{"event": "start", "run": "pilot_E_s0", "condition": "E", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "parity_b", "p3_objective": "none", "freeze_in_p3": true, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 1.4}
{"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 1.4}
{"event": "probe", "step": 0, "phase": 1, "objective": "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": 3.5}
{"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": 7.806, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 3.8}
{"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6328.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.363615, "cps": 49.4375, "mean_p": 0.496549, "boundary_entropy": 0.686989, "bpb": 8.13796}, "de": {"chunks": 8596.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.418643, "cps": 67.15625, "mean_p": 0.498781, "boundary_entropy": 0.687819, "bpb": 8.13665}, "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": 8.072292}, "tr": {"chunks": 5243.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.289621, "cps": 40.960938, "mean_p": 0.489423, "boundary_entropy": 0.68728, "bpb": 8.156685}, "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": 8.196344}, "ru": {"chunks": 10502.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.554195, "cps": 82.046875, "mean_p": 0.493306, "boundary_entropy": 0.689882, "bpb": 8.059552}, "el": {"chunks": 15040.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.732836, "cps": 117.5, "mean_p": 0.497468, "boundary_entropy": 0.690234, "bpb": 8.140715}, "hi": {"chunks": 2500.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.148588, "cps": 19.53125, "mean_p": 0.484438, "boundary_entropy": 0.690242, "bpb": 8.249479}, "ta": {"chunks": 37854.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.024819, "cps": 295.734375, "mean_p": 0.512695, "boundary_entropy": 0.690391, "bpb": 8.028618}, "th": {"chunks": 33855.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.104756, "cps": 264.492188, "mean_p": 0.51177, "boundary_entropy": 0.690558, "bpb": 8.229618}, "zh": {"chunks": 11819.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.001185, "cps": 92.335938, "mean_p": 0.520021, "boundary_entropy": 0.685675, "bpb": 8.12062}, "my": {"chunks": 48158.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.541588, "cps": 376.234375, "mean_p": 0.521508, "boundary_entropy": 0.69021, "bpb": 7.893447}}, "bpb": {"en": 8.13796, "de": 8.13665, "vi": 8.07229, "tr": 8.15668, "fi": 8.19634, "ru": 8.05955, "el": 8.14072, "hi": 8.24948, "ta": 8.02862, "th": 8.22962, "zh": 8.12062, "my": 7.89345}, "gini_cpc": 0.425843, "gini_cps": 0.427294, "bpic": 1.902194, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1513348466897018, "vi": 1.9427886084191666, "tr": 0.7965021565388616, "fi": 1.2922570202455042, "ru": 1.5241245175773628, "el": 2.015415715943378, "hi": 0.408641608887503, "ta": 5.568573521520691, "th": 5.788411565533451, "zh": 5.503575640149103, "my": 6.989767273771027}, "t": 5.2}
{"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": 0.496201, "boundary_entropy": 0.686976, "bpb": 8.088584}, "de": {"chunks": 8518.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.414844, "cps": 66.546875, "mean_p": 0.498576, "boundary_entropy": 0.687814, "bpb": 8.09281}, "vi": {"chunks": 12839.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.696447, "cps": 100.304688, "mean_p": 0.505085, "boundary_entropy": 0.68839, "bpb": 8.038729}, "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": 8.12213}, "fi": {"chunks": 8750.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.465153, "cps": 68.359375, "mean_p": 0.50014, "boundary_entropy": 0.687503, "bpb": 8.154251}, "ru": {"chunks": 10058.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.530765, "cps": 78.578125, "mean_p": 0.492759, "boundary_entropy": 0.689865, "bpb": 7.993326}, "el": {"chunks": 15093.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.735419, "cps": 117.914062, "mean_p": 0.497514, "boundary_entropy": 0.690235, "bpb": 8.120838}, "hi": {"chunks": 2405.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.142942, "cps": 18.789062, "mean_p": 0.484162, "boundary_entropy": 0.690223, "bpb": 8.217298}, "ta": {"chunks": 37476.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.0046, "cps": 292.78125, "mean_p": 0.512238, "boundary_entropy": 0.690413, "bpb": 7.98783}, "th": {"chunks": 33649.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.091949, "cps": 262.882812, "mean_p": 0.511606, "boundary_entropy": 0.690564, "bpb": 8.202604}, "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": 8.114115}, "my": {"chunks": 48087.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.53784, "cps": 375.679688, "mean_p": 0.521422, "boundary_entropy": 0.690216, "bpb": 7.860126}}, "bpb": {"en": 8.08858, "de": 8.09281, "vi": 8.03873, "tr": 8.12213, "fi": 8.15425, "ru": 7.99333, "el": 8.12084, "hi": 8.2173, "ta": 7.98783, "th": 8.2026, "zh": 8.11412, "my": 7.86013}, "gini_cpc": 0.428216, "gini_cps": 0.429762, "bpic": 1.917917, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1553107757943555, "vi": 1.9395530038264217, "tr": 0.7951875581317117, "fi": 1.2954175161247852, "ru": 1.4781415073905229, "el": 2.0480866296888456, "hi": 0.39808297410795235, "ta": 5.58266228079242, "th": 5.825922357089102, "zh": 5.568908092683926, "my": 7.067696683274264}, "t": 6.8}
{"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5919.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.340114, "cps": 46.242188, "mean_p": 0.494973, "boundary_entropy": 0.686928, "bpb": 7.918834}, "de": {"chunks": 8280.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.403253, "cps": 64.6875, "mean_p": 0.497821, "boundary_entropy": 0.687798, "bpb": 7.939083}, "vi": {"chunks": 12276.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.665907, "cps": 95.90625, "mean_p": 0.503598, "boundary_entropy": 0.688396, "bpb": 7.92694}, "tr": {"chunks": 4900.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.270673, "cps": 38.28125, "mean_p": 0.488079, "boundary_entropy": 0.687201, "bpb": 8.003465}, "fi": {"chunks": 8334.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.443039, "cps": 65.109375, "mean_p": 0.498782, "boundary_entropy": 0.687477, "bpb": 8.007484}, "ru": {"chunks": 8691.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.458628, "cps": 67.898438, "mean_p": 0.49085, "boundary_entropy": 0.689793, "bpb": 7.771058}, "el": {"chunks": 15240.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.742581, "cps": 119.0625, "mean_p": 0.497599, "boundary_entropy": 0.690236, "bpb": 8.033798}, "hi": {"chunks": 2072.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.12315, "cps": 16.1875, "mean_p": 0.482793, "boundary_entropy": 0.690125, "bpb": 8.096535}, "ta": {"chunks": 35897.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.920139, "cps": 280.445312, "mean_p": 0.510437, "boundary_entropy": 0.690491, "bpb": 7.841583}, "th": {"chunks": 32617.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.02779, "cps": 254.820312, "mean_p": 0.510752, "boundary_entropy": 0.690595, "bpb": 8.088114}, "zh": {"chunks": 11780.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.994582, "cps": 92.03125, "mean_p": 0.519761, "boundary_entropy": 0.68569, "bpb": 8.087305}, "my": {"chunks": 47876.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.526705, "cps": 374.03125, "mean_p": 0.521122, "boundary_entropy": 0.690237, "bpb": 7.783255}}, "bpb": {"en": 7.91883, "de": 7.93908, "vi": 7.92694, "tr": 8.00347, "fi": 8.00748, "ru": 7.77106, "el": 8.0338, "hi": 8.09654, "ta": 7.84158, "th": 8.08811, "zh": 8.08731, "my": 7.78326}, "gini_cpc": 0.43724, "gini_cps": 0.438161, "bpic": 1.979457, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1856423673520349, "vi": 1.9578955442823107, "tr": 0.7958318367395848, "fi": 1.302618953231875, "ru": 1.348454558706121, "el": 2.183332613435894, "hi": 0.36208493714175544, "ta": 5.645578698359926, "th": 5.962092759558987, "zh": 5.8644545934971415, "my": 7.428998359850808}, "t": 8.5}
{"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5086.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.292248, "cps": 39.734375, "mean_p": 0.491682, "boundary_entropy": 0.686769, "bpb": 7.437741}, "de": {"chunks": 7671.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.373594, "cps": 59.929688, "mean_p": 0.495829, "boundary_entropy": 0.68775, "bpb": 7.496115}, "vi": {"chunks": 10394.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.563819, "cps": 81.203125, "mean_p": 0.498908, "boundary_entropy": 0.688358, "bpb": 7.58607}, "tr": {"chunks": 4096.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.226261, "cps": 32.0, "mean_p": 0.485082, "boundary_entropy": 0.687, "bpb": 7.643642}, "fi": {"chunks": 7121.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.378555, "cps": 55.632812, "mean_p": 0.494863, "boundary_entropy": 0.687364, "bpb": 7.577082}, "ru": {"chunks": 5209.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.274881, "cps": 40.695312, "mean_p": 0.485326, "boundary_entropy": 0.689504, "bpb": 7.189171}, "el": {"chunks": 14861.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.724114, "cps": 116.101562, "mean_p": 0.497136, "boundary_entropy": 0.69023, "bpb": 7.711294}, "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": 7.696343}, "ta": {"chunks": 28684.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.534314, "cps": 224.09375, "mean_p": 0.503576, "boundary_entropy": 0.690676, "bpb": 7.366198}, "th": {"chunks": 27423.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.70488, "cps": 214.242188, "mean_p": 0.506238, "boundary_entropy": 0.690718, "bpb": 7.547739}, "zh": {"chunks": 11595.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.963258, "cps": 90.585938, "mean_p": 0.518923, "boundary_entropy": 0.685739, "bpb": 7.998409}, "my": {"chunks": 47096.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.485539, "cps": 367.9375, "mean_p": 0.52, "boundary_entropy": 0.690311, "bpb": 7.472399}}, "bpb": {"en": 7.43774, "de": 7.49612, "vi": 7.58607, "tr": 7.64364, "fi": 7.57708, "ru": 7.18917, "el": 7.71129, "hi": 7.69634, "ta": 7.3662, "th": 7.54774, "zh": 7.99841, "my": 7.4724}, "gini_cpc": 0.469752, "gini_cps": 0.46352, "bpic": 2.252633, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2783428301130746, "vi": 1.9292447846591116, "tr": 0.7742071198345971, "fi": 1.2953193901218223, "ru": 0.9405738656584323, "el": 2.47773555767563, "hi": 0.23062466146909125, "ta": 5.250032703085954, "th": 5.833667374377982, "zh": 6.717769135515261, "my": 8.504884324015782}, "t": 10.2}
{"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": 0.488084, "boundary_entropy": 0.686662, "bpb": 6.812284}, "de": {"chunks": 6917.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.336872, "cps": 54.039062, "mean_p": 0.493873, "boundary_entropy": 0.687797, "bpb": 6.904783}, "vi": {"chunks": 7193.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.390182, "cps": 56.195312, "mean_p": 0.490796, "boundary_entropy": 0.688163, "bpb": 7.158029}, "tr": {"chunks": 3499.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.193283, "cps": 27.335938, "mean_p": 0.483492, "boundary_entropy": 0.686945, "bpb": 7.131209}, "fi": {"chunks": 5524.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.293658, "cps": 43.15625, "mean_p": 0.490059, "boundary_entropy": 0.687261, "bpb": 7.002}, "ru": {"chunks": 2616.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.138047, "cps": 20.4375, "mean_p": 0.478855, "boundary_entropy": 0.689013, "bpb": 6.360667}, "el": {"chunks": 5629.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.274278, "cps": 43.976562, "mean_p": 0.487042, "boundary_entropy": 0.689872, "bpb": 7.040732}, "hi": {"chunks": 404.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.024012, "cps": 3.15625, "mean_p": 0.466984, "boundary_entropy": 0.688445, "bpb": 7.056598}, "ta": {"chunks": 14610.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.781492, "cps": 114.140625, "mean_p": 0.49018, "boundary_entropy": 0.690531, "bpb": 6.621792}, "th": {"chunks": 17447.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.084675, "cps": 136.304688, "mean_p": 0.496557, "boundary_entropy": 0.690724, "bpb": 6.370019}, "zh": {"chunks": 10879.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.842025, "cps": 84.992188, "mean_p": 0.515377, "boundary_entropy": 0.685938, "bpb": 7.770503}, "my": {"chunks": 42518.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.243931, "cps": 332.171875, "mean_p": 0.515696, "boundary_entropy": 0.690552, "bpb": 6.672612}}, "bpb": {"en": 6.81228, "de": 6.90478, "vi": 7.15803, "tr": 7.13121, "fi": 7.002, "ru": 6.36067, "el": 7.04073, "hi": 7.0566, "ta": 6.62179, "th": 6.37002, "zh": 7.7705, "my": 6.67261}, "gini_cpc": 0.523296, "gini_cps": 0.49206, "bpic": 3.167163, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.488344640464893, "vi": 1.7238721668999848, "tr": 0.8539482059143753, "fi": 1.2974180207550612, "ru": 0.6099112789298008, "el": 1.2117932886696066, "hi": 0.10608755286318436, "ta": 3.452732126958041, "th": 4.7922320043390965, "zh": 8.138299595413187, "my": 9.913969783980882}, "t": 12.1}
{"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4870.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.279837, "cps": 38.046875, "mean_p": 0.495237, "boundary_entropy": 0.687434, "bpb": 6.360852}, "de": {"chunks": 8531.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.415478, "cps": 66.648438, "mean_p": 0.49897, "boundary_entropy": 0.68833, "bpb": 6.444319}, "vi": {"chunks": 5671.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.307621, "cps": 44.304688, "mean_p": 0.489076, "boundary_entropy": 0.688362, "bpb": 6.904147}, "tr": {"chunks": 6805.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.375905, "cps": 53.164062, "mean_p": 0.496495, "boundary_entropy": 0.687979, "bpb": 6.718635}, "fi": {"chunks": 7159.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.380575, "cps": 55.929688, "mean_p": 0.497718, "boundary_entropy": 0.688012, "bpb": 6.521123}, "ru": {"chunks": 5338.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.281689, "cps": 41.703125, "mean_p": 0.487966, "boundary_entropy": 0.689828, "bpb": 5.791418}, "el": {"chunks": 485.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.023632, "cps": 3.789062, "mean_p": 0.469279, "boundary_entropy": 0.68835, "bpb": 6.573441}, "hi": {"chunks": 2676.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.159049, "cps": 20.90625, "mean_p": 0.485673, "boundary_entropy": 0.690385, "bpb": 5.948979}, "ta": {"chunks": 5524.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.29548, "cps": 43.15625, "mean_p": 0.482429, "boundary_entropy": 0.690238, "bpb": 5.809561}, "th": {"chunks": 3723.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.231458, "cps": 29.085938, "mean_p": 0.487887, "boundary_entropy": 0.690626, "bpb": 5.621182}, "zh": {"chunks": 7412.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.254995, "cps": 57.90625, "mean_p": 0.501043, "boundary_entropy": 0.686376, "bpb": 7.331264}, "my": {"chunks": 25506.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.346105, "cps": 199.265625, "mean_p": 0.496512, "boundary_entropy": 0.690916, "bpb": 5.630075}}, "bpb": {"en": 6.36085, "de": 6.44432, "vi": 6.90415, "tr": 6.71864, "fi": 6.52112, "ru": 5.79142, "el": 6.57344, "hi": 5.94898, "ta": 5.80956, "th": 5.62118, "zh": 7.33126, "my": 5.63008}, "gini_cpc": 0.41259, "gini_cps": 0.368122, "bpic": 4.585197, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.484713624214479, "vi": 1.0992884483971377, "tr": 1.3432991412021629, "fi": 1.35998975871189, "ru": 1.006617587621158, "el": 0.08444929968390251, "hi": 0.5683635198672163, "ta": 1.055901384650279, "th": 0.827117348225131, "zh": 4.484738521574482, "my": 4.810321883863283}, "t": 14.4}
{"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.22433, "bpb": 6.09442, "chunk_loss": 0.031183, "F": 0.18106, "G": 0.4876, "chunks": 185.41, "bytes_per_chunk": 5.523, "boundary_entropy": 0.69168, "router_noise": 0.0, "grad_norm": 1.5909, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.237, "alpha": 0.03, "t": 15.7}
{"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4690.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.269494, "cps": 36.640625, "mean_p": 0.495912, "boundary_entropy": 0.687587, "bpb": 5.820111}, "de": {"chunks": 6099.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.297034, "cps": 47.648438, "mean_p": 0.49574, "boundary_entropy": 0.688412, "bpb": 5.934562}, "vi": {"chunks": 2873.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.155845, "cps": 22.445312, "mean_p": 0.480869, "boundary_entropy": 0.687901, "bpb": 6.313429}, "tr": {"chunks": 5022.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.277413, "cps": 39.234375, "mean_p": 0.493876, "boundary_entropy": 0.688057, "bpb": 6.227108}, "fi": {"chunks": 5382.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.286109, "cps": 42.046875, "mean_p": 0.495016, "boundary_entropy": 0.6881, "bpb": 5.973641}, "ru": {"chunks": 7552.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.398522, "cps": 59.0, "mean_p": 0.493521, "boundary_entropy": 0.6902, "bpb": 5.415087}, "el": {"chunks": 4168.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.203089, "cps": 32.5625, "mean_p": 0.490307, "boundary_entropy": 0.690334, "bpb": 5.661142}, "hi": {"chunks": 425.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.02526, "cps": 3.320312, "mean_p": 0.4737, "boundary_entropy": 0.689293, "bpb": 5.441916}, "ta": {"chunks": 2812.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.150415, "cps": 21.96875, "mean_p": 0.483562, "boundary_entropy": 0.690386, "bpb": 5.20753}, "th": {"chunks": 1539.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.095679, "cps": 12.023438, "mean_p": 0.48435, "boundary_entropy": 0.690448, "bpb": 5.100715}, "zh": {"chunks": 4653.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.787843, "cps": 36.351562, "mean_p": 0.49305, "boundary_entropy": 0.686644, "bpb": 6.993661}, "my": {"chunks": 21877.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.154581, "cps": 170.914062, "mean_p": 0.493964, "boundary_entropy": 0.690996, "bpb": 4.996093}}, "bpb": {"en": 5.82011, "de": 5.93456, "vi": 6.31343, "tr": 6.22711, "fi": 5.97364, "ru": 5.41509, "el": 5.66114, "hi": 5.44192, "ta": 5.20753, "th": 5.10072, "zh": 6.99366, "my": 4.99609}, "gini_cpc": 0.433841, "gini_cps": 0.407701, "bpic": 5.72022, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.102192632443463, "vi": 0.5782874422494062, "tr": 1.0293840493635862, "fi": 1.061654212892493, "ru": 1.478781621481736, "el": 0.7535952330564274, "hi": 0.09373140664750965, "ta": 0.5581373991996994, "th": 0.35503308980211074, "zh": 2.923417802863194, "my": 4.284258511567739}, "t": 17.5}
{"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.37965, "bpb": 4.87581, "chunk_loss": 0.030057, "F": 0.16687, "G": 0.47919, "chunks": 170.88, "bytes_per_chunk": 5.993, "boundary_entropy": 0.69024, "router_noise": 0.0, "grad_norm": 1.6762, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.079, "alpha": 0.03, "t": 19.7}
