, "th": 2.4191470478766384, "zh": 2.4909946550947666, "my": 2.3731193938110753}, "t": 235.5} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.07514, "de": 2.26568, "vi": 1.94746, "tr": 2.26479, "fi": 2.43749, "ru": 1.37957, "el": 1.43677, "hi": 1.17435, "ta": 0.97575, "th": 1.13963, "zh": 3.19292, "my": 1.12724}, "bytes_per_chunk": {"en": 5.095, "de": 4.969, "vi": 6.444, "tr": 5.408, "fi": 4.68, "ru": 5.764, "el": 5.389, "hi": 5.756, "ta": 6.66, "th": 6.112, "zh": 5.661, "my": 5.911}, "bpb_mean": 1.78473, "t": 236.8} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 236.8} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.18051, "bpb": 1.70312, "chunk_loss": 0.030341, "F": 0.18167, "G": 0.29736, "chunks": 186.03, "bytes_per_chunk": 5.504, "boundary_entropy": 0.42028, "router_noise": 0.0, "grad_norm": 0.6309, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 236.8} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.23951, "bpb": 1.78824, "chunk_loss": 0.030407, "F": 0.1861, "G": 0.27535, "chunks": 190.56, "bytes_per_chunk": 5.374, "boundary_entropy": 0.37226, "router_noise": 0.0, "grad_norm": 0.6946, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 238.9} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.4759, "bpb": 2.12928, "chunk_loss": 0.030316, "F": 0.18719, "G": 0.26203, "chunks": 191.69, "bytes_per_chunk": 5.342, "boundary_entropy": 0.33218, "router_noise": 0.0, "grad_norm": 0.7224, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 240.9} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.45898, "bpb": 2.10487, "chunk_loss": 0.030355, "F": 0.18759, "G": 0.26184, "chunks": 192.09, "bytes_per_chunk": 5.331, "boundary_entropy": 0.3326, "router_noise": 0.0, "grad_norm": 0.7558, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 242.9} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3369.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193587, "cps": 26.320312, "mean_p": 0.221901, "boundary_entropy": 0.23061, "bpb": 2.054636}, "de": {"chunks": 4122.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.20075, "cps": 32.203125, "mean_p": 0.236331, "boundary_entropy": 0.247439, "bpb": 2.250712}, "vi": {"chunks": 3988.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.216328, "cps": 31.15625, "mean_p": 0.26619, "boundary_entropy": 0.401669, "bpb": 1.939985}, "tr": {"chunks": 3586.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.198089, "cps": 28.015625, "mean_p": 0.228592, "boundary_entropy": 0.281307, "bpb": 2.237527}, "fi": {"chunks": 4001.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.212695, "cps": 31.257812, "mean_p": 0.238271, "boundary_entropy": 0.266729, "bpb": 2.407938}, "ru": {"chunks": 5785.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305277, "cps": 45.195312, "mean_p": 0.307416, "boundary_entropy": 0.45649, "bpb": 1.388461}, "el": {"chunks": 6793.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.330994, "cps": 53.070312, "mean_p": 0.27716, "boundary_entropy": 0.396456, "bpb": 1.432488}, "hi": {"chunks": 7922.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.470847, "cps": 61.890625, "mean_p": 0.31542, "boundary_entropy": 0.531599, "bpb": 1.163539}, "ta": {"chunks": 9206.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.492431, "cps": 71.921875, "mean_p": 0.362767, "boundary_entropy": 0.569921, "bpb": 0.951296}, "th": {"chunks": 8243.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.512465, "cps": 64.398438, "mean_p": 0.387165, "boundary_entropy": 0.614444, "bpb": 1.112319}, "zh": {"chunks": 2832.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.479512, "cps": 22.125, "mean_p": 0.284323, "boundary_entropy": 0.382224, "bpb": 3.174916}, "my": {"chunks": 10223.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.539529, "cps": 79.867188, "mean_p": 0.44592, "boundary_entropy": 0.667366, "bpb": 1.104123}}, "bpb": {"en": 2.05464, "de": 2.25071, "vi": 1.93999, "tr": 2.23753, "fi": 2.40794, "ru": 1.38846, "el": 1.43249, "hi": 1.16354, "ta": 0.9513, "th": 1.11232, "zh": 3.17492, "my": 1.10412}, "gini_cpc": 0.216734, "gini_cps": 0.233269, "bpic": 5.477109, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0369998402762288, "vi": 1.1174680551942868, "tr": 1.0232525674667896, "fi": 1.0987016510914944, "ru": 1.5769475843305132, "el": 1.7097943540442366, "hi": 2.432220106671395, "ta": 2.5437159330176473, "th": 2.6472035942182712, "zh": 2.4769823705852962, "my": 2.787007220962428}, "t": 245.8} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.06594, "de": 2.25595, "vi": 1.94135, "tr": 2.24456, "fi": 2.41829, "ru": 1.36456, "el": 1.42431, "hi": 1.15818, "ta": 0.95935, "th": 1.13024, "zh": 3.17407, "my": 1.11278}, "bytes_per_chunk": {"en": 5.167, "de": 4.957, "vi": 6.226, "tr": 5.264, "fi": 4.667, "ru": 5.931, "el": 5.541, "hi": 5.258, "ta": 5.549, "th": 5.602, "zh": 5.733, "my": 5.339}, "bpb_mean": 1.7708, "t": 247.1} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.4735, "bpb": 2.12581, "chunk_loss": 0.030439, "F": 0.18756, "G": 0.26677, "chunks": 192.06, "bytes_per_chunk": 5.332, "boundary_entropy": 0.34255, "router_noise": 0.0, "grad_norm": 0.7487, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 247.2} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.48689, "bpb": 2.14512, "chunk_loss": 0.030495, "F": 0.1861, "G": 0.28556, "chunks": 190.56, "bytes_per_chunk": 5.374, "boundary_entropy": 0.37719, "router_noise": 0.0, "grad_norm": 0.7098, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 249.2} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.41109, "bpb": 2.03578, "chunk_loss": 0.030359, "F": 0.18781, "G": 0.26159, "chunks": 192.31, "bytes_per_chunk": 5.325, "boundary_entropy": 0.32636, "router_noise": 0.0, "grad_norm": 0.7576, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 251.2} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.20835, "bpb": 1.74329, "chunk_loss": 0.030367, "F": 0.18719, "G": 0.2774, "chunks": 191.69, "bytes_per_chunk": 5.342, "boundary_entropy": 0.36703, "router_noise": 0.0, "grad_norm": 0.6985, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 253.2} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3381.