| {"event": "start", "run": "pilot_C_s0", "condition": "C", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "parity_a", "p3_objective": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 1.4} |
| {"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": 1.5} |
| {"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": 3.3} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.63916, "bpb": 8.13559, "chunk_loss": 0.05247, "F": 0.47223, "G": 0.4995, "chunks": 483.56, "bytes_per_chunk": 2.118, "boundary_entropy": 0.69158, "router_noise": 0.0, "grad_norm": 7.806, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.004, "alpha": 0.03, "t": 3.6} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6331.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.363788, "cps": 49.460938, "mean_p": 0.496549, "boundary_entropy": 0.686989, "bpb": 8.137938}, "de": {"chunks": 8595.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.418594, "cps": 67.148438, "mean_p": 0.498781, "boundary_entropy": 0.687819, "bpb": 8.13671}, "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": 8.072259}, "tr": {"chunks": 5244.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.289676, "cps": 40.96875, "mean_p": 0.489423, "boundary_entropy": 0.68728, "bpb": 8.156751}, "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": 8.196431}, "ru": {"chunks": 10505.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.554354, "cps": 82.070312, "mean_p": 0.493306, "boundary_entropy": 0.689882, "bpb": 8.059463}, "el": {"chunks": 15043.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.732983, "cps": 117.523438, "mean_p": 0.497467, "boundary_entropy": 0.690234, "bpb": 8.140753}, "hi": {"chunks": 2502.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.148707, "cps": 19.546875, "mean_p": 0.484437, "boundary_entropy": 0.690242, "bpb": 8.249664}, "ta": {"chunks": 37854.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.024819, "cps": 295.734375, "mean_p": 0.512694, "boundary_entropy": 0.690391, "bpb": 8.028656}, "th": {"chunks": 33861.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.105129, "cps": 264.539062, "mean_p": 0.511769, "boundary_entropy": 0.690558, "bpb": 8.229598}, "zh": {"chunks": 11820.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.001355, "cps": 92.34375, "mean_p": 0.520021, "boundary_entropy": 0.685675, "bpb": 8.120651}, "my": {"chunks": 48157.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.541535, "cps": 376.226562, "mean_p": 0.521509, "boundary_entropy": 0.69021, "bpb": 7.893469}}, "bpb": {"en": 8.13794, "de": 8.13671, "vi": 8.07226, "tr": 8.15675, "fi": 8.19643, "ru": 8.05946, "el": 8.14075, "hi": 8.24966, "ta": 8.02866, "th": 8.2296, "zh": 8.12065, "my": 7.89347}, "gini_cpc": 0.425813, "gini_cps": 0.427257, "bpic": 1.902034, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1506554015782662, "vi": 1.9418680009597988, "tr": 0.7962765722561626, "fi": 1.2916446728974176, "ru": 1.5238374726864377, "el": 2.0148625141077643, "hi": 0.40877472884551425, "ta": 5.565934804009309, "th": 5.786694049102493, "zh": 5.501433156732364, "my": 6.986310037493565}, "t": 4.9} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6247.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.358961, "cps": 48.804688, "mean_p": 0.4962, "boundary_entropy": 0.686976, "bpb": 8.088614}, "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": 8.09271}, "vi": {"chunks": 12841.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.696555, "cps": 100.320312, "mean_p": 0.505084, "boundary_entropy": 0.68839, "bpb": 8.038676}, "tr": {"chunks": 5169.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.285533, "cps": 40.382812, "mean_p": 0.489117, "boundary_entropy": 0.687262, "bpb": 8.122119}, "fi": {"chunks": 8749.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.4651, "cps": 68.351562, "mean_p": 0.50014, "boundary_entropy": 0.687503, "bpb": 8.154274}, "ru": {"chunks": 10058.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.530765, "cps": 78.578125, "mean_p": 0.492759, "boundary_entropy": 0.689864, "bpb": 7.993316}, "el": {"chunks": 15092.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.73537, "cps": 117.90625, "mean_p": 0.497513, "boundary_entropy": 0.690235, "bpb": 8.120841}, "hi": {"chunks": 2408.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.14312, "cps": 18.8125, "mean_p": 0.484161, "boundary_entropy": 0.690223, "bpb": 8.21746}, "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": 7.987951}, "th": {"chunks": 33648.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.091887, "cps": 262.875, "mean_p": 0.511606, "boundary_entropy": 0.690564, "bpb": 8.202592}, "zh": {"chunks": 11809.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.999492, "cps": 92.257812, "mean_p": 0.519978, "boundary_entropy": 0.685678, "bpb": 8.114115}, "my": {"chunks": 48088.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.537893, "cps": 375.6875, "mean_p": 0.521422, "boundary_entropy": 0.690216, "bpb": 7.860046}}, "bpb": {"en": 8.08861, "de": 8.09271, "vi": 8.03868, "tr": 8.12212, "fi": 8.15427, "ru": 7.99332, "el": 8.12084, "hi": 8.21746, "ta": 7.98795, "th": 8.20259, "zh": 8.11412, "my": 7.86005}, "gini_cpc": 0.428207, "gini_cps": 0.429757, "bpic": 1.917917, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1556806527835644, "vi": 1.940476190166072, "tr": 0.7954421403497784, "fi": 1.295684153687481, "ru": 1.478614739824456, "el": 2.048606590920987, "hi": 0.3987071500115003, "ta": 5.584449590630997, "th": 5.827614354050653, "zh": 5.570219306350692, "my": 7.07010645667235}, "t": 6.4} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5922.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.340286, "cps": 46.265625, "mean_p": 0.494976, "boundary_entropy": 0.686928, "bpb": 7.918764}, "de": {"chunks": 8279.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.403205, "cps": 64.679688, "mean_p": 0.497824, "boundary_entropy": 0.687798, "bpb": 7.939157}, "vi": {"chunks": 12271.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.665636, "cps": 95.867188, "mean_p": 0.503598, "boundary_entropy": 0.688396, "bpb": 7.926948}, "tr": {"chunks": 4898.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.270563, "cps": 38.265625, "mean_p": 0.488082, "boundary_entropy": 0.687201, "bpb": 8.003735}, "fi": {"chunks": 8332.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.442932, "cps": 65.09375, "mean_p": 0.498785, "boundary_entropy": 0.687477, "bpb": 8.007467}, "ru": {"chunks": 8688.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.45847, "cps": 67.875, "mean_p": 0.49085, "boundary_entropy": 0.689793, "bpb": 7.770983}, "el": {"chunks": 15251.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.743117, "cps": 119.148438, "mean_p": 0.497602, "boundary_entropy": 0.690236, "bpb": 8.034001}, "hi": {"chunks": 2068.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.122912, "cps": 16.15625, "mean_p": 0.48279, "boundary_entropy": 0.690124, "bpb": 8.096883}, "ta": {"chunks": 35898.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.920193, "cps": 280.453125, "mean_p": 0.510438, "boundary_entropy": 0.690491, "bpb": 7.841648}, "th": {"chunks": 32623.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.028163, "cps": 254.867188, "mean_p": 0.510754, "boundary_entropy": 0.690595, "bpb": 8.088094}, "zh": {"chunks": 11778.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.994243, "cps": 92.015625, "mean_p": 0.51976, "boundary_entropy": 0.68569, "bpb": 8.087311}, "my": {"chunks": 47875.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.526652, "cps": 374.023438, "mean_p": 0.521124, "boundary_entropy": 0.690237, "bpb": 7.783238}}, "bpb": {"en": 7.91876, "de": 7.93916, "vi": 7.92695, "tr": 8.00374, "fi": 8.00747, "ru": 7.77098, "el": 8.034, "hi": 8.09688, "ta": 7.84165, "th": 8.08809, "zh": 8.08731, "my": 7.78324}, "gini_cpc": 0.437283, "gini_cps": 0.438216, "bpic": 1.979446, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1848986170037594, "vi": 1.9561066580490778, "tr": 0.7951040150134927, "fi": 1.3016466200309513, "ru": 1.347306220978258, "el": 2.1838016664625317, "hi": 0.36120285855798484, "ta": 5.642875921344962, "th": 5.960168639311359, "zh": 5.860488586641579, "my": 7.425079843111387}, "t": 8.0} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5085.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.292191, "cps": 39.726562, "mean_p": 0.49168, "boundary_entropy": 0.686769, "bpb": 7.437782}, "de": {"chunks": 7672.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.373642, "cps": 59.9375, "mean_p": 0.495828, "boundary_entropy": 0.687749, "bpb": 7.496108}, "vi": {"chunks": 10396.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.563927, "cps": 81.21875, "mean_p": 0.498905, "boundary_entropy": 0.688358, "bpb": 7.585994}, "tr": {"chunks": 4095.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.226206, "cps": 31.992188, "mean_p": 0.48508, "boundary_entropy": 0.686999, "bpb": 7.643549}, "fi": {"chunks": 7123.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.378661, "cps": 55.648438, "mean_p": 0.49486, "boundary_entropy": 0.687364, "bpb": 7.577092}, "ru": {"chunks": 5205.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.27467, "cps": 40.664062, "mean_p": 0.485319, "boundary_entropy": 0.689504, "bpb": 7.189295}, "el": {"chunks": 14860.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.724066, "cps": 116.09375, "mean_p": 0.497135, "boundary_entropy": 0.69023, "bpb": 7.711179}, "hi": {"chunks": 1135.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.067459, "cps": 8.867188, "mean_p": 0.477095, "boundary_entropy": 0.689634, "bpb": 7.697652}, "ta": {"chunks": 28672.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.533672, "cps": 224.0, "mean_p": 0.503565, "boundary_entropy": 0.690676, "bpb": 7.36612}, "th": {"chunks": 27403.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.703637, "cps": 214.085938, "mean_p": 0.506239, "boundary_entropy": 0.690718, "bpb": 7.54752}, "zh": {"chunks": 11598.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.963766, "cps": 90.609375, "mean_p": 0.518921, "boundary_entropy": 0.685739, "bpb": 7.998394}, "my": {"chunks": 47091.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.485275, "cps": 367.898438, "mean_p": 0.520003, "boundary_entropy": 0.690311, "bpb": 7.472396}}, "bpb": {"en": 7.43778, "de": 7.49611, "vi": 7.58599, "tr": 7.64355, "fi": 7.57709, "ru": 7.1893, "el": 7.71118, "hi": 7.69765, "ta": 7.36612, "th": 7.54752, "zh": 7.99839, "my": 7.4724}, "gini_cpc": 0.469757, "gini_cps": 0.463478, "bpic": 2.253095, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2787609039165242, "vi": 1.929995479613055, "tr": 0.7741703203742204, "fi": 1.295937997741546, "ru": 0.9400364256193522, "el": 2.4780560611398363, "hi": 0.2308734280967972, "ta": 5.248868365889007, "th": 5.8305591882358, "zh": 6.720828675802918, "my": 8.505653759427291}, "t": 9.7} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3938.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.226283, "cps": 30.765625, "mean_p": 0.488092, "boundary_entropy": 0.686662, "bpb": 6.812397}, "de": {"chunks": 6919.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.33697, "cps": 54.054688, "mean_p": 0.493884, "boundary_entropy": 0.687798, "bpb": 6.904788}, "vi": {"chunks": 7199.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.390507, "cps": 56.242188, "mean_p": 0.490791, "boundary_entropy": 0.688163, "bpb": 7.158164}, "tr": {"chunks": 3501.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.193393, "cps": 27.351562, "mean_p": 0.483501, "boundary_entropy": 0.686946, "bpb": 7.13126}, "fi": {"chunks": 5524.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.293658, "cps": 43.15625, "mean_p": 0.490062, "boundary_entropy": 0.687262, "bpb": 7.002106}, "ru": {"chunks": 2629.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.138734, "cps": 20.539062, "mean_p": 0.478877, "boundary_entropy": 0.689014, "bpb": 6.360144}, "el": {"chunks": 5620.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.273839, "cps": 43.90625, "mean_p": 0.487025, "boundary_entropy": 0.689871, "bpb": 7.04131}, "hi": {"chunks": 404.