{"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3122.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.179394, "cps": 24.390625, "mean_p": 0.447977, "boundary_entropy": 0.675496, "bpb": 4.500981}, "de": {"chunks": 3626.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.176594, "cps": 28.328125, "mean_p": 0.449602, "boundary_entropy": 0.677613, "bpb": 4.70209}, "vi": {"chunks": 6160.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.334147, "cps": 48.125, "mean_p": 0.466085, "boundary_entropy": 0.682856, "bpb": 4.984991}, "tr": {"chunks": 3657.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.202011, "cps": 28.570312, "mean_p": 0.456302, "boundary_entropy": 0.679539, "bpb": 5.012739}, "fi": {"chunks": 4337.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.230557, "cps": 33.882812, "mean_p": 0.453561, "boundary_entropy": 0.677777, "bpb": 4.720152}, "ru": {"chunks": 5573.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.29409, "cps": 43.539062, "mean_p": 0.466158, "boundary_entropy": 0.685417, "bpb": 3.672206}, "el": {"chunks": 5613.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.273498, "cps": 43.851562, "mean_p": 0.471631, "boundary_entropy": 0.687395, "bpb": 4.599649}, "hi": {"chunks": 10003.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.594532, "cps": 78.148438, "mean_p": 0.490201, "boundary_entropy": 0.68966, "bpb": 3.647087}, "ta": {"chunks": 6326.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.338379, "cps": 49.421875, "mean_p": 0.484174, "boundary_entropy": 0.689589, "bpb": 3.372903}, "th": {"chunks": 12891.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.80143, "cps": 100.710938, "mean_p": 0.471254, "boundary_entropy": 0.687321, "bpb": 3.148725}, "zh": {"chunks": 2992.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506603, "cps": 23.375, "mean_p": 0.480834, "boundary_entropy": 0.685183, "bpb": 6.312779}, "my": {"chunks": 13266.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.700127, "cps": 103.640625, "mean_p": 0.470281, "boundary_entropy": 0.688191, "bpb": 4.012571}}, "bpb": {"en": 4.50098, "de": 4.70209, "vi": 4.98499, "tr": 5.01274, "fi": 4.72015, "ru": 3.67221, "el": 4.59965, "hi": 3.64709, "ta": 3.3729, "th": 3.14873, "zh": 6.31278, "my": 4.01257}, "gini_cpc": 0.286496, "gini_cps": 0.284089, "bpic": 4.947799, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9843886872253546, "vi": 1.8626394275852256, "tr": 1.126070627316116, "fi": 1.2851942095978774, "ru": 1.639347603778783, "el": 1.5245631508630977, "hi": 3.314106170750071, "ta": 1.8862312666375287, "th": 4.46741979916511, "zh": 2.8239653785887073, "my": 3.902723993120793}, "t": 22.0}
{"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.89272, "bpb": 4.17332, "chunk_loss": 0.029735, "F": 0.16159, "G": 0.43193, "chunks": 165.47, "bytes_per_chunk": 6.188, "boundary_entropy": 0.67168, "router_noise": 0.0, "grad_norm": 1.6704, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.071, "alpha": 0.03, "t": 23.2}
{"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2188.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.125725, "cps": 17.09375, "mean_p": 0.37003, "boundary_entropy": 0.623888, "bpb": 3.888371}, "de": {"chunks": 2523.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.122875, "cps": 19.710938, "mean_p": 0.374745, "boundary_entropy": 0.631069, "bpb": 4.112904}, "vi": {"chunks": 4328.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234771, "cps": 33.8125, "mean_p": 0.399445, "boundary_entropy": 0.621135, "bpb": 3.658745}, "tr": {"chunks": 3010.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.166271, "cps": 23.515625, "mean_p": 0.393531, "boundary_entropy": 0.64135, "bpb": 4.166064}, "fi": {"chunks": 3334.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177237, "cps": 26.046875, "mean_p": 0.37481, "boundary_entropy": 0.627603, "bpb": 4.131318}, "ru": {"chunks": 4316.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.227757, "cps": 33.71875, "mean_p": 0.453139, "boundary_entropy": 0.67522, "bpb": 2.827726}, "el": {"chunks": 6069.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.295717, "cps": 47.414062, "mean_p": 0.40451, "boundary_entropy": 0.65476, "bpb": 3.066706}, "hi": {"chunks": 10035.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.596434, "cps": 78.398438, "mean_p": 0.408294, "boundary_entropy": 0.648582, "bpb": 2.396659}, "ta": {"chunks": 8466.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.452848, "cps": 66.140625, "mean_p": 0.465579, "boundary_entropy": 0.68025, "bpb": 2.304827}, "th": {"chunks": 5694.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.353994, "cps": 44.484375, "mean_p": 0.446164, "boundary_entropy": 0.678737, "bpb": 2.318348}, "zh": {"chunks": 3613.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.611751, "cps": 28.226562, "mean_p": 0.471466, "boundary_entropy": 0.681717, "bpb": 5.296948}, "my": {"chunks": 5142.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.271374, "cps": 40.171875, "mean_p": 0.448927, "boundary_entropy": 0.683107, "bpb": 2.453691}}, "bpb": {"en": 3.88837, "de": 4.1129, "vi": 3.65874, "tr": 4.16606, "fi": 4.13132, "ru": 2.82773, "el": 3.06671, "hi": 2.39666, "ta": 2.30483, "th": 2.31835, "zh": 5.29695, "my": 2.45369}, "gini_cpc": 0.292026, "gini_cps": 0.252075, "bpic": 6.536003, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9773309358609509, "vi": 1.8673293041562602, "tr": 1.3224910654965325, "fi": 1.4097123828807232, "ru": 1.8115445726992516, "el": 2.3520854472722936, "hi": 4.743939114065373, "ta": 3.6018829639459167, "th": 2.815614545442146, "zh": 4.865767143505411, "my": 2.158467424943239}, "t": 26.5}
{"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.89304, "de": 4.11455, "vi": 3.66409, "tr": 4.17539, "fi": 4.13399, "ru": 2.81262, "el": 3.06067, "hi": 2.39258, "ta": 2.32289, "th": 2.35101, "zh": 5.30543, "my": 2.47118}, "bytes_per_chunk": {"en": 7.167, "de": 8.059, "vi": 5.593, "tr": 7.084, "fi": 6.149, "ru": 7.773, "el": 5.976, "hi": 4.311, "ta": 5.865, "th": 7.253, "zh": 5.955, "my": 9.798}, "bpb_mean": 3.39145, "t": 27.8}
{"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.49464, "bpb": 3.599, "chunk_loss": 0.028899, "F": 0.14569, "G": 0.40243, "chunks": 149.19, "bytes_per_chunk": 6.864, "boundary_entropy": 0.64892, "router_noise": 0.0, "grad_norm": 1.4946, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 27.9}
{"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.5629, "bpb": 3.69748, "chunk_loss": 0.032516, "F": 0.2131, "G": 0.40823, "chunks": 218.22, "bytes_per_chunk": 4.693, "boundary_entropy": 0.64359, "router_noise": 0.0, "grad_norm": 1.302, "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": 2358.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.135494, "cps": 18.421875, "mean_p": 0.346566, "boundary_entropy": 0.597902, "bpb": 3.640094}, "de": {"chunks": 2900.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.141236, "cps": 22.65625, "mean_p": 0.355521, "boundary_entropy": 0.60953, "bpb": 3.907296}, "vi": {"chunks": 4380.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237592, "cps": 34.21875, "mean_p": 0.420122, "boundary_entropy": 0.6379, "bpb": 3.22282}, "tr": {"chunks": 3585.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.198033, "cps": 28.007812, "mean_p": 0.373009, "boundary_entropy": 0.610943, "bpb": 3.855173}, "fi": {"chunks": 2480.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.131838, "cps": 19.375, "mean_p": 0.340538, "boundary_entropy": 0.592614, "bpb": 3.889235}, "ru": {"chunks": 4388.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.231557, "cps": 34.28125, "mean_p": 0.450752, "boundary_entropy": 0.671687, "bpb": 2.648602}, "el": {"chunks": 5147.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.250792, "cps": 40.210938, "mean_p": 0.386985, "boundary_entropy": 0.646324, "bpb": 2.758954}, "hi": {"chunks": 8233.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.489331, "cps": 64.320312, "mean_p": 0.374216, "boundary_entropy": 0.614455, "bpb": 2.153349}, "ta": {"chunks": 9069.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.485103, "cps": 70.851562, "mean_p": 0.41857, "boundary_entropy": 0.649786, "bpb": 1.90803}, "th": {"chunks": 4499.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.279702, "cps": 35.148438, "mean_p": 0.437455, "boundary_entropy": 0.673805, "bpb": 2.082161}, "zh": {"chunks": 4115.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.696749, "cps": 32.148438, "mean_p": 0.4648, "boundary_entropy": 0.674754, "bpb": 4.972556}, "my": {"chunks": 10446.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.551298, "cps": 81.609375, "mean_p": 0.425627, "boundary_entropy": 0.669391, "bpb": 2.161537}}, "bpb": {"en": 3.64009, "de": 3.9073, "vi": 3.22282, "tr": 3.85517, "fi": 3.88924, "ru": 2.6486, "el": 2.75895, "hi": 2.15335, "ta": 1.90803, "th": 2.08216, "zh": 4.97256, "my": 2.16154}, "gini_cpc": 0.304408, "gini_cps": 0.263217, "bpic": 6.230211, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.042379616304369, "vi": 1.7535222786062854, "tr": 1.4615676706614777, "fi": 0.9730163026532984, "ru": 1.708982926812893, "el": 1.850945547470541, "hi": 3.611464597771552, "ta": 3.580257406458436, "th": 2.064311573139199, "zh": 5.14229179107114, "my": 4.068805827761636}, "t": 31.3}
{"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.35271, "bpb": 3.39424, "chunk_loss": 0.030821, "F": 0.1853, "G": 0.3771, "chunks": 189.75, "bytes_per_chunk": 5.397, "boundary_entropy": 0.61464, "router_noise": 0.0, "grad_norm": 1.0568, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.067, "alpha": 0.03, "t": 33.4}
{"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.04441, "bpb": 2.94946, "chunk_loss": 0.032543, "F": 0.21442, "G": 0.36369, "chunks": 219.56, "bytes_per_chunk": 4.664, "boundary_entropy": 0.59207, "router_noise": 0.0, "grad_norm": 0.8266, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 35.8}
{"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3543.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203586, "cps": 27.679688, "mean_p": 0.324582, "boundary_entropy": 0.538032, "bpb": 3.260117}, "de": {"chunks": 4724.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.230069, "cps": 36.90625, "mean_p": 0.335579, "boundary_entropy": 0.543647, "bpb": 3.530933}, "vi": {"chunks": 5874.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.318633, "cps": 45.890625, "mean_p": 0.338713, "boundary_entropy": 0.56275, "bpb": 2.794747}, "tr": {"chunks": 4728.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.261172, "cps": 36.9375, "mean_p": 0.341092, "boundary_entropy": 0.550266, "bpb": 3.432649}, "fi": {"chunks": 4423.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.235128, "cps": 34.554688, "mean_p": 0.344438, "boundary_entropy": 0.565396, "bpb": 3.579724}, "ru": {"chunks": 5633.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.297256, "cps": 44.007812, "mean_p": 0.301917, "boundary_entropy": 0.490772, "bpb": 2.514333}, "el": {"chunks": 5868.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.285923, "cps": 45.84375, "mean_p": 0.34776, "boundary_entropy": 0.585417, "bpb": 2.549509}, "hi": {"chunks": 5782.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.343655, "cps": 45.171875, "mean_p": 0.345364, "boundary_entropy": 0.597991, "bpb": 1.799663}, "ta": {"chunks": 4906.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.262423, "cps": 38.328125, "mean_p": 0.379208, "boundary_entropy": 0.633757, "bpb": 1.596363}, "th": {"chunks": 5141.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.319615, "cps": 40.164062, "mean_p": 0.326515, "boundary_entropy": 0.586577, "bpb": 1.903608}, "zh": {"chunks": 2793.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.472909, "cps": 21.820312, "mean_p": 0.343059, "boundary_entropy": 0.584308, "bpb": 4.151039}, "my": {"chunks": 6895.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.363891, "cps": 53.867188, "mean_p": 0.342983, "boundary_entropy": 0.589187, "bpb": 1.929362}}, "bpb": {"en": 3.26012, "de": 3.53093, "vi": 2.79475, "tr": 3.43265, "fi": 3.57972, "ru": 2.51433, "el": 2.54951, "hi": 1.79966, "ta": 1.59636, "th": 1.90361, "zh": 4.15104, "my": 1.92936}, "gini_cpc": 0.125172, "gini_cps": 0.116697, "bpic": 6.363472, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1300832805727365, "vi": 1.5651060423066878, "tr": 1.2828618326320977, "fi": 1.1549362793339206, "ru": 1.4601030386573697, "el": 1.4044368881603808, "hi": 1.688013759211286, "ta": 1.2890063067300548, "th": 1.5699271729858075, "zh": 2.322899716214484, "my": 1.7874086787190826}, "t": 38.9}
{"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.26805, "de": 3.52898, "vi": 2.80045, "tr": 3.43973, "fi": 3.57193, "ru": 2.49803, "el": 2.54297, "hi": 1.79606, "ta": 1.612, "th": 1.93286, "zh": 4.1592, "my": 1.94476}, "bytes_per_chunk": {"en": 4.772, "de": 4.393, "vi": 4.121, "tr": 4.129, "fi": 4.304, "ru": 6.225, "el": 6.243, "hi": 6.845, "ta": 8.317, "th": 7.023, "zh": 4.85, "my": 6.856}, "bpb_mean": 2.75792, "t": 40.3}
{"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.05453, "bpb": 2.96407, "chunk_loss": 0.031226, "F": 0.19855, "G": 0.33132, "chunks": 203.31, "bytes_per_chunk": 5.037, "boundary_entropy": 0.54967, "router_noise": 0.0, "grad_norm": 1.1058, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 40.3}
{"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.12714, "bpb": 3.06882, "chunk_loss": 0.030446, "F": 0.17886, "G": 0.32096, "chunks": 183.16, "bytes_per_chunk": 5.591, "boundary_entropy": 0.54539, "router_noise": 0.0, "grad_norm": 0.7715, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 42.5}
{"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.95705, "bpb": 2.82343, "chunk_loss": 0.031042, "F": 0.19601, "G": 0.32729, "chunks": 200.72, "bytes_per_chunk": 5.102, "boundary_entropy": 0.54123, "router_noise": 0.0, "grad_norm": 0.8817, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 44.7}
{"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.91907, "bpb": 2.76863, "chunk_loss": 0.030575, "F": 0.18433, "G": 0.30782, "chunks": 188.75, "bytes_per_chunk": 5.425, "boundary_entropy": 0.51658, "router_noise": 0.0, "grad_norm": 0.9193, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 46.9}
{"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2957.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.169913, "cps": 23.101562, "mean_p": 0.278056, "boundary_entropy": 0.467246, "bpb": 2.947112}, "de": {"chunks": 3691.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.179759, "cps": 28.835938, "mean_p": 0.280616, "boundary_entropy": 0.459734, "bpb": 3.27217}, "vi": {"chunks": 5436.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.294874, "cps": 42.46875, "mean_p": 0.293575, "boundary_entropy": 0.491472, "bpb": 2.570139}, "tr": {"chunks": 4558.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.251781, "cps": 35.609375, "mean_p": 0.313868, "boundary_entropy": 0.495129, "bpb": 3.160186}, "fi": {"chunks": 3529.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.187603, "cps": 27.570312, "mean_p": 0.285615, "boundary_entropy": 0.486486, "bpb": 3.381178}, "ru": {"chunks": 6191.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.326702, "cps": 48.367188, "mean_p": 0.32386, "boundary_entropy": 0.536824, "bpb": 2.23425}, "el": {"chunks": 6425.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.313063, "cps": 50.195312, "mean_p": 0.353304, "boundary_entropy": 0.593019, "bpb": 2.272318}, "hi": {"chunks": 6070.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.360773, "cps": 47.421875, "mean_p": 0.332949, "boundary_entropy": 0.564491, "bpb": 1.637105}, "ta": {"chunks": 5327.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.284942, "cps": 41.617188, "mean_p": 0.346999, "boundary_entropy": 0.594374, "bpb": 1.401118}, "th": {"chunks": 4060.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.252409, "cps": 31.71875, "mean_p": 0.292405, "boundary_entropy": 0.540962, "bpb": 1.755693}, "zh": {"chunks": 2868.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.485608, "cps": 22.40625, "mean_p": 0.348334, "boundary_entropy": 0.585704, "bpb": 3.879023}, "my": {"chunks": 13268.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.700232, "cps": 103.65625, "mean_p": 0.382878, "boundary_entropy": 0.598282, "bpb": 1.654737}}, "bpb": {"en": 2.94711, "de": 3.27217, "vi": 2.57014, "tr": 3.16019, "fi": 3.38118, "ru": 2.23425, "el": 2.27232, "hi": 1.63711, "ta": 1.40112, "th": 1.75569, "zh": 3.87902, "my": 1.65474}, "gini_cpc": 0.226842, "gini_cps": 0.234335, "bpic": 5.961184, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0579482723860258, "vi": 1.7354379961438058, "tr": 1.4818237974707948, "fi": 1.1041105777714049, "ru": 1.922756930248246, "el": 1.8424897585560653, "hi": 2.1232758191834322, "ta": 1.676988262628205, "th": 1.4855174714898627, "zh": 2.8579754904391548, "my": 4.121116411439199}, "t": 50.0}
{"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.95609, "de": 3.27018, "vi": 2.57805, "tr": 3.17224, "fi": 3.36971, "ru": 2.21834, "el": 2.26482, "hi": 1.63157, "ta": 1.41561, "th": 1.78504, "zh": 3.89187, "my": 1.6662}, "bytes_per_chunk": {"en": 5.814, "de": 5.511, "vi": 4.847, "tr": 4.104, "fi": 5.9, "ru": 5.562, "el": 5.655, "hi": 5.95, "ta": 7.414, "th": 9.08, "zh": 5.117, "my": 4.334}, "bpb_mean": 2.51831, "t": 51.3}
{"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.94835, "bpb": 2.81088, "chunk_loss": 0.030376, "F": 0.17639, "G": 0.30611, "chunks": 180.62, "bytes_per_chunk": 5.669, "boundary_entropy": 0.51013, "router_noise": 0.0, "grad_norm": 0.8575, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 51.4}
{"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.73875, "bpb": 2.50849, "chunk_loss": 0.030166, "F": 0.17108, "G": 0.31685, "chunks": 175.19, "bytes_per_chunk": 5.845, "boundary_entropy": 0.52452, "router_noise": 0.0, "grad_norm": 0.8181, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 53.5}
{"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.75936, "bpb": 2.53822, "chunk_loss": 0.030625, "F": 0.18356, "G": 0.31583, "chunks": 187.97, "bytes_per_chunk": 5.448, "boundary_entropy": 0.51815, "router_noise": 0.0, "grad_norm": 0.6683, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 55.7}
{"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.69882, "bpb": 2.45088, "chunk_loss": 0.029936, "F": 0.16568, "G": 0.31532, "chunks": 169.66, "bytes_per_chunk": 6.036, "boundary_entropy": 0.52223, "router_noise": 0.0, "grad_norm": 0.6439, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 57.9}
{"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3126.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.179624, "cps": 24.421875, "mean_p": 0.276437, "boundary_entropy": 0.441525, "bpb": 2.73885}, "de": {"chunks": 3554.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.173087, "cps": 27.765625, "mean_p": 0.273974, "boundary_entropy": 0.443329, "bpb": 3.078063}, "vi": {"chunks": 4796.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.260157, "cps": 37.46875, "mean_p": 0.276364, "boundary_entropy": 0.47003, "bpb": 2.431378}, "tr": {"chunks": 3907.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215821, "cps": 30.523438, "mean_p": 0.313483, "boundary_entropy": 0.485981, "bpb": 3.019434}, "fi": {"chunks": 3429.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182287, "cps": 26.789062, "mean_p": 0.278516, "boundary_entropy": 0.458572, "bpb": 3.248636}, "ru": {"chunks": 5838.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.308074, "cps": 45.609375, "mean_p": 0.346447, "boundary_entropy": 0.580501, "bpb": 2.066938}, "el": {"chunks": 6051.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.29484, "cps": 47.273438, "mean_p": 0.318043, "boundary_entropy": 0.553578, "bpb": 2.179537}, "hi": {"chunks": 6878.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.408796, "cps": 53.734375, "mean_p": 0.317382, "boundary_entropy": 0.536503, "bpb": 1.556079}, "ta": {"chunks": 7556.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.404172, "cps": 59.03125, "mean_p": 0.399894, "boundary_entropy": 0.633397, "bpb": 1.290381}, "th": {"chunks": 4778.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.297047, "cps": 37.328125, "mean_p": 0.358374, "boundary_entropy": 0.605574, "bpb": 1.641762}, "zh": {"chunks": 2962.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.501524, "cps": 23.140625, "mean_p": 0.341127, "boundary_entropy": 0.572626, "bpb": 3.757687}, "my": {"chunks": 12122.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.639751, "cps": 94.703125, "mean_p": 0.380825, "boundary_entropy": 0.608739, "bpb": 1.479215}}, "bpb": {"en": 2.73885, "de": 3.07806, "vi": 2.43138, "tr": 3.01943, "fi": 3.24864, "ru": 2.06694, "el": 2.17954, "hi": 1.55608, "ta": 1.29038, "th": 1.64176, "zh": 3.75769, "my": 1.47922}, "gini_cpc": 0.230771, "gini_cps": 0.234715, "bpic": 5.904596, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9636074806035283, "vi": 1.4483421806242507, "tr": 1.20151170583597, "fi": 1.0148240437176312, "ru": 1.7151022743962037, "el": 1.6414265517186233, "hi": 2.2758427187401074, "ta": 2.2500989980077195, "th": 1.653713328365324, "zh": 2.7920729301604865, "my": 3.561607442046664}, "t": 60.9}