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194277, "cps": 26.414062, "mean_p": 0.222649, "boundary_entropy": 0.229954, "bpb": 2.043376}, "de": {"chunks": 4099.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19963, "cps": 32.023438, "mean_p": 0.235366, "boundary_entropy": 0.245117, "bpb": 2.24187}, "vi": {"chunks": 4072.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.220884, "cps": 31.8125, "mean_p": 0.268228, "boundary_entropy": 0.40425, "bpb": 1.933581}, "tr": {"chunks": 3618.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.199856, "cps": 28.265625, "mean_p": 0.228409, "boundary_entropy": 0.277735, "bpb": 2.227568}, "fi": {"chunks": 3950.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.209984, "cps": 30.859375, "mean_p": 0.23599, "boundary_entropy": 0.260305, "bpb": 2.386739}, "ru": {"chunks": 5943.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.313615, "cps": 46.429688, "mean_p": 0.309028, "boundary_entropy": 0.452023, "bpb": 1.377663}, "el": {"chunks": 6553.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.3193, "cps": 51.195312, "mean_p": 0.272531, "boundary_entropy": 0.398443, "bpb": 1.427929}, "hi": {"chunks": 7248.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.430788, "cps": 56.625, "mean_p": 0.309991, "boundary_entropy": 0.529346, "bpb": 1.159107}, "ta": {"chunks": 8057.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.430971, "cps": 62.945312, "mean_p": 0.350684, "boundary_entropy": 0.564349, "bpb": 0.950302}, "th": {"chunks": 7981.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.496177, "cps": 62.351562, "mean_p": 0.382296, "boundary_entropy": 0.6109, "bpb": 1.105378}, "zh": {"chunks": 2886.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.488656, "cps": 22.546875, "mean_p": 0.287474, "boundary_entropy": 0.383855, "bpb": 3.156765}, "my": {"chunks": 9039.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.477042, "cps": 70.617188, "mean_p": 0.440386, "boundary_entropy": 0.666988, "bpb": 1.09959}}, "bpb": {"en": 2.04338, "de": 2.24187, "vi": 1.93358, "tr": 2.22757, "fi": 2.38674, "ru": 1.37766, "el": 1.42793, "hi": 1.15911, "ta": 0.9503, "th": 1.10538, "zh": 3.15676, "my": 1.09959}, "gini_cpc": 0.201003, "gini_cps": 0.207108, "bpic": 5.742903, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0275535419404718, "vi": 1.1369557877277632, "tr": 1.028719472293047, "fi": 1.0808468510632339, "ru": 1.6142673593470886, "el": 1.6435324083555747, "hi": 2.2173898804304977, "ta": 2.2183335299071896, "th": 2.5539664917878873, "zh": 2.515253914401639, "my": 2.455477504491347}, "t": 256.2} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.05366, "de": 2.24665, "vi": 1.93269, "tr": 2.2341, "fi": 2.39897, "ru": 1.3549, "el": 1.41993, "hi": 1.15383, "ta": 0.95887, "th": 1.12156, "zh": 3.15437, "my": 1.10843}, "bytes_per_chunk": {"en": 5.134, "de": 4.962, "vi": 6.187, "tr": 5.236, "fi": 4.709, "ru": 5.783, "el": 5.708, "hi": 5.765, "ta": 6.204, "th": 5.792, "zh": 5.666, "my": 5.778}, "bpb_mean": 1.76149, "t": 257.5} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.2878, "bpb": 1.8579, "chunk_loss": 0.03054, "F": 0.18781, "G": 0.2881, "chunks": 192.31, "bytes_per_chunk": 5.325, "boundary_entropy": 0.3914, "router_noise": 0.0, "grad_norm": 0.7179, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 257.5} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.39671, "bpb": 2.01503, "chunk_loss": 0.030155, "F": 0.17712, "G": 0.27309, "chunks": 181.38, "bytes_per_chunk": 5.646, "boundary_entropy": 0.36794, "router_noise": 0.0, "grad_norm": 0.6748, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 259.5} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.31664, "bpb": 1.89951, "chunk_loss": 0.030188, "F": 0.18143, "G": 0.26469, "chunks": 185.78, "bytes_per_chunk": 5.512, "boundary_entropy": 0.34671, "router_noise": 0.0, "grad_norm": 0.7529, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 261.5} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.32844, "bpb": 1.91653, "chunk_loss": 0.03022, "F": 0.18369, "G": 0.27059, "chunks": 188.09, "bytes_per_chunk": 5.444, "boundary_entropy": 0.34927, "router_noise": 0.0, "grad_norm": 0.7274, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 263.6} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3403.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195541, "cps": 26.585938, "mean_p": 0.222805, "boundary_entropy": 0.225498, "bpb": 2.036491}, "de": {"chunks": 4099.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19963, "cps": 32.023438, "mean_p": 0.23364, "boundary_entropy": 0.241526, "bpb": 2.229923}, "vi": {"chunks": 4021.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.218118, "cps": 31.414062, "mean_p": 0.264478, "boundary_entropy": 0.396258, "bpb": 1.927279}, "tr": {"chunks": 3542.