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.024012, "cps": 3.15625, "mean_p": 0.46693, "boundary_entropy": 0.688438, "bpb": 7.052669}, "ta": {"chunks": 14593.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.780583, "cps": 114.007812, "mean_p": 0.490155, "boundary_entropy": 0.690531, "bpb": 6.62259}, "th": {"chunks": 17460.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.085483, "cps": 136.40625, "mean_p": 0.496574, "boundary_entropy": 0.690724, "bpb": 6.369857}, "zh": {"chunks": 10873.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.841009, "cps": 84.945312, "mean_p": 0.515368, "boundary_entropy": 0.685938, "bpb": 7.770677}, "my": {"chunks": 42537.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.244934, "cps": 332.320312, "mean_p": 0.515717, "boundary_entropy": 0.690551, "bpb": 6.67284}}, "bpb": {"en": 6.8124, "de": 6.90479, "vi": 7.15816, "tr": 7.13126, "fi": 7.00211, "ru": 6.36014, "el": 7.04131, "hi": 7.05267, "ta": 6.62259, "th": 6.36986, "zh": 7.77068, "my": 6.67284}, "gini_cpc": 0.523242, "gini_cps": 0.492026, "bpic": 3.166588, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.4891530380172862, "vi": 1.7257482434198756, "tr": 0.8546532879249423, "fi": 1.2977474819081225, "ru": 0.6130978316946979, "el": 1.2101630231310054, "hi": 0.10611449231287029, "ta": 3.449590326987265, "th": 4.7970205894188656, "zh": 8.135876617541415, "my": 9.920918674219239}, "t": 11.5} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4849.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.27863, "cps": 37.882812, "mean_p": 0.4952, "boundary_entropy": 0.687433, "bpb": 6.360866}, "de": {"chunks": 8506.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.41426, "cps": 66.453125, "mean_p": 0.498941, "boundary_entropy": 0.68833, "bpb": 6.444291}, "vi": {"chunks": 5639.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.305886, "cps": 44.054688, "mean_p": 0.488977, "boundary_entropy": 0.688357, "bpb": 6.904337}, "tr": {"chunks": 6795.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.375352, "cps": 53.085938, "mean_p": 0.496457, "boundary_entropy": 0.687979, "bpb": 6.718509}, "fi": {"chunks": 7139.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.379512, "cps": 55.773438, "mean_p": 0.497677, "boundary_entropy": 0.688011, "bpb": 6.521004}, "ru": {"chunks": 5373.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.283536, "cps": 41.976562, "mean_p": 0.488009, "boundary_entropy": 0.68983, "bpb": 5.791019}, "el": {"chunks": 475.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.023145, "cps": 3.710938, "mean_p": 0.469267, "boundary_entropy": 0.688348, "bpb": 6.571728}, "hi": {"chunks": 2720.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.161664, "cps": 21.25, "mean_p": 0.485759, "boundary_entropy": 0.69039, "bpb": 5.945729}, "ta": {"chunks": 5478.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.29302, "cps": 42.796875, "mean_p": 0.482371, "boundary_entropy": 0.690233, "bpb": 5.80941}, "th": {"chunks": 3752.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.233261, "cps": 29.3125, "mean_p": 0.48793, "boundary_entropy": 0.690629, "bpb": 5.623819}, "zh": {"chunks": 7406.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.253979, "cps": 57.859375, "mean_p": 0.501021, "boundary_entropy": 0.686375, "bpb": 7.331352}, "my": {"chunks": 25651.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.353758, "cps": 200.398438, "mean_p": 0.496587, "boundary_entropy": 0.690917, "bpb": 5.629907}}, "bpb": {"en": 6.36087, "de": 6.44429, "vi": 6.90434, "tr": 6.71851, "fi": 6.521, "ru": 5.79102, "el": 6.57173, "hi": 5.94573, "ta": 5.80941, "th": 5.62382, "zh": 7.33135, "my": 5.62991}, "gini_cpc": 0.412735, "gini_cps": 0.368477, "bpic": 4.580655, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.4867738272661937, "vi": 1.0978193686937037, "tr": 1.3471341501126088, "fi": 1.3620637495081789, "ru": 1.0176057735158515, "el": 0.0830662683107315, "hi": 0.580210744881759, "ta": 1.0516433852492368, "th": 0.8371700839393834, "zh": 4.500514871852603, "my": 4.858619184749936}, "t": 13.9} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.22418, "bpb": 6.0942, "chunk_loss": 0.030942, "F": 0.17761, "G": 0.48749, "chunks": 181.88, "bytes_per_chunk": 5.63, "boundary_entropy": 0.69167, "router_noise": 0.0, "grad_norm": 1.5904, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.23, "alpha": 0.03, "t": 15.0} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4744.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.272597, "cps": 37.0625, "mean_p": 0.496013, "boundary_entropy": 0.68759, "bpb": 5.81998}, "de": {"chunks": 6154.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.299713, "cps": 48.078125, "mean_p": 0.495808, "boundary_entropy": 0.688414, "bpb": 5.934664}, "vi": {"chunks": 2888.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.156659, "cps": 22.5625, "mean_p": 0.480961, "boundary_entropy": 0.68791, "bpb": 6.312616}, "tr": {"chunks": 5057.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.279346, "cps": 39.507812, "mean_p": 0.49393, "boundary_entropy": 0.688059, "bpb": 6.227124}, "fi": {"chunks": 5436.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.28898, "cps": 42.46875, "mean_p": 0.495109, "boundary_entropy": 0.688103, "bpb": 5.973507}, "ru": {"chunks": 7681.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.40533, "cps": 60.007812, "mean_p": 0.493634, "boundary_entropy": 0.690204, "bpb": 5.414592}, "el": {"chunks": 4099.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.199727, "cps": 32.023438, "mean_p": 0.490193, "boundary_entropy": 0.690329, "bpb": 5.662245}, "hi": {"chunks": 417.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.024785, "cps": 3.257812, "mean_p": 0.47361, "boundary_entropy": 0.689283, "bpb": 5.447409}, "ta": {"chunks": 2724.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.145707, "cps": 21.28125, "mean_p": 0.483387, "boundary_entropy": 0.690372, "bpb": 5.207946}, "th": {"chunks": 1476.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.091763, "cps": 11.53125, "mean_p": 0.484083, "boundary_entropy": 0.690429, "bpb": 5.096268}, "zh": {"chunks": 4618.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.781917, "cps": 36.078125, "mean_p": 0.492891, "boundary_entropy": 0.686638, "bpb": 6.993826}, "my": {"chunks": 22156.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.169305, "cps": 173.09375, "mean_p": 0.494184, "boundary_entropy": 0.691001, "bpb": 4.996376}}, "bpb": {"en": 5.81998, "de": 5.93466, "vi": 6.31262, "tr": 6.22712, "fi": 5.97351, "ru": 5.41459, "el": 5.66225, "hi": 5.44741, "ta": 5.20795, "th": 5.09627, "zh": 6.99383, "my": 4.99638}, "gini_cpc": 0.437503, "gini_cps": 0.413546, "bpic": 5.689859, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.099472888171051, "vi": 0.5746897970216671, "tr": 1.0247592379152495, "fi": 1.060100414857189, "ru": 1.4869213270922415, "el": 0.732683671484771, "hi": 0.09092020827434483, "ta": 0.5345164477971809, "th": 0.33662373321325734, "zh": 2.8684013998968663, "my": 4.2895073888205975}, "t": 16.8} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.39089, "bpb": 4.89202, "chunk_loss": 0.029001, "F": 0.15131, "G": 0.47798, "chunks": 154.94, "bytes_per_chunk": 6.609, "boundary_entropy": 0.69011, "router_noise": 0.0, "grad_norm": 2.3766, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.077, "alpha": 0.03, "t": 18.9} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3244.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.186405, "cps": 25.34375, "mean_p": 0.45614, "boundary_entropy": 0.678127, "bpb": 4.497678}, "de": {"chunks": 3753.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182779, "cps": 29.320312, "mean_p": 0.45721, "boundary_entropy": 0.679941, "bpb": 4.728091}, "vi": {"chunks": 6693.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.363059, "cps": 52.289062, "mean_p": 0.472052, "boundary_entropy": 0.68416, "bpb": 5.080639}, "tr": {"chunks": 3775.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.208529, "cps": 29.492188, "mean_p": 0.461181, "boundary_entropy": 0.681161, "bpb": 5.039903}, "fi": {"chunks": 4537.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.241189, "cps": 35.445312, "mean_p": 0.460832, "boundary_entropy": 0.680009, "bpb": 4.759946}, "ru": {"chunks": 5183.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.273509, "cps": 40.492188, "mean_p": 0.462435, "boundary_entropy": 0.684448, "bpb": 3.672107}, "el": {"chunks": 7585.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.369585, "cps": 59.257812, "mean_p": 0.479789, "boundary_entropy": 0.688285, "bpb": 4.612136}, "hi": {"chunks": 10444.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.620743, "cps": 81.59375, "mean_p": 0.491211, "boundary_entropy": 0.689756, "bpb": 3.634012}, "ta": {"chunks": 6476.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.346403, "cps": 50.59375, "mean_p": 0.484463, "boundary_entropy": 0.689674, "bpb": 3.418924}, "th": {"chunks": 4514.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.280634, "cps": 35.265625, "mean_p": 0.462536, "boundary_entropy": 0.685972, "bpb": 3.173085}, "zh": {"chunks": 3898.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.660007, "cps": 30.453125, "mean_p": 0.486563, "boundary_entropy": 0.685754, "bpb": 6.326721}, "my": {"chunks": 16722.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.882521, "cps": 130.640625, "mean_p": 0.479138, "boundary_entropy": 0.68943, "bpb": 3.990218}}, "bpb": {"en": 4.49768, "de": 4.72809, "vi": 5.08064, "tr": 5.0399, "fi": 4.75995, "ru": 3.67211, "el": 4.61214, "hi": 3.63401, "ta": 3.41892, "th": 3.17309, "zh": 6.32672, "my": 3.99022}, "gini_cpc": 0.285667, "gini_cps": 0.277316, "bpic": 4.995587, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9805492953119945, "vi": 1.9476950374177675, "tr": 1.1186898023523837, "fi": 1.2938983840881726, "ru": 1.467287673773217, "el": 1.9827046020305148, "hi": 3.3300830519436495, "ta": 1.8583377331417532, "th": 1.5055104145785316, "zh": 3.5407206740371033, "my": 4.734434556045008}, "t": 21.1} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.89416, "bpb": 4.17538, "chunk_loss": 0.030118, "F": 0.16861, "G": 0.43307, "chunks": 172.66, "bytes_per_chunk": 5.931, "boundary_entropy": 0.67163, "router_noise": 0.0, "grad_norm": 1.4288, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.068, "alpha": 0.03, "t": 22.3} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 1899.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.109119, "cps": 14.835938, "mean_p": 0.35173, "boundary_entropy": 0.609218, "bpb": 3.889848}, "de": {"chunks": 2340.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.113963, "cps": 18.28125, "mean_p": 0.35968, "boundary_entropy": 0.619984, "bpb": 4.121612}, "vi": {"chunks": 4208.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228261, "cps": 32.875, "mean_p": 0.398854, "boundary_entropy": 0.618797, "bpb": 3.670366}, "tr": {"chunks": 2458.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.135779, "cps": 19.203125, "mean_p": 0.379827, "boundary_entropy": 0.632415, "bpb": 4.190826}, "fi": {"chunks": 2357.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.125299, "cps": 18.414062, "mean_p": 0.353669, "boundary_entropy": 0.612639, "bpb": 4.138419}, "ru": {"chunks": 4537.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.23942, "cps": 35.445312, "mean_p": 0.455244, "boundary_entropy": 0.677545, "bpb": 2.834601}, "el": {"chunks": 4848.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.236223, "cps": 37.875, "mean_p": 0.401451, "boundary_entropy": 0.657179, "bpb": 3.082594}, "hi": {"chunks": 9613.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.571352, "cps": 75.101562, "mean_p": 0.404878, "boundary_entropy": 0.649302, "bpb": 2.403709}, "ta": {"chunks": 8224.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.439904, "cps": 64.25, "mean_p": 0.452668, "boundary_entropy": 0.679128, "bpb": 2.283818}, "th": {"chunks": 4691.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.291638, "cps": 36.648438, "mean_p": 0.454202, "boundary_entropy": 0.681554, "bpb": 2.326199}, "zh": {"chunks": 4075.