{"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.74901, "de": 3.07841, "vi": 2.437, "tr": 3.03126, "fi": 3.24598, "ru": 2.05147, "el": 2.16821, "hi": 1.55095, "ta": 1.3006, "th": 1.66915, "zh": 3.77001, "my": 1.49069}, "bytes_per_chunk": {"en": 5.644, "de": 5.703, "vi": 5.069, "tr": 4.757, "fi": 5.662, "ru": 5.89, "el": 6.307, "hi": 5.772, "ta": 5.984, "th": 8.584, "zh": 5.385, "my": 5.112}, "bpb_mean": 2.37856, "t": 62.3}
{"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.82188, "bpb": 2.62842, "chunk_loss": 0.030073, "F": 0.17209, "G": 0.30712, "chunks": 176.22, "bytes_per_chunk": 5.811, "boundary_entropy": 0.50603, "router_noise": 0.0, "grad_norm": 0.7776, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 62.3}
{"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.87209, "bpb": 2.70085, "chunk_loss": 0.030413, "F": 0.18039, "G": 0.30903, "chunks": 184.72, "bytes_per_chunk": 5.544, "boundary_entropy": 0.49886, "router_noise": 0.0, "grad_norm": 0.8689, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 64.5}
{"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.68278, "bpb": 2.42773, "chunk_loss": 0.030215, "F": 0.1781, "G": 0.30603, "chunks": 182.38, "bytes_per_chunk": 5.615, "boundary_entropy": 0.48759, "router_noise": 0.0, "grad_norm": 0.9587, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 66.7}
{"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.81677, "bpb": 2.62104, "chunk_loss": 0.030381, "F": 0.18005, "G": 0.30648, "chunks": 184.38, "bytes_per_chunk": 5.554, "boundary_entropy": 0.48323, "router_noise": 0.0, "grad_norm": 0.81, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 68.8}
{"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3445.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197954, "cps": 26.914062, "mean_p": 0.275047, "boundary_entropy": 0.406401, "bpb": 2.607226}, "de": {"chunks": 3806.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18536, "cps": 29.734375, "mean_p": 0.276753, "boundary_entropy": 0.427319, "bpb": 2.958297}, "vi": {"chunks": 4074.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.220993, "cps": 31.828125, "mean_p": 0.260113, "boundary_entropy": 0.457583, "bpb": 2.328269}, "tr": {"chunks": 3660.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.202176, "cps": 28.59375, "mean_p": 0.279574, "boundary_entropy": 0.429128, "bpb": 2.923251}, "fi": {"chunks": 3958.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.210409, "cps": 30.921875, "mean_p": 0.289111, "boundary_entropy": 0.43479, "bpb": 3.140842}, "ru": {"chunks": 6664.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.351662, "cps": 52.0625, "mean_p": 0.361047, "boundary_entropy": 0.58656, "bpb": 1.965126}, "el": {"chunks": 6580.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.320616, "cps": 51.40625, "mean_p": 0.322643, "boundary_entropy": 0.541983, "bpb": 2.044101}, "hi": {"chunks": 7740.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.46003, "cps": 60.46875, "mean_p": 0.36305, "boundary_entropy": 0.59052, "bpb": 1.503433}, "ta": {"chunks": 8409.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449799, "cps": 65.695312, "mean_p": 0.426726, "boundary_entropy": 0.646252, "bpb": 1.225617}, "th": {"chunks": 8743.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.54355, "cps": 68.304688, "mean_p": 0.434856, "boundary_entropy": 0.655763, "bpb": 1.532675}, "zh": {"chunks": 2490.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.421605, "cps": 19.453125, "mean_p": 0.339574, "boundary_entropy": 0.569309, "bpb": 3.636351}, "my": {"chunks": 13019.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.687091, "cps": 101.710938, "mean_p": 0.396442, "boundary_entropy": 0.633225, "bpb": 1.417753}}, "bpb": {"en": 2.60723, "de": 2.9583, "vi": 2.32827, "tr": 2.92325, "fi": 3.14084, "ru": 1.96513, "el": 2.0441, "hi": 1.50343, "ta": 1.22562, "th": 1.53267, "zh": 3.63635, "my": 1.41775}, "gini_cpc": 0.240404, "gini_cps": 0.260647, "bpic": 5.287114, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9363781493163648, "vi": 1.1163818744161713, "tr": 1.0213284457499165, "fi": 1.0629156472770658, "ru": 1.7764814135433942, "el": 1.6196454019126971, "hi": 2.323917903285982, "ta": 2.272237782352181, "th": 2.745834181888399, "zh": 2.1298096889979785, "my": 3.4709562922648587}, "t": 72.0}
{"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.61585, "de": 2.96069, "vi": 2.3337, "tr": 2.93231, "fi": 3.14271, "ru": 1.94913, "el": 2.03301, "hi": 1.49951, "ta": 1.23788, "th": 1.55825, "zh": 3.64673, "my": 1.42821}, "bytes_per_chunk": {"en": 5.194, "de": 5.387, "vi": 6.054, "tr": 5.255, "fi": 4.834, "ru": 5.253, "el": 5.609, "hi": 5.361, "ta": 5.775, "th": 5.043, "zh": 6.272, "my": 4.535}, "bpb_mean": 2.27816, "t": 73.3}
{"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.74602, "bpb": 2.51898, "chunk_loss": 0.03067, "F": 0.18683, "G": 0.32267, "chunks": 191.31, "bytes_per_chunk": 5.352, "boundary_entropy": 0.50818, "router_noise": 0.0, "grad_norm": 0.8245, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 73.3}
{"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.61192, "bpb": 2.32551, "chunk_loss": 0.030824, "F": 0.19119, "G": 0.32967, "chunks": 195.78, "bytes_per_chunk": 5.23, "boundary_entropy": 0.51187, "router_noise": 0.0, "grad_norm": 0.6976, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 75.5}
{"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.80436, "bpb": 2.60313, "chunk_loss": 0.030477, "F": 0.18298, "G": 0.31784, "chunks": 187.38, "bytes_per_chunk": 5.465, "boundary_entropy": 0.4949, "router_noise": 0.0, "grad_norm": 0.7333, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 77.6}
{"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.66816, "bpb": 2.40665, "chunk_loss": 0.030437, "F": 0.18167, "G": 0.31013, "chunks": 186.03, "bytes_per_chunk": 5.504, "boundary_entropy": 0.47778, "router_noise": 0.0, "grad_norm": 0.8999, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 79.8}
{"event": "probe", "step": 1200, "phase": 1, "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": 0.257766, "boundary_entropy": 0.381752, "bpb": 2.505447}, "de": {"chunks": 3521.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.17148, "cps": 27.507812, "mean_p": 0.255634, "boundary_entropy": 0.399986, "bpb": 2.838305}, "vi": {"chunks": 4314.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234011, "cps": 33.703125, "mean_p": 0.277481, "boundary_entropy": 0.464441, "bpb": 2.295492}, "tr": {"chunks": 3604.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.199083, "cps": 28.15625, "mean_p": 0.268855, "boundary_entropy": 0.412431, "bpb": 2.80636}, "fi": {"chunks": 3664.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19478, "cps": 28.625, "mean_p": 0.273851, "boundary_entropy": 0.417528, "bpb": 3.059714}, "ru": {"chunks": 5261.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.277625, "cps": 41.101562, "mean_p": 0.327939, "boundary_entropy": 0.55962, "bpb": 1.872103}, "el": {"chunks": 6136.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.298982, "cps": 47.9375, "mean_p": 0.289499, "boundary_entropy": 0.51269, "bpb": 1.977267}, "hi": {"chunks": 7577.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.450342, "cps": 59.195312, "mean_p": 0.396795, "boundary_entropy": 0.626943, "bpb": 1.465463}, "ta": {"chunks": 8001.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.427975, "cps": 62.507812, "mean_p": 0.423123, "boundary_entropy": 0.64666, "bpb": 1.203699}, "th": {"chunks": 7745.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.481505, "cps": 60.507812, "mean_p": 0.408779, "boundary_entropy": 0.637867, "bpb": 1.477365}, "zh": {"chunks": 2712.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.459194, "cps": 21.1875, "mean_p": 0.34072, "boundary_entropy": 0.561285, "bpb": 3.578722}, "my": {"chunks": 8188.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.43213, "cps": 63.96875, "mean_p": 0.419619, "boundary_entropy": 0.655024, "bpb": 1.335951}}, "bpb": {"en": 2.50545, "de": 2.83831, "vi": 2.29549, "tr": 2.80636, "fi": 3.05971, "ru": 1.8721, "el": 1.97727, "hi": 1.46546, "ta": 1.2037, "th": 1.47736, "zh": 3.57872, "my": 1.33595}, "gini_cpc": 0.204671, "gini_cps": 0.208253, "bpic": 5.991616, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8961764033550913, "vi": 1.2229730462765451, "tr": 1.0404329977660365, "fi": 1.0179430187723209, "ru": 1.4509049894221397, "el": 1.5625157096871758, "hi": 2.3535428028325502, "ta": 2.2366533904859662, "th": 2.516403285473848, "zh": 2.399805969193033, "my": 2.2583660925016216}, "t": 82.8}
{"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.51611, "de": 2.84404, "vi": 2.29948, "tr": 2.81683, "fi": 3.06379, "ru": 1.85404, "el": 1.96652, "hi": 1.46455, "ta": 1.21144, "th": 1.50184, "zh": 3.58909, "my": 1.34763}, "bytes_per_chunk": {"en": 5.32, "de": 5.731, "vi": 5.733, "tr": 5.187, "fi": 5.133, "ru": 6.684, "el": 6.011, "hi": 5.417, "ta": 6.173, "th": 5.96, "zh": 6.046, "my": 7.034}, "bpb_mean": 2.20628, "t": 84.1}
{"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.72464, "bpb": 2.48813, "chunk_loss": 0.03034, "F": 0.17999, "G": 0.29338, "chunks": 184.31, "bytes_per_chunk": 5.556, "boundary_entropy": 0.45704, "router_noise": 0.0, "grad_norm": 0.8759, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 84.2}
{"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.8365, "bpb": 2.64951, "chunk_loss": 0.030441, "F": 0.18359, "G": 0.29657, "chunks": 188.0, "bytes_per_chunk": 5.447, "boundary_entropy": 0.45483, "router_noise": 0.0, "grad_norm": 0.8221, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 86.3}
{"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.52288, "bpb": 2.19705, "chunk_loss": 0.030251, "F": 0.17514, "G": 0.31822, "chunks": 179.34, "bytes_per_chunk": 5.71, "boundary_entropy": 0.49673, "router_noise": 0.0, "grad_norm": 0.7643, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 88.3}
{"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.52045, "bpb": 2.19355, "chunk_loss": 0.030471, "F": 0.18149, "G": 0.3149, "chunks": 185.84, "bytes_per_chunk": 5.51, "boundary_entropy": 0.4876, "router_noise": 0.0, "grad_norm": 0.7941, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 90.3}
{"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3419.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19646, "cps": 26.710938, "mean_p": 0.258718, "boundary_entropy": 0.366825, "bpb": 2.43401}, "de": {"chunks": 3840.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187016, "cps": 30.0, "mean_p": 0.261722, "boundary_entropy": 0.3888, "bpb": 2.759716}, "vi": {"chunks": 3751.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.203472, "cps": 29.304688, "mean_p": 0.271897, "boundary_entropy": 0.45753, "bpb": 2.208197}, "tr": {"chunks": 3546.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195879, "cps": 27.703125, "mean_p": 0.255019, "boundary_entropy": 0.382016, "bpb": 2.730885}, "fi": {"chunks": 3322.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176599, "cps": 25.953125, "mean_p": 0.249227, "boundary_entropy": 0.378669, "bpb": 3.002421}, "ru": {"chunks": 6254.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.330026, "cps": 48.859375, "mean_p": 0.355766, "boundary_entropy": 0.56893, "bpb": 1.805643}, "el": {"chunks": 5578.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.271793, "cps": 43.578125, "mean_p": 0.309351, "boundary_entropy": 0.543689, "bpb": 1.903299}, "hi": {"chunks": 6683.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.397207, "cps": 52.210938, "mean_p": 0.396701, "boundary_entropy": 0.633787, "bpb": 1.432353}, "ta": {"chunks": 7200.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.38513, "cps": 56.25, "mean_p": 0.405911, "boundary_entropy": 0.639508, "bpb": 1.167362}, "th": {"chunks": 6497.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.403917, "cps": 50.757812, "mean_p": 0.384545, "boundary_entropy": 0.624112, "bpb": 1.429609}, "zh": {"chunks": 2730.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.462242, "cps": 21.328125, "mean_p": 0.339315, "boundary_entropy": 0.561665, "bpb": 3.544054}, "my": {"chunks": 9666.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.510133, "cps": 75.515625, "mean_p": 0.41829, "boundary_entropy": 0.651057, "bpb": 1.30033}}, "bpb": {"en": 2.43401, "de": 2.75972, "vi": 2.2082, "tr": 2.73088, "fi": 3.00242, "ru": 1.80564, "el": 1.9033, "hi": 1.43235, "ta": 1.16736, "th": 1.42961, "zh": 3.54405, "my": 1.30033}, "gini_cpc": 0.206304, "gini_cps": 0.210778, "bpic": 6.141872, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9519274197278389, "vi": 1.0356879935058738, "tr": 0.9970414170113021, "fi": 0.8989028638800748, "ru": 1.679862293670093, "el": 1.3834475041000447, "hi": 2.0218149689676648, "ta": 1.9603429101302206, "th": 2.055970225448437, "zh": 2.3528499083872947, "my": 2.596620216978436}, "t": 93.4}
{"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.44047, "de": 2.76419, "vi": 2.21118, "tr": 2.73983, "fi": 3.0007, "ru": 1.78769, "el": 1.89139, "hi": 1.4291, "ta": 1.17608, "th": 1.45037, "zh": 3.5512, "my": 1.31114}, "bytes_per_chunk": {"en": 5.154, "de": 5.308, "vi": 6.416, "tr": 5.303, "fi": 5.551, "ru": 5.587, "el": 6.674, "hi": 6.01, "ta": 6.681, "th": 7.021, "zh": 6.087, "my": 6.067}, "bpb_mean": 2.14611, "t": 94.7}
{"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.5747, "bpb": 2.27181, "chunk_loss": 0.030365, "F": 0.1839, "G": 0.31124, "chunks": 188.31, "bytes_per_chunk": 5.438, "boundary_entropy": 0.47221, "router_noise": 0.0, "grad_norm": 0.7232, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 94.8}
{"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.67193, "bpb": 2.41209, "chunk_loss": 0.030499, "F": 0.1839, "G": 0.31143, "chunks": 188.31, "bytes_per_chunk": 5.438, "boundary_entropy": 0.47917, "router_noise": 0.0, "grad_norm": 0.7367, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 96.9}
{"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.74535, "bpb": 2.51801, "chunk_loss": 0.030506, "F": 0.18265, "G": 0.31297, "chunks": 187.03, "bytes_per_chunk": 5.475, "boundary_entropy": 0.49268, "router_noise": 0.0, "grad_norm": 0.695, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 99.1}
{"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.60881, "bpb": 2.32103, "chunk_loss": 0.030096, "F": 0.17554, "G": 0.29645, "chunks": 179.75, "bytes_per_chunk": 5.697, "boundary_entropy": 0.45756, "router_noise": 0.0, "grad_norm": 0.7856, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 101.3}
{"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3300.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189622, "cps": 25.78125, "mean_p": 0.246271, "boundary_entropy": 0.34514, "bpb": 2.376888}, "de": {"chunks": 3782.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.184191, "cps": 29.546875, "mean_p": 0.254915, "boundary_entropy": 0.369177, "bpb": 2.663505}, "vi": {"chunks": 4301.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.233306, "cps": 33.601562, "mean_p": 0.273038, "boundary_entropy": 0.448958, "bpb": 2.167378}, "tr": {"chunks": 3893.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215047, "cps": 30.414062, "mean_p": 0.257062, "boundary_entropy": 0.366628, "bpb": 2.656147}, "fi": {"chunks": 3657.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194408, "cps": 28.570312, "mean_p": 0.260136, "boundary_entropy": 0.372167, "bpb": 2.882762}, "ru": {"chunks": 6287.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.331768, "cps": 49.117188, "mean_p": 0.349386, "boundary_entropy": 0.555685, "bpb": 1.748073}, "el": {"chunks": 6450.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.314282, "cps": 50.390625, "mean_p": 0.327708, "boundary_entropy": 0.544786, "bpb": 1.85989}, "hi": {"chunks": 6375.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.3789, "cps": 49.804688, "mean_p": 0.364835, "boundary_entropy": 0.608628, "bpb": 1.409249}, "ta": {"chunks": 9367.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.501043, "cps": 73.179688, "mean_p": 0.402786, "boundary_entropy": 0.633979, "bpb": 1.162191}, "th": {"chunks": 8470.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.526578, "cps": 66.171875, "mean_p": 0.385987, "boundary_entropy": 0.618543, "bpb": 1.372577}, "zh": {"chunks": 2648.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.448358, "cps": 20.6875, "mean_p": 0.335957, "boundary_entropy": 0.554682, "bpb": 3.513941}, "my": {"chunks": 7189.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.379407, "cps": 56.164062, "mean_p": 0.403153, "boundary_entropy": 0.64585, "bpb": 1.262484}}, "bpb": {"en": 2.37689, "de": 2.6635, "vi": 2.16738, "tr": 2.65615, "fi": 2.88276, "ru": 1.74807, "el": 1.85989, "hi": 1.40925, "ta": 1.16219, "th": 1.37258, "zh": 3.51394, "my": 1.26248}, "gini_cpc": 0.205777, "gini_cps": 0.212946, "bpic": 5.839727, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9713579470741113, "vi": 1.2303721182533225, "tr": 1.1340808906610154, "fi": 1.0252346064440052, "ru": 1.749622787239146, "el": 1.6574065460924108, "hi": 1.9981831689855463, "ta": 2.6423188990736466, "th": 2.7769785514454455, "zh": 2.3644749561308993, "my": 2.00085348417679}, "t": 104.4}
{"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.38543, "de": 2.67132, "vi": 2.17371, "tr": 2.6678, "fi": 2.88939, "ru": 1.73012, "el": 1.84927, "hi": 1.40541, "ta": 1.17108, "th": 1.39125, "zh": 3.51963, "my": 1.27308}, "bytes_per_chunk": {"en": 5.298, "de": 5.409, "vi": 5.449, "tr": 4.923, "fi": 5.071, "ru": 5.448, "el": 5.775, "hi": 6.395, "ta": 5.195, "th": 5.578, "zh": 6.37, "my": 7.561}, "bpb_mean": 2.09396, "t": 105.7}
{"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.60302, "bpb": 2.31267, "chunk_loss": 0.030577, "F": 0.18689, "G": 0.30674, "chunks": 191.38, "bytes_per_chunk": 5.351, "boundary_entropy": 0.46434, "router_noise": 0.0, "grad_norm": 0.7745, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 105.7}
{"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.58619, "bpb": 2.28839, "chunk_loss": 0.030277, "F": 0.17938, "G": 0.29838, "chunks": 183.69, "bytes_per_chunk": 5.575, "boundary_entropy": 0.45088, "router_noise": 0.0, "grad_norm": 0.8317, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 107.9}
{"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.54447, "bpb": 2.2282, "chunk_loss": 0.0304, "F": 0.18265, "G": 0.30516, "chunks": 187.03, "bytes_per_chunk": 5.475, "boundary_entropy": 0.46292, "router_noise": 0.0, "grad_norm": 0.7086, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 110.0}
{"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.56352, "bpb": 2.25568, "chunk_loss": 0.030699, "F": 0.18878, "G": 0.30134, "chunks": 193.31, "bytes_per_chunk": 5.297, "boundary_entropy": 0.45045, "router_noise": 0.0, "grad_norm": 0.6916, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 112.2}
{"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3291.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189105, "cps": 25.710938, "mean_p": 0.238836, "boundary_entropy": 0.33295, "bpb": 2.329736}, "de": {"chunks": 3838.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186919, "cps": 29.984375, "mean_p": 0.251906, "boundary_entropy": 0.358287, "bpb": 2.626912}, "vi": {"chunks": 4159.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225603, "cps": 32.492188, "mean_p": 0.2793, "boundary_entropy": 0.448414, "bpb": 2.135864}, "tr": {"chunks": 3515.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.194167, "cps": 27.460938, "mean_p": 0.245401, "boundary_entropy": 0.356507, "bpb": 2.625674}, "fi": {"chunks": 3439.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182819, "cps": 26.867188, "mean_p": 0.245406, "boundary_entropy": 0.35702, "bpb": 2.840941}, "ru": {"chunks": 5622.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.296675, "cps": 43.921875, "mean_p": 0.320243, "boundary_entropy": 0.532445, "bpb": 1.708568}, "el": {"chunks": 5314.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.258929, "cps": 41.515625, "mean_p": 0.305668, "boundary_entropy": 0.523089, "bpb": 1.812597}, "hi": {"chunks": 5968.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.35471, "cps": 46.625, "mean_p": 0.370756, "boundary_entropy": 0.610927, "bpb": 1.380061}, "ta": {"chunks": 7462.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.399144, "cps": 58.296875, "mean_p": 0.375885, "boundary_entropy": 0.618176, "bpb": 1.108149}, "th": {"chunks": 8810.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.547715, "cps": 68.828125, "mean_p": 0.384608, "boundary_entropy": 0.613502, "bpb": 1.323137}, "zh": {"chunks": 2630.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.44531, "cps": 20.546875, "mean_p": 0.321855, "boundary_entropy": 0.535009, "bpb": 3.472733}, "my": {"chunks": 8162.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.430758, "cps": 63.765625, "mean_p": 0.399271, "boundary_entropy": 0.64663, "bpb": 1.245166}}, "bpb": {"en": 2.32974, "de": 2.62691, "vi": 2.13586, "tr": 2.62567, "fi": 2.84094, "ru": 1.70857, "el": 1.8126, "hi": 1.38006, "ta": 1.10815, "th": 1.32314, "zh": 3.47273, "my": 1.24517}, "gini_cpc": 0.212852, "gini_cps": 0.212656, "bpic": 6.169121, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9884365612666733, "vi": 1.1930043200394398, "tr": 1.0267649046217908, "fi": 0.9667552301556329, "ru": 1.5688371500109435, "el": 1.3692316924280563, "hi": 1.8757285098763208, "ta": 2.1106975843818256, "th": 2.8963502632012195, "zh": 2.3548244897807984, "my": 2.2778727136726356}, "t": 115.3}