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195658, "cps": 27.671875, "mean_p": 0.224523, "boundary_entropy": 0.268079, "bpb": 2.214688}, "fi": {"chunks": 3781.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.200999, "cps": 29.539062, "mean_p": 0.228218, "boundary_entropy": 0.250845, "bpb": 2.379012}, "ru": {"chunks": 5879.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.310237, "cps": 45.929688, "mean_p": 0.306231, "boundary_entropy": 0.4469, "bpb": 1.365675}, "el": {"chunks": 7040.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.34303, "cps": 55.0, "mean_p": 0.279744, "boundary_entropy": 0.395678, "bpb": 1.4063}, "hi": {"chunks": 8203.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.487548, "cps": 64.085938, "mean_p": 0.311895, "boundary_entropy": 0.521333, "bpb": 1.158784}, "ta": {"chunks": 8867.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.474298, "cps": 69.273438, "mean_p": 0.359879, "boundary_entropy": 0.568075, "bpb": 0.943269}, "th": {"chunks": 8259.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.51346, "cps": 64.523438, "mean_p": 0.388331, "boundary_entropy": 0.616542, "bpb": 1.102446}, "zh": {"chunks": 2842.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.481206, "cps": 22.203125, "mean_p": 0.283198, "boundary_entropy": 0.377509, "bpb": 3.152589}, "my": {"chunks": 8476.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.44733, "cps": 66.21875, "mean_p": 0.434891, "boundary_entropy": 0.66345, "bpb": 1.090297}}, "bpb": {"en": 2.03649, "de": 2.22992, "vi": 1.92728, "tr": 2.21469, "fi": 2.37901, "ru": 1.36567, "el": 1.4063, "hi": 1.15878, "ta": 0.94327, "th": 1.10245, "zh": 3.15259, "my": 1.0903}, "gini_cpc": 0.208629, "gini_cps": 0.216699, "bpic": 5.609849, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0209105275641301, "vi": 1.1154576909714708, "tr": 1.0005992580835275, "fi": 1.0279144353065064, "ru": 1.5865596939531084, "el": 1.7542600978534164, "hi": 2.4933302710352154, "ta": 2.4255677521926007, "th": 2.6258418877599907, "zh": 2.4608934030539578, "my": 2.287650858302764}, "t": 266.5} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.04789, "de": 2.23536, "vi": 1.92705, "tr": 2.22046, "fi": 2.38984, "ru": 1.34192, "el": 1.39857, "hi": 1.15202, "ta": 0.95229, "th": 1.11873, "zh": 3.14954, "my": 1.09864}, "bytes_per_chunk": {"en": 5.106, "de": 4.98, "vi": 6.132, "tr": 5.302, "fi": 4.879, "ru": 5.84, "el": 5.363, "hi": 5.027, "ta": 5.802, "th": 5.542, "zh": 5.733, "my": 6.16}, "bpb_mean": 1.75269, "t": 267.8} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.25031, "bpb": 1.80382, "chunk_loss": 0.030374, "F": 0.18628, "G": 0.27751, "chunks": 190.75, "bytes_per_chunk": 5.368, "boundary_entropy": 0.36882, "router_noise": 0.0, "grad_norm": 0.7176, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 267.9} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.48271, "bpb": 2.1391, "chunk_loss": 0.030228, "F": 0.18488, "G": 0.25472, "chunks": 189.31, "bytes_per_chunk": 5.409, "boundary_entropy": 0.31672, "router_noise": 0.0, "grad_norm": 0.7666, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 269.9} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.39697, "bpb": 2.0154, "chunk_loss": 0.030116, "F": 0.17639, "G": 0.2788, "chunks": 180.62, "bytes_per_chunk": 5.669, "boundary_entropy": 0.37966, "router_noise": 0.0, "grad_norm": 0.6894, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 271.9} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.50999, "bpb": 2.17845, "chunk_loss": 0.03057, "F": 0.19409, "G": 0.26126, "chunks": 198.75, "bytes_per_chunk": 5.152, "boundary_entropy": 0.31862, "router_noise": 0.0, "grad_norm": 0.7443, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 273.9} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3437.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197495, "cps": 26.851562, "mean_p": 0.22492, "boundary_entropy": 0.224325, "bpb": 2.027194}, "de": {"chunks": 4147.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201968, "cps": 32.398438, "mean_p": 0.235097, "boundary_entropy": 0.240286, "bpb": 2.216332}, "vi": {"chunks": 4338.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.235313, "cps": 33.890625, "mean_p": 0.28014, "boundary_entropy": 0.405226, "bpb": 1.921906}, "tr": {"chunks": 3641.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201127, "cps": 28.445312, "mean_p": 0.226459, "boundary_entropy": 0.268749, "bpb": 2.207809}, "fi": {"chunks": 4011.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.213226, "cps": 31.335938, "mean_p": 0.236027, "boundary_entropy": 0.254525, "bpb": 2.358925}, "ru": {"chunks": 5932.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.313034, "cps": 46.34375, "mean_p": 0.304577, "boundary_entropy": 0.441946, "bpb": 1.359983}, "el": {"chunks": 6846.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.333577, "cps": 53.484375, "mean_p": 0.275021, "boundary_entropy": 0.386992, "bpb": 1.398109}, "hi": {"chunks": 6673.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.396612, "cps": 52.132812, "mean_p": 0.300158, "boundary_entropy": 0.523066, "bpb": 1.153386}, "ta": {"chunks": 8407.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449692, "cps": 65.679688, "mean_p": 0.355699, "boundary_entropy": 0.565537, "bpb": 0.9387}, "th": {"chunks": 7140.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.443892, "cps": 55.78125, "mean_p": 0.376775, "boundary_entropy": 0.610426, "bpb": 1.095701}, "zh": {"chunks": 2873.