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.689976, "cps": 31.835938, "mean_p": 0.474103, "boundary_entropy": 0.681901, "bpb": 5.299672}, "my": {"chunks": 18882.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.996517, "cps": 147.515625, "mean_p": 0.484945, "boundary_entropy": 0.688059, "bpb": 2.429866}}, "bpb": {"en": 3.88985, "de": 4.12161, "vi": 3.67037, "tr": 4.19083, "fi": 4.13842, "ru": 2.8346, "el": 3.08259, "hi": 2.40371, "ta": 2.28382, "th": 2.3262, "zh": 5.29967, "my": 2.42987}, "gini_cpc": 0.394615, "gini_cps": 0.376067, "bpic": 5.632904, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.044389761666747, "vi": 2.0918558052265954, "tr": 1.2443154185558696, "fi": 1.1482774985527862, "ru": 2.1941116349252003, "el": 2.164815662833954, "hi": 5.236040834818044, "ta": 4.03140831853479, "th": 2.6726588262100877, "zh": 6.323147691934034, "my": 9.132375761250296}, "t": 25.4} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.89326, "de": 4.12183, "vi": 3.67599, "tr": 4.1999, "fi": 4.1422, "ru": 2.81918, "el": 3.07522, "hi": 2.39961, "ta": 2.3017, "th": 2.35873, "zh": 5.30907, "my": 2.44781}, "bytes_per_chunk": {"en": 7.736, "de": 8.496, "vi": 5.773, "tr": 8.577, "fi": 8.786, "ru": 7.311, "el": 7.205, "hi": 4.485, "ta": 6.045, "th": 8.833, "zh": 4.708, "my": 3.03}, "bpb_mean": 3.39538, "t": 26.7} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.49724, "bpb": 3.60276, "chunk_loss": 0.028886, "F": 0.14017, "G": 0.39401, "chunks": 143.53, "bytes_per_chunk": 7.134, "boundary_entropy": 0.64195, "router_noise": 0.0, "grad_norm": 1.3854, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 26.8} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.56799, "bpb": 3.70483, "chunk_loss": 0.032406, "F": 0.21252, "G": 0.40492, "chunks": 217.62, "bytes_per_chunk": 4.705, "boundary_entropy": 0.63783, "router_noise": 0.0, "grad_norm": 1.8365, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 28.9} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2882.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165604, "cps": 22.515625, "mean_p": 0.360091, "boundary_entropy": 0.603091, "bpb": 3.616343}, "de": {"chunks": 3472.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.169094, "cps": 27.125, "mean_p": 0.370083, "boundary_entropy": 0.616464, "bpb": 3.899177}, "vi": {"chunks": 4308.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.233686, "cps": 33.65625, "mean_p": 0.42073, "boundary_entropy": 0.632904, "bpb": 3.225343}, "tr": {"chunks": 3538.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195437, "cps": 27.640625, "mean_p": 0.377827, "boundary_entropy": 0.616266, "bpb": 3.840126}, "fi": {"chunks": 2309.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.122747, "cps": 18.039062, "mean_p": 0.342508, "boundary_entropy": 0.595168, "bpb": 3.879647}, "ru": {"chunks": 4586.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.242005, "cps": 35.828125, "mean_p": 0.453236, "boundary_entropy": 0.673995, "bpb": 2.659949}, "el": {"chunks": 5150.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.250938, "cps": 40.234375, "mean_p": 0.376327, "boundary_entropy": 0.637795, "bpb": 2.779199}, "hi": {"chunks": 10216.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.607192, "cps": 79.8125, "mean_p": 0.365557, "boundary_entropy": 0.602136, "bpb": 2.107525}, "ta": {"chunks": 9031.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.48307, "cps": 70.554688, "mean_p": 0.40635, "boundary_entropy": 0.644226, "bpb": 1.915775}, "th": {"chunks": 11516.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.715947, "cps": 89.96875, "mean_p": 0.459994, "boundary_entropy": 0.679841, "bpb": 2.083082}, "zh": {"chunks": 3638.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.615984, "cps": 28.421875, "mean_p": 0.45659, "boundary_entropy": 0.672616, "bpb": 4.974531}, "my": {"chunks": 11920.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.62909, "cps": 93.125, "mean_p": 0.427773, "boundary_entropy": 0.668066, "bpb": 2.148286}}, "bpb": {"en": 3.61634, "de": 3.89918, "vi": 3.22534, "tr": 3.84013, "fi": 3.87965, "ru": 2.65995, "el": 2.7792, "hi": 2.10753, "ta": 1.91578, "th": 2.08308, "zh": 4.97453, "my": 2.14829}, "gini_cpc": 0.312974, "gini_cps": 0.301925, "bpic": 5.288716, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0210745533002796, "vi": 1.4111159357850331, "tr": 1.1801505710460956, "fi": 0.7412115757125438, "ru": 1.4613524762011139, "el": 1.515292688154787, "hi": 3.6665360133554277, "ta": 2.917027453558428, "th": 4.323253827849404, "zh": 3.719627036855215, "my": 3.798770205230957}, "t": 30.1} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.34905, "bpb": 3.38896, "chunk_loss": 0.031726, "F": 0.203, "G": 0.3798, "chunks": 207.88, "bytes_per_chunk": 4.926, "boundary_entropy": 0.61187, "router_noise": 0.0, "grad_norm": 1.2127, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.064, "alpha": 0.03, "t": 32.1} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.05228, "bpb": 2.96081, "chunk_loss": 0.032459, "F": 0.21658, "G": 0.36098, "chunks": 221.78, "bytes_per_chunk": 4.617, "boundary_entropy": 0.58745, "router_noise": 0.0, "grad_norm": 1.1751, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 34.3} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3139.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.180371, "cps": 24.523438, "mean_p": 0.304439, "boundary_entropy": 0.522379, "bpb": 3.253446}, "de": {"chunks": 4576.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.222861, "cps": 35.75, "mean_p": 0.327434, "boundary_entropy": 0.532709, "bpb": 3.5421}, "vi": {"chunks": 5438.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.294982, "cps": 42.484375, "mean_p": 0.334379, "boundary_entropy": 0.55764, "bpb": 2.807061}, "tr": {"chunks": 4081.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.225432, "cps": 31.882812, "mean_p": 0.335131, "boundary_entropy": 0.546544, "bpb": 3.448702}, "fi": {"chunks": 3295.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175163, "cps": 25.742188, "mean_p": 0.315955, "boundary_entropy": 0.54686, "bpb": 3.56214}, "ru": {"chunks": 6194.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.32686, "cps": 48.390625, "mean_p": 0.305975, "boundary_entropy": 0.488904, "bpb": 2.496125}, "el": {"chunks": 4846.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.236125, "cps": 37.859375, "mean_p": 0.312454, "boundary_entropy": 0.555602, "bpb": 2.562713}, "hi": {"chunks": 5575.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.331352, "cps": 43.554688, "mean_p": 0.329357, "boundary_entropy": 0.57868, "bpb": 1.82987}, "ta": {"chunks": 4312.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.23065, "cps": 33.6875, "mean_p": 0.353751, "boundary_entropy": 0.615966, "bpb": 1.601053}, "th": {"chunks": 4656.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.289462, "cps": 36.375, "mean_p": 0.291607, "boundary_entropy": 0.543026, "bpb": 1.899435}, "zh": {"chunks": 2824.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.478158, "cps": 22.0625, "mean_p": 0.372133, "boundary_entropy": 0.610145, "bpb": 4.15557}, "my": {"chunks": 5513.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.290954, "cps": 43.070312, "mean_p": 0.323987, "boundary_entropy": 0.578244, "bpb": 1.927992}}, "bpb": {"en": 3.25345, "de": 3.5421, "vi": 2.80706, "tr": 3.4487, "fi": 3.56214, "ru": 2.49612, "el": 2.56271, "hi": 1.82987, "ta": 1.60105, "th": 1.89944, "zh": 4.15557, "my": 1.92799}, "gini_cpc": 0.152374, "gini_cps": 0.12701, "bpic": 7.048449, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2355673204605389, "vi": 1.635418347771066, "tr": 1.249823964906346, "fi": 0.9711276956985084, "ru": 1.8121527031195754, "el": 1.3091076753224926, "hi": 1.8370568796812565, "ta": 1.2787512968911843, "th": 1.6048140240952735, "zh": 2.650965368343185, "my": 1.6130856246599148}, "t": 37.4} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.26099, "de": 3.53913, "vi": 2.81321, "tr": 3.45441, "fi": 3.55638, "ru": 2.47939, "el": 2.55678, "hi": 1.82653, "ta": 1.61711, "th": 1.92776, "zh": 4.16417, "my": 1.94375}, "bytes_per_chunk": {"en": 5.487, "de": 4.611, "vi": 4.464, "tr": 4.794, "fi": 5.911, "ru": 5.515, "el": 7.544, "hi": 7.643, "ta": 8.637, "th": 8.011, "zh": 4.625, "my": 8.439}, "bpb_mean": 2.76163, "t": 38.7} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.05346, "bpb": 2.96251, "chunk_loss": 0.030449, "F": 0.17734, "G": 0.31493, "chunks": 181.59, "bytes_per_chunk": 5.639, "boundary_entropy": 0.53513, "router_noise": 0.0, "grad_norm": 0.9106, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 38.7} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.11998, "bpb": 3.05849, "chunk_loss": 0.030087, "F": 0.16727, "G": 0.31494, "chunks": 171.28, "bytes_per_chunk": 5.978, "boundary_entropy": 0.5425, "router_noise": 0.0, "grad_norm": 0.7228, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 40.8} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.95233, "bpb": 2.81661, "chunk_loss": 0.031069, "F": 0.1944, "G": 0.31914, "chunks": 199.06, "bytes_per_chunk": 5.144, "boundary_entropy": 0.53485, "router_noise": 0.0, "grad_norm": 0.8196, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 42.9} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.91689, "bpb": 2.76549, "chunk_loss": 0.030549, "F": 0.18362, "G": 0.31002, "chunks": 188.03, "bytes_per_chunk": 5.446, "boundary_entropy": 0.5197, "router_noise": 0.0, "grad_norm": 1.0087, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 45.0} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3115.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.178992, "cps": 24.335938, "mean_p": 0.278372, "boundary_entropy": 0.466855, "bpb": 2.948744}, "de": {"chunks": 3760.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18312, "cps": 29.375, "mean_p": 0.280357, "boundary_entropy": 0.459289, "bpb": 3.269493}, "vi": {"chunks": 5929.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.321616, "cps": 46.320312, "mean_p": 0.308915, "boundary_entropy": 0.495007, "bpb": 2.583216}, "tr": {"chunks": 4256.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.235099, "cps": 33.25, "mean_p": 0.312161, "boundary_entropy": 0.489039, "bpb": 3.171986}, "fi": {"chunks": 3622.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.192547, "cps": 28.296875, "mean_p": 0.287677, "boundary_entropy": 0.485738, "bpb": 3.377927}, "ru": {"chunks": 6912.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.364749, "cps": 54.0, "mean_p": 0.362613, "boundary_entropy": 0.572177, "bpb": 2.227839}, "el": {"chunks": 6629.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.323003, "cps": 51.789062, "mean_p": 0.353838, "boundary_entropy": 0.594714, "bpb": 2.258323}, "hi": {"chunks": 8534.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.507221, "cps": 66.671875, "mean_p": 0.333718, "boundary_entropy": 0.548389, "bpb": 1.623492}, "ta": {"chunks": 7341.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.392672, "cps": 57.351562, "mean_p": 0.275824, "boundary_entropy": 0.497651, "bpb": 1.395907}, "th": {"chunks": 7813.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.485732, "cps": 61.039062, "mean_p": 0.292555, "boundary_entropy": 0.532208, "bpb": 1.732641}, "zh": {"chunks": 3280.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.555367, "cps": 25.625, "mean_p": 0.357764, "boundary_entropy": 0.590158, "bpb": 3.867807}, "my": {"chunks": 8594.