{"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.3383, "de": 2.63198, "vi": 2.14126, "tr": 2.63599, "fi": 2.8428, "ru": 1.69061, "el": 1.80361, "hi": 1.37619, "ta": 1.11646, "th": 1.3465, "zh": 3.47664, "my": 1.25376}, "bytes_per_chunk": {"en": 5.25, "de": 5.31, "vi": 5.859, "tr": 5.406, "fi": 5.313, "ru": 6.052, "el": 6.9, "hi": 6.754, "ta": 6.654, "th": 5.193, "zh": 6.21, "my": 7.166}, "bpb_mean": 2.05451, "t": 116.6}
{"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.53041, "bpb": 2.20791, "chunk_loss": 0.030126, "F": 0.17456, "G": 0.29653, "chunks": 178.75, "bytes_per_chunk": 5.729, "boundary_entropy": 0.46337, "router_noise": 0.0, "grad_norm": 0.7378, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 116.6}
{"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.58464, "bpb": 2.28615, "chunk_loss": 0.030482, "F": 0.18658, "G": 0.28398, "chunks": 191.06, "bytes_per_chunk": 5.36, "boundary_entropy": 0.42113, "router_noise": 0.0, "grad_norm": 0.7315, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 118.8}
{"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.63394, "bpb": 2.35727, "chunk_loss": 0.03094, "F": 0.19415, "G": 0.30252, "chunks": 198.81, "bytes_per_chunk": 5.151, "boundary_entropy": 0.44107, "router_noise": 0.0, "grad_norm": 0.737, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 120.9}
{"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.48195, "bpb": 2.138, "chunk_loss": 0.029938, "F": 0.16855, "G": 0.30256, "chunks": 172.59, "bytes_per_chunk": 5.933, "boundary_entropy": 0.46997, "router_noise": 0.0, "grad_norm": 0.7261, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 123.1}
{"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3286.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188818, "cps": 25.671875, "mean_p": 0.233855, "boundary_entropy": 0.314994, "bpb": 2.294023}, "de": {"chunks": 3999.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19476, "cps": 31.242188, "mean_p": 0.255228, "boundary_entropy": 0.353307, "bpb": 2.563604}, "vi": {"chunks": 4248.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230431, "cps": 33.1875, "mean_p": 0.273109, "boundary_entropy": 0.429052, "bpb": 2.108491}, "tr": {"chunks": 3712.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205049, "cps": 29.0, "mean_p": 0.245315, "boundary_entropy": 0.340562, "bpb": 2.559391}, "fi": {"chunks": 3366.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.178938, "cps": 26.296875, "mean_p": 0.242045, "boundary_entropy": 0.345894, "bpb": 2.775525}, "ru": {"chunks": 5410.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.285488, "cps": 42.265625, "mean_p": 0.316622, "boundary_entropy": 0.52542, "bpb": 1.677421}, "el": {"chunks": 6371.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.310432, "cps": 49.773438, "mean_p": 0.329345, "boundary_entropy": 0.540245, "bpb": 1.737231}, "hi": {"chunks": 6995.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.41575, "cps": 54.648438, "mean_p": 0.392039, "boundary_entropy": 0.618097, "bpb": 1.351307}, "ta": {"chunks": 6724.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.359668, "cps": 52.53125, "mean_p": 0.38954, "boundary_entropy": 0.627455, "bpb": 1.104455}, "th": {"chunks": 7863.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.488841, "cps": 61.429688, "mean_p": 0.391752, "boundary_entropy": 0.62349, "bpb": 1.302197}, "zh": {"chunks": 2641.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.447172, "cps": 20.632812, "mean_p": 0.334736, "boundary_entropy": 0.54752, "bpb": 3.437204}, "my": {"chunks": 9342.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.493034, "cps": 72.984375, "mean_p": 0.412225, "boundary_entropy": 0.646724, "bpb": 1.222558}}, "bpb": {"en": 2.29402, "de": 2.5636, "vi": 2.10849, "tr": 2.55939, "fi": 2.77553, "ru": 1.67742, "el": 1.73723, "hi": 1.35131, "ta": 1.10446, "th": 1.3022, "zh": 3.4372, "my": 1.22256}, "gini_cpc": 0.205972, "gini_cps": 0.213223, "bpic": 6.00061, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0314675225982288, "vi": 1.2203879961193453, "tr": 1.0859603710141894, "fi": 0.9476736151679295, "ru": 1.5119750055002674, "el": 1.6440814187985993, "hi": 2.201857490782183, "ta": 1.9048412897756322, "th": 2.588950646585934, "zh": 2.3682716690104804, "my": 2.611157377123745}, "t": 126.2}
{"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.30187, "de": 2.56934, "vi": 2.11599, "tr": 2.56691, "fi": 2.78614, "ru": 1.655, "el": 1.72798, "hi": 1.34866, "ta": 1.11107, "th": 1.32149, "zh": 3.43962, "my": 1.23066}, "bytes_per_chunk": {"en": 5.32, "de": 5.15, "vi": 5.61, "tr": 5.192, "fi": 5.483, "ru": 6.472, "el": 5.851, "hi": 6.07, "ta": 7.438, "th": 5.839, "zh": 6.311, "my": 5.897}, "bpb_mean": 2.01456, "t": 127.5}
{"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.59774, "bpb": 2.30506, "chunk_loss": 0.030283, "F": 0.18237, "G": 0.28494, "chunks": 186.75, "bytes_per_chunk": 5.483, "boundary_entropy": 0.42186, "router_noise": 0.0, "grad_norm": 0.7495, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 127.5}
{"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.61211, "bpb": 2.32579, "chunk_loss": 0.03062, "F": 0.18756, "G": 0.30742, "chunks": 192.06, "bytes_per_chunk": 5.332, "boundary_entropy": 0.46283, "router_noise": 0.0, "grad_norm": 0.7714, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 129.7}
{"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.54944, "bpb": 2.23536, "chunk_loss": 0.030439, "F": 0.18314, "G": 0.29077, "chunks": 187.53, "bytes_per_chunk": 5.46, "boundary_entropy": 0.43107, "router_noise": 0.0, "grad_norm": 0.7544, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 131.8}
{"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.50705, "bpb": 2.17421, "chunk_loss": 0.030367, "F": 0.18286, "G": 0.29348, "chunks": 187.25, "bytes_per_chunk": 5.469, "boundary_entropy": 0.4344, "router_noise": 0.0, "grad_norm": 0.7054, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 134.0}
{"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3401.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195426, "cps": 26.570312, "mean_p": 0.236227, "boundary_entropy": 0.311591, "bpb": 2.253811}, "de": {"chunks": 3887.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189305, "cps": 30.367188, "mean_p": 0.249513, "boundary_entropy": 0.349754, "bpb": 2.523271}, "vi": {"chunks": 4103.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222566, "cps": 32.054688, "mean_p": 0.266149, "boundary_entropy": 0.427743, "bpb": 2.093082}, "tr": {"chunks": 3780.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.208805, "cps": 29.53125, "mean_p": 0.245979, "boundary_entropy": 0.334774, "bpb": 2.509146}, "fi": {"chunks": 3684.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.195843, "cps": 28.78125, "mean_p": 0.254455, "boundary_entropy": 0.348943, "bpb": 2.721436}, "ru": {"chunks": 5716.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.301636, "cps": 44.65625, "mean_p": 0.324644, "boundary_entropy": 0.527171, "bpb": 1.641631}, "el": {"chunks": 6368.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.310286, "cps": 49.75, "mean_p": 0.323933, "boundary_entropy": 0.529162, "bpb": 1.705685}, "hi": {"chunks": 6658.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.395721, "cps": 52.015625, "mean_p": 0.390184, "boundary_entropy": 0.617248, "bpb": 1.330981}, "ta": {"chunks": 7986.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.427173, "cps": 62.390625, "mean_p": 0.377893, "boundary_entropy": 0.615504, "bpb": 1.081544}, "th": {"chunks": 7945.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.493938, "cps": 62.070312, "mean_p": 0.406719, "boundary_entropy": 0.637073, "bpb": 1.26721}, "zh": {"chunks": 3082.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.521842, "cps": 24.078125, "mean_p": 0.333487, "boundary_entropy": 0.53319, "bpb": 3.439677}, "my": {"chunks": 7505.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.396084, "cps": 58.632812, "mean_p": 0.417226, "boundary_entropy": 0.653258, "bpb": 1.208808}}, "bpb": {"en": 2.25381, "de": 2.52327, "vi": 2.09308, "tr": 2.50915, "fi": 2.72144, "ru": 1.64163, "el": 1.70568, "hi": 1.33098, "ta": 1.08154, "th": 1.26721, "zh": 3.43968, "my": 1.20881}, "gini_cpc": 0.20362, "gini_cps": 0.188506, "bpic": 5.985822, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9686784133165747, "vi": 1.1388744850566854, "tr": 1.0684611528110497, "fi": 1.002132683350066, "ru": 1.5434781792410892, "el": 1.587741139619417, "hi": 2.024912242837414, "ta": 2.185854875369397, "th": 2.5274951131142562, "zh": 2.67027924397846, "my": 2.026771593647065}, "t": 137.1}
{"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.26362, "de": 2.53012, "vi": 2.09542, "tr": 2.51612, "fi": 2.73188, "ru": 1.61866, "el": 1.6952, "hi": 1.32636, "ta": 1.08767, "th": 1.28465, "zh": 3.44392, "my": 1.21739}, "bytes_per_chunk": {"en": 5.167, "de": 5.208, "vi": 5.863, "tr": 4.929, "fi": 4.979, "ru": 6.022, "el": 5.949, "hi": 6.289, "ta": 6.346, "th": 5.966, "zh": 5.377, "my": 7.545}, "bpb_mean": 1.98425, "t": 138.4}
{"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.4538, "bpb": 2.09739, "chunk_loss": 0.030337, "F": 0.18524, "G": 0.2975, "chunks": 189.69, "bytes_per_chunk": 5.398, "boundary_entropy": 0.43999, "router_noise": 0.0, "grad_norm": 0.6894, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 138.4}
{"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.54089, "bpb": 2.22303, "chunk_loss": 0.030236, "F": 0.17938, "G": 0.29221, "chunks": 183.69, "bytes_per_chunk": 5.575, "boundary_entropy": 0.4324, "router_noise": 0.0, "grad_norm": 0.7344, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 140.6}
{"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.4664, "bpb": 2.11557, "chunk_loss": 0.030638, "F": 0.19043, "G": 0.30156, "chunks": 195.0, "bytes_per_chunk": 5.251, "boundary_entropy": 0.44476, "router_noise": 0.0, "grad_norm": 0.7043, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 142.7}
{"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.48956, "bpb": 2.14898, "chunk_loss": 0.030071, "F": 0.17654, "G": 0.28801, "chunks": 180.78, "bytes_per_chunk": 5.664, "boundary_entropy": 0.42715, "router_noise": 0.0, "grad_norm": 0.7084, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 144.9}
{"event": "probe", "step": 2400, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3447.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198069, "cps": 26.929688, "mean_p": 0.237721, "boundary_entropy": 0.306082, "bpb": 2.223984}, "de": {"chunks": 3968.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19325, "cps": 31.0, "mean_p": 0.251081, "boundary_entropy": 0.341303, "bpb": 2.481891}, "vi": {"chunks": 3987.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.216273, "cps": 31.148438, "mean_p": 0.2673, "boundary_entropy": 0.421052, "bpb": 2.078738}, "tr": {"chunks": 3811.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210518, "cps": 29.773438, "mean_p": 0.246663, "boundary_entropy": 0.331522, "bpb": 2.471774}, "fi": {"chunks": 3790.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.201478, "cps": 29.609375, "mean_p": 0.256597, "boundary_entropy": 0.3441, "bpb": 2.675764}, "ru": {"chunks": 5750.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.30343, "cps": 44.921875, "mean_p": 0.33294, "boundary_entropy": 0.534227, "bpb": 1.608151}, "el": {"chunks": 6217.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.302928, "cps": 48.570312, "mean_p": 0.315851, "boundary_entropy": 0.521201, "bpb": 1.688765}, "hi": {"chunks": 6600.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.392273, "cps": 51.5625, "mean_p": 0.379784, "boundary_entropy": 0.610709, "bpb": 1.304982}, "ta": {"chunks": 10329.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.552501, "cps": 80.695312, "mean_p": 0.404711, "boundary_entropy": 0.623426, "bpb": 1.0629}, "th": {"chunks": 7816.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.485919, "cps": 61.0625, "mean_p": 0.382287, "boundary_entropy": 0.61787, "bpb": 1.249947}, "zh": {"chunks": 2627.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.444802, "cps": 20.523438, "mean_p": 0.325087, "boundary_entropy": 0.529766, "bpb": 3.381845}, "my": {"chunks": 10831.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.571617, "cps": 84.617188, "mean_p": 0.421477, "boundary_entropy": 0.650732, "bpb": 1.181123}}, "bpb": {"en": 2.22398, "de": 2.48189, "vi": 2.07874, "tr": 2.47177, "fi": 2.67576, "ru": 1.60815, "el": 1.68877, "hi": 1.30498, "ta": 1.0629, "th": 1.24995, "zh": 3.38184, "my": 1.18112}, "gini_cpc": 0.225934, "gini_cps": 0.244338, "bpic": 5.548133, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9756680716520058, "vi": 1.0919077047930028, "tr": 1.0628481822923435, "fi": 1.0172089280391141, "ru": 1.5319389810113648, "el": 1.5294062442922622, "hi": 1.98048564744587, "ta": 2.7894311389999005, "th": 2.4532755030458566, "zh": 2.2456882514190255, "my": 2.885944777084937}, "t": 148.0}
{"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.23338, "de": 2.48855, "vi": 2.08274, "tr": 2.48175, "fi": 2.68832, "ru": 1.5845, "el": 1.6795, "hi": 1.30192, "ta": 1.06981, "th": 1.26833, "zh": 3.3867, "my": 1.19058}, "bytes_per_chunk": {"en": 5.076, "de": 5.139, "vi": 6.214, "tr": 4.937, "fi": 4.818, "ru": 6.039, "el": 6.025, "hi": 6.202, "ta": 4.992, "th": 5.926, "zh": 6.204, "my": 5.303}, "bpb_mean": 1.95467, "t": 149.3}
{"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 149.3}
{"event": "train", "step": 2400, "phase": 2, "objective": "parity_b", "lm_loss": 1.37321, "bpb": 1.98113, "chunk_loss": 0.064431, "F": 0.19147, "G": 0.29902, "chunks": 196.06, "bytes_per_chunk": 5.223, "boundary_entropy": 0.43556, "router_noise": 0.0, "grad_norm": 0.7912, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.09, "worst_lang": "my", "n_target_mean": 21.145, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 149.4}
{"event": "train", "step": 2450, "phase": 2, "objective": "parity_b", "lm_loss": 1.47809, "bpb": 2.13244, "chunk_loss": 0.031173, "F": 0.13437, "G": 0.24668, "chunks": 137.59, "bytes_per_chunk": 7.442, "boundary_entropy": 0.41838, "router_noise": 0.0, "grad_norm": 0.6918, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.046, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.25, "fallback": 0.0, "alpha": 0.03, "t": 151.6}
{"event": "train", "step": 2500, "phase": 2, "objective": "parity_b", "lm_loss": 1.50773, "bpb": 2.1752, "chunk_loss": 0.030474, "F": 0.14792, "G": 0.2357, "chunks": 151.47, "bytes_per_chunk": 6.76, "boundary_entropy": 0.37246, "router_noise": 0.0, "grad_norm": 0.7156, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.045, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 153.9}
{"event": "train", "step": 2550, "phase": 2, "objective": "parity_b", "lm_loss": 1.57234, "bpb": 2.2684, "chunk_loss": 0.031121, "F": 0.15692, "G": 0.2375, "chunks": 160.69, "bytes_per_chunk": 6.373, "boundary_entropy": 0.3619, "router_noise": 0.0, "grad_norm": 0.7328, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.045, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 156.1}
{"event": "probe", "step": 2600, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3499.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201057, "cps": 27.335938, "mean_p": 0.237593, "boundary_entropy": 0.290806, "bpb": 2.199798}, "de": {"chunks": 3617.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.176155, "cps": 28.257812, "mean_p": 0.2148, "boundary_entropy": 0.28122, "bpb": 2.456964}, "vi": {"chunks": 3298.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.178899, "cps": 25.765625, "mean_p": 0.232452, "boundary_entropy": 0.379423, "bpb": 2.054643}, "tr": {"chunks": 3282.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181296, "cps": 25.640625, "mean_p": 0.2145, "boundary_entropy": 0.286433, "bpb": 2.456722}, "fi": {"chunks": 3185.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.169316, "cps": 24.882812, "mean_p": 0.208087, "boundary_entropy": 0.281544, "bpb": 2.655108}, "ru": {"chunks": 2649.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.139789, "cps": 20.695312, "mean_p": 0.177039, "boundary_entropy": 0.369737, "bpb": 1.634094}, "el": {"chunks": 3620.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.176387, "cps": 28.28125, "mean_p": 0.208631, "boundary_entropy": 0.417438, "bpb": 1.684746}, "hi": {"chunks": 2557.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.151976, "cps": 19.976562, "mean_p": 0.246091, "boundary_entropy": 0.504335, "bpb": 1.353983}, "ta": {"chunks": 1312.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.070179, "cps": 10.25, "mean_p": 0.177462, "boundary_entropy": 0.426139, "bpb": 1.099543}, "th": {"chunks": 2063.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.128256, "cps": 16.117188, "mean_p": 0.203655, "boundary_entropy": 0.447871, "bpb": 1.268319}, "zh": {"chunks": 2758.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.466983, "cps": 21.546875, "mean_p": 0.322946, "boundary_entropy": 0.524923, "bpb": 3.35211}, "my": {"chunks": 1138.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.060059, "cps": 8.890625, "mean_p": 0.18014, "boundary_entropy": 0.443433, "bpb": 1.204573}}, "bpb": {"en": 2.1998, "de": 2.45696, "vi": 2.05464, "tr": 2.45672, "fi": 2.65511, "ru": 1.63409, "el": 1.68475, "hi": 1.35398, "ta": 1.09954, "th": 1.26832, "zh": 3.35211, "my": 1.20457}, "gini_cpc": 0.249996, "gini_cps": 0.162052, "bpic": 11.637486, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.876145589989623, "vi": 0.8897903411203619, "tr": 0.9017127343633681, "fi": 0.8421272698508192, "ru": 0.6952690893214118, "el": 0.8772995429312002, "hi": 0.7558851839222159, "ta": 0.34905072407880877, "th": 0.6379084285638518, "zh": 2.322635164398478, "my": 0.29871639793582566}, "t": 159.2}
{"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.20943, "de": 2.46332, "vi": 2.05831, "tr": 2.46286, "fi": 2.66467, "ru": 1.61114, "el": 1.67459, "hi": 1.3492, "ta": 1.10477, "th": 1.28726, "zh": 3.35538, "my": 1.214}, "bytes_per_chunk": {"en": 4.966, "de": 5.593, "vi": 7.488, "tr": 5.705, "fi": 5.659, "ru": 12.5, "el": 10.18, "hi": 14.944, "ta": 36.178, "th": 20.807, "zh": 6.051, "my": 44.169}, "bpb_mean": 1.95458, "t": 160.5}
{"event": "train", "step": 2600, "phase": 2, "objective": "parity_b", "lm_loss": 1.60175, "bpb": 2.31084, "chunk_loss": 0.029848, "F": 0.15869, "G": 0.23762, "chunks": 162.5, "bytes_per_chunk": 6.302, "boundary_entropy": 0.35892, "router_noise": 0.0, "grad_norm": 0.7436, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 160.6}
{"event": "train", "step": 2650, "phase": 2, "objective": "parity_b", "lm_loss": 1.50389, "bpb": 2.16966, "chunk_loss": 0.029805, "F": 0.14868, "G": 0.23851, "chunks": 152.25, "bytes_per_chunk": 6.726, "boundary_entropy": 0.37467, "router_noise": 0.0, "grad_norm": 0.6449, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 162.8}
{"event": "train", "step": 2700, "phase": 2, "objective": "parity_b", "lm_loss": 1.52026, "bpb": 2.19327, "chunk_loss": 0.030496, "F": 0.15424, "G": 0.22773, "chunks": 157.94, "bytes_per_chunk": 6.484, "boundary_entropy": 0.3453, "router_noise": 0.0, "grad_norm": 0.686, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 165.0}