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.486454, "cps": 22.445312, "mean_p": 0.282762, "boundary_entropy": 0.370559, "bpb": 3.154527}, "my": {"chunks": 8995.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.47472, "cps": 70.273438, "mean_p": 0.441885, "boundary_entropy": 0.666625, "bpb": 1.08913}}, "bpb": {"en": 2.02719, "de": 2.21633, "vi": 1.92191, "tr": 2.20781, "fi": 2.35892, "ru": 1.35998, "el": 1.39811, "hi": 1.15339, "ta": 0.9387, "th": 1.0957, "zh": 3.15453, "my": 1.08913}, "gini_cpc": 0.190704, "gini_cps": 0.200572, "bpic": 5.776355, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0226481010769142, "vi": 1.1914916243161517, "tr": 1.0183913821551718, "fi": 1.0796559104766126, "ru": 1.5850264581540476, "el": 1.6890427341317817, "hi": 2.008217003926763, "ta": 2.2769849807909037, "th": 2.2476140311114516, "zh": 2.4631268103793884, "my": 2.4037117126645677}, "t": 276.8} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.03763, "de": 2.22182, "vi": 1.92129, "tr": 2.21491, "fi": 2.37123, "ru": 1.33613, "el": 1.3908, "hi": 1.14935, "ta": 0.94754, "th": 1.11386, "zh": 3.14937, "my": 1.09784}, "bytes_per_chunk": {"en": 5.037, "de": 4.924, "vi": 5.838, "tr": 5.207, "fi": 4.672, "ru": 5.792, "el": 5.47, "hi": 6.143, "ta": 5.917, "th": 6.23, "zh": 5.68, "my": 5.911}, "bpb_mean": 1.74598, "t": 278.1} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.27471, "bpb": 1.83902, "chunk_loss": 0.030471, "F": 0.18994, "G": 0.27809, "chunks": 194.5, "bytes_per_chunk": 5.265, "boundary_entropy": 0.36023, "router_noise": 0.0, "grad_norm": 0.7541, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 278.2} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.24754, "bpb": 1.79982, "chunk_loss": 0.030302, "F": 0.18488, "G": 0.26352, "chunks": 189.31, "bytes_per_chunk": 5.409, "boundary_entropy": 0.34182, "router_noise": 0.0, "grad_norm": 0.696, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 280.2} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.29796, "bpb": 1.87257, "chunk_loss": 0.030093, "F": 0.17816, "G": 0.26561, "chunks": 182.44, "bytes_per_chunk": 5.613, "boundary_entropy": 0.35169, "router_noise": 0.0, "grad_norm": 0.7387, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 282.2} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.32556, "bpb": 1.91238, "chunk_loss": 0.030291, "F": 0.18414, "G": 0.2571, "chunks": 188.56, "bytes_per_chunk": 5.431, "boundary_entropy": 0.32661, "router_noise": 0.0, "grad_norm": 0.7479, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 284.2} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3436.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197437, "cps": 26.84375, "mean_p": 0.223818, "boundary_entropy": 0.220876, "bpb": 2.021539}, "de": {"chunks": 4185.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.203818, "cps": 32.695312, "mean_p": 0.237165, "boundary_entropy": 0.239129, "bpb": 2.208622}, "vi": {"chunks": 4327.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234717, "cps": 33.804688, "mean_p": 0.274622, "boundary_entropy": 0.397704, "bpb": 1.917625}, "tr": {"chunks": 3708.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.204828, "cps": 28.96875, "mean_p": 0.229821, "boundary_entropy": 0.270565, "bpb": 2.198029}, "fi": {"chunks": 4071.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.216416, "cps": 31.804688, "mean_p": 0.238973, "boundary_entropy": 0.255283, "bpb": 2.349364}, "ru": {"chunks": 5896.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.311135, "cps": 46.0625, "mean_p": 0.298181, "boundary_entropy": 0.433517, "bpb": 1.352417}, "el": {"chunks": 6726.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.32773, "cps": 52.546875, "mean_p": 0.265034, "boundary_entropy": 0.37002, "bpb": 1.391752}, "hi": {"chunks": 7195.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.427637, "cps": 56.210938, "mean_p": 0.289385, "boundary_entropy": 0.49885, "bpb": 1.138294}, "ta": {"chunks": 7967.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.426157, "cps": 62.242188, "mean_p": 0.348082, "boundary_entropy": 0.559314, "bpb": 0.935258}, "th": {"chunks": 7916.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.492136, "cps": 61.84375, "mean_p": 0.37574, "boundary_entropy": 0.605833, "bpb": 1.083235}, "zh": {"chunks": 2851.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482729, "cps": 22.273438, "mean_p": 0.282629, "boundary_entropy": 0.37147, "bpb": 3.125365}, "my": {"chunks": 7768.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.409964, "cps": 60.6875, "mean_p": 0.430617, "boundary_entropy": 0.662629, "bpb": 1.083418}}, "bpb": {"en": 2.02154, "de": 2.20862, "vi": 1.91763, "tr": 2.19803, "fi": 2.34936, "ru": 1.35242, "el": 1.39175, "hi": 1.13829, "ta": 0.93526, "th": 1.08323, "zh": 3.12536, "my": 1.08342}, "gini_cpc": 0.188549, "gini_cps": 0.189045, "bpic": 5.810814, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0323192365825604, "vi": 1.1888162101181101, "tr": 1.0374331925638038, "fi": 1.0961252546546434, "ru": 1.575865782449372, "el": 1.6599193318243646, "hi": 2.1659413395790055, "ta": 2.158441650027508, "th": 2.492617761506375, "zh": 2.444976784131053, "my": 2.076427661957252}, "t": 287.1} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.03138, "de": 2.21338, "vi": 1.91742, "tr": 2.20352, "fi": 2.3609, "ru": 1.32838, "el": 1.38434, "hi": 1.13322, "ta": 0.94494, "th": 1.10229, "zh": 