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.453557, "cps": 67.140625, "mean_p": 0.373566, "boundary_entropy": 0.616042, "bpb": 1.664325}}, "bpb": {"en": 2.94874, "de": 3.26949, "vi": 2.58322, "tr": 3.17199, "fi": 3.37793, "ru": 2.22784, "el": 2.25832, "hi": 1.62349, "ta": 1.39591, "th": 1.73264, "zh": 3.86781, "my": 1.66432}, "gini_cpc": 0.207784, "gini_cps": 0.195201, "bpic": 5.499477, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0230609474205834, "vi": 1.7968191916329161, "tr": 1.3134608640817496, "fi": 1.075728393268752, "ru": 2.037795431925698, "el": 1.8045679634851992, "hi": 2.8337637134223272, "ta": 2.1937938960168664, "th": 2.7137062086397954, "zh": 3.10274775658674, "my": 2.5339500079799375}, "t": 48.0} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.95685, "de": 3.26714, "vi": 2.59044, "tr": 3.18215, "fi": 3.36815, "ru": 2.21186, "el": 2.2506, "hi": 1.61844, "ta": 1.40827, "th": 1.7636, "zh": 3.87967, "my": 1.67545}, "bytes_per_chunk": {"en": 5.57, "de": 5.428, "vi": 4.521, "tr": 4.354, "fi": 5.462, "ru": 5.102, "el": 5.483, "hi": 4.566, "ta": 5.504, "th": 4.87, "zh": 4.995, "my": 6.71}, "bpb_mean": 2.51439, "t": 49.3} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.94358, "bpb": 2.80399, "chunk_loss": 0.030841, "F": 0.19211, "G": 0.31443, "chunks": 196.72, "bytes_per_chunk": 5.205, "boundary_entropy": 0.51597, "router_noise": 0.0, "grad_norm": 0.8283, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 49.3} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.73664, "bpb": 2.50544, "chunk_loss": 0.030417, "F": 0.17664, "G": 0.31324, "chunks": 180.88, "bytes_per_chunk": 5.661, "boundary_entropy": 0.51528, "router_noise": 0.0, "grad_norm": 0.839, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 51.4} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.75805, "bpb": 2.53633, "chunk_loss": 0.030387, "F": 0.17831, "G": 0.3014, "chunks": 182.59, "bytes_per_chunk": 5.608, "boundary_entropy": 0.503, "router_noise": 0.0, "grad_norm": 0.7098, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 53.4} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.69437, "bpb": 2.44445, "chunk_loss": 0.030251, "F": 0.17615, "G": 0.31131, "chunks": 180.38, "bytes_per_chunk": 5.677, "boundary_entropy": 0.51019, "router_noise": 0.0, "grad_norm": 0.6367, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 55.4} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3335.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.191634, "cps": 26.054688, "mean_p": 0.270477, "boundary_entropy": 0.42035, "bpb": 2.737276}, "de": {"chunks": 3648.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.177665, "cps": 28.5, "mean_p": 0.2654, "boundary_entropy": 0.420277, "bpb": 3.078056}, "vi": {"chunks": 4531.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.245782, "cps": 35.398438, "mean_p": 0.284426, "boundary_entropy": 0.464367, "bpb": 2.43295}, "tr": {"chunks": 3601.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.198917, "cps": 28.132812, "mean_p": 0.280918, "boundary_entropy": 0.44057, "bpb": 3.015493}, "fi": {"chunks": 3607.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19175, "cps": 28.179688, "mean_p": 0.277304, "boundary_entropy": 0.440861, "bpb": 3.255749}, "ru": {"chunks": 5725.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.302111, "cps": 44.726562, "mean_p": 0.314608, "boundary_entropy": 0.540025, "bpb": 2.059868}, "el": {"chunks": 6769.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.329825, "cps": 52.882812, "mean_p": 0.308599, "boundary_entropy": 0.538331, "bpb": 2.137206}, "hi": {"chunks": 7693.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.457236, "cps": 60.101562, "mean_p": 0.315557, "boundary_entropy": 0.525014, "bpb": 1.540068}, "ta": {"chunks": 8877.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.474833, "cps": 69.351562, "mean_p": 0.34886, "boundary_entropy": 0.592112, "bpb": 1.266607}, "th": {"chunks": 6399.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.397824, "cps": 49.992188, "mean_p": 0.334689, "boundary_entropy": 0.58168, "bpb": 1.606606}, "zh": {"chunks": 2921.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.494582, "cps": 22.820312, "mean_p": 0.354187, "boundary_entropy": 0.585025, "bpb": 3.746326}, "my": {"chunks": 11170.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.589508, "cps": 87.265625, "mean_p": 0.405498, "boundary_entropy": 0.637271, "bpb": 1.462571}}, "bpb": {"en": 2.73728, "de": 3.07806, "vi": 2.43295, "tr": 3.01549, "fi": 3.25575, "ru": 2.05987, "el": 2.13721, "hi": 1.54007, "ta": 1.26661, "th": 1.60661, "zh": 3.74633, "my": 1.46257}, "gini_cpc": 0.226529, "gini_cps": 0.237416, "bpic": 5.621023, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9271088022929557, "vi": 1.2825644622765915, "tr": 1.0380083614492814, "fi": 1.000604705355578, "ru": 1.5765021183339727, "el": 1.7211231688643325, "hi": 2.3859917708307807, "ta": 2.4778158821097604, "th": 2.075961652372446, "zh": 2.580871585879935, "my": 3.076224870465337}, "t": 58.3} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.74759, "de": 3.07982, "vi": 2.43912, "tr": 3.0272, "fi": 3.25389, "ru": 2.04462, "el": 2.1256, "hi": 1.53455, "ta": 1.27725, "th": 1.63416, "zh": 3.75939, "my": 1.47329}, "bytes_per_chunk": {"en": 5.309, "de": 5.517, "vi": 5.3, "tr": 5.243, "fi": 5.186, "ru": 6.135, "el": 5.224, "hi": 5.16, "ta": 5.156, "th": 6.913, "zh": 5.757, "my": 5.243}, "bpb_mean": 2.36637, "t": 59.6} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.81773, "bpb": 2.62244, "chunk_loss": 0.03032, "F": 0.17874, "G": 0.29839, "chunks": 183.03, "bytes_per_chunk": 5.595, "boundary_entropy": 0.48668, "router_noise": 0.0, "grad_norm": 0.8063, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 59.6} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.86077, "bpb": 2.68452, "chunk_loss": 0.030515, "F": 0.18344, "G": 0.29943, "chunks": 187.84, "bytes_per_chunk": 5.451, "boundary_entropy": 0.48021, "router_noise": 0.0, "grad_norm": 0.7945, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 61.7} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.67461, "bpb": 2.41596, "chunk_loss": 0.030098, "F": 0.17651, "G": 0.29806, "chunks": 180.75, "bytes_per_chunk": 5.665, "boundary_entropy": 0.47169, "router_noise": 0.0, "grad_norm": 0.9031, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 63.7} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.80709, "bpb": 2.60708, "chunk_loss": 0.030766, "F": 0.18854, "G": 0.30554, "chunks": 193.06, "bytes_per_chunk": 5.304, "boundary_entropy": 0.47128, "router_noise": 0.0, "grad_norm": 0.797, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 65.7} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3506.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20146, "cps": 27.390625, "mean_p": 0.273279, "boundary_entropy": 0.392264, "bpb": 2.59213}, "de": {"chunks": 3964.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.193055, "cps": 30.96875, "mean_p": 0.27319, "boundary_entropy": 0.411085, "bpb": 2.948932}, "vi": {"chunks": 4752.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.257771, "cps": 37.125, "mean_p": 0.281608, "boundary_entropy": 0.45352, "bpb": 2.324703}, "tr": {"chunks": 3739.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.20654, "cps": 29.210938, "mean_p": 0.273867, "boundary_entropy": 0.412136, "bpb": 2.91601}, "fi": {"chunks": 3845.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.204402, "cps": 30.039062, "mean_p": 0.281095, "boundary_entropy": 0.424805, "bpb": 3.141738}, "ru": {"chunks": 5755.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.303694, "cps": 44.960938, "mean_p": 0.338189, "boundary_entropy": 0.552635, "bpb": 1.953689}, "el": {"chunks": 4850.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.23632, "cps": 37.890625, "mean_p": 0.283806, "boundary_entropy": 0.513932, "bpb": 2.061328}, "hi": {"chunks": 5517.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.327905, "cps": 43.101562, "mean_p": 0.394597, "boundary_entropy": 0.640668, "bpb": 1.516594}, "ta": {"chunks": 6229.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.333191, "cps": 48.664062, "mean_p": 0.421605, "boundary_entropy": 0.653118, "bpb": 1.247919}, "th": {"chunks": 8307.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.516444, "cps": 64.898438, "mean_p": 0.413265, "boundary_entropy": 0.652075, "bpb": 1.529424}, "zh": {"chunks": 2692.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.455808, "cps": 21.03125, "mean_p": 0.354727, "boundary_entropy": 0.582939, "bpb": 3.627292}, "my": {"chunks": 9605.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.506914, "cps": 75.039062, "mean_p": 0.406826, "boundary_entropy": 0.650371, "bpb": 1.396015}}, "bpb": {"en": 2.59213, "de": 2.94893, "vi": 2.3247, "tr": 2.91601, "fi": 3.14174, "ru": 1.95369, "el": 2.06133, "hi": 1.51659, "ta": 1.24792, "th": 1.52942, "zh": 3.62729, "my": 1.39602}, "gini_cpc": 0.200795, "gini_cps": 0.200481, "bpic": 6.11496, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9582822570287353, "vi": 1.2795153443799936, "tr": 1.0252201235882938, "fi": 1.0146042312451875, "ru": 1.5074687644400027, "el": 1.1730407571608819, "hi": 1.6276466155662677, "ta": 1.653884091566866, "th": 2.5635119935610184, "zh": 2.262527264469849, "my": 2.516206046126662}, "t": 68.6} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.60106, "de": 2.95156, "vi": 2.32812, "tr": 2.92436, "fi": 3.14505, "ru": 1.93797, "el": 2.04885, "hi": 1.51022, "ta": 1.25882, "th": 1.55327, "zh": 3.63771, "my": 1.40714}, "bytes_per_chunk": {"en": 5.099, "de": 5.127, "vi": 5.149, "tr": 5.151, "fi": 4.885, "ru": 6.01, "el": 7.891, "hi": 7.361, "ta": 7.497, "th": 5.566, "zh": 6.044, "my": 6.096}, "bpb_mean": 2.27534, "t": 69.9} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.74368, "bpb": 2.51559, "chunk_loss": 0.030183, "F": 0.17633, "G": 0.31816, "chunks": 180.56, "bytes_per_chunk": 5.671, "boundary_entropy": 0.49973, "router_noise": 0.0, "grad_norm": 0.8114, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 69.9} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.60318, "bpb": 2.31289, "chunk_loss": 0.030646, "F": 0.186, "G": 0.32634, "chunks": 190.47, "bytes_per_chunk": 5.376, "boundary_entropy": 0.50021, "router_noise": 0.0, "grad_norm": 0.7505, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 71.9} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.80482, "bpb": 2.6038, "chunk_loss": 0.030583, "F": 0.1875, "G": 0.31768, "chunks": 192.0, "bytes_per_chunk": 5.333, "boundary_entropy": 0.48874, "router_noise": 0.0, "grad_norm": 0.8374, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 73.9} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.6566, "bpb": 2.38997, "chunk_loss": 0.030406, "F": 0.1817, "G": 0.30737, "chunks": 186.06, "bytes_per_chunk": 5.504, "boundary_entropy": 0.46649, "router_noise": 0.0, "grad_norm": 0.8231, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 75.9} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3230.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.1856, "cps": 25.234375, "mean_p": 0.250158, "boundary_entropy": 0.356651, "bpb": 2.488525}, "de": {"chunks": 3552.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.17299, "cps": 27.75, "mean_p": 0.249083, "boundary_entropy": 0.375425, "bpb": 2.81655}, "vi": {"chunks": 4445.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.241117, "cps": 34.726562, "mean_p": 0.272747, "boundary_entropy": 0.445853, "bpb": 2.280258}, "tr": {"chunks": 3588.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.198199, "cps": 28.03125, "mean_p": 0.256899, "boundary_entropy": 0.387152, "bpb": 2.791426}, "fi": {"chunks": 3565.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.189517, "cps": 27.851562, "mean_p": 0.259167, "boundary_entropy": 0.391127, "bpb": 3.04856}, "ru": {"chunks": 5960.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.314512, "cps": 46.5625, "mean_p": 0.33785, "boundary_entropy": 0.546181, "bpb": 1.864109}, "el": {"chunks": 5454.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.265751, "cps": 42.609375, "mean_p": 0.301075, "boundary_entropy": 0.53926, "bpb": 1.990027}, "hi": {"chunks": 6162.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.366241, "cps": 48.140625, "mean_p": 0.388912, "boundary_entropy": 0.629357, "bpb": 1.467956}, "ta": {"chunks": 5242.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.280396, "cps": 40.953125, "mean_p": 0.381578, "boundary_entropy": 0.631959, "bpb": 1.221897}, "th": {"chunks": 8084.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.50258, "cps": 63.15625, "mean_p": 0.40395, "boundary_entropy": 0.644363, "bpb": 1.472534}, "zh": {"chunks": 2558.