{"event": "train", "step": 2750, "phase": 2, "objective": "parity_b", "lm_loss": 1.41185, "bpb": 2.03686, "chunk_loss": 0.030269, "F": 0.15109, "G": 0.21399, "chunks": 154.72, "bytes_per_chunk": 6.618, "boundary_entropy": 0.31455, "router_noise": 0.0, "grad_norm": 0.7337, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 167.2}
{"event": "probe", "step": 2800, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3410.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195943, "cps": 26.640625, "mean_p": 0.229668, "boundary_entropy": 0.277609, "bpb": 2.183696}, "de": {"chunks": 3441.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.167584, "cps": 26.882812, "mean_p": 0.202459, "boundary_entropy": 0.266178, "bpb": 2.435738}, "vi": {"chunks": 2750.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.149173, "cps": 21.484375, "mean_p": 0.191425, "boundary_entropy": 0.337928, "bpb": 2.051121}, "tr": {"chunks": 3045.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168204, "cps": 23.789062, "mean_p": 0.205589, "boundary_entropy": 0.276435, "bpb": 2.428041}, "fi": {"chunks": 3179.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.168997, "cps": 24.835938, "mean_p": 0.204935, "boundary_entropy": 0.270695, "bpb": 2.634959}, "ru": {"chunks": 3278.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.172982, "cps": 25.609375, "mean_p": 0.187958, "boundary_entropy": 0.36249, "bpb": 1.579017}, "el": {"chunks": 2874.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.140038, "cps": 22.453125, "mean_p": 0.177814, "boundary_entropy": 0.378301, "bpb": 1.664446}, "hi": {"chunks": 2747.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.163269, "cps": 21.460938, "mean_p": 0.220151, "boundary_entropy": 0.457778, "bpb": 1.325429}, "ta": {"chunks": 1706.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.091254, "cps": 13.328125, "mean_p": 0.181826, "boundary_entropy": 0.421338, "bpb": 1.114541}, "th": {"chunks": 2122.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.131924, "cps": 16.578125, "mean_p": 0.196407, "boundary_entropy": 0.429315, "bpb": 1.259075}, "zh": {"chunks": 3012.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50999, "cps": 23.53125, "mean_p": 0.318755, "boundary_entropy": 0.511299, "bpb": 3.342942}, "my": {"chunks": 605.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.031929, "cps": 4.726562, "mean_p": 0.183007, "boundary_entropy": 0.444818, "bpb": 1.195874}}, "bpb": {"en": 2.1837, "de": 2.43574, "vi": 2.05112, "tr": 2.42804, "fi": 2.63496, "ru": 1.57902, "el": 1.66445, "hi": 1.32543, "ta": 1.11454, "th": 1.25908, "zh": 3.34294, "my": 1.19587}, "gini_cpc": 0.26676, "gini_cps": 0.149748, "bpic": 11.93015, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8552675737061848, "vi": 0.7613060710254862, "tr": 0.8584331625985746, "fi": 0.8624787144104086, "ru": 0.882814537407439, "el": 0.7146866336788581, "hi": 0.8332461730858893, "ta": 0.46571829468101544, "th": 0.6732774291998976, "zh": 2.6027428739400165, "my": 0.16295276716582566}, "t": 170.2}
{"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.1925, "de": 2.44182, "vi": 2.05354, "tr": 2.43782, "fi": 2.64254, "ru": 1.55718, "el": 1.6536, "hi": 1.32067, "ta": 1.11962, "th": 1.27658, "zh": 3.34459, "my": 1.20462}, "bytes_per_chunk": {"en": 5.098, "de": 5.836, "vi": 9.052, "tr": 6.04, "fi": 5.657, "ru": 10.252, "el": 12.57, "hi": 14.126, "ta": 29.29, "th": 20.851, "zh": 5.541, "my": 65.032}, "bpb_mean": 1.93709, "t": 171.5}
{"event": "train", "step": 2800, "phase": 2, "objective": "parity_b", "lm_loss": 1.47911, "bpb": 2.1339, "chunk_loss": 0.029908, "F": 0.14529, "G": 0.21918, "chunks": 148.78, "bytes_per_chunk": 6.883, "boundary_entropy": 0.33823, "router_noise": 0.0, "grad_norm": 0.7522, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.088, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 171.6}
{"event": "train", "step": 2850, "phase": 2, "objective": "parity_b", "lm_loss": 1.4427, "bpb": 2.08137, "chunk_loss": 0.0, "F": 0.1416, "G": 0.21411, "chunks": 145.0, "bytes_per_chunk": 7.062, "boundary_entropy": 0.33317, "router_noise": 0.0, "grad_norm": 0.6601, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "fi", "n_target_mean": 6.6232, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 173.8}
{"event": "train", "step": 2900, "phase": 2, "objective": "parity_b", "lm_loss": 1.46062, "bpb": 2.10723, "chunk_loss": 0.029281, "F": 0.15073, "G": 0.22302, "chunks": 154.34, "bytes_per_chunk": 6.635, "boundary_entropy": 0.33525, "router_noise": 0.0, "grad_norm": 0.6887, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "th", "n_target_mean": 16.5724, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 175.9}
{"event": "train", "step": 2950, "phase": 2, "objective": "parity_b", "lm_loss": 1.36566, "bpb": 1.97024, "chunk_loss": 0.030585, "F": 0.14026, "G": 0.21886, "chunks": 143.62, "bytes_per_chunk": 7.13, "boundary_entropy": 0.34619, "router_noise": 0.0, "grad_norm": 0.665, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 178.1}
{"event": "probe", "step": 3000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3359.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193013, "cps": 26.242188, "mean_p": 0.222107, "boundary_entropy": 0.264589, "bpb": 2.159609}, "de": {"chunks": 3402.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.165685, "cps": 26.578125, "mean_p": 0.189619, "boundary_entropy": 0.242832, "bpb": 2.414192}, "vi": {"chunks": 3392.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.183998, "cps": 26.5, "mean_p": 0.222267, "boundary_entropy": 0.360946, "bpb": 2.026333}, "tr": {"chunks": 3047.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168315, "cps": 23.804688, "mean_p": 0.200968, "boundary_entropy": 0.265289, "bpb": 2.397568}, "fi": {"chunks": 3184.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.169263, "cps": 24.875, "mean_p": 0.204645, "boundary_entropy": 0.265121, "bpb": 2.587103}, "ru": {"chunks": 3346.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.17657, "cps": 26.140625, "mean_p": 0.179261, "boundary_entropy": 0.339842, "bpb": 1.558586}, "el": {"chunks": 3140.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.152999, "cps": 24.53125, "mean_p": 0.176981, "boundary_entropy": 0.368985, "bpb": 1.631552}, "hi": {"chunks": 2478.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.147281, "cps": 19.359375, "mean_p": 0.207096, "boundary_entropy": 0.439536, "bpb": 1.300054}, "ta": {"chunks": 1492.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.079807, "cps": 11.65625, "mean_p": 0.180788, "boundary_entropy": 0.427908, "bpb": 1.076234}, "th": {"chunks": 2844.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.176811, "cps": 22.21875, "mean_p": 0.216617, "boundary_entropy": 0.452609, "bpb": 1.232417}, "zh": {"chunks": 3467.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.58703, "cps": 27.085938, "mean_p": 0.336214, "boundary_entropy": 0.520334, "bpb": 3.318863}, "my": {"chunks": 1366.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.072092, "cps": 10.671875, "mean_p": 0.180738, "boundary_entropy": 0.436329, "bpb": 1.180775}}, "bpb": {"en": 2.15961, "de": 2.41419, "vi": 2.02633, "tr": 2.39757, "fi": 2.5871, "ru": 1.55859, "el": 1.63155, "hi": 1.30005, "ta": 1.07623, "th": 1.23242, "zh": 3.31886, "my": 1.18078}, "gini_cpc": 0.260056, "gini_cps": 0.121233, "bpic": 11.118608, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8584124710870311, "vi": 0.9532939086689124, "tr": 0.8720392227486864, "fi": 0.8769509279636586, "ru": 0.9148098637850033, "el": 0.7926891600079186, "hi": 0.7630629118639374, "ta": 0.41348282042129575, "th": 0.916057306825723, "zh": 3.0414068193702932, "my": 0.37350931708051927}, "t": 181.2}
{"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.16984, "de": 2.41887, "vi": 2.02646, "tr": 2.40815, "fi": 2.59621, "ru": 1.53744, "el": 1.62127, "hi": 1.29643, "ta": 1.08082, "th": 1.25072, "zh": 3.31974, "my": 1.19025}, "bytes_per_chunk": {"en": 5.188, "de": 5.975, "vi": 7.209, "tr": 6.025, "fi": 5.566, "ru": 10.112, "el": 11.703, "hi": 15.627, "ta": 33.725, "th": 15.005, "zh": 4.746, "my": 35.372}, "bpb_mean": 1.90968, "t": 182.5}
{"event": "train", "step": 3000, "phase": 2, "objective": "parity_b", "lm_loss": 1.44539, "bpb": 2.08526, "chunk_loss": 0.030749, "F": 0.14136, "G": 0.22189, "chunks": 144.75, "bytes_per_chunk": 7.074, "boundary_entropy": 0.35303, "router_noise": 0.0, "grad_norm": 0.6974, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.088, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 182.5}
{"event": "train", "step": 3050, "phase": 2, "objective": "parity_b", "lm_loss": 1.40261, "bpb": 2.02354, "chunk_loss": 0.030534, "F": 0.13971, "G": 0.21144, "chunks": 143.06, "bytes_per_chunk": 7.158, "boundary_entropy": 0.33016, "router_noise": 0.0, "grad_norm": 0.6523, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 184.7}
{"event": "train", "step": 3100, "phase": 2, "objective": "parity_b", "lm_loss": 1.41229, "bpb": 2.03751, "chunk_loss": 0.030472, "F": 0.15646, "G": 0.21741, "chunks": 160.22, "bytes_per_chunk": 6.391, "boundary_entropy": 0.31805, "router_noise": 0.0, "grad_norm": 0.7259, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "tr", "n_target_mean": 6.7176, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 186.9}
{"event": "train", "step": 3150, "phase": 2, "objective": "parity_b", "lm_loss": 1.35994, "bpb": 1.96199, "chunk_loss": 0.029968, "F": 0.13586, "G": 0.20731, "chunks": 139.12, "bytes_per_chunk": 7.36, "boundary_entropy": 0.32521, "router_noise": 0.0, "grad_norm": 0.6622, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "tr", "n_target_mean": 6.7176, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 189.1}
{"event": "probe", "step": 3200, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3439.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19761, "cps": 26.867188, "mean_p": 0.22451, "boundary_entropy": 0.253975, "bpb": 2.14754}, "de": {"chunks": 3425.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.166805, "cps": 26.757812, "mean_p": 0.194572, "boundary_entropy": 0.242235, "bpb": 2.378321}, "vi": {"chunks": 3132.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.169894, "cps": 24.46875, "mean_p": 0.202038, "boundary_entropy": 0.33607, "bpb": 2.010079}, "tr": {"chunks": 3151.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.17406, "cps": 24.617188, "mean_p": 0.194083, "boundary_entropy": 0.241664, "bpb": 2.371127}, "fi": {"chunks": 3194.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.169794, "cps": 24.953125, "mean_p": 0.197398, "boundary_entropy": 0.247052, "bpb": 2.550209}, "ru": {"chunks": 3059.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.161425, "cps": 23.898438, "mean_p": 0.157722, "boundary_entropy": 0.302374, "bpb": 1.539338}, "el": {"chunks": 3558.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.173366, "cps": 27.796875, "mean_p": 0.193771, "boundary_entropy": 0.377533, "bpb": 1.605646}, "hi": {"chunks": 2500.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.148588, "cps": 19.53125, "mean_p": 0.200999, "boundary_entropy": 0.427463, "bpb": 1.298313}, "ta": {"chunks": 1641.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.087777, "cps": 12.820312, "mean_p": 0.177015, "boundary_entropy": 0.415251, "bpb": 1.060152}, "th": {"chunks": 2245.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.139571, "cps": 17.539062, "mean_p": 0.18628, "boundary_entropy": 0.409857, "bpb": 1.256174}, "zh": {"chunks": 3625.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.613783, "cps": 28.320312, "mean_p": 0.33106, "boundary_entropy": 0.510121, "bpb": 3.28215}, "my": {"chunks": 1046.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.055204, "cps": 8.171875, "mean_p": 0.156444, "boundary_entropy": 0.399305, "bpb": 1.170081}}, "bpb": {"en": 2.14754, "de": 2.37832, "vi": 2.01008, "tr": 2.37113, "fi": 2.55021, "ru": 1.53934, "el": 1.60565, "hi": 1.29831, "ta": 1.06015, "th": 1.25617, "zh": 3.28215, "my": 1.17008}, "gini_cpc": 0.277789, "gini_cps": 0.144615, "bpic": 11.282699, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8441120753042213, "vi": 0.8597467757837338, "tr": 0.8808253318700685, "fi": 0.8592409621432981, "ru": 0.8168874181839386, "el": 0.8773180061938607, "hi": 0.751929078522861, "ta": 0.44419642194014336, "th": 0.7062967769996972, "zh": 3.1060361974116746, "my": 0.27935744683280356}, "t": 192.2}
{"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.15603, "de": 2.38407, "vi": 2.01194, "tr": 2.38007, "fi": 2.55961, "ru": 1.51746, "el": 1.59748, "hi": 1.29427, "ta": 1.06654, "th": 1.27471, "zh": 3.28442, "my": 1.17785}, "bytes_per_chunk": {"en": 5.105, "de": 5.915, "vi": 7.932, "tr": 5.888, "fi": 5.58, "ru": 10.882, "el": 10.515, "hi": 15.502, "ta": 31.049, "th": 19.0, "zh": 4.703, "my": 45.283}, "bpb_mean": 1.89204, "t": 193.5}
{"event": "train", "step": 3200, "phase": 2, "objective": "parity_b", "lm_loss": 1.39076, "bpb": 2.00644, "chunk_loss": 0.030996, "F": 0.13718, "G": 0.21259, "chunks": 140.47, "bytes_per_chunk": 7.29, "boundary_entropy": 0.34116, "router_noise": 0.0, "grad_norm": 0.6543, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 193.5}
{"event": "train", "step": 3250, "phase": 2, "objective": "parity_b", "lm_loss": 1.45052, "bpb": 2.09266, "chunk_loss": 0.030471, "F": 0.14978, "G": 0.21441, "chunks": 153.38, "bytes_per_chunk": 6.676, "boundary_entropy": 0.31767, "router_noise": 0.0, "grad_norm": 0.6617, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "tr", "n_target_mean": 6.7176, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 195.7}
{"event": "train", "step": 3300, "phase": 2, "objective": "parity_b", "lm_loss": 1.41611, "bpb": 2.04302, "chunk_loss": 0.030359, "F": 0.15714, "G": 0.21777, "chunks": 160.91, "bytes_per_chunk": 6.364, "boundary_entropy": 0.31136, "router_noise": 0.0, "grad_norm": 0.7754, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 197.8}
{"event": "train", "step": 3350, "phase": 2, "objective": "parity_b", "lm_loss": 1.47589, "bpb": 2.12926, "chunk_loss": 0.030297, "F": 0.15082, "G": 0.22155, "chunks": 154.44, "bytes_per_chunk": 6.631, "boundary_entropy": 0.33137, "router_noise": 0.0, "grad_norm": 0.6821, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "en", "n_target_mean": 6.0, "frac_in_batch": 0.2812, "fallback": 0.0, "alpha": 0.03, "t": 200.0}
{"event": "probe", "step": 3400, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3465.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199104, "cps": 27.070312, "mean_p": 0.221395, "boundary_entropy": 0.247175, "bpb": 2.126072}, "de": {"chunks": 3461.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.168558, "cps": 27.039062, "mean_p": 0.194932, "boundary_entropy": 0.236117, "bpb": 2.352847}, "vi": {"chunks": 3308.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.179441, "cps": 25.84375, "mean_p": 0.208918, "boundary_entropy": 0.335813, "bpb": 1.995101}, "tr": {"chunks": 3193.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.17638, "cps": 24.945312, "mean_p": 0.197238, "boundary_entropy": 0.240277, "bpb": 2.344574}, "fi": {"chunks": 3154.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.167668, "cps": 24.640625, "mean_p": 0.193722, "boundary_entropy": 0.237136, "bpb": 2.542283}, "ru": {"chunks": 2867.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.151293, "cps": 22.398438, "mean_p": 0.150879, "boundary_entropy": 0.295318, "bpb": 1.527323}, "el": {"chunks": 2888.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.14072, "cps": 22.5625, "mean_p": 0.176681, "boundary_entropy": 0.365338, "bpb": 1.614438}, "hi": {"chunks": 2529.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.150312, "cps": 19.757812, "mean_p": 0.205653, "boundary_entropy": 0.432003, "bpb": 1.290732}, "ta": {"chunks": 1450.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.077561, "cps": 11.328125, "mean_p": 0.168948, "boundary_entropy": 0.402869, "bpb": 1.053052}, "th": {"chunks": 2772.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.172334, "cps": 21.65625, "mean_p": 0.192435, "boundary_entropy": 0.414524, "bpb": 1.201396}, "zh": {"chunks": 3536.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.598713, "cps": 27.625, "mean_p": 0.320239, "boundary_entropy": 0.492157, "bpb": 3.259648}, "my": {"chunks": 1186.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.062592, "cps": 9.265625, "mean_p": 0.157036, "boundary_entropy": 0.398184, "bpb": 1.15906}}, "bpb": {"en": 2.12607, "de": 2.35285, "vi": 1.9951, "tr": 2.34457, "fi": 2.54228, "ru": 1.52732, "el": 1.61444, "hi": 1.29073, "ta": 1.05305, "th": 1.2014, "zh": 3.25965, "my": 1.15906}, "gini_cpc": 0.276826, "gini_cps": 0.134168, "bpic": 11.351445, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8465840333468054, "vi": 0.9012457716770166, "tr": 0.885868477947139, "fi": 0.8421136108765688, "ru": 0.7598701070257798, "el": 0.7067685648010162, "hi": 0.7549438258621023, "ta": 0.38955018414767006, "th": 0.8655517562946846, "zh": 3.0070433911951016, "my": 0.31437079562475384}, "t": 203.1}
{"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.13445, "de": 2.36006, "vi": 1.9963, "tr": 2.35534, "fi": 2.55205, "ru": 1.50677, "el": 1.60529, "hi": 1.28664, "ta": 1.05875, "th": 1.22088, "zh": 3.26168, "my": 1.16753}, "bytes_per_chunk": {"en": 5.113, "de": 5.894, "vi": 7.359, "tr": 5.773, "fi": 5.683, "ru": 11.53, "el": 12.933, "hi": 15.268, "ta": 34.953, "th": 15.318, "zh": 4.678, "my": 39.438}, "bpb_mean": 1.87548, "t": 204.4}
{"event": "train", "step": 3400, "phase": 2, "objective": "parity_b", "lm_loss": 1.40303, "bpb": 2.02414, "chunk_loss": 0.030904, "F": 0.14188, "G": 0.21456, "chunks": 145.28, "bytes_per_chunk": 7.048, "boundary_entropy": 0.34074, "router_noise": 0.0, "grad_norm": 0.6656, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.088, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1875, "fallback": 0.0, "alpha": 0.03, "t": 204.4}
{"event": "train", "step": 3450, "phase": 2, "objective": "parity_b", "lm_loss": 1.26731, "bpb": 1.82835, "chunk_loss": 0.030603, "F": 0.13266, "G": 0.20206, "chunks": 135.84, "bytes_per_chunk": 7.538, "boundary_entropy": 0.31411, "router_noise": 0.0, "grad_norm": 0.6753, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 206.6}
{"event": "train", "step": 3500, "phase": 2, "objective": "parity_b", "lm_loss": 1.4545, "bpb": 2.09839, "chunk_loss": 0.030097, "F": 0.14856, "G": 0.20966, "chunks": 152.12, "bytes_per_chunk": 6.731, "boundary_entropy": 0.30145, "router_noise": 0.0, "grad_norm": 0.7913, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "tr", "n_target_mean": 6.7176, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 208.7}
{"event": "train", "step": 3550, "phase": 2, "objective": "parity_b", "lm_loss": 1.38504, "bpb": 1.99818, "chunk_loss": 0.030669, "F": 0.1304, "G": 0.19755, "chunks": 133.53, "bytes_per_chunk": 7.669, "boundary_entropy": 0.31114, "router_noise": 0.0, "grad_norm": 0.6711, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "fi", "n_target_mean": 6.6232, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 210.9}