3.12351, "my": 1.09226}, "bytes_per_chunk": {"en": 5.044, "de": 4.896, "vi": 5.792, "tr": 5.115, "fi": 4.609, "ru": 5.873, "el": 5.555, "hi": 5.803, "ta": 6.234, "th": 5.827, "zh": 5.711, "my": 6.543}, "bpb_mean": 1.73629, "t": 288.4} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.27416, "bpb": 1.83822, "chunk_loss": 0.030486, "F": 0.18896, "G": 0.27356, "chunks": 193.5, "bytes_per_chunk": 5.292, "boundary_entropy": 0.34735, "router_noise": 0.0, "grad_norm": 0.7692, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 288.5} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.36832, "bpb": 1.97407, "chunk_loss": 0.030341, "F": 0.18527, "G": 0.26883, "chunks": 189.72, "bytes_per_chunk": 5.397, "boundary_entropy": 0.35137, "router_noise": 0.0, "grad_norm": 0.7304, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 290.5} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.36203, "bpb": 1.96499, "chunk_loss": 0.03054, "F": 0.19266, "G": 0.27198, "chunks": 197.28, "bytes_per_chunk": 5.191, "boundary_entropy": 0.34045, "router_noise": 0.0, "grad_norm": 0.7402, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 292.5} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.41323, "bpb": 2.03886, "chunk_loss": 0.030549, "F": 0.19431, "G": 0.26895, "chunks": 198.97, "bytes_per_chunk": 5.147, "boundary_entropy": 0.33529, "router_noise": 0.0, "grad_norm": 0.7538, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 294.5} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3357.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.192898, "cps": 26.226562, "mean_p": 0.220066, "boundary_entropy": 0.21941, "bpb": 2.016579}, "de": {"chunks": 4128.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201042, "cps": 32.25, "mean_p": 0.234459, "boundary_entropy": 0.237575, "bpb": 2.202382}, "vi": {"chunks": 4033.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.218769, "cps": 31.507812, "mean_p": 0.266807, "boundary_entropy": 0.395366, "bpb": 1.910789}, "tr": {"chunks": 3604.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.199083, "cps": 28.15625, "mean_p": 0.225712, "boundary_entropy": 0.266334, "bpb": 2.189781}, "fi": {"chunks": 3985.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.211844, "cps": 31.132812, "mean_p": 0.234774, "boundary_entropy": 0.250033, "bpb": 2.342094}, "ru": {"chunks": 5631.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.29715, "cps": 43.992188, "mean_p": 0.285704, "boundary_entropy": 0.420133, "bpb": 1.348792}, "el": {"chunks": 6662.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.324611, "cps": 52.046875, "mean_p": 0.265097, "boundary_entropy": 0.375745, "bpb": 1.380962}, "hi": {"chunks": 7020.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.417236, "cps": 54.84375, "mean_p": 0.281916, "boundary_entropy": 0.489185, "bpb": 1.133624}, "ta": {"chunks": 8043.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.430222, "cps": 62.835938, "mean_p": 0.346487, "boundary_entropy": 0.554231, "bpb": 0.927864}, "th": {"chunks": 7222.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.44899, "cps": 56.421875, "mean_p": 0.354443, "boundary_entropy": 0.590113, "bpb": 1.081113}, "zh": {"chunks": 2842.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.481206, "cps": 22.203125, "mean_p": 0.28269, "boundary_entropy": 0.371402, "bpb": 3.124562}, "my": {"chunks": 8003.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.422366, "cps": 62.523438, "mean_p": 0.431972, "boundary_entropy": 0.661062, "bpb": 1.078334}}, "bpb": {"en": 2.01658, "de": 2.20238, "vi": 1.91079, "tr": 2.18978, "fi": 2.34209, "ru": 1.34879, "el": 1.38096, "hi": 1.13362, "ta": 0.92786, "th": 1.08111, "zh": 3.12456, "my": 1.07833}, "gini_cpc": 0.188542, "gini_cps": 0.192308, "bpic": 5.947327, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.042221577793623, "vi": 1.1341169963398197, "tr": 1.0320649039502239, "fi": 1.098219678188218, "ru": 1.5404552689100002, "el": 1.6828157270922135, "hi": 2.162991526875381, "ta": 2.23031075246998, "th": 2.3276045516682347, "zh": 2.4946143135515695, "my": 2.189587069866016}, "t": 297.4} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.02723, "de": 2.20697, "vi": 1.90983, "tr": 2.19569, "fi": 2.35374, "ru": 1.32476, "el": 1.37405, "hi": 1.12849, "ta": 0.93776, "th": 1.09934, "zh": 3.12138, "my": 1.08679}, "bytes_per_chunk": {"en": 5.194, "de": 4.955, "vi": 6.069, "tr": 5.221, "fi": 4.714, "ru": 6.097, "el": 5.635, "hi": 5.854, "ta": 6.22, "th": 6.309, "zh": 5.721, "my": 6.488}, "bpb_mean": 1.7305, "t": 298.7} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.11212, "bpb": 1.60446, "chunk_loss": 0.030223, "F": 0.17871, "G": 0.27596, "chunks": 183.0, "bytes_per_chunk": 5.596, "boundary_entropy": 0.38107, "router_noise": 0.0, "grad_norm": 0.6783, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 298.8} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.17595, "bpb": 1.69654, "chunk_loss": 0.030352, "F": 0.18436, "G": 0.27496, "chunks": 188.78, "bytes_per_chunk": 5.424, "boundary_entropy": 0.36834, "router_noise": 0.0, "grad_norm": 0.7405, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 300.8} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.38537, "bpb": 1.99867, "chunk_loss": 0.03039, "F": 0.18906, "G": 0.26743, "chunks": 193.59, "bytes_per_chunk": 5.289, "boundary_entropy": 0.33084, "router_noise": 0.0, "grad_norm": 0.763, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 302.8} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.39649, "bpb": 2.01472, "chunk_loss": 0.030322, "F": 0.186, "G": 0.25699, "chunks": 190.47, "bytes_per_chunk": 5.376, "boundary_entropy": 0.32551, "router_noise": 0.0, "grad_norm": 0.7669, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 304.8} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3437.