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.433119, "cps": 19.984375, "mean_p": 0.339871, "boundary_entropy": 0.562557, "bpb": 3.588825}, "my": {"chunks": 6304.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.3327, "cps": 49.25, "mean_p": 0.383921, "boundary_entropy": 0.638291, "bpb": 1.326477}}, "bpb": {"en": 2.48852, "de": 2.81655, "vi": 2.28026, "tr": 2.79143, "fi": 3.04856, "ru": 1.86411, "el": 1.99003, "hi": 1.46796, "ta": 1.2219, "th": 1.47253, "zh": 3.58882, "my": 1.32648}, "gini_cpc": 0.191363, "gini_cps": 0.179285, "bpic": 6.600526, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9320563032339846, "vi": 1.299122849002562, "tr": 1.0678825277236261, "fi": 1.0211022863409245, "ru": 1.6945666043114929, "el": 1.4318446011414514, "hi": 1.9732777499206455, "ta": 1.5107518856136926, "th": 2.707863931070522, "zh": 2.333612247187359, "my": 1.7925629929982725}, "t": 78.8} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.49976, "de": 2.82166, "vi": 2.28492, "tr": 2.80162, "fi": 3.05324, "ru": 1.84694, "el": 1.97997, "hi": 1.46438, "ta": 1.23009, "th": 1.49574, "zh": 3.59945, "my": 1.33696}, "bytes_per_chunk": {"en": 5.417, "de": 5.602, "vi": 5.482, "tr": 5.257, "fi": 5.207, "ru": 5.918, "el": 6.941, "hi": 6.691, "ta": 9.227, "th": 5.877, "zh": 6.509, "my": 8.577}, "bpb_mean": 2.20123, "t": 80.1} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.72005, "bpb": 2.48151, "chunk_loss": 0.030115, "F": 0.1749, "G": 0.28746, "chunks": 179.09, "bytes_per_chunk": 5.718, "boundary_entropy": 0.44243, "router_noise": 0.0, "grad_norm": 0.8279, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 80.1} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.83139, "bpb": 2.64213, "chunk_loss": 0.030121, "F": 0.17615, "G": 0.29077, "chunks": 180.38, "bytes_per_chunk": 5.677, "boundary_entropy": 0.44325, "router_noise": 0.0, "grad_norm": 0.8313, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 82.2} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.51663, "bpb": 2.18803, "chunk_loss": 0.030422, "F": 0.1814, "G": 0.32093, "chunks": 185.75, "bytes_per_chunk": 5.513, "boundary_entropy": 0.48791, "router_noise": 0.0, "grad_norm": 0.8075, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 84.2} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.51301, "bpb": 2.18281, "chunk_loss": 0.030223, "F": 0.1767, "G": 0.31349, "chunks": 180.94, "bytes_per_chunk": 5.659, "boundary_entropy": 0.47932, "router_noise": 0.0, "grad_norm": 0.7403, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 86.2} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3440.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197667, "cps": 26.875, "mean_p": 0.255406, "boundary_entropy": 0.350188, "bpb": 2.425007}, "de": {"chunks": 3817.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.185896, "cps": 29.820312, "mean_p": 0.257441, "boundary_entropy": 0.369047, "bpb": 2.745182}, "vi": {"chunks": 4596.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.249308, "cps": 35.90625, "mean_p": 0.279202, "boundary_entropy": 0.439882, "bpb": 2.197788}, "tr": {"chunks": 3462.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.191239, "cps": 27.046875, "mean_p": 0.244533, "boundary_entropy": 0.361895, "bpb": 2.720795}, "fi": {"chunks": 3172.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.168625, "cps": 24.78125, "mean_p": 0.234747, "boundary_entropy": 0.357197, "bpb": 3.002962}, "ru": {"chunks": 5824.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.307335, "cps": 45.5, "mean_p": 0.345347, "boundary_entropy": 0.553273, "bpb": 1.802361}, "el": {"chunks": 5693.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.277396, "cps": 44.476562, "mean_p": 0.31146, "boundary_entropy": 0.550721, "bpb": 1.890764}, "hi": {"chunks": 6743.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.400773, "cps": 52.679688, "mean_p": 0.401409, "boundary_entropy": 0.635134, "bpb": 1.431856}, "ta": {"chunks": 8727.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.466809, "cps": 68.179688, "mean_p": 0.418956, "boundary_entropy": 0.643923, "bpb": 1.149945}, "th": {"chunks": 8030.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.499223, "cps": 62.734375, "mean_p": 0.403345, "boundary_entropy": 0.643362, "bpb": 1.408611}, "zh": {"chunks": 2682.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.454114, "cps": 20.953125, "mean_p": 0.337791, "boundary_entropy": 0.556447, "bpb": 3.545591}, "my": {"chunks": 8708.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.459574, "cps": 68.03125, "mean_p": 0.435153, "boundary_entropy": 0.665782, "bpb": 1.284318}}, "bpb": {"en": 2.42501, "de": 2.74518, "vi": 2.19779, "tr": 2.7208, "fi": 3.00296, "ru": 1.80236, "el": 1.89076, "hi": 1.43186, "ta": 1.14995, "th": 1.40861, "zh": 3.54559, "my": 1.28432}, "gini_cpc": 0.211893, "gini_cps": 0.220837, "bpic": 5.913967, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9404493928661632, "vi": 1.261253997388688, "tr": 0.9674804317413661, "fi": 0.8530745246781632, "ru": 1.5548118058538378, "el": 1.403350055921377, "hi": 2.0275135457341302, "ta": 2.3615937136530722, "th": 2.5255743470371788, "zh": 2.2973703328896904, "my": 2.3249880333446247}, "t": 89.1} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.43243, "de": 2.74923, "vi": 2.2007, "tr": 2.73145, "fi": 3.00142, "ru": 1.7836, "el": 1.87835, "hi": 1.42845, "ta": 1.15775, "th": 1.42959, "zh": 3.55196, "my": 1.29521}, "bytes_per_chunk": {"en": 5.183, "de": 5.272, "vi": 5.316, "tr": 5.357, "fi": 5.766, "ru": 6.103, "el": 6.627, "hi": 6.178, "ta": 5.738, "th": 5.831, "zh": 6.108, "my": 6.343}, "bpb_mean": 2.13668, "t": 90.3} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.56724, "bpb": 2.26105, "chunk_loss": 0.030284, "F": 0.18143, "G": 0.30804, "chunks": 185.78, "bytes_per_chunk": 5.512, "boundary_entropy": 0.46225, "router_noise": 0.0, "grad_norm": 0.6732, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 90.4} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.66444, "bpb": 2.40127, "chunk_loss": 0.030298, "F": 0.17877, "G": 0.30552, "chunks": 183.06, "bytes_per_chunk": 5.594, "boundary_entropy": 0.46692, "router_noise": 0.0, "grad_norm": 0.708, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 92.4} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.74212, "bpb": 2.51335, "chunk_loss": 0.030147, "F": 0.17468, "G": 0.30609, "chunks": 178.88, "bytes_per_chunk": 5.725, "boundary_entropy": 0.48052, "router_noise": 0.0, "grad_norm": 0.7345, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 94.4} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.6041, "bpb": 2.31423, "chunk_loss": 0.03019, "F": 0.18039, "G": 0.30272, "chunks": 184.72, "bytes_per_chunk": 5.544, "boundary_entropy": 0.45086, "router_noise": 0.0, "grad_norm": 0.8048, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 96.4} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3476.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199736, "cps": 27.15625, "mean_p": 0.25893, "boundary_entropy": 0.342973, "bpb": 2.370527}, "de": {"chunks": 4152.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.202211, "cps": 32.4375, "mean_p": 0.269824, "boundary_entropy": 0.372691, "bpb": 2.658282}, "vi": {"chunks": 4352.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.236073, "cps": 34.0, "mean_p": 0.277919, "boundary_entropy": 0.435399, "bpb": 2.163931}, "tr": {"chunks": 4180.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.230901, "cps": 32.65625, "mean_p": 0.265857, "boundary_entropy": 0.356276, "bpb": 2.647505}, "fi": {"chunks": 3928.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.208814, "cps": 30.6875, "mean_p": 0.269797, "boundary_entropy": 0.373649, "bpb": 2.887413}, "ru": {"chunks": 5976.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.315356, "cps": 46.6875, "mean_p": 0.34273, "boundary_entropy": 0.539565, "bpb": 1.739653}, "el": {"chunks": 5447.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.26541, "cps": 42.554688, "mean_p": 0.31408, "boundary_entropy": 0.553312, "bpb": 1.857693}, "hi": {"chunks": 7931.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.471382, "cps": 61.960938, "mean_p": 0.404496, "boundary_entropy": 0.632064, "bpb": 1.397669}, "ta": {"chunks": 10667.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.57058, "cps": 83.335938, "mean_p": 0.422364, "boundary_entropy": 0.641287, "bpb": 1.163565}, "th": {"chunks": 7865.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.488965, "cps": 61.445312, "mean_p": 0.386169, "boundary_entropy": 0.63188, "bpb": 1.38071}, "zh": {"chunks": 2391.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.404843, "cps": 18.679688, "mean_p": 0.334221, "boundary_entropy": 0.554882, "bpb": 3.51435}, "my": {"chunks": 9656.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.509605, "cps": 75.4375, "mean_p": 0.428242, "boundary_entropy": 0.658583, "bpb": 1.253641}}, "bpb": {"en": 2.37053, "de": 2.65828, "vi": 2.16393, "tr": 2.64751, "fi": 2.88741, "ru": 1.73965, "el": 1.85769, "hi": 1.39767, "ta": 1.16356, "th": 1.38071, "zh": 3.51435, "my": 1.25364}, "gini_cpc": 0.214544, "gini_cps": 0.237697, "bpic": 5.480941, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0123933647018128, "vi": 1.181925485088497, "tr": 1.1560326038694195, "fi": 1.0454516396797677, "ru": 1.5788676518364904, "el": 1.3288038646696145, "hi": 2.3600284010758554, "ta": 2.8566772621246335, "th": 2.448059753760678, "zh": 2.026891427531519, "my": 2.5513981332958253}, "t": 99.3} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.37677, "de": 2.66448, "vi": 2.16974, "tr": 2.65934, "fi": 2.89453, "ru": 1.72204, "el": 1.84682, "hi": 1.39433, "ta": 1.17112, "th": 1.39849, "zh": 3.52075, "my": 1.26377}, "bytes_per_chunk": {"en": 5.073, "de": 4.881, "vi": 5.665, "tr": 4.542, "fi": 4.769, "ru": 6.027, "el": 6.817, "hi": 5.473, "ta": 4.581, "th": 6.119, "zh": 6.777, "my": 5.424}, "bpb_mean": 2.09018, "t": 100.6} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.59698, "bpb": 2.30395, "chunk_loss": 0.030672, "F": 0.18927, "G": 0.31181, "chunks": 193.81, "bytes_per_chunk": 5.283, "boundary_entropy": 0.46128, "router_noise": 0.0, "grad_norm": 0.7331, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 100.6} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.57663, "bpb": 2.27459, "chunk_loss": 0.030453, "F": 0.18726, "G": 0.30023, "chunks": 191.75, "bytes_per_chunk": 5.34, "boundary_entropy": 0.44537, "router_noise": 0.0, "grad_norm": 0.7793, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 102.6} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.54169, "bpb": 2.22418, "chunk_loss": 0.030383, "F": 0.18369, "G": 0.3101, "chunks": 188.09, "bytes_per_chunk": 5.444, "boundary_entropy": 0.45985, "router_noise": 0.0, "grad_norm": 0.737, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 104.6} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.56293, "bpb": 2.25483, "chunk_loss": 0.030107, "F": 0.17731, "G": 0.29422, "chunks": 181.56, "bytes_per_chunk": 5.64, "boundary_entropy": 0.44034, "router_noise": 0.0, "grad_norm": 0.7342, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 106.6} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3446.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198012, "cps": 26.921875, "mean_p": 0.252776, "boundary_entropy": 0.328635, "bpb": 2.325992}, "de": {"chunks": 4042.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196854, "cps": 31.578125, "mean_p": 0.256335, "boundary_entropy": 0.346479, "bpb": 2.625012}, "vi": {"chunks": 4788.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.259723, "cps": 37.40625, "mean_p": 0.28966, "boundary_entropy": 0.442917, "bpb": 2.11938}, "tr": {"chunks": 3975.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.219577, "cps": 31.054688, "mean_p": 0.254347, "boundary_entropy": 0.342048, "bpb": 2.601516}, "fi": {"chunks": 3482.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.185104, "cps": 27.203125, "mean_p": 0.240408, "boundary_entropy": 0.335681, "bpb": 2.833845}, "ru": {"chunks": 6147.