{"event": "probe", "step": 3600, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3493.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200713, "cps": 27.289062, "mean_p": 0.225594, "boundary_entropy": 0.243967, "bpb": 2.110494}, "de": {"chunks": 3422.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.166659, "cps": 26.734375, "mean_p": 0.192205, "boundary_entropy": 0.229779, "bpb": 2.344768}, "vi": {"chunks": 3277.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.17776, "cps": 25.601562, "mean_p": 0.216403, "boundary_entropy": 0.347401, "bpb": 1.99601}, "tr": {"chunks": 3158.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174446, "cps": 24.671875, "mean_p": 0.194745, "boundary_entropy": 0.237416, "bpb": 2.32315}, "fi": {"chunks": 3128.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.166286, "cps": 24.4375, "mean_p": 0.193092, "boundary_entropy": 0.233782, "bpb": 2.498605}, "ru": {"chunks": 3346.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.17657, "cps": 26.140625, "mean_p": 0.164628, "boundary_entropy": 0.300817, "bpb": 1.503314}, "el": {"chunks": 2780.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.135458, "cps": 21.71875, "mean_p": 0.167317, "boundary_entropy": 0.348748, "bpb": 1.595953}, "hi": {"chunks": 2777.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.165052, "cps": 21.695312, "mean_p": 0.205072, "boundary_entropy": 0.420643, "bpb": 1.268022}, "ta": {"chunks": 1524.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.081519, "cps": 11.90625, "mean_p": 0.169237, "boundary_entropy": 0.402048, "bpb": 1.051778}, "th": {"chunks": 2946.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.183152, "cps": 23.015625, "mean_p": 0.209136, "boundary_entropy": 0.439041, "bpb": 1.187676}, "zh": {"chunks": 2852.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482899, "cps": 22.28125, "mean_p": 0.303259, "boundary_entropy": 0.488953, "bpb": 3.264468}, "my": {"chunks": 1556.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.082119, "cps": 12.15625, "mean_p": 0.181013, "boundary_entropy": 0.429163, "bpb": 1.156828}}, "bpb": {"en": 2.11049, "de": 2.34477, "vi": 1.99601, "tr": 2.32315, "fi": 2.49861, "ru": 1.50331, "el": 1.59595, "hi": 1.26802, "ta": 1.05178, "th": 1.18768, "zh": 3.26447, "my": 1.15683}, "gini_cpc": 0.224941, "gini_cps": 0.111944, "bpic": 11.202341, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8303345947866958, "vi": 0.8856432847096029, "tr": 0.8691347388869484, "fi": 0.8284768843564388, "ru": 0.8797155260388851, "el": 0.6748845488918823, "hi": 0.8223303920513496, "ta": 0.4061486668748799, "th": 0.9125091161051175, "zh": 2.405922386110327, "my": 0.40913982130550797}, "t": 214.0}
{"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.12131, "de": 2.35109, "vi": 1.99628, "tr": 2.33308, "fi": 2.51081, "ru": 1.48041, "el": 1.58694, "hi": 1.26454, "ta": 1.05729, "th": 1.2068, "zh": 3.26587, "my": 1.1645}, "bytes_per_chunk": {"en": 4.999, "de": 5.894, "vi": 7.46, "tr": 5.89, "fi": 5.678, "ru": 9.933, "el": 13.243, "hi": 13.683, "ta": 33.407, "th": 13.986, "zh": 5.964, "my": 31.063}, "bpb_mean": 1.86158, "t": 215.3}
{"event": "train", "step": 3600, "phase": 2, "objective": "parity_b", "lm_loss": 1.64582, "bpb": 2.37442, "chunk_loss": 0.029476, "F": 0.16019, "G": 0.23508, "chunks": 164.03, "bytes_per_chunk": 6.243, "boundary_entropy": 0.34205, "router_noise": 0.0, "grad_norm": 0.7565, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.2812, "fallback": 0.0, "alpha": 0.03, "t": 215.3}
{"event": "train", "step": 3650, "phase": 2, "objective": "parity_b", "lm_loss": 1.44664, "bpb": 2.08707, "chunk_loss": 0.0, "F": 0.14749, "G": 0.21161, "chunks": 151.03, "bytes_per_chunk": 6.78, "boundary_entropy": 0.31248, "router_noise": 0.0, "grad_norm": 0.752, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 217.5}
{"event": "train", "step": 3700, "phase": 2, "objective": "parity_b", "lm_loss": 1.36334, "bpb": 1.96688, "chunk_loss": 0.030378, "F": 0.13214, "G": 0.20184, "chunks": 135.31, "bytes_per_chunk": 7.568, "boundary_entropy": 0.31675, "router_noise": 0.0, "grad_norm": 0.718, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "en", "n_target_mean": 6.0, "frac_in_batch": 0.1875, "fallback": 0.0, "alpha": 0.03, "t": 219.7}
{"event": "train", "step": 3750, "phase": 2, "objective": "parity_b", "lm_loss": 1.40661, "bpb": 2.02931, "chunk_loss": 0.0, "F": 0.14728, "G": 0.20927, "chunks": 150.81, "bytes_per_chunk": 6.79, "boundary_entropy": 0.30898, "router_noise": 0.0, "grad_norm": 0.7006, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "fi", "n_target_mean": 6.6232, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 221.9}
{"event": "probe", "step": 3800, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3452.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198357, "cps": 26.96875, "mean_p": 0.222107, "boundary_entropy": 0.235077, "bpb": 2.102237}, "de": {"chunks": 3405.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.165831, "cps": 26.601562, "mean_p": 0.19063, "boundary_entropy": 0.224413, "bpb": 2.32023}, "vi": {"chunks": 3043.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.165066, "cps": 23.773438, "mean_p": 0.200244, "boundary_entropy": 0.324702, "bpb": 1.982225}, "tr": {"chunks": 3215.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177595, "cps": 25.117188, "mean_p": 0.192765, "boundary_entropy": 0.227747, "bpb": 2.304625}, "fi": {"chunks": 3159.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.167934, "cps": 24.679688, "mean_p": 0.188203, "boundary_entropy": 0.223433, "bpb": 2.48569}, "ru": {"chunks": 3156.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.166544, "cps": 24.65625, "mean_p": 0.150874, "boundary_entropy": 0.274137, "bpb": 1.491168}, "el": {"chunks": 2769.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.134922, "cps": 21.632812, "mean_p": 0.16299, "boundary_entropy": 0.337275, "bpb": 1.566348}, "hi": {"chunks": 2477.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.147221, "cps": 19.351562, "mean_p": 0.177451, "boundary_entropy": 0.382928, "bpb": 1.267239}, "ta": {"chunks": 1404.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.0751, "cps": 10.96875, "mean_p": 0.161422, "boundary_entropy": 0.388629, "bpb": 1.043731}, "th": {"chunks": 2401.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.14927, "cps": 18.757812, "mean_p": 0.175846, "boundary_entropy": 0.387818, "bpb": 1.179917}, "zh": {"chunks": 3358.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.568574, "cps": 26.234375, "mean_p": 0.311381, "boundary_entropy": 0.483233, "bpb": 3.227283}, "my": {"chunks": 1185.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.06254, "cps": 9.257812, "mean_p": 0.153427, "boundary_entropy": 0.38598, "bpb": 1.147379}}, "bpb": {"en": 2.10224, "de": 2.32023, "vi": 1.98222, "tr": 2.30462, "fi": 2.48569, "ru": 1.49117, "el": 1.56635, "hi": 1.26724, "ta": 1.04373, "th": 1.17992, "zh": 3.22728, "my": 1.14738}, "gini_cpc": 0.272705, "gini_cps": 0.137824, "bpic": 11.621275, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8360226470923018, "vi": 0.8321701691546605, "tr": 0.8953312828376364, "fi": 0.8466249746748977, "ru": 0.8396167874842927, "el": 0.6801981458329346, "hi": 0.7422056683981894, "ta": 0.3786125202486583, "th": 0.7525310569465301, "zh": 2.866424995850401, "my": 0.31528862855981277}, "t": 225.0}
{"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.11241, "de": 2.32501, "vi": 1.98281, "tr": 2.31278, "fi": 2.49451, "ru": 1.46975, "el": 1.55818, "hi": 1.26285, "ta": 1.04896, "th": 1.19849, "zh": 3.22632, "my": 1.15508}, "bytes_per_chunk": {"en": 5.112, "de": 5.937, "vi": 8.163, "tr": 5.786, "fi": 5.678, "ru": 10.476, "el": 13.391, "hi": 16.608, "ta": 36.891, "th": 17.453, "zh": 5.128, "my": 38.534}, "bpb_mean": 1.8456, "t": 226.3}
{"event": "train", "step": 3800, "phase": 2, "objective": "parity_b", "lm_loss": 1.47987, "bpb": 2.13499, "chunk_loss": 0.030366, "F": 0.14056, "G": 0.21161, "chunks": 143.94, "bytes_per_chunk": 7.114, "boundary_entropy": 0.32715, "router_noise": 0.0, "grad_norm": 0.7074, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.088, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.2188, "fallback": 0.0, "alpha": 0.03, "t": 226.3}
{"event": "train", "step": 3850, "phase": 2, "objective": "parity_b", "lm_loss": 1.47854, "bpb": 2.13309, "chunk_loss": 0.030239, "F": 0.1474, "G": 0.21479, "chunks": 150.94, "bytes_per_chunk": 6.784, "boundary_entropy": 0.3225, "router_noise": 0.0, "grad_norm": 0.7225, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "en", "n_target_mean": 6.0, "frac_in_batch": 0.2188, "fallback": 0.0, "alpha": 0.03, "t": 228.5}
{"event": "train", "step": 3900, "phase": 2, "objective": "parity_b", "lm_loss": 1.47885, "bpb": 2.13353, "chunk_loss": 0.030469, "F": 0.15771, "G": 0.20674, "chunks": 161.5, "bytes_per_chunk": 6.341, "boundary_entropy": 0.27833, "router_noise": 0.0, "grad_norm": 0.7345, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 230.7}
{"event": "train", "step": 3950, "phase": 2, "objective": "parity_b", "lm_loss": 1.55373, "bpb": 2.24156, "chunk_loss": 0.030307, "F": 0.15292, "G": 0.22304, "chunks": 156.59, "bytes_per_chunk": 6.539, "boundary_entropy": 0.33476, "router_noise": 0.0, "grad_norm": 0.7103, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 232.9}
{"event": "probe", "step": 4000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3430.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197092, "cps": 26.796875, "mean_p": 0.218078, "boundary_entropy": 0.228824, "bpb": 2.087138}, "de": {"chunks": 3425.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.166805, "cps": 26.757812, "mean_p": 0.191409, "boundary_entropy": 0.219027, "bpb": 2.30549}, "vi": {"chunks": 3160.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.171413, "cps": 24.6875, "mean_p": 0.209132, "boundary_entropy": 0.335576, "bpb": 1.969436}, "tr": {"chunks": 3170.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175109, "cps": 24.765625, "mean_p": 0.193787, "boundary_entropy": 0.22684, "bpb": 2.289212}, "fi": {"chunks": 3124.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.166073, "cps": 24.40625, "mean_p": 0.184272, "boundary_entropy": 0.2132, "bpb": 2.461134}, "ru": {"chunks": 2961.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.156253, "cps": 23.132812, "mean_p": 0.138669, "boundary_entropy": 0.259258, "bpb": 1.47817}, "el": {"chunks": 3000.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.146177, "cps": 23.4375, "mean_p": 0.169353, "boundary_entropy": 0.342239, "bpb": 1.550634}, "hi": {"chunks": 2833.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.16838, "cps": 22.132812, "mean_p": 0.19508, "boundary_entropy": 0.40429, "bpb": 1.252012}, "ta": {"chunks": 1162.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.062156, "cps": 9.078125, "mean_p": 0.162616, "boundary_entropy": 0.39336, "bpb": 1.039436}, "th": {"chunks": 2694.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.167485, "cps": 21.046875, "mean_p": 0.18318, "boundary_entropy": 0.396249, "bpb": 1.166582}, "zh": {"chunks": 3446.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583474, "cps": 26.921875, "mean_p": 0.311157, "boundary_entropy": 0.479419, "bpb": 3.207036}, "my": {"chunks": 1296.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.068398, "cps": 10.125, "mean_p": 0.148466, "boundary_entropy": 0.375993, "bpb": 1.143275}}, "bpb": {"en": 2.08714, "de": 2.30549, "vi": 1.96944, "tr": 2.28921, "fi": 2.46113, "ru": 1.47817, "el": 1.55063, "hi": 1.25201, "ta": 1.03944, "th": 1.16658, "zh": 3.20704, "my": 1.14327}, "gini_cpc": 0.268404, "gini_cps": 0.128218, "bpic": 11.387822, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8463269466388388, "vi": 0.8697089529755876, "tr": 0.8884617001289512, "fi": 0.8426149143862076, "ru": 0.792791879812611, "el": 0.7416694829622287, "hi": 0.8543218260345955, "ta": 0.31536294218141925, "th": 0.8497800406187609, "zh": 2.9604097824123117, "my": 0.347033680024077}, "t": 236.0}
{"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.09705, "de": 2.31151, "vi": 1.96908, "tr": 2.29801, "fi": 2.47133, "ru": 1.45605, "el": 1.54268, "hi": 1.24916, "ta": 1.04562, "th": 1.18404, "zh": 3.2078, "my": 1.15147}, "bytes_per_chunk": {"en": 5.135, "de": 5.897, "vi": 7.941, "tr": 5.841, "fi": 5.752, "ru": 11.246, "el": 12.409, "hi": 14.088, "ta": 43.546, "th": 15.94, "zh": 4.954, "my": 36.118}, "bpb_mean": 1.83198, "t": 237.3}
{"event": "train", "step": 4000, "phase": 2, "objective": "parity_b", "lm_loss": 1.34921, "bpb": 1.9465, "chunk_loss": 0.030677, "F": 0.15317, "G": 0.20736, "chunks": 156.84, "bytes_per_chunk": 6.529, "boundary_entropy": 0.29165, "router_noise": 0.0, "grad_norm": 0.7079, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 237.3}
{"event": "train", "step": 4050, "phase": 2, "objective": "parity_b", "lm_loss": 1.54223, "bpb": 2.22497, "chunk_loss": 0.030463, "F": 0.15851, "G": 0.22258, "chunks": 162.31, "bytes_per_chunk": 6.309, "boundary_entropy": 0.31551, "router_noise": 0.0, "grad_norm": 0.7155, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "fi", "n_target_mean": 6.6232, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 239.5}
{"event": "train", "step": 4100, "phase": 2, "objective": "parity_b", "lm_loss": 1.41294, "bpb": 2.03844, "chunk_loss": 0.032462, "F": 0.15332, "G": 0.21082, "chunks": 157.0, "bytes_per_chunk": 6.522, "boundary_entropy": 0.29835, "router_noise": 0.0, "grad_norm": 0.6787, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 241.7}
{"event": "train", "step": 4150, "phase": 2, "objective": "parity_b", "lm_loss": 1.39129, "bpb": 2.00721, "chunk_loss": 0.030351, "F": 0.15817, "G": 0.20393, "chunks": 161.97, "bytes_per_chunk": 6.322, "boundary_entropy": 0.26549, "router_noise": 0.0, "grad_norm": 0.765, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "en", "n_target_mean": 6.0, "frac_in_batch": 0.4375, "fallback": 0.0, "alpha": 0.03, "t": 244.0}
{"event": "probe", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3393.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194966, "cps": 26.507812, "mean_p": 0.214299, "boundary_entropy": 0.224667, "bpb": 2.0749}, "de": {"chunks": 3357.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.163493, "cps": 26.226562, "mean_p": 0.187611, "boundary_entropy": 0.218759, "bpb": 2.28626}, "vi": {"chunks": 3272.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.177488, "cps": 25.5625, "mean_p": 0.213694, "boundary_entropy": 0.333195, "bpb": 1.959067}, "tr": {"chunks": 2983.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164779, "cps": 23.304688, "mean_p": 0.180828, "boundary_entropy": 0.213481, "bpb": 2.275072}, "fi": {"chunks": 3099.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.164744, "cps": 24.210938, "mean_p": 0.184942, "boundary_entropy": 0.213402, "bpb": 2.44528}, "ru": {"chunks": 3109.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.164063, "cps": 24.289062, "mean_p": 0.14858, "boundary_entropy": 0.270131, "bpb": 1.463041}, "el": {"chunks": 2685.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.130829, "cps": 20.976562, "mean_p": 0.159963, "boundary_entropy": 0.337914, "bpb": 1.542416}, "hi": {"chunks": 2249.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.13367, "cps": 17.570312, "mean_p": 0.171421, "boundary_entropy": 0.374193, "bpb": 1.24769}, "ta": {"chunks": 1471.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.078684, "cps": 11.492188, "mean_p": 0.162095, "boundary_entropy": 0.390033, "bpb": 1.023633}, "th": {"chunks": 2690.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.167237, "cps": 21.015625, "mean_p": 0.17851, "boundary_entropy": 0.388084, "bpb": 1.15477}, "zh": {"chunks": 3521.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.596173, "cps": 27.507812, "mean_p": 0.310811, "boundary_entropy": 0.472438, "bpb": 3.197963}, "my": {"chunks": 1229.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.064862, "cps": 9.601562, "mean_p": 0.148829, "boundary_entropy": 0.373315, "bpb": 1.142298}}, "bpb": {"en": 2.0749, "de": 2.28626, "vi": 1.95907, "tr": 2.27507, "fi": 2.44528, "ru": 1.46304, "el": 1.54242, "hi": 1.24769, "ta": 1.02363, "th": 1.15477, "zh": 3.19796, "my": 1.1423}, "gini_cpc": 0.279797, "gini_cps": 0.137571, "bpic": 11.609323, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8385697552791063, "vi": 0.910354228072586, "tr": 0.8451678388324438, "fi": 0.8449868558480723, "ru": 0.8414954426582122, "el": 0.671032732764986, "hi": 0.6856060528428207, "ta": 0.40357798141273343, "th": 0.8577712291602937, "zh": 3.0578265205879434, "my": 0.33268158916757984}, "t": 247.1}
{"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.0857, "de": 2.29386, "vi": 1.96035, "tr": 2.28452, "fi": 2.45438, "ru": 1.44178, "el": 1.53547, "hi": 1.24432, "ta": 1.02924, "th": 1.17199, "zh": 3.19609, "my": 1.1499}, "bytes_per_chunk": {"en": 5.185, "de": 6.033, "vi": 7.493, "tr": 6.226, "fi": 5.752, "ru": 10.625, "el": 13.626, "hi": 17.967, "ta": 34.999, "th": 15.999, "zh": 4.814, "my": 38.097}, "bpb_mean": 1.82063, "t": 248.4}
{"event": "phase_transition", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584, "t": 248.4}
{"event": "train", "step": 4200, "phase": 3, "objective": "none", "lm_loss": 1.20897, "bpb": 1.74417, "chunk_loss": 0.0, "F": 0.12421, "G": 0.19954, "chunks": 127.19, "bytes_per_chunk": 8.051, "boundary_entropy": 0.32159, "router_noise": 0.0, "grad_norm": 0.6384, "lr": 9.12e-05, "router_trainable": false, "sec_per_step": 0.089, "alpha": 0.0, "t": 248.4}
{"event": "train", "step": 4250, "phase": 3, "objective": "none", "lm_loss": 1.25667, "bpb": 1.813, "chunk_loss": 0.0, "F": 0.13574, "G": 0.19862, "chunks": 139.0, "bytes_per_chunk": 7.367, "boundary_entropy": 0.29346, "router_noise": 0.0, "grad_norm": 0.7054, "lr": 8.81e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 250.5}
{"event": "train", "step": 4300, "phase": 3, "objective": "none", "lm_loss": 1.48256, "bpb": 2.13888, "chunk_loss": 0.0, "F": 0.16199, "G": 0.21811, "chunks": 165.88, "bytes_per_chunk": 6.173, "boundary_entropy": 0.29133, "router_noise": 0.0, "grad_norm": 0.7249, "lr": 8.5e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 252.7}
{"event": "train", "step": 4350, "phase": 3, "objective": "none", "lm_loss": 1.47619, "bpb": 2.12969, "chunk_loss": 0.0, "F": 0.15961, "G": 0.21634, "chunks": 163.44, "bytes_per_chunk": 6.265, "boundary_entropy": 0.29429, "router_noise": 0.0, "grad_norm": 0.7644, "lr": 8.21e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 254.8}
{"event": "probe", "step": 4400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3517.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.202092, "cps": 27.476562, "mean_p": 0.224189, "boundary_entropy": 0.23118, "bpb": 2.059525}, "de": {"chunks": 3439.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.167486, "cps": 26.867188, "mean_p": 0.194134, "boundary_entropy": 0.222813, "bpb": 2.278631}, "vi": {"chunks": 3401.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.184486, "cps": 26.570312, "mean_p": 0.216899, "boundary_entropy": 0.33394, "bpb": 1.951913}, "tr": {"chunks": 3253.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179694, "cps": 25.414062, "mean_p": 0.196949, "boundary_entropy": 0.225666, "bpb": 2.247981}, "fi": {"chunks": 3296.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175217, "cps": 25.75, "mean_p": 0.197804, "boundary_entropy": 0.224985, "bpb": 2.428577}, "ru": {"chunks": 3241.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.171029, "cps": 25.320312, "mean_p": 0.158902, "boundary_entropy": 0.285688, "bpb": 1.448884}, "el": {"chunks": 3233.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.157531, "cps": 25.257812, "mean_p": 0.18368, "boundary_entropy": 0.362669, "bpb": 1.528353}, "hi": {"chunks": 2832.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.168321, "cps": 22.125, "mean_p": 0.185935, "boundary_entropy": 0.390585, "bpb": 1.234465}, "ta": {"chunks": 2107.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.112704, "cps": 16.460938, "mean_p": 0.180676, "boundary_entropy": 0.411513, "bpb": 1.01964}, "th": {"chunks": 2651.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.164812, "cps": 20.710938, "mean_p": 0.183758, "boundary_entropy": 0.399156, "bpb": 1.148805}, "zh": {"chunks": 3431.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.580935, "cps": 26.804688, "mean_p": 0.310434, "boundary_entropy": 0.476474, "bpb": 3.176238}, "my": {"chunks": 1662.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.087714, "cps": 12.984375, "mean_p": 0.160821, "boundary_entropy": 0.391644, "bpb": 1.129422}}, "bpb": {"en": 2.05952, "de": 2.27863, "vi": 1.95191, "tr": 2.24798, "fi": 2.42858, "ru": 1.44888, "el": 1.52835, "hi": 1.23446, "ta": 1.01964, "th": 1.14881, "zh": 3.17624, "my": 1.12942}, "gini_cpc": 0.230969, "gini_cps": 0.094508, "bpic": 10.64196, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8287652264478905, "vi": 0.9128832570291503, "tr": 0.8891709464423642, "fi": 0.8670159189581194, "ru": 0.8462945977434894, "el": 0.7795008827381211, "hi": 0.8328943729480371, "ta": 0.5576873814216707, "th": 0.8155308878082204, "zh": 2.8746106305197885, "my": 0.43402964664707366}, "t": 257.8}
{"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.07019, "de": 2.28454, "vi": 1.95304, "tr": 2.25799, "fi": 2.43738, "ru": 1.42659, "el": 1.51991, "hi": 1.23033, "ta": 1.02502, "th": 1.16706, "zh": 3.17526, "my": 1.1372}, "bytes_per_chunk": {"en": 5.028, "de": 5.876, "vi": 7.101, "tr": 5.736, "fi": 5.461, "ru": 10.167, "el": 11.483, "hi": 14.384, "ta": 25.247, "th": 16.045, "zh": 5.019, "my": 28.921}, "bpb_mean": 1.80704, "t": 259.1}