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197495, "cps": 26.851562, "mean_p": 0.223762, "boundary_entropy": 0.218628, "bpb": 2.010093}, "de": {"chunks": 4125.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.200896, "cps": 32.226562, "mean_p": 0.232545, "boundary_entropy": 0.233775, "bpb": 2.194178}, "vi": {"chunks": 4090.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.221861, "cps": 31.953125, "mean_p": 0.268783, "boundary_entropy": 0.396202, "bpb": 1.908915}, "tr": {"chunks": 3628.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200409, "cps": 28.34375, "mean_p": 0.226759, "boundary_entropy": 0.266612, "bpb": 2.180227}, "fi": {"chunks": 4013.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.213333, "cps": 31.351562, "mean_p": 0.235202, "boundary_entropy": 0.252756, "bpb": 2.33168}, "ru": {"chunks": 5918.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.312296, "cps": 46.234375, "mean_p": 0.295647, "boundary_entropy": 0.428267, "bpb": 1.342732}, "el": {"chunks": 6856.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.334064, "cps": 53.5625, "mean_p": 0.268924, "boundary_entropy": 0.376284, "bpb": 1.371962}, "hi": {"chunks": 6580.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.391085, "cps": 51.40625, "mean_p": 0.280064, "boundary_entropy": 0.492985, "bpb": 1.130965}, "ta": {"chunks": 8250.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.441294, "cps": 64.453125, "mean_p": 0.346338, "boundary_entropy": 0.552471, "bpb": 0.923876}, "th": {"chunks": 7435.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.462232, "cps": 58.085938, "mean_p": 0.360712, "boundary_entropy": 0.594526, "bpb": 1.07876}, "zh": {"chunks": 2822.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.477819, "cps": 22.046875, "mean_p": 0.279371, "boundary_entropy": 0.368465, "bpb": 3.117928}, "my": {"chunks": 8504.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.448807, "cps": 66.4375, "mean_p": 0.431423, "boundary_entropy": 0.662567, "bpb": 1.079318}}, "bpb": {"en": 2.01009, "de": 2.19418, "vi": 1.90891, "tr": 2.18023, "fi": 2.33168, "ru": 1.34273, "el": 1.37196, "hi": 1.13097, "ta": 0.92388, "th": 1.07876, "zh": 3.11793, "my": 1.07932}, "gini_cpc": 0.18891, "gini_cps": 0.197927, "bpic": 5.845152, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0172229122117846, "vi": 1.123374998490793, "tr": 1.0147552690082295, "fi": 1.080194257976227, "ru": 1.5812856674571232, "el": 1.6915099306467274, "hi": 1.9802289653586245, "ta": 2.2344624826364883, "th": 2.340477636038325, "zh": 2.419402665816441, "my": 2.272502991050526}, "t": 307.8} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.02151, "de": 2.19992, "vi": 1.90796, "tr": 2.18692, "fi": 2.34397, "ru": 1.31826, "el": 1.36437, "hi": 1.12711, "ta": 0.93268, "th": 1.09673, "zh": 3.1149, "my": 1.0877}, "bytes_per_chunk": {"en": 5.054, "de": 4.95, "vi": 6.099, "tr": 5.187, "fi": 4.664, "ru": 5.804, "el": 5.468, "hi": 6.24, "ta": 6.036, "th": 6.129, "zh": 5.746, "my": 6.22}, "bpb_mean": 1.72517, "t": 309.1} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.45367, "bpb": 2.0972, "chunk_loss": 0.030353, "F": 0.18597, "G": 0.25805, "chunks": 190.44, "bytes_per_chunk": 5.377, "boundary_entropy": 0.3165, "router_noise": 0.0, "grad_norm": 0.7577, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 309.1} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.25468, "bpb": 1.81012, "chunk_loss": 0.030443, "F": 0.18875, "G": 0.27755, "chunks": 193.28, "bytes_per_chunk": 5.298, "boundary_entropy": 0.3646, "router_noise": 0.0, "grad_norm": 0.7251, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 311.1} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.30791, "bpb": 1.88692, "chunk_loss": 0.030427, "F": 0.18991, "G": 0.26301, "chunks": 194.47, "bytes_per_chunk": 5.266, "boundary_entropy": 0.32866, "router_noise": 0.0, "grad_norm": 0.7586, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 313.2} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.18896, "bpb": 1.7153, "chunk_loss": 0.030388, "F": 0.18793, "G": 0.26835, "chunks": 192.44, "bytes_per_chunk": 5.321, "boundary_entropy": 0.35086, "router_noise": 0.0, "grad_norm": 0.7223, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 315.2} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3382.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194334, "cps": 26.421875, "mean_p": 0.220916, "boundary_entropy": 0.217176, "bpb": 2.007243}, "de": {"chunks": 4131.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201188, "cps": 32.273438, "mean_p": 0.232349, "boundary_entropy": 0.232808, "bpb": 2.19091}, "vi": {"chunks": 3980.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.215894, "cps": 31.09375, "mean_p": 0.264781, "boundary_entropy": 0.393921, "bpb": 1.902916}, "tr": {"chunks": 3600.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.198862, "cps": 28.125, "mean_p": 0.223856, "boundary_entropy": 0.263063, "bpb": 2.172771}, "fi": {"chunks": 3868.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.205624, "cps": 30.21875, "mean_p": 0.228688, "boundary_entropy": 0.246032, "bpb": 2.328984}, "ru": {"chunks": 5797.