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.32438, "cps": 48.023438, "mean_p": 0.347988, "boundary_entropy": 0.545182, "bpb": 1.702564}, "el": {"chunks": 6544.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.318862, "cps": 51.125, "mean_p": 0.334762, "boundary_entropy": 0.568585, "bpb": 1.795367}, "hi": {"chunks": 6582.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.391204, "cps": 51.421875, "mean_p": 0.390957, "boundary_entropy": 0.624588, "bpb": 1.36976}, "ta": {"chunks": 8504.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.454881, "cps": 66.4375, "mean_p": 0.383294, "boundary_entropy": 0.615875, "bpb": 1.108058}, "th": {"chunks": 8115.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.504507, "cps": 63.398438, "mean_p": 0.394166, "boundary_entropy": 0.633499, "bpb": 1.34494}, "zh": {"chunks": 2547.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.431256, "cps": 19.898438, "mean_p": 0.327996, "boundary_entropy": 0.538036, "bpb": 3.469596}, "my": {"chunks": 8636.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.455774, "cps": 67.46875, "mean_p": 0.428314, "boundary_entropy": 0.664753, "bpb": 1.244503}}, "bpb": {"en": 2.32599, "de": 2.62501, "vi": 2.11938, "tr": 2.60152, "fi": 2.83384, "ru": 1.70256, "el": 1.79537, "hi": 1.36976, "ta": 1.10806, "th": 1.34494, "zh": 3.4696, "my": 1.2445}, "gini_cpc": 0.192467, "gini_cps": 0.208099, "bpic": 5.744537, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9941519051236944, "vi": 1.3116556878719774, "tr": 1.108907775392047, "fi": 0.9348151248113585, "ru": 1.6381846261276505, "el": 1.6103166841760057, "hi": 1.9756574756872967, "ta": 2.297241372137919, "th": 2.547864372581023, "zh": 2.1779321677061345, "my": 2.3017497501485877}, "t": 109.5} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.33134, "de": 2.62856, "vi": 2.12596, "tr": 2.61191, "fi": 2.83697, "ru": 1.68332, "el": 1.78654, "hi": 1.36777, "ta": 1.11465, "th": 1.36812, "zh": 3.47517, "my": 1.25361}, "bytes_per_chunk": {"en": 5.086, "de": 5.024, "vi": 5.126, "tr": 4.663, "fi": 5.231, "ru": 5.776, "el": 5.65, "hi": 6.717, "ta": 5.888, "th": 5.783, "zh": 6.251, "my": 6.618}, "bpb_mean": 2.04866, "t": 110.7} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.52724, "bpb": 2.20334, "chunk_loss": 0.030411, "F": 0.18497, "G": 0.31129, "chunks": 189.41, "bytes_per_chunk": 5.406, "boundary_entropy": 0.4676, "router_noise": 0.0, "grad_norm": 0.6749, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 110.8} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.57734, "bpb": 2.27562, "chunk_loss": 0.030431, "F": 0.185, "G": 0.28485, "chunks": 189.44, "bytes_per_chunk": 5.405, "boundary_entropy": 0.41736, "router_noise": 0.0, "grad_norm": 0.6859, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 112.8} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.62951, "bpb": 2.35088, "chunk_loss": 0.030895, "F": 0.1936, "G": 0.30484, "chunks": 198.25, "bytes_per_chunk": 5.165, "boundary_entropy": 0.43876, "router_noise": 0.0, "grad_norm": 0.7242, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 114.8} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.46878, "bpb": 2.11901, "chunk_loss": 0.030581, "F": 0.18347, "G": 0.31265, "chunks": 187.88, "bytes_per_chunk": 5.45, "boundary_entropy": 0.46863, "router_noise": 0.0, "grad_norm": 0.6943, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 116.8} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3380.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.194219, "cps": 26.40625, "mean_p": 0.240308, "boundary_entropy": 0.3091, "bpb": 2.286451}, "de": {"chunks": 3955.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.192617, "cps": 30.898438, "mean_p": 0.246396, "boundary_entropy": 0.327831, "bpb": 2.567409}, "vi": {"chunks": 4095.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.222132, "cps": 31.992188, "mean_p": 0.27071, "boundary_entropy": 0.421472, "bpb": 2.103161}, "tr": {"chunks": 3790.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.209358, "cps": 29.609375, "mean_p": 0.238569, "boundary_entropy": 0.317302, "bpb": 2.548258}, "fi": {"chunks": 3400.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180745, "cps": 26.5625, "mean_p": 0.240006, "boundary_entropy": 0.339044, "bpb": 2.764873}, "ru": {"chunks": 5409.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.285435, "cps": 42.257812, "mean_p": 0.346406, "boundary_entropy": 0.548312, "bpb": 1.667135}, "el": {"chunks": 6014.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.293037, "cps": 46.984375, "mean_p": 0.328428, "boundary_entropy": 0.564129, "bpb": 1.754874}, "hi": {"chunks": 8475.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.503715, "cps": 66.210938, "mean_p": 0.405464, "boundary_entropy": 0.626557, "bpb": 1.357528}, "ta": {"chunks": 8892.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.475635, "cps": 69.46875, "mean_p": 0.403581, "boundary_entropy": 0.626975, "bpb": 1.093981}, "th": {"chunks": 6529.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.405906, "cps": 51.007812, "mean_p": 0.376052, "boundary_entropy": 0.625958, "bpb": 1.313269}, "zh": {"chunks": 2646.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.448019, "cps": 20.671875, "mean_p": 0.334622, "boundary_entropy": 0.542657, "bpb": 3.432327}, "my": {"chunks": 8106.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.427802, "cps": 63.328125, "mean_p": 0.431743, "boundary_entropy": 0.664317, "bpb": 1.22353}}, "bpb": {"en": 2.28645, "de": 2.56741, "vi": 2.10316, "tr": 2.54826, "fi": 2.76487, "ru": 1.66714, "el": 1.75487, "hi": 1.35753, "ta": 1.09398, "th": 1.31327, "zh": 3.43233, "my": 1.22353}, "gini_cpc": 0.206836, "gini_cps": 0.217301, "bpic": 5.932525, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9917483819523313, "vi": 1.1437159667021344, "tr": 1.0779436995470038, "fi": 0.93062443949609, "ru": 1.4696542911117703, "el": 1.5087941743525162, "hi": 2.593534645717752, "ta": 2.4489583804801778, "th": 2.0899361748698206, "zh": 2.3067674634360404, "my": 2.2026761188828226}, "t": 119.7} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.29272, "de": 2.57353, "vi": 2.10836, "tr": 2.55434, "fi": 2.77554, "ru": 1.64421, "el": 1.74625, "hi": 1.35506, "ta": 1.10031, "th": 1.33219, "zh": 3.43633, "my": 1.23247}, "bytes_per_chunk": {"en": 5.216, "de": 5.144, "vi": 5.789, "tr": 5.053, "fi": 5.453, "ru": 6.489, "el": 6.286, "hi": 5.539, "ta": 5.944, "th": 7.131, "zh": 6.179, "my": 6.723}, "bpb_mean": 2.01261, "t": 121.0} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.59748, "bpb": 2.30468, "chunk_loss": 0.030102, "F": 0.17902, "G": 0.28738, "chunks": 183.31, "bytes_per_chunk": 5.586, "boundary_entropy": 0.41948, "router_noise": 0.0, "grad_norm": 0.7794, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 121.0} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.60515, "bpb": 2.31575, "chunk_loss": 0.030796, "F": 0.19354, "G": 0.31034, "chunks": 198.19, "bytes_per_chunk": 5.167, "boundary_entropy": 0.46055, "router_noise": 0.0, "grad_norm": 0.7193, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 123.0} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.54898, "bpb": 2.23471, "chunk_loss": 0.030522, "F": 0.18591, "G": 0.29135, "chunks": 190.38, "bytes_per_chunk": 5.379, "boundary_entropy": 0.4234, "router_noise": 0.0, "grad_norm": 0.7212, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 125.0} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.50962, "bpb": 2.17792, "chunk_loss": 0.03043, "F": 0.18417, "G": 0.29309, "chunks": 188.59, "bytes_per_chunk": 5.43, "boundary_entropy": 0.42933, "router_noise": 0.0, "grad_norm": 0.7062, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 127.0} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3452.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198357, "cps": 26.96875, "mean_p": 0.242549, "boundary_entropy": 0.303246, "bpb": 2.261465}, "de": {"chunks": 3933.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.191545, "cps": 30.726562, "mean_p": 0.243896, "boundary_entropy": 0.325557, "bpb": 2.519711}, "vi": {"chunks": 4472.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.242582, "cps": 34.9375, "mean_p": 0.280728, "boundary_entropy": 0.429749, "bpb": 2.077366}, "tr": {"chunks": 3825.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.211291, "cps": 29.882812, "mean_p": 0.239989, "boundary_entropy": 0.316729, "bpb": 2.497455}, "fi": {"chunks": 3865.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.205465, "cps": 30.195312, "mean_p": 0.258145, "boundary_entropy": 0.339955, "bpb": 2.71272}, "ru": {"chunks": 5594.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.295198, "cps": 43.703125, "mean_p": 0.3378, "boundary_entropy": 0.533863, "bpb": 1.637327}, "el": {"chunks": 5858.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.285436, "cps": 45.765625, "mean_p": 0.315954, "boundary_entropy": 0.54835, "bpb": 1.716543}, "hi": {"chunks": 7140.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.424368, "cps": 55.78125, "mean_p": 0.383034, "boundary_entropy": 0.614346, "bpb": 1.322774}, "ta": {"chunks": 6913.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.369778, "cps": 54.007812, "mean_p": 0.384314, "boundary_entropy": 0.619442, "bpb": 1.078186}, "th": {"chunks": 6938.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.431334, "cps": 54.203125, "mean_p": 0.38549, "boundary_entropy": 0.630206, "bpb": 1.277271}, "zh": {"chunks": 2564.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.434135, "cps": 20.03125, "mean_p": 0.319615, "boundary_entropy": 0.523667, "bpb": 3.436641}, "my": {"chunks": 8292.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.437619, "cps": 64.78125, "mean_p": 0.418188, "boundary_entropy": 0.65567, "bpb": 1.201495}}, "bpb": {"en": 2.26147, "de": 2.51971, "vi": 2.07737, "tr": 2.49746, "fi": 2.71272, "ru": 1.63733, "el": 1.71654, "hi": 1.32277, "ta": 1.07819, "th": 1.27727, "zh": 3.43664, "my": 1.2015}, "gini_cpc": 0.175888, "gini_cps": 0.185809, "bpic": 6.106689, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9656614011788615, "vi": 1.2229592495759583, "tr": 1.0652075137959438, "fi": 1.0358358743648244, "ru": 1.4882180954331854, "el": 1.4390035168975555, "hi": 2.1394220720076995, "ta": 1.8642082282613783, "th": 2.174535807203259, "zh": 2.1886580373318725, "my": 2.2062222008590444}, "t": 129.9} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.26862, "de": 2.52442, "vi": 2.08024, "tr": 2.50344, "fi": 2.72465, "ru": 1.61508, "el": 1.70681, "hi": 1.31994, "ta": 1.08494, "th": 1.29525, "zh": 3.44178, "my": 1.21091}, "bytes_per_chunk": {"en": 5.182, "de": 5.192, "vi": 5.448, "tr": 4.873, "fi": 4.753, "ru": 6.348, "el": 6.402, "hi": 5.904, "ta": 7.188, "th": 6.72, "zh": 6.139, "my": 6.537}, "bpb_mean": 1.98134, "t": 131.1} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.44937, "bpb": 2.091, "chunk_loss": 0.030216, "F": 0.18152, "G": 0.29724, "chunks": 185.88, "bytes_per_chunk": 5.509, "boundary_entropy": 0.43557, "router_noise": 0.0, "grad_norm": 0.6934, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 131.2} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.53436, "bpb": 2.21361, "chunk_loss": 0.030326, "F": 0.18237, "G": 0.29296, "chunks": 186.75, "bytes_per_chunk": 5.483, "boundary_entropy": 0.42792, "router_noise": 0.0, "grad_norm": 0.6995, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 133.2} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.47478, "bpb": 2.12766, "chunk_loss": 0.030277, "F": 0.18243, "G": 0.30628, "chunks": 186.81, "bytes_per_chunk": 5.481, "boundary_entropy": 0.44799, "router_noise": 0.0, "grad_norm": 0.7592, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 135.2} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.4905, "bpb": 2.15034, "chunk_loss": 0.03047, "F": 0.18689, "G": 0.29965, "chunks": 191.38, "bytes_per_chunk": 5.351, "boundary_entropy": 0.43204, "router_noise": 0.0, "grad_norm": 0.7127, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 137.2} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3370.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193645, "cps": 26.328125, "mean_p": 0.234203, "boundary_entropy": 0.292877, "bpb": 2.229226}, "de": {"chunks": 4033.