{"event": "train", "step": 4400, "phase": 3, "objective": "none", "lm_loss": 1.48745, "bpb": 2.14594, "chunk_loss": 0.0, "F": 0.16565, "G": 0.22455, "chunks": 169.62, "bytes_per_chunk": 6.037, "boundary_entropy": 0.30649, "router_noise": 0.0, "grad_norm": 0.758, "lr": 7.92e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 259.2}
{"event": "train", "step": 4450, "phase": 3, "objective": "none", "lm_loss": 1.49445, "bpb": 2.15604, "chunk_loss": 0.0, "F": 0.16162, "G": 0.231, "chunks": 165.5, "bytes_per_chunk": 6.187, "boundary_entropy": 0.33539, "router_noise": 0.0, "grad_norm": 0.7027, "lr": 7.63e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 261.2}
{"event": "train", "step": 4500, "phase": 3, "objective": "none", "lm_loss": 1.42521, "bpb": 2.05615, "chunk_loss": 0.0, "F": 0.15955, "G": 0.21248, "chunks": 163.38, "bytes_per_chunk": 6.268, "boundary_entropy": 0.28083, "router_noise": 0.0, "grad_norm": 0.7462, "lr": 7.36e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 263.4}
{"event": "train", "step": 4550, "phase": 3, "objective": "none", "lm_loss": 1.23337, "bpb": 1.77938, "chunk_loss": 0.0, "F": 0.14838, "G": 0.2039, "chunks": 151.94, "bytes_per_chunk": 6.74, "boundary_entropy": 0.28855, "router_noise": 0.0, "grad_norm": 0.7248, "lr": 7.09e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 265.5}
{"event": "probe", "step": 4600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3539.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203356, "cps": 27.648438, "mean_p": 0.228657, "boundary_entropy": 0.2361, "bpb": 2.050309}, "de": {"chunks": 3537.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.172259, "cps": 27.632812, "mean_p": 0.20033, "boundary_entropy": 0.226768, "bpb": 2.26411}, "vi": {"chunks": 3348.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.181611, "cps": 26.15625, "mean_p": 0.223735, "boundary_entropy": 0.345031, "bpb": 1.945379}, "tr": {"chunks": 3375.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.186433, "cps": 26.367188, "mean_p": 0.201501, "boundary_entropy": 0.23002, "bpb": 2.232593}, "fi": {"chunks": 3297.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.17527, "cps": 25.757812, "mean_p": 0.200791, "boundary_entropy": 0.230291, "bpb": 2.405031}, "ru": {"chunks": 3362.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.177414, "cps": 26.265625, "mean_p": 0.15957, "boundary_entropy": 0.285289, "bpb": 1.440445}, "el": {"chunks": 3165.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.154217, "cps": 24.726562, "mean_p": 0.183269, "boundary_entropy": 0.36167, "bpb": 1.518216}, "hi": {"chunks": 2965.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.176226, "cps": 23.164062, "mean_p": 0.191751, "boundary_entropy": 0.399721, "bpb": 1.22681}, "ta": {"chunks": 1264.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.067612, "cps": 9.875, "mean_p": 0.167345, "boundary_entropy": 0.401549, "bpb": 1.020306}, "th": {"chunks": 2924.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.181784, "cps": 22.84375, "mean_p": 0.188511, "boundary_entropy": 0.400039, "bpb": 1.141374}, "zh": {"chunks": 3378.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.571961, "cps": 26.390625, "mean_p": 0.311893, "boundary_entropy": 0.480337, "bpb": 3.159299}, "my": {"chunks": 1336.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.070509, "cps": 10.4375, "mean_p": 0.154639, "boundary_entropy": 0.383767, "bpb": 1.124135}}, "bpb": {"en": 2.05031, "de": 2.26411, "vi": 1.94538, "tr": 2.23259, "fi": 2.40503, "ru": 1.44044, "el": 1.51822, "hi": 1.22681, "ta": 1.02031, "th": 1.14137, "zh": 3.1593, "my": 1.12413}, "gini_cpc": 0.252906, "gini_cps": 0.119057, "bpic": 10.813779, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8470834762755939, "vi": 0.8930707487944974, "tr": 0.9167834892600762, "fi": 0.8618875776077403, "ru": 0.872432935380431, "el": 0.7583617616618448, "hi": 0.8665890205802715, "ta": 0.3324797214336019, "th": 0.8939224835425464, "zh": 2.812611577806469, "my": 0.34672618381873843}, "t": 268.5}
{"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.06093, "de": 2.2707, "vi": 1.94598, "tr": 2.2437, "fi": 2.41474, "ru": 1.41834, "el": 1.50936, "hi": 1.22334, "ta": 1.02578, "th": 1.15786, "zh": 3.1594, "my": 1.13114}, "bytes_per_chunk": {"en": 4.974, "de": 5.686, "vi": 7.358, "tr": 5.562, "fi": 5.455, "ru": 9.921, "el": 11.516, "hi": 13.479, "ta": 38.28, "th": 14.073, "zh": 5.023, "my": 35.444}, "bpb_mean": 1.79677, "t": 269.8}
{"event": "train", "step": 4600, "phase": 3, "objective": "none", "lm_loss": 1.31314, "bpb": 1.89447, "chunk_loss": 0.0, "F": 0.13303, "G": 0.20836, "chunks": 136.22, "bytes_per_chunk": 7.517, "boundary_entropy": 0.32343, "router_noise": 0.0, "grad_norm": 0.7306, "lr": 6.82e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 269.9}
{"event": "train", "step": 4650, "phase": 3, "objective": "none", "lm_loss": 1.41312, "bpb": 2.03871, "chunk_loss": 0.0, "F": 0.15112, "G": 0.22508, "chunks": 154.75, "bytes_per_chunk": 6.617, "boundary_entropy": 0.33167, "router_noise": 0.0, "grad_norm": 0.6796, "lr": 6.57e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 272.0}
{"event": "train", "step": 4700, "phase": 3, "objective": "none", "lm_loss": 1.33166, "bpb": 1.92118, "chunk_loss": 0.0, "F": 0.15558, "G": 0.21519, "chunks": 159.31, "bytes_per_chunk": 6.428, "boundary_entropy": 0.29668, "router_noise": 0.0, "grad_norm": 0.7471, "lr": 6.32e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 274.1}
{"event": "train", "step": 4750, "phase": 3, "objective": "none", "lm_loss": 1.34323, "bpb": 1.93788, "chunk_loss": 0.0, "F": 0.15875, "G": 0.21943, "chunks": 162.56, "bytes_per_chunk": 6.299, "boundary_entropy": 0.3008, "router_noise": 0.0, "grad_norm": 0.7333, "lr": 6.08e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 276.2}
{"event": "probe", "step": 4800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3558.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.204448, "cps": 27.796875, "mean_p": 0.230502, "boundary_entropy": 0.235189, "bpb": 2.042404}, "de": {"chunks": 3568.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.173769, "cps": 27.875, "mean_p": 0.202893, "boundary_entropy": 0.228192, "bpb": 2.251648}, "vi": {"chunks": 3513.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.190561, "cps": 27.445312, "mean_p": 0.226701, "boundary_entropy": 0.342642, "bpb": 1.941208}, "tr": {"chunks": 3350.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185052, "cps": 26.171875, "mean_p": 0.204049, "boundary_entropy": 0.23351, "bpb": 2.223959}, "fi": {"chunks": 3313.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.17612, "cps": 25.882812, "mean_p": 0.20218, "boundary_entropy": 0.23075, "bpb": 2.38956}, "ru": {"chunks": 3638.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.191979, "cps": 28.421875, "mean_p": 0.171989, "boundary_entropy": 0.296078, "bpb": 1.425233}, "el": {"chunks": 3533.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.172148, "cps": 27.601562, "mean_p": 0.194109, "boundary_entropy": 0.369468, "bpb": 1.503916}, "hi": {"chunks": 3208.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.190669, "cps": 25.0625, "mean_p": 0.201587, "boundary_entropy": 0.408824, "bpb": 1.222687}, "ta": {"chunks": 2154.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.115218, "cps": 16.828125, "mean_p": 0.190564, "boundary_entropy": 0.428769, "bpb": 1.009065}, "th": {"chunks": 3558.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.2212, "cps": 27.796875, "mean_p": 0.202226, "boundary_entropy": 0.415396, "bpb": 1.133275}, "zh": {"chunks": 3437.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.581951, "cps": 26.851562, "mean_p": 0.314134, "boundary_entropy": 0.48082, "bpb": 3.14977}, "my": {"chunks": 1929.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.101805, "cps": 15.070312, "mean_p": 0.16941, "boundary_entropy": 0.401874, "bpb": 1.118759}}, "bpb": {"en": 2.0424, "de": 2.25165, "vi": 1.94121, "tr": 2.22396, "fi": 2.38956, "ru": 1.42523, "el": 1.50392, "hi": 1.22269, "ta": 1.00906, "th": 1.13328, "zh": 3.14977, "my": 1.11876}, "gini_cpc": 0.219626, "gini_cps": 0.077025, "bpic": 9.901726, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8499445921339638, "vi": 0.9320799893523163, "tr": 0.9051330688856598, "fi": 0.8614453457007278, "ru": 0.9390131125220804, "el": 0.8420172111427815, "hi": 0.9326044065358855, "ta": 0.5635577232281584, "th": 1.0819396953683555, "zh": 2.846454630127192, "my": 0.49795148059426686}, "t": 279.2}
{"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.05351, "de": 2.26027, "vi": 1.94199, "tr": 2.23291, "fi": 2.39876, "ru": 1.40382, "el": 1.49569, "hi": 1.21857, "ta": 1.01466, "th": 1.14992, "zh": 3.14716, "my": 1.12591}, "bytes_per_chunk": {"en": 4.949, "de": 5.648, "vi": 7.076, "tr": 5.607, "fi": 5.422, "ru": 9.209, "el": 10.529, "hi": 12.528, "ta": 24.703, "th": 11.88, "zh": 4.929, "my": 24.385}, "bpb_mean": 1.78693, "t": 280.5}
{"event": "train", "step": 4800, "phase": 3, "objective": "none", "lm_loss": 1.26664, "bpb": 1.82737, "chunk_loss": 0.0, "F": 0.14557, "G": 0.21805, "chunks": 149.06, "bytes_per_chunk": 6.87, "boundary_entropy": 0.31677, "router_noise": 0.0, "grad_norm": 0.6988, "lr": 5.85e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 280.5}
{"event": "train", "step": 4850, "phase": 3, "objective": "none", "lm_loss": 1.49642, "bpb": 2.15888, "chunk_loss": 0.0, "F": 0.16919, "G": 0.22951, "chunks": 173.25, "bytes_per_chunk": 5.911, "boundary_entropy": 0.30446, "router_noise": 0.0, "grad_norm": 0.7555, "lr": 5.62e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 282.7}
{"event": "train", "step": 4900, "phase": 3, "objective": "none", "lm_loss": 1.41208, "bpb": 2.0372, "chunk_loss": 0.0, "F": 0.15515, "G": 0.23502, "chunks": 158.88, "bytes_per_chunk": 6.445, "boundary_entropy": 0.35571, "router_noise": 0.0, "grad_norm": 0.6989, "lr": 5.41e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 284.8}
{"event": "train", "step": 4950, "phase": 3, "objective": "none", "lm_loss": 1.52336, "bpb": 2.19774, "chunk_loss": 0.0, "F": 0.18234, "G": 0.23713, "chunks": 186.72, "bytes_per_chunk": 5.484, "boundary_entropy": 0.30764, "router_noise": 0.0, "grad_norm": 0.7699, "lr": 5.2e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 286.9}
{"event": "probe", "step": 5000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3598.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.206746, "cps": 28.109375, "mean_p": 0.23264, "boundary_entropy": 0.236387, "bpb": 2.031415}, "de": {"chunks": 3614.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.176009, "cps": 28.234375, "mean_p": 0.204603, "boundary_entropy": 0.228037, "bpb": 2.237947}, "vi": {"chunks": 3653.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.198156, "cps": 28.539062, "mean_p": 0.231511, "boundary_entropy": 0.34426, "bpb": 1.93474}, "tr": {"chunks": 3441.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190079, "cps": 26.882812, "mean_p": 0.20732, "boundary_entropy": 0.233262, "bpb": 2.218395}, "fi": {"chunks": 3428.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182234, "cps": 26.78125, "mean_p": 0.20718, "boundary_entropy": 0.23242, "bpb": 2.376167}, "ru": {"chunks": 3668.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.193562, "cps": 28.65625, "mean_p": 0.172896, "boundary_entropy": 0.296106, "bpb": 1.415391}, "el": {"chunks": 3339.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.162696, "cps": 26.085938, "mean_p": 0.189318, "boundary_entropy": 0.367481, "bpb": 1.49769}, "hi": {"chunks": 3111.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.184903, "cps": 24.304688, "mean_p": 0.198656, "boundary_entropy": 0.405969, "bpb": 1.212196}, "ta": {"chunks": 1815.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.097085, "cps": 14.179688, "mean_p": 0.186487, "boundary_entropy": 0.426264, "bpb": 1.003845}, "th": {"chunks": 3291.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.204601, "cps": 25.710938, "mean_p": 0.198361, "boundary_entropy": 0.412291, "bpb": 1.129363}, "zh": {"chunks": 3491.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.591094, "cps": 27.273438, "mean_p": 0.314815, "boundary_entropy": 0.476956, "bpb": 3.142876}, "my": {"chunks": 1566.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.082647, "cps": 12.234375, "mean_p": 0.165414, "boundary_entropy": 0.40093, "bpb": 1.113474}}, "bpb": {"en": 2.03141, "de": 2.23795, "vi": 1.93474, "tr": 2.2184, "fi": 2.37617, "ru": 1.41539, "el": 1.49769, "hi": 1.2122, "ta": 1.00384, "th": 1.12936, "zh": 3.14288, "my": 1.11347}, "gini_cpc": 0.236964, "gini_cps": 0.09906, "bpic": 10.095515, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8513314986082275, "vi": 0.9584500656046596, "tr": 0.919384298303909, "fi": 0.88143824638222, "ru": 0.9362311222446947, "el": 0.7869344084107212, "hi": 0.8943507991479499, "ta": 0.46958489761555144, "th": 0.9896229954022396, "zh": 2.8590343113070116, "my": 0.3997526421159317}, "t": 289.9}
{"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.04206, "de": 2.24504, "vi": 1.9357, "tr": 2.22715, "fi": 2.38558, "ru": 1.39269, "el": 1.49024, "hi": 1.20848, "ta": 1.00966, "th": 1.14629, "zh": 3.14064, "my": 1.12117}, "bytes_per_chunk": {"en": 4.869, "de": 5.562, "vi": 6.766, "tr": 5.436, "fi": 5.285, "ru": 9.139, "el": 11.027, "hi": 12.472, "ta": 27.185, "th": 12.801, "zh": 4.833, "my": 29.881}, "bpb_mean": 1.77872, "t": 291.2}
{"event": "train", "step": 5000, "phase": 3, "objective": "none", "lm_loss": 1.29467, "bpb": 1.86781, "chunk_loss": 0.0, "F": 0.15457, "G": 0.22485, "chunks": 158.28, "bytes_per_chunk": 6.469, "boundary_entropy": 0.32517, "router_noise": 0.0, "grad_norm": 0.7555, "lr": 5e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 291.3}
{"event": "train", "step": 5050, "phase": 3, "objective": "none", "lm_loss": 1.26939, "bpb": 1.83134, "chunk_loss": 0.0, "F": 0.15244, "G": 0.21397, "chunks": 156.09, "bytes_per_chunk": 6.56, "boundary_entropy": 0.29772, "router_noise": 0.0, "grad_norm": 0.6853, "lr": 4.81e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 293.4}
{"event": "train", "step": 5100, "phase": 3, "objective": "none", "lm_loss": 1.31025, "bpb": 1.89029, "chunk_loss": 0.0, "F": 0.15472, "G": 0.22009, "chunks": 158.44, "bytes_per_chunk": 6.463, "boundary_entropy": 0.30769, "router_noise": 0.0, "grad_norm": 0.7225, "lr": 4.63e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 295.5}
{"event": "train", "step": 5150, "phase": 3, "objective": "none", "lm_loss": 1.33785, "bpb": 1.93012, "chunk_loss": 0.0, "F": 0.16058, "G": 0.21993, "chunks": 164.44, "bytes_per_chunk": 6.227, "boundary_entropy": 0.30176, "router_noise": 0.0, "grad_norm": 0.7256, "lr": 4.45e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 297.6}
{"event": "probe", "step": 5200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3588.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.206171, "cps": 28.03125, "mean_p": 0.232738, "boundary_entropy": 0.236338, "bpb": 2.027512}, "de": {"chunks": 3644.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.17747, "cps": 28.46875, "mean_p": 0.205684, "boundary_entropy": 0.230019, "bpb": 2.231537}, "vi": {"chunks": 3590.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.194738, "cps": 28.046875, "mean_p": 0.22802, "boundary_entropy": 0.34258, "bpb": 1.929313}, "tr": {"chunks": 3424.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.18914, "cps": 26.75, "mean_p": 0.205293, "boundary_entropy": 0.231273, "bpb": 2.203813}, "fi": {"chunks": 3376.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179469, "cps": 26.375, "mean_p": 0.206007, "boundary_entropy": 0.232162, "bpb": 2.364552}, "ru": {"chunks": 3696.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.19504, "cps": 28.875, "mean_p": 0.173884, "boundary_entropy": 0.297699, "bpb": 1.407761}, "el": {"chunks": 3379.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.164645, "cps": 26.398438, "mean_p": 0.19085, "boundary_entropy": 0.366883, "bpb": 1.491172}, "hi": {"chunks": 3078.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.182942, "cps": 24.046875, "mean_p": 0.201126, "boundary_entropy": 0.411394, "bpb": 1.203692}, "ta": {"chunks": 1669.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.089275, "cps": 13.039062, "mean_p": 0.177995, "boundary_entropy": 0.411506, "bpb": 1.003166}, "th": {"chunks": 3207.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.199378, "cps": 25.054688, "mean_p": 0.198404, "boundary_entropy": 0.415083, "bpb": 1.12055}, "zh": {"chunks": 3512.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.59465, "cps": 27.4375, "mean_p": 0.317952, "boundary_entropy": 0.480311, "bpb": 3.125624}, "my": {"chunks": 1521.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.080272, "cps": 11.882812, "mean_p": 0.164301, "boundary_entropy": 0.398174, "bpb": 1.11211}}, "bpb": {"en": 2.02751, "de": 2.23154, "vi": 1.92931, "tr": 2.20381, "fi": 2.36455, "ru": 1.40776, "el": 1.49117, "hi": 1.20369, "ta": 1.00317, "th": 1.12055, "zh": 3.12562, "my": 1.11211}, "gini_cpc": 0.240952, "gini_cps": 0.105354, "bpic": 10.18419, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.860790860034919, "vi": 0.9445457373960575, "tr": 0.9173918770646673, "fi": 0.8704869011211508, "ru": 0.9460071831346353, "el": 0.7985811090159551, "hi": 0.8873301296545693, "ta": 0.4330146132110155, "th": 0.9670514501243915, "zh": 2.884248997957975, "my": 0.38934761957210123}, "t": 300.6}
{"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.03734, "de": 2.23774, "vi": 1.93027, "tr": 2.21333, "fi": 2.37453, "ru": 1.38526, "el": 1.48445, "hi": 1.20163, "ta": 1.00886, "th": 1.13734, "zh": 3.12347, "my": 1.11975}, "bytes_per_chunk": {"en": 4.879, "de": 5.527, "vi": 6.881, "tr": 5.493, "fi": 5.34, "ru": 9.096, "el": 10.777, "hi": 13.028, "ta": 29.922, "th": 13.13, "zh": 4.829, "my": 30.492}, "bpb_mean": 1.77116, "t": 301.9}
{"event": "train", "step": 5200, "phase": 3, "objective": "none", "lm_loss": 1.29661, "bpb": 1.87061, "chunk_loss": 0.0, "F": 0.14697, "G": 0.21558, "chunks": 150.5, "bytes_per_chunk": 6.804, "boundary_entropy": 0.31167, "router_noise": 0.0, "grad_norm": 0.789, "lr": 4.29e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 301.9}
{"event": "train", "step": 5250, "phase": 3, "objective": "none", "lm_loss": 1.38109, "bpb": 1.9925, "chunk_loss": 0.0, "F": 0.16479, "G": 0.2305, "chunks": 168.75, "bytes_per_chunk": 6.068, "boundary_entropy": 0.32026, "router_noise": 0.0, "grad_norm": 0.7231, "lr": 4.14e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 304.1}
{"event": "train", "step": 5300, "phase": 3, "objective": "none", "lm_loss": 1.3832, "bpb": 1.99554, "chunk_loss": 0.0, "F": 0.16345, "G": 0.23039, "chunks": 167.38, "bytes_per_chunk": 6.118, "boundary_entropy": 0.31758, "router_noise": 0.0, "grad_norm": 0.7482, "lr": 3.99e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 306.2}
{"event": "train", "step": 5350, "phase": 3, "objective": "none", "lm_loss": 1.43306, "bpb": 2.06746, "chunk_loss": 0.0, "F": 0.16412, "G": 0.22818, "chunks": 168.06, "bytes_per_chunk": 6.093, "boundary_entropy": 0.31379, "router_noise": 0.0, "grad_norm": 0.7815, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 308.3}