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.30591, "cps": 45.289062, "mean_p": 0.290132, "boundary_entropy": 0.423085, "bpb": 1.338325}, "el": {"chunks": 6741.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.328461, "cps": 52.664062, "mean_p": 0.266105, "boundary_entropy": 0.374428, "bpb": 1.368512}, "hi": {"chunks": 7371.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.438098, "cps": 57.585938, "mean_p": 0.284373, "boundary_entropy": 0.487749, "bpb": 1.122099}, "ta": {"chunks": 8734.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.467184, "cps": 68.234375, "mean_p": 0.349608, "boundary_entropy": 0.551479, "bpb": 0.916472}, "th": {"chunks": 8154.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.506932, "cps": 63.703125, "mean_p": 0.365764, "boundary_entropy": 0.594612, "bpb": 1.070831}, "zh": {"chunks": 2851.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482729, "cps": 22.273438, "mean_p": 0.281069, "boundary_entropy": 0.368991, "bpb": 3.113753}, "my": {"chunks": 7369.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.388906, "cps": 57.570312, "mean_p": 0.421205, "boundary_entropy": 0.658482, "bpb": 1.076506}}, "bpb": {"en": 2.00724, "de": 2.19091, "vi": 1.90292, "tr": 2.17277, "fi": 2.32898, "ru": 1.33832, "el": 1.36851, "hi": 1.1221, "ta": 0.91647, "th": 1.07083, "zh": 3.11375, "my": 1.07651}, "gini_cpc": 0.202487, "gini_cps": 0.203997, "bpic": 5.816803, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0352692265033947, "vi": 1.1109395983811294, "tr": 1.0232988033877894, "fi": 1.0580960641163912, "ru": 1.5741445242467884, "el": 1.6901840496100882, "hi": 2.2543526562704574, "ta": 2.404020877957957, "th": 2.6085559657106954, "zh": 2.4840154436056467, "my": 2.001223976054731}, "t": 318.1} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.01793, "de": 2.19718, "vi": 1.90236, "tr": 2.18006, "fi": 2.34098, "ru": 1.31435, "el": 1.36128, "hi": 1.11677, "ta": 0.92555, "th": 1.08942, "zh": 3.11071, "my": 1.08554}, "bytes_per_chunk": {"en": 5.12, "de": 4.948, "vi": 6.189, "tr": 5.226, "fi": 4.8, "ru": 5.913, "el": 5.558, "hi": 5.586, "ta": 5.709, "th": 5.633, "zh": 5.717, "my": 6.963}, "bpb_mean": 1.72018, "t": 319.4} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.24521, "bpb": 1.79646, "chunk_loss": 0.030324, "F": 0.18619, "G": 0.26612, "chunks": 190.66, "bytes_per_chunk": 5.371, "boundary_entropy": 0.3422, "router_noise": 0.0, "grad_norm": 0.6963, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 319.5} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.35789, "bpb": 1.95902, "chunk_loss": 0.030369, "F": 0.18579, "G": 0.27198, "chunks": 190.25, "bytes_per_chunk": 5.382, "boundary_entropy": 0.35395, "router_noise": 0.0, "grad_norm": 0.7093, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 321.5} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.3115, "bpb": 1.8921, "chunk_loss": 0.030364, "F": 0.18558, "G": 0.25835, "chunks": 190.03, "bytes_per_chunk": 5.389, "boundary_entropy": 0.32444, "router_noise": 0.0, "grad_norm": 0.7529, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 323.5} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.2707, "bpb": 1.83323, "chunk_loss": 0.030278, "F": 0.1817, "G": 0.26967, "chunks": 186.06, "bytes_per_chunk": 5.504, "boundary_entropy": 0.34836, "router_noise": 0.0, "grad_norm": 0.7452, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 325.5} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3414.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196173, "cps": 26.671875, "mean_p": 0.221905, "boundary_entropy": 0.216095, "bpb": 2.003143}, "de": {"chunks": 4134.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201334, "cps": 32.296875, "mean_p": 0.23215, "boundary_entropy": 0.230947, "bpb": 2.187151}, "vi": {"chunks": 4041.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.219203, "cps": 31.570312, "mean_p": 0.267189, "boundary_entropy": 0.394022, "bpb": 1.90063}, "tr": {"chunks": 3645.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201348, "cps": 28.476562, "mean_p": 0.22538, "boundary_entropy": 0.261141, "bpb": 2.17271}, "fi": {"chunks": 4019.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.213652, "cps": 31.398438, "mean_p": 0.234518, "boundary_entropy": 0.247597, "bpb": 2.325915}, "ru": {"chunks": 5803.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306227, "cps": 45.335938, "mean_p": 0.286072, "boundary_entropy": 0.415335, "bpb": 1.336446}, "el": {"chunks": 6686.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325781, "cps": 52.234375, "mean_p": 0.261607, "boundary_entropy": 0.368575, "bpb": 1.363756}, "hi": {"chunks": 7271.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.432155, "cps": 56.804688, "mean_p": 0.28339, "boundary_entropy": 0.486893, "bpb": 1.120083}, "ta": {"chunks": 7973.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.426478, "cps": 62.289062, "mean_p": 0.344087, "boundary_entropy": 0.550694, "bpb": 0.91767}, "th": {"chunks": 7854.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.488281, "cps": 61.359375, "mean_p": 0.359722, "boundary_entropy": 0.590272, "bpb": 1.064836}, "zh": {"chunks": 2839.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.480698, "cps": 22.179688, "mean_p": 0.277257, "boundary_entropy": 0.363413, "bpb": 3.105139}, "my": {"chunks": 9165.