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196416, "cps": 31.507812, "mean_p": 0.248061, "boundary_entropy": 0.322901, "bpb": 2.47962}, "vi": {"chunks": 3941.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.213778, "cps": 30.789062, "mean_p": 0.272843, "boundary_entropy": 0.421233, "bpb": 2.07285}, "tr": {"chunks": 3959.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.218693, "cps": 30.929688, "mean_p": 0.243918, "boundary_entropy": 0.313724, "bpb": 2.47225}, "fi": {"chunks": 3785.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.201212, "cps": 29.570312, "mean_p": 0.252505, "boundary_entropy": 0.332494, "bpb": 2.670726}, "ru": {"chunks": 6061.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319842, "cps": 47.351562, "mean_p": 0.358878, "boundary_entropy": 0.548768, "bpb": 1.603815}, "el": {"chunks": 6333.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.308581, "cps": 49.476562, "mean_p": 0.32501, "boundary_entropy": 0.554163, "bpb": 1.696141}, "hi": {"chunks": 7691.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.457117, "cps": 60.085938, "mean_p": 0.384693, "boundary_entropy": 0.607785, "bpb": 1.291987}, "ta": {"chunks": 9972.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.533405, "cps": 77.90625, "mean_p": 0.406934, "boundary_entropy": 0.624547, "bpb": 1.05629}, "th": {"chunks": 8106.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.503948, "cps": 63.328125, "mean_p": 0.377409, "boundary_entropy": 0.617378, "bpb": 1.253772}, "zh": {"chunks": 2794.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.473078, "cps": 21.828125, "mean_p": 0.330796, "boundary_entropy": 0.529903, "bpb": 3.375156}, "my": {"chunks": 9108.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.480684, "cps": 71.15625, "mean_p": 0.414513, "boundary_entropy": 0.654941, "bpb": 1.180181}}, "bpb": {"en": 2.22923, "de": 2.47962, "vi": 2.07285, "tr": 2.47225, "fi": 2.67073, "ru": 1.60382, "el": 1.69614, "hi": 1.29199, "ta": 1.05629, "th": 1.25377, "zh": 3.37516, "my": 1.18018}, "gini_cpc": 0.213079, "gini_cps": 0.227451, "bpic": 5.549738, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0143084281639125, "vi": 1.1039706113145955, "tr": 1.1293515952484883, "fi": 1.0390781673769705, "ru": 1.651692850935227, "el": 1.5935396069344505, "hi": 2.360597581139413, "ta": 2.7545522811523098, "th": 2.60243417526881, "zh": 2.4430208780287814, "my": 2.4822977018471293}, "t": 140.1} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.23708, "de": 2.48582, "vi": 2.07711, "tr": 2.48024, "fi": 2.68262, "ru": 1.5809, "el": 1.68638, "hi": 1.29103, "ta": 1.06288, "th": 1.27149, "zh": 3.38015, "my": 1.19026}, "bytes_per_chunk": {"en": 5.299, "de": 5.074, "vi": 6.323, "tr": 4.755, "fi": 4.874, "ru": 5.856, "el": 5.889, "hi": 5.513, "ta": 5.1, "th": 5.799, "zh": 5.828, "my": 5.958}, "bpb_mean": 1.95216, "t": 141.4} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 141.4} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "parity_a", "lm_loss": 1.3722, "bpb": 1.97966, "chunk_loss": 0.037461, "F": 0.18881, "G": 0.3018, "chunks": 193.34, "bytes_per_chunk": 5.296, "boundary_entropy": 0.43333, "router_noise": 0.0, "grad_norm": 0.7131, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.085, "n_target_mean": 10.027, "alpha": 0.03, "t": 141.4} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "parity_a", "lm_loss": 1.47118, "bpb": 2.12247, "chunk_loss": 0.031053, "F": 0.1452, "G": 0.25159, "chunks": 148.69, "bytes_per_chunk": 6.887, "boundary_entropy": 0.41209, "router_noise": 0.0, "grad_norm": 0.6844, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 9.8405, "alpha": 0.03, "t": 143.5} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "parity_a", "lm_loss": 1.50749, "bpb": 2.17485, "chunk_loss": 0.030711, "F": 0.15802, "G": 0.24439, "chunks": 161.81, "bytes_per_chunk": 6.328, "boundary_entropy": 0.37177, "router_noise": 0.0, "grad_norm": 0.703, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.0154, "alpha": 0.03, "t": 145.5} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "parity_a", "lm_loss": 1.56977, "bpb": 2.2647, "chunk_loss": 0.030628, "F": 0.15921, "G": 0.23701, "chunks": 163.03, "bytes_per_chunk": 6.281, "boundary_entropy": 0.34838, "router_noise": 0.0, "grad_norm": 0.7185, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 8.354, "alpha": 0.03, "t": 147.5} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3479.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199908, "cps": 27.179688, "mean_p": 0.236093, "boundary_entropy": 0.276527, "bpb": 2.199516}, "de": {"chunks": 3978.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.193737, "cps": 31.078125, "mean_p": 0.234606, "boundary_entropy": 0.290875, "bpb": 2.442372}, "vi": {"chunks": 3556.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.192894, "cps": 27.78125, "mean_p": 0.231282, "boundary_entropy": 0.364705, "bpb": 2.051389}, "tr": {"chunks": 3678.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203171, "cps": 28.734375, "mean_p": 0.227064, "boundary_entropy": 0.28761, "bpb": 2.43409}, "fi": {"chunks": 3567.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.189623, "cps": 27.867188, "mean_p": 0.232914, "boundary_entropy": 0.299703, "bpb": 2.642083}, "ru": {"chunks": 3458.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.18248, "cps": 27.015625, "mean_p": 0.203974, "boundary_entropy": 0.381944, "bpb": 1.603864}, "el": {"chunks": 3017.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.147006, "cps": 23.570312, "mean_p": 0.223428, "boundary_entropy": 0.444649, "bpb": 1.71277}, "hi": {"chunks": 2860.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.169985, "cps": 22.34375, "mean_p": 0.240365, "boundary_entropy": 0.491246, "bpb": 1.349119}, "ta": {"chunks": 2705.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.144691, "cps": 21.132812, "mean_p": 0.188282, "boundary_entropy": 0.42203, "bpb": 1.086509}, "th": {"chunks": 2710.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.16848, "cps": 21.171875, "mean_p": 0.208398, "boundary_entropy": 0.4462, "bpb": 1.272142}, "zh": {"chunks": 2943.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.498307, "cps": 22.992188, "mean_p": 0.321715, "boundary_entropy": 0.510254, "bpb": 3.345692}, "my": {"chunks": 2207.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.116477, "cps": 17.242188, "mean_p": 0.2908, "boundary_entropy": 0.573133, "bpb": 1.193051}}, "bpb": {"en": 2.19952, "de": 2.44237, "vi": 2.05139, "tr": 2.43409, "fi": 2.64208, "ru": 1.60386, "el": 1.71277, "hi": 1.34912, "ta": 1.08651, "th": 1.27214, "zh": 3.34569, "my": 1.19305}, "gini_cpc": 0.183474, "gini_cps": 0.086845, "bpic": 10.057681, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9691300567946995, "vi": 0.964913320443409, "tr": 1.0163209193370941, "fi": 0.9485515048109691, "ru": 0.9128206707262041, "el": 0.7353670333652257, "hi": 0.8503165885570267, "ta": 0.7237881881900042, "th": 0.8427871728796331, "zh": 2.492679895346147, "my": 0.5826512044309646}, "t": 150.4} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.20698, "de": 2.44849, "vi": 2.0537, "tr": 2.44061, "fi": 2.65097, "ru": 1.58104, "el": 1.70439, "hi": 1.34481, "ta": 1.09156, "th": 1.28867, "zh": 3.34842, "my": 1.20245}, "bytes_per_chunk": {"en": 5.093, "de": 5.106, "vi": 6.941, "tr": 5.1, "fi": 5.137, "ru": 9.802, "el": 11.973, "hi": 15.266, "ta": 18.866, "th": 17.03, "zh": 5.481, "my": 22.251}, "bpb_mean": 1.94684, "t": 151.6} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "parity_a", "lm_loss": 1.59874, "bpb": 2.3065, "chunk_loss": 0.030729, "F": 0.16983, "G": 0.24375, "chunks": 173.91, "bytes_per_chunk": 5.888, "boundary_entropy": 0.35289, "router_noise": 0.0, "grad_norm": 0.7493, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.084, "n_target_mean": 8.0814, "alpha": 0.03, "t": 151.7} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "parity_a", "lm_loss": 1.49919, "bpb": 2.16288, "chunk_loss": 0.030713, "F": 0.14709, "G": 0.23208, "chunks": 150.62, "bytes_per_chunk": 6.798, "boundary_entropy": 0.35607, "router_noise": 0.0, "grad_norm": 0.639, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 9.2749, "alpha": 0.03, "t": 153.7} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "parity_a", "lm_loss": 1.51487, "bpb": 2.18549, "chunk_loss": 0.030527, "F": 0.16116, "G": 0.24156, "chunks": 165.03, "bytes_per_chunk": 6.205, "boundary_entropy": 0.3519, "router_noise": 0.0, "grad_norm": 0.6898, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 8.6096, "alpha": 0.03, "t": 155.8} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "parity_a", "lm_loss": 1.40314, "bpb": 2.02431, "chunk_loss": 0.03074, "F": 0.1619, "G": 0.22735, "chunks": 165.78, "bytes_per_chunk": 6.177, "boundary_entropy": 0.31822, "router_noise": 0.0, "grad_norm": 0.7017, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.5736, "alpha": 0.03, "t": 157.8} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3328.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.191231, "cps": 26.0, "mean_p": 0.227845, "boundary_entropy": 0.265653, "bpb": 2.181759}, "de": {"chunks": 3959.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.192812, "cps": 30.929688, "mean_p": 0.232142, "boundary_entropy": 0.282811, "bpb": 2.420215}, "vi": {"chunks": 3353.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.181882, "cps": 26.195312, "mean_p": 0.219552, "boundary_entropy": 0.348603, "bpb": 2.038249}, "tr": {"chunks": 3612.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.199525, "cps": 28.21875, "mean_p": 0.22276, "boundary_entropy": 0.275553, "bpb": 2.412602}, "fi": {"chunks": 3659.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194514, "cps": 28.585938, "mean_p": 0.237163, "boundary_entropy": 0.298915, "bpb": 2.613239}, "ru": {"chunks": 3267.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.172401, "cps": 25.523438, "mean_p": 0.184417, "boundary_entropy": 0.348927, "bpb": 1.570216}, "el": {"chunks": 3221.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.156946, "cps": 25.164062, "mean_p": 0.208847, "boundary_entropy": 0.417465, "bpb": 1.66631}, "hi": {"chunks": 3080.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.183061, "cps": 24.0625, "mean_p": 0.237602, "boundary_entropy": 0.477088, "bpb": 1.326004}, "ta": {"chunks": 2790.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.149238, "cps": 21.796875, "mean_p": 0.203857, "boundary_entropy": 0.442336, "bpb": 1.096397}, "th": {"chunks": 2645.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.164439, "cps": 20.664062, "mean_p": 0.192693, "boundary_entropy": 0.418052, "bpb": 1.262881}, "zh": {"chunks": 2830.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.479174, "cps": 22.109375, "mean_p": 0.311599, "boundary_entropy": 0.496597, "bpb": 3.335837}, "my": {"chunks": 1999.