{"event": "probe", "step": 5400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3611.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.207493, "cps": 28.210938, "mean_p": 0.235129, "boundary_entropy": 0.238044, "bpb": 2.019799}, "de": {"chunks": 3702.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.180295, "cps": 28.921875, "mean_p": 0.209288, "boundary_entropy": 0.232545, "bpb": 2.225208}, "vi": {"chunks": 3773.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.204665, "cps": 29.476562, "mean_p": 0.237028, "boundary_entropy": 0.350933, "bpb": 1.922462}, "tr": {"chunks": 3429.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.189416, "cps": 26.789062, "mean_p": 0.208007, "boundary_entropy": 0.234669, "bpb": 2.198923}, "fi": {"chunks": 3419.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181755, "cps": 26.710938, "mean_p": 0.207388, "boundary_entropy": 0.23254, "bpb": 2.3556}, "ru": {"chunks": 3599.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.189921, "cps": 28.117188, "mean_p": 0.171735, "boundary_entropy": 0.296412, "bpb": 1.404459}, "el": {"chunks": 3492.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.170151, "cps": 27.28125, "mean_p": 0.193194, "boundary_entropy": 0.368424, "bpb": 1.481868}, "hi": {"chunks": 3155.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.187519, "cps": 24.648438, "mean_p": 0.200004, "boundary_entropy": 0.406736, "bpb": 1.201686}, "ta": {"chunks": 1929.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.103183, "cps": 15.070312, "mean_p": 0.185096, "boundary_entropy": 0.419095, "bpb": 0.997633}, "th": {"chunks": 3277.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.20373, "cps": 25.601562, "mean_p": 0.203085, "boundary_entropy": 0.420118, "bpb": 1.11526}, "zh": {"chunks": 3532.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.598036, "cps": 27.59375, "mean_p": 0.317807, "boundary_entropy": 0.480004, "bpb": 3.121571}, "my": {"chunks": 2042.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.107769, "cps": 15.953125, "mean_p": 0.175841, "boundary_entropy": 0.411308, "bpb": 1.108414}}, "bpb": {"en": 2.0198, "de": 2.22521, "vi": 1.92246, "tr": 2.19892, "fi": 2.3556, "ru": 1.40446, "el": 1.48187, "hi": 1.20169, "ta": 0.99763, "th": 1.11526, "zh": 3.12157, "my": 1.10841}, "gini_cpc": 0.222077, "gini_cps": 0.086473, "bpic": 9.850642, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8689216916394914, "vi": 0.9863709995608303, "tr": 0.9128797332559604, "fi": 0.8759591412020057, "ru": 0.915312227589548, "el": 0.8200305300793014, "hi": 0.9037346262843063, "ta": 0.4972827447913492, "th": 0.9818655169543343, "zh": 2.882198474897459, "my": 0.519384510316009}, "t": 311.4}
{"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.03033, "de": 2.22982, "vi": 1.92228, "tr": 2.20865, "fi": 2.36531, "ru": 1.38222, "el": 1.47512, "hi": 1.19813, "ta": 1.00339, "th": 1.13221, "zh": 3.11944, "my": 1.1167}, "bytes_per_chunk": {"en": 4.865, "de": 5.45, "vi": 6.536, "tr": 5.456, "fi": 5.325, "ru": 9.239, "el": 10.42, "hi": 12.555, "ta": 27.07, "th": 12.876, "zh": 4.83, "my": 24.228}, "bpb_mean": 1.7653, "t": 312.7}
{"event": "train", "step": 5400, "phase": 3, "objective": "none", "lm_loss": 1.14478, "bpb": 1.65157, "chunk_loss": 0.0, "F": 0.13562, "G": 0.21258, "chunks": 138.88, "bytes_per_chunk": 7.374, "boundary_entropy": 0.32648, "router_noise": 0.0, "grad_norm": 0.697, "lr": 3.73e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 312.7}
{"event": "train", "step": 5450, "phase": 3, "objective": "none", "lm_loss": 1.20468, "bpb": 1.73799, "chunk_loss": 0.0, "F": 0.1394, "G": 0.21302, "chunks": 142.75, "bytes_per_chunk": 7.173, "boundary_entropy": 0.31958, "router_noise": 0.0, "grad_norm": 0.7387, "lr": 3.62e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 314.8}
{"event": "train", "step": 5500, "phase": 3, "objective": "none", "lm_loss": 1.39666, "bpb": 2.01495, "chunk_loss": 0.0, "F": 0.16959, "G": 0.22574, "chunks": 173.66, "bytes_per_chunk": 5.897, "boundary_entropy": 0.29535, "router_noise": 0.0, "grad_norm": 0.7566, "lr": 3.51e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 317.0}
{"event": "train", "step": 5550, "phase": 3, "objective": "none", "lm_loss": 1.41649, "bpb": 2.04357, "chunk_loss": 0.0, "F": 0.16516, "G": 0.22919, "chunks": 169.12, "bytes_per_chunk": 6.055, "boundary_entropy": 0.31192, "router_noise": 0.0, "grad_norm": 0.8019, "lr": 3.41e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 319.1}
{"event": "probe", "step": 5600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3624.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20824, "cps": 28.3125, "mean_p": 0.234935, "boundary_entropy": 0.237861, "bpb": 2.012658}, "de": {"chunks": 3690.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.179711, "cps": 28.828125, "mean_p": 0.208152, "boundary_entropy": 0.231309, "bpb": 2.215031}, "vi": {"chunks": 3682.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.199729, "cps": 28.765625, "mean_p": 0.23511, "boundary_entropy": 0.349115, "bpb": 1.920815}, "tr": {"chunks": 3482.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.192344, "cps": 27.203125, "mean_p": 0.208532, "boundary_entropy": 0.234477, "bpb": 2.185789}, "fi": {"chunks": 3427.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182181, "cps": 26.773438, "mean_p": 0.207696, "boundary_entropy": 0.233309, "bpb": 2.349226}, "ru": {"chunks": 3594.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.189657, "cps": 28.078125, "mean_p": 0.172289, "boundary_entropy": 0.298025, "bpb": 1.400823}, "el": {"chunks": 3452.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.168202, "cps": 26.96875, "mean_p": 0.196191, "boundary_entropy": 0.374004, "bpb": 1.474418}, "hi": {"chunks": 3179.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.188945, "cps": 24.835938, "mean_p": 0.20186, "boundary_entropy": 0.411786, "bpb": 1.194882}, "ta": {"chunks": 1925.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.102969, "cps": 15.039062, "mean_p": 0.182826, "boundary_entropy": 0.417432, "bpb": 0.996884}, "th": {"chunks": 3307.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.205595, "cps": 25.835938, "mean_p": 0.202234, "boundary_entropy": 0.41826, "bpb": 1.114788}, "zh": {"chunks": 3445.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583305, "cps": 26.914062, "mean_p": 0.315712, "boundary_entropy": 0.480274, "bpb": 3.116291}, "my": {"chunks": 1918.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.101224, "cps": 14.984375, "mean_p": 0.173334, "boundary_entropy": 0.409586, "bpb": 1.104861}}, "bpb": {"en": 2.01266, "de": 2.21503, "vi": 1.92082, "tr": 2.18579, "fi": 2.34923, "ru": 1.40082, "el": 1.47442, "hi": 1.19488, "ta": 0.99688, "th": 1.11479, "zh": 3.11629, "my": 1.10486}, "gini_cpc": 0.22126, "gini_cps": 0.086135, "bpic": 9.91042, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8629982006029383, "vi": 0.9591280083773847, "tr": 0.9236642750296042, "fi": 0.8748591777883942, "ru": 0.910761764140558, "el": 0.8077293671161381, "hi": 0.9073427767336134, "ta": 0.4944714208034517, "th": 0.9872998268717306, "zh": 2.8011200025566096, "my": 0.48609500452266013}, "t": 322.2}
{"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.0242, "de": 2.22157, "vi": 1.92018, "tr": 2.19531, "fi": 2.35777, "ru": 1.3774, "el": 1.46773, "hi": 1.19185, "ta": 1.00233, "th": 1.13087, "zh": 3.1135, "my": 1.11318}, "bytes_per_chunk": {"en": 4.859, "de": 5.462, "vi": 6.765, "tr": 5.417, "fi": 5.31, "ru": 9.284, "el": 10.564, "hi": 12.679, "ta": 27.216, "th": 12.669, "zh": 4.947, "my": 25.588}, "bpb_mean": 1.75966, "t": 323.5}
{"event": "train", "step": 5600, "phase": 3, "objective": "none", "lm_loss": 1.47297, "bpb": 2.12505, "chunk_loss": 0.0, "F": 0.17374, "G": 0.23373, "chunks": 177.91, "bytes_per_chunk": 5.756, "boundary_entropy": 0.31213, "router_noise": 0.0, "grad_norm": 0.789, "lr": 3.33e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 323.5}
{"event": "train", "step": 5650, "phase": 3, "objective": "none", "lm_loss": 1.27922, "bpb": 1.84552, "chunk_loss": 0.0, "F": 0.15106, "G": 0.22182, "chunks": 154.69, "bytes_per_chunk": 6.62, "boundary_entropy": 0.32101, "router_noise": 0.0, "grad_norm": 0.7343, "lr": 3.25e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 325.7}
{"event": "train", "step": 5700, "phase": 3, "objective": "none", "lm_loss": 1.32613, "bpb": 1.9132, "chunk_loss": 0.0, "F": 0.16602, "G": 0.22388, "chunks": 170.0, "bytes_per_chunk": 6.024, "boundary_entropy": 0.29682, "router_noise": 0.0, "grad_norm": 0.7576, "lr": 3.18e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 327.8}
{"event": "train", "step": 5750, "phase": 3, "objective": "none", "lm_loss": 1.21106, "bpb": 1.74719, "chunk_loss": 0.0, "F": 0.15005, "G": 0.21468, "chunks": 153.66, "bytes_per_chunk": 6.664, "boundary_entropy": 0.3013, "router_noise": 0.0, "grad_norm": 0.717, "lr": 3.13e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 329.9}
{"event": "probe", "step": 5800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3676.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.211228, "cps": 28.71875, "mean_p": 0.237971, "boundary_entropy": 0.237598, "bpb": 2.008663}, "de": {"chunks": 3752.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18273, "cps": 29.3125, "mean_p": 0.210658, "boundary_entropy": 0.230189, "bpb": 2.21128}, "vi": {"chunks": 3755.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.203689, "cps": 29.335938, "mean_p": 0.235655, "boundary_entropy": 0.346957, "bpb": 1.915524}, "tr": {"chunks": 3544.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195769, "cps": 27.6875, "mean_p": 0.211567, "boundary_entropy": 0.236474, "bpb": 2.179072}, "fi": {"chunks": 3458.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183829, "cps": 27.015625, "mean_p": 0.210898, "boundary_entropy": 0.235298, "bpb": 2.341234}, "ru": {"chunks": 3607.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.190343, "cps": 28.179688, "mean_p": 0.173147, "boundary_entropy": 0.298603, "bpb": 1.394838}, "el": {"chunks": 3388.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.165083, "cps": 26.46875, "mean_p": 0.190727, "boundary_entropy": 0.366817, "bpb": 1.467901}, "hi": {"chunks": 3389.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.201426, "cps": 26.476562, "mean_p": 0.205548, "boundary_entropy": 0.414395, "bpb": 1.19047}, "ta": {"chunks": 1821.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.097406, "cps": 14.226562, "mean_p": 0.185477, "boundary_entropy": 0.422269, "bpb": 0.992911}, "th": {"chunks": 3703.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.230214, "cps": 28.929688, "mean_p": 0.212568, "boundary_entropy": 0.430371, "bpb": 1.110327}, "zh": {"chunks": 3573.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.604978, "cps": 27.914062, "mean_p": 0.320183, "boundary_entropy": 0.480786, "bpb": 3.109915}, "my": {"chunks": 1896.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.100063, "cps": 14.8125, "mean_p": 0.173233, "boundary_entropy": 0.40883, "bpb": 1.104676}}, "bpb": {"en": 2.00866, "de": 2.21128, "vi": 1.91552, "tr": 2.17907, "fi": 2.34123, "ru": 1.39484, "el": 1.4679, "hi": 1.19047, "ta": 0.99291, "th": 1.11033, "zh": 3.10991, "my": 1.10468}, "gini_cpc": 0.235079, "gini_cps": 0.088667, "bpic": 9.700748, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8650855145624811, "vi": 0.9643072165149262, "tr": 0.9268122796545003, "fi": 0.8702854504422868, "ru": 0.901126051891898, "el": 0.7815399342026464, "hi": 0.95359752104733, "ta": 0.4611403170009612, "th": 1.0898864776792765, "zh": 2.8641000907581122, "my": 0.4737220221059733}, "t": 333.0}
{"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.01925, "de": 2.21785, "vi": 1.91491, "tr": 2.18882, "fi": 2.35073, "ru": 1.37258, "el": 1.4618, "hi": 1.18731, "ta": 0.99849, "th": 1.12602, "zh": 3.10791, "my": 1.11282}, "bytes_per_chunk": {"en": 4.79, "de": 5.386, "vi": 6.657, "tr": 5.342, "fi": 5.244, "ru": 9.256, "el": 10.765, "hi": 11.603, "ta": 27.682, "th": 11.412, "zh": 4.775, "my": 25.688}, "bpb_mean": 1.75487, "t": 334.3}
{"event": "train", "step": 5800, "phase": 3, "objective": "none", "lm_loss": 1.26247, "bpb": 1.82137, "chunk_loss": 0.0, "F": 0.15659, "G": 0.22445, "chunks": 160.34, "bytes_per_chunk": 6.386, "boundary_entropy": 0.31414, "router_noise": 0.0, "grad_norm": 0.7215, "lr": 3.08e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 334.3}
{"event": "train", "step": 5850, "phase": 3, "objective": "none", "lm_loss": 1.37357, "bpb": 1.98164, "chunk_loss": 0.0, "F": 0.16183, "G": 0.22971, "chunks": 165.72, "bytes_per_chunk": 6.179, "boundary_entropy": 0.32188, "router_noise": 0.0, "grad_norm": 0.7014, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 336.5}
{"event": "train", "step": 5900, "phase": 3, "objective": "none", "lm_loss": 1.33052, "bpb": 1.91953, "chunk_loss": 0.0, "F": 0.16428, "G": 0.22332, "chunks": 168.22, "bytes_per_chunk": 6.087, "boundary_entropy": 0.30014, "router_noise": 0.0, "grad_norm": 0.7651, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 338.6}
{"event": "train", "step": 5950, "phase": 3, "objective": "none", "lm_loss": 1.28787, "bpb": 1.85801, "chunk_loss": 0.0, "F": 0.1499, "G": 0.22145, "chunks": 153.5, "bytes_per_chunk": 6.671, "boundary_entropy": 0.31842, "router_noise": 0.0, "grad_norm": 0.7141, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 340.8}
{"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3637.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.208987, "cps": 28.414062, "mean_p": 0.235227, "boundary_entropy": 0.235231, "bpb": 2.005763}, "de": {"chunks": 3628.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.176691, "cps": 28.34375, "mean_p": 0.205738, "boundary_entropy": 0.227873, "bpb": 2.209956}, "vi": {"chunks": 3600.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.195281, "cps": 28.125, "mean_p": 0.231494, "boundary_entropy": 0.346194, "bpb": 1.914264}, "tr": {"chunks": 3473.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.191847, "cps": 27.132812, "mean_p": 0.209226, "boundary_entropy": 0.235501, "bpb": 2.17907}, "fi": {"chunks": 3443.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183031, "cps": 26.898438, "mean_p": 0.208423, "boundary_entropy": 0.231517, "bpb": 2.339462}, "ru": {"chunks": 3732.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.196939, "cps": 29.15625, "mean_p": 0.177415, "boundary_entropy": 0.300712, "bpb": 1.391678}, "el": {"chunks": 3423.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.166788, "cps": 26.742188, "mean_p": 0.193502, "boundary_entropy": 0.370145, "bpb": 1.463996}, "hi": {"chunks": 3246.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.192927, "cps": 25.359375, "mean_p": 0.204301, "boundary_entropy": 0.415585, "bpb": 1.19246}, "ta": {"chunks": 1600.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.085584, "cps": 12.5, "mean_p": 0.182408, "boundary_entropy": 0.421223, "bpb": 0.992586}, "th": {"chunks": 3263.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.20286, "cps": 25.492188, "mean_p": 0.206003, "boundary_entropy": 0.425167, "bpb": 1.106602}, "zh": {"chunks": 3660.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.619709, "cps": 28.59375, "mean_p": 0.322644, "boundary_entropy": 0.480944, "bpb": 3.105653}, "my": {"chunks": 1881.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.099272, "cps": 14.695312, "mean_p": 0.172898, "boundary_entropy": 0.407868, "bpb": 1.100703}}, "bpb": {"en": 2.00576, "de": 2.20996, "vi": 1.91426, "tr": 2.17907, "fi": 2.33946, "ru": 1.39168, "el": 1.464, "hi": 1.19246, "ta": 0.99259, "th": 1.1066, "zh": 3.10565, "my": 1.1007}, "gini_cpc": 0.239901, "gini_cps": 0.096291, "bpic": 9.94612, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8454651103762486, "vi": 0.9344158100241328, "tr": 0.9179838716867316, "fi": 0.8758020520083962, "ru": 0.9423521807166596, "el": 0.7980808189731832, "hi": 0.9231542251279875, "ta": 0.40952020344821577, "th": 0.9706816678545352, "zh": 2.9652988004444296, "my": 0.4750138128930667}, "t": 343.8}
{"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.0183, "de": 2.21731, "vi": 1.91294, "tr": 2.18754, "fi": 2.34871, "ru": 1.36915, "el": 1.45771, "hi": 1.1899, "ta": 0.99858, "th": 1.12313, "zh": 3.10239, "my": 1.1091}, "bytes_per_chunk": {"en": 4.838, "de": 5.531, "vi": 6.966, "tr": 5.434, "fi": 5.29, "ru": 8.921, "el": 10.693, "hi": 12.217, "ta": 31.459, "th": 12.857, "zh": 4.651, "my": 26.094}, "bpb_mean": 1.7529, "t": 345.1}
{"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3637.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.208987, "cps": 28.414062, "mean_p": 0.235227, "boundary_entropy": 0.235231, "bpb": 2.005763}, "de": {"chunks": 3628.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.176691, "cps": 28.34375, "mean_p": 0.205738, "boundary_entropy": 0.227873, "bpb": 2.209956}, "vi": {"chunks": 3600.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.195281, "cps": 28.125, "mean_p": 0.231494, "boundary_entropy": 0.346194, "bpb": 1.914264}, "tr": {"chunks": 3473.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.191847, "cps": 27.132812, "mean_p": 0.209226, "boundary_entropy": 0.235501, "bpb": 2.17907}, "fi": {"chunks": 3443.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.183031, "cps": 26.898438, "mean_p": 0.208423, "boundary_entropy": 0.231517, "bpb": 2.339462}, "ru": {"chunks": 3732.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.196939, "cps": 29.15625, "mean_p": 0.177415, "boundary_entropy": 0.300712, "bpb": 1.391678}, "el": {"chunks": 3423.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.166788, "cps": 26.742188, "mean_p": 0.193502, "boundary_entropy": 0.370145, "bpb": 1.463996}, "hi": {"chunks": 3246.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.192927, "cps": 25.359375, "mean_p": 0.204301, "boundary_entropy": 0.415585, "bpb": 1.19246}, "ta": {"chunks": 1600.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.085584, "cps": 12.5, "mean_p": 0.182408, "boundary_entropy": 0.421223, "bpb": 0.992586}, "th": {"chunks": 3263.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.20286, "cps": 25.492188, "mean_p": 0.206003, "boundary_entropy": 0.425167, "bpb": 1.106602}, "zh": {"chunks": 3660.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.619709, "cps": 28.59375, "mean_p": 0.322644, "boundary_entropy": 0.480944, "bpb": 3.105653}, "my": {"chunks": 1881.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.099272, "cps": 14.695312, "mean_p": 0.172898, "boundary_entropy": 0.407868, "bpb": 1.100703}}, "bpb": {"en": 2.00576, "de": 2.20996, "vi": 1.91426, "tr": 2.17907, "fi": 2.33946, "ru": 1.39168, "el": 1.464, "hi": 1.19246, "ta": 0.99259, "th": 1.1066, "zh": 3.10565, "my": 1.1007}, "gini_cpc": 0.239901, "gini_cps": 0.096291, "bpic": 9.94612, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8454651103762486, "vi": 0.9344158100241328, "tr": 0.9179838716867316, "fi": 0.8758020520083962, "ru": 0.9423521807166596, "el": 0.7980808189731832, "hi": 0.9231542251279875, "ta": 0.40952020344821577, "th": 0.9706816678545352, "zh": 2.9652988004444296, "my": 0.4750138128930667}, "t": 346.0}
{"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s0/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s0/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s0/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s0/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s0/th0.70"], "t": 348.0}
{"event": "final", "run": "pilot_E_s0", "condition": "E", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.20898695627190714, "cps": 28.4140625, "bpb": 2.027534847106518, "bytes_per_chunk": 4.792351075401505}, "de": {"cpc": 0.1766911800516242, "cps": 28.34375, "bpb": 2.209753539885548, "bytes_per_chunk": 5.571071842225504}, "vi": {"cpc": 0.19528071602929212, "cps": 28.125, "bpb": 1.9175251448329642, "bytes_per_chunk": 6.9727493981992525}, "tr": {"cpc": 0.19184665525051098, "cps": 27.1328125, "bpb": 2.1790023647295897, "bytes_per_chunk": 5.453437210705333}, "fi": {"cpc": 0.18303120514592525, "cps": 26.8984375, "bpb": 2.3653980538947033, "bytes_per_chunk": 5.287773194420631}, "ru": {"cpc": 0.19693931398416886, "cps": 29.15625, "bpb": 1.3694184250825936, "bytes_per_chunk": 8.874054264484352}, "el": {"cpc": 0.16678848121619647, "cps": 26.7421875, "bpb": 1.4537136629856469, "bytes_per_chunk": 10.662544996034248}, "hi": {"cpc": 0.19292719167904904, "cps": 25.359375, "bpb": 1.1981697263057156, "bytes_per_chunk": 12.130960919965755}, "ta": {"cpc": 0.08558438085049479, "cps": 12.5, "bpb": 0.9989697245544682, "bytes_per_chunk": 32.08224207563334}, "th": {"cpc": 0.20285980727385763, "cps": 25.4921875, "bpb": 1.1227061361153114, "bytes_per_chunk": 13.105889411058895}, "zh": {"cpc": 0.6197087707416187, "cps": 28.59375, "bpb": 3.112132933082202, "bytes_per_chunk": 4.639224153187272}, "my": {"cpc": 0.09927169094363521, "cps": 14.6953125, "bpb": 1.1179256096116879, "bytes_per_chunk": 25.561308566135246}}, "gini_cpc": 0.2399013066286535, "gini_cps": 0.09629140102627898, "bpic": 9.946120354532733, "seed": 0, "mean_boundary_entropy": 0.3112594485282898, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584}], "t": 350.4}