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.483692, "cps": 71.601562, "mean_p": 0.438641, "boundary_entropy": 0.664609, "bpb": 1.075344}}, "bpb": {"en": 2.00314, "de": 2.18715, "vi": 1.90063, "tr": 2.17271, "fi": 2.32592, "ru": 1.33645, "el": 1.36376, "hi": 1.12008, "ta": 0.91767, "th": 1.06484, "zh": 3.10514, "my": 1.07534}, "gini_cpc": 0.197981, "gini_cps": 0.20688, "bpic": 5.741443, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0263102552477357, "vi": 1.1173939405206055, "tr": 1.0263785910869134, "fi": 1.0890974123411878, "ru": 1.5610037978029316, "el": 1.660680669730827, "hi": 2.2029248153090455, "ta": 2.17398677597226, "th": 2.4890317420688532, "zh": 2.4503750021574078, "my": 2.4656401684677296}, "t": 328.5} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.01506, "de": 2.19365, "vi": 1.90005, "tr": 2.17909, "fi": 2.3368, "ru": 1.31215, "el": 1.35626, "hi": 1.11525, "ta": 0.92676, "th": 1.08411, "zh": 3.10295, "my": 1.08422}, "bytes_per_chunk": {"en": 5.109, "de": 4.953, "vi": 6.139, "tr": 5.174, "fi": 4.67, "ru": 5.933, "el": 5.603, "hi": 5.689, "ta": 6.23, "th": 5.761, "zh": 5.735, "my": 5.743}, "bpb_mean": 1.71719, "t": 329.8} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3414.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196173, "cps": 26.671875, "mean_p": 0.221905, "boundary_entropy": 0.216095, "bpb": 2.003143}, "de": {"chunks": 4134.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.201334, "cps": 32.296875, "mean_p": 0.23215, "boundary_entropy": 0.230947, "bpb": 2.187151}, "vi": {"chunks": 4041.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.219203, "cps": 31.570312, "mean_p": 0.267189, "boundary_entropy": 0.394022, "bpb": 1.90063}, "tr": {"chunks": 3645.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201348, "cps": 28.476562, "mean_p": 0.22538, "boundary_entropy": 0.261141, "bpb": 2.17271}, "fi": {"chunks": 4019.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.213652, "cps": 31.398438, "mean_p": 0.234518, "boundary_entropy": 0.247597, "bpb": 2.325915}, "ru": {"chunks": 5803.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306227, "cps": 45.335938, "mean_p": 0.286072, "boundary_entropy": 0.415335, "bpb": 1.336446}, "el": {"chunks": 6686.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.325781, "cps": 52.234375, "mean_p": 0.261607, "boundary_entropy": 0.368575, "bpb": 1.363756}, "hi": {"chunks": 7271.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.432155, "cps": 56.804688, "mean_p": 0.28339, "boundary_entropy": 0.486893, "bpb": 1.120083}, "ta": {"chunks": 7973.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.426478, "cps": 62.289062, "mean_p": 0.344087, "boundary_entropy": 0.550694, "bpb": 0.91767}, "th": {"chunks": 7854.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.488281, "cps": 61.359375, "mean_p": 0.359722, "boundary_entropy": 0.590272, "bpb": 1.064836}, "zh": {"chunks": 2839.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.480698, "cps": 22.179688, "mean_p": 0.277257, "boundary_entropy": 0.363413, "bpb": 3.105139}, "my": {"chunks": 9165.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.483692, "cps": 71.601562, "mean_p": 0.438641, "boundary_entropy": 0.664609, "bpb": 1.075344}}, "bpb": {"en": 2.00314, "de": 2.18715, "vi": 1.90063, "tr": 2.17271, "fi": 2.32592, "ru": 1.33645, "el": 1.36376, "hi": 1.12008, "ta": 0.91767, "th": 1.06484, "zh": 3.10514, "my": 1.07534}, "gini_cpc": 0.197981, "gini_cps": 0.20688, "bpic": 5.741443, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0263102552477357, "vi": 1.1173939405206055, "tr": 1.0263785910869134, "fi": 1.0890974123411878, "ru": 1.5610037978029316, "el": 1.660680669730827, "hi": 2.2029248153090455, "ta": 2.17398677597226, "th": 2.4890317420688532, "zh": 2.4503750021574078, "my": 2.4656401684677296}, "t": 330.7} |
| {"event": "final", "run": "pilot_Apc_s0", "condition": "A", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.19617307360799863, "cps": 26.671875, "bpb": 2.024522936439789, "bytes_per_chunk": 5.06984615087078}, "de": {"cpc": 0.20133443724735792, "cps": 32.296875, "bpb": 2.1852193010821352, "bytes_per_chunk": 4.95490114542774}, "vi": {"cpc": 0.2192026037428804, "cps": 31.5703125, "bpb": 1.9042950676967878, "bytes_per_chunk": 6.127293551176869}, "tr": {"cpc": 0.201347842898967, "cps": 28.4765625, "bpb": 2.169624838708644, "bytes_per_chunk": 5.202302044056361}, "fi": {"cpc": 0.2136515868374887, "cps": 31.3984375, "bpb": 2.3537633308235932, "bytes_per_chunk": 4.656695207304509}, "ru": {"cpc": 0.3062269129287599, "cps": 45.3359375, "bpb": 1.312258523841349, "bytes_per_chunk": 5.931866266900492}, "el": {"cpc": 0.32578083126248597, "cps": 52.234375, "bpb": 1.3533775709898723, "bytes_per_chunk": 5.588530618770985}, "hi": {"cpc": 0.43215453194650816, "cps": 56.8046875, "bpb": 1.123992947930272, "bytes_per_chunk": 5.68666753438327}, "ta": {"cpc": 0.4264776678256218, "cps": 62.2890625, "bpb": 0.9281541374862824, "bytes_per_chunk": 6.228029745076144}, "th": {"cpc": 0.48828100714951816, "cps": 61.359375, "bpb": 1.082798357402357, "bytes_per_chunk": 5.766856589744154}, "zh": {"cpc": 0.48069759566542497, "cps": 22.1796875, "bpb": 3.1126575876564906, "bytes_per_chunk": 5.735188588430909}, "my": {"cpc": 0.48369221025965803, "cps": 71.6015625, "bpb": 1.095115566751127, "bytes_per_chunk": 5.752367159299124}}, "gini_cpc": 0.19798115513813697, "gini_cps": 0.20687969999800537, "bpic": 5.741442762252409, "seed": 0, "mean_boundary_entropy": 0.32927095890045166, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 333.2} |
|
|