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.105499, "cps": 15.617188, "mean_p": 0.256694, "boundary_entropy": 0.539544, "bpb": 1.187531}}, "bpb": {"en": 2.18176, "de": 2.42021, "vi": 2.03825, "tr": 2.4126, "fi": 2.61324, "ru": 1.57022, "el": 1.66631, "hi": 1.326, "ta": 1.0964, "th": 1.26288, "zh": 3.33584, "my": 1.18753}, "gini_cpc": 0.176652, "gini_cps": 0.087974, "bpic": 10.16827, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0082631553006605, "vi": 0.9511110209206985, "tr": 1.0433691531152933, "fi": 1.01716481418912, "ru": 0.9015311199766592, "el": 0.8207118086790056, "hi": 0.9572744027888901, "ta": 0.7804040891743989, "th": 0.8598949802431793, "zh": 2.505727246900154, "my": 0.5516837865220198}, "t": 160.7} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.18963, "de": 2.4283, "vi": 2.04, "tr": 2.4193, "fi": 2.62417, "ru": 1.54953, "el": 1.65745, "hi": 1.32153, "ta": 1.10183, "th": 1.27878, "zh": 3.33714, "my": 1.1962}, "bytes_per_chunk": {"en": 5.284, "de": 5.175, "vi": 7.233, "tr": 5.247, "fi": 4.986, "ru": 10.306, "el": 11.131, "hi": 14.087, "ta": 18.544, "th": 17.347, "zh": 5.695, "my": 23.964}, "bpb_mean": 1.92866, "t": 161.9} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "parity_a", "lm_loss": 1.47557, "bpb": 2.1288, "chunk_loss": 0.030504, "F": 0.15121, "G": 0.23154, "chunks": 154.84, "bytes_per_chunk": 6.613, "boundary_entropy": 0.34357, "router_noise": 0.0, "grad_norm": 0.7796, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.2272, "alpha": 0.03, "t": 161.9} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "parity_a", "lm_loss": 1.43816, "bpb": 2.07483, "chunk_loss": 0.03065, "F": 0.15244, "G": 0.22554, "chunks": 156.09, "bytes_per_chunk": 6.56, "boundary_entropy": 0.33282, "router_noise": 0.0, "grad_norm": 0.6825, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.3574, "alpha": 0.03, "t": 163.9} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "parity_a", "lm_loss": 1.4546, "bpb": 2.09855, "chunk_loss": 0.030378, "F": 0.14972, "G": 0.22113, "chunks": 153.31, "bytes_per_chunk": 6.679, "boundary_entropy": 0.32431, "router_noise": 0.0, "grad_norm": 0.7007, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.9121, "alpha": 0.03, "t": 165.9} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "parity_a", "lm_loss": 1.35695, "bpb": 1.95767, "chunk_loss": 0.030617, "F": 0.14883, "G": 0.22656, "chunks": 152.41, "bytes_per_chunk": 6.719, "boundary_entropy": 0.34281, "router_noise": 0.0, "grad_norm": 0.6438, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.8268, "alpha": 0.03, "t": 167.9} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3344.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.192151, "cps": 26.125, "mean_p": 0.221959, "boundary_entropy": 0.249741, "bpb": 2.157646}, "de": {"chunks": 3812.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.185652, "cps": 29.78125, "mean_p": 0.218224, "boundary_entropy": 0.261132, "bpb": 2.389328}, "vi": {"chunks": 3204.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.1738, "cps": 25.03125, "mean_p": 0.209932, "boundary_entropy": 0.330291, "bpb": 2.022208}, "tr": {"chunks": 3490.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.192786, "cps": 27.265625, "mean_p": 0.215934, "boundary_entropy": 0.261722, "bpb": 2.379017}, "fi": {"chunks": 3543.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188347, "cps": 27.679688, "mean_p": 0.2271, "boundary_entropy": 0.283251, "bpb": 2.566198}, "ru": {"chunks": 3238.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.170871, "cps": 25.296875, "mean_p": 0.167963, "boundary_entropy": 0.305449, "bpb": 1.55276}, "el": {"chunks": 2660.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.129611, "cps": 20.78125, "mean_p": 0.192415, "boundary_entropy": 0.397807, "bpb": 1.644873}, "hi": {"chunks": 2981.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.177177, "cps": 23.289062, "mean_p": 0.235716, "boundary_entropy": 0.469872, "bpb": 1.306037}, "ta": {"chunks": 2014.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.107729, "cps": 15.734375, "mean_p": 0.189701, "boundary_entropy": 0.426929, "bpb": 1.069409}, "th": {"chunks": 2386.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.148337, "cps": 18.640625, "mean_p": 0.178002, "boundary_entropy": 0.398486, "bpb": 1.246234}, "zh": {"chunks": 2896.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.490349, "cps": 22.625, "mean_p": 0.308613, "boundary_entropy": 0.490824, "bpb": 3.308805}, "my": {"chunks": 2603.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.137376, "cps": 20.335938, "mean_p": 0.231847, "boundary_entropy": 0.50172, "bpb": 1.1718}}, "bpb": {"en": 2.15765, "de": 2.38933, "vi": 2.02221, "tr": 2.37902, "fi": 2.5662, "ru": 1.55276, "el": 1.64487, "hi": 1.30604, "ta": 1.06941, "th": 1.24623, "zh": 3.3088, "my": 1.1718}, "gini_cpc": 0.19646, "gini_cps": 0.094618, "bpic": 10.610185, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9661806516619179, "vi": 0.9044971794083181, "tr": 1.003304423359474, "fi": 0.980205466795235, "ru": 0.8892532918407797, "el": 0.6745259199213299, "hi": 0.9220718307656213, "ta": 0.5606500279608062, "th": 0.7719820912526605, "zh": 2.5518959714960214, "my": 0.7149384319765099}, "t": 170.8} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.16796, "de": 2.39321, "vi": 2.02187, "tr": 2.38792, "fi": 2.57675, "ru": 1.53157, "el": 1.63554, "hi": 1.30301, "ta": 1.07529, "th": 1.2629, "zh": 3.30948, "my": 1.18143}, "bytes_per_chunk": {"en": 5.29, "de": 5.34, "vi": 7.422, "tr": 5.343, "fi": 5.138, "ru": 10.338, "el": 13.309, "hi": 13.864, "ta": 26.238, "th": 19.719, "zh": 5.548, "my": 19.865}, "bpb_mean": 1.90391, "t": 172.0} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "parity_a", "lm_loss": 1.43758, "bpb": 2.07399, "chunk_loss": 0.030198, "F": 0.14124, "G": 0.2203, "chunks": 144.62, "bytes_per_chunk": 7.08, "boundary_entropy": 0.34074, "router_noise": 0.0, "grad_norm": 0.6915, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.5059, "alpha": 0.03, "t": 172.1} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "parity_a", "lm_loss": 1.39636, "bpb": 2.01453, "chunk_loss": 0.030443, "F": 0.14282, "G": 0.2241, "chunks": 146.25, "bytes_per_chunk": 7.002, "boundary_entropy": 0.33851, "router_noise": 0.0, "grad_norm": 0.6459, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.6662, "alpha": 0.03, "t": 174.1} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "parity_a", "lm_loss": 1.41065, "bpb": 2.03514, "chunk_loss": 0.030431, "F": 0.15781, "G": 0.22261, "chunks": 161.59, "bytes_per_chunk": 6.337, "boundary_entropy": 0.31384, "router_noise": 0.0, "grad_norm": 0.7212, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.4014, "alpha": 0.03, "t": 176.1} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "parity_a", "lm_loss": 1.35611, "bpb": 1.95645, "chunk_loss": 0.030234, "F": 0.1362, "G": 0.21272, "chunks": 139.47, "bytes_per_chunk": 7.342, "boundary_entropy": 0.32399, "router_noise": 0.0, "grad_norm": 0.6696, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.7269, "alpha": 0.03, "t": 178.0} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3375.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.193932, "cps": 26.367188, "mean_p": 0.2249, "boundary_entropy": 0.244763, "bpb": 2.140027}, "de": {"chunks": 3812.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.185652, "cps": 29.78125, "mean_p": 0.215935, "boundary_entropy": 0.254821, "bpb": 2.359227}, "vi": {"chunks": 3073.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.166694, "cps": 24.007812, "mean_p": 0.197099, "boundary_entropy": 0.313761, "bpb": 2.012408}, "tr": {"chunks": 3501.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.193393, "cps": 27.351562, "mean_p": 0.213968, "boundary_entropy": 0.250063, "bpb": 2.35031}, "fi": {"chunks": 3666.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194886, "cps": 28.640625, "mean_p": 0.231397, "boundary_entropy": 0.274509, "bpb": 2.532573}, "ru": {"chunks": 3090.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.163061, "cps": 24.140625, "mean_p": 0.163954, "boundary_entropy": 0.300445, "bpb": 1.527371}, "el": {"chunks": 2641.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.128685, "cps": 20.632812, "mean_p": 0.196105, "boundary_entropy": 0.399629, "bpb": 1.622462}, "hi": {"chunks": 2444.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.14526, "cps": 19.09375, "mean_p": 0.235841, "boundary_entropy": 0.479899, "bpb": 1.304961}, "ta": {"chunks": 2187.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.116983, "cps": 17.085938, "mean_p": 0.192613, "boundary_entropy": 0.431522, "bpb": 1.056618}, "th": {"chunks": 2225.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.138328, "cps": 17.382812, "mean_p": 0.180451, "boundary_entropy": 0.402545, "bpb": 1.243688}, "zh": {"chunks": 2905.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.491873, "cps": 22.695312, "mean_p": 0.299852, "boundary_entropy": 0.474507, "bpb": 3.273764}, "my": {"chunks": 2279.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.120277, "cps": 17.804688, "mean_p": 0.25983, "boundary_entropy": 0.53851, "bpb": 1.161856}}, "bpb": {"en": 2.14003, "de": 2.35923, "vi": 2.01241, "tr": 2.35031, "fi": 2.53257, "ru": 1.52737, "el": 1.62246, "hi": 1.30496, "ta": 1.05662, "th": 1.24369, "zh": 3.27376, "my": 1.16186}, "gini_cpc": 0.210365, "gini_cps": 0.107885, "bpic": 10.903489, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9573061034540604, "vi": 0.8595472621523069, "tr": 0.9972221178810142, "fi": 1.0049186823312601, "ru": 0.8408133685136323, "el": 0.663556486760112, "hi": 0.7490252732375764, "ta": 0.6032171168761701, "th": 0.7132787621317307, "zh": 2.5363141061820373, "my": 0.6201993291581639}, "t": 180.9} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.14795, "de": 2.36552, "vi": 2.01414, "tr": 2.35898, "fi": 2.54279, "ru": 1.50481, "el": 1.61485, "hi": 1.3004, "ta": 1.06241, "th": 1.25981, "zh": 3.27526, "my": 1.16935}, "bytes_per_chunk": {"en": 5.284, "de": 5.396, "vi": 7.859, "tr": 5.389, "fi": 4.999, "ru": 10.803, "el": 13.592, "hi": 17.198, "ta": 24.518, "th": 20.12, "zh": 5.633, "my": 22.797}, "bpb_mean": 1.88469, "t": 182.2} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "parity_a", "lm_loss": 1.38805, "bpb": 2.00253, "chunk_loss": 0.030063, "F": 0.13257, "G": 0.21258, "chunks": 135.75, "bytes_per_chunk": 7.543, "boundary_entropy": 0.33454, "router_noise": 0.0, "grad_norm": 0.6725, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.469, "alpha": 0.03, "t": 182.2} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "parity_a", "lm_loss": 1.44957, "bpb": 2.09128, "chunk_loss": 0.030551, "F": 0.15164, "G": 0.22171, "chunks": 155.28, "bytes_per_chunk": 6.594, "boundary_entropy": 0.31691, "router_noise": 0.0, "grad_norm": 0.6798, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.943, "alpha": 0.03, "t": 184.2} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "parity_a", "lm_loss": 1.40714, "bpb": 2.03007, "chunk_loss": 0.030476, "F": 0.16046, "G": 0.22171, "chunks": 164.31, "bytes_per_chunk": 6.232, "boundary_entropy": 0.30865, "router_noise": 0.0, "grad_norm": 0.7073, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.4233, "alpha": 0.03, "t": 186.2} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "parity_a", "lm_loss": 1.47306, "bpb": 2.12517, "chunk_loss": 0.03039, "F": 0.15384, "G": 0.22815, "chunks": 157.53, "bytes_per_chunk": 6.5, "boundary_entropy": 0.32967, "router_noise": 0.0, "grad_norm": 0.6841, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.9185, "alpha": 0.03, "t": 188.2} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3418.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196403, "cps": 26.703125, "mean_p": 0.225489, "boundary_entropy": 0.240103, "bpb": 2.12067}, "de": {"chunks": 3831.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.186578, "cps": 29.929688, "mean_p": 0.217728, "boundary_entropy": 0.247743, "bpb": 2.333224}, "vi": {"chunks": 3133.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.169948, "cps": 24.476562, "mean_p": 0.200973, "boundary_entropy": 0.308867, "bpb": 1.993467}, "tr": {"chunks": 3358.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185494, "cps": 26.234375, "mean_p": 0.208417, "boundary_entropy": 0.2461, "bpb": 2.328287}, "fi": {"chunks": 3505.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.186327, "cps": 27.382812, "mean_p": 0.220031, "boundary_entropy": 0.25983, "bpb": 2.520739}, "ru": {"chunks": 3172.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.167388, "cps": 24.78125, "mean_p": 0.163311, "boundary_entropy": 0.293503, "bpb": 1.506601}, "el": {"chunks": 2701.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.131608, "cps": 21.101562, "mean_p": 0.189139, "boundary_entropy": 0.383664, "bpb": 1.620572}, "hi": {"chunks": 2668.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.158574, "cps": 20.84375, "mean_p": 0.235144, "boundary_entropy": 0.473957, "bpb": 1.286868}, "ta": {"chunks": 2593.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.1387, "cps": 20.257812, "mean_p": 0.199148, "boundary_entropy": 0.436732, "bpb": 1.043448}, "th": {"chunks": 2816.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.17507, "cps": 22.0, "mean_p": 0.177344, "boundary_entropy": 0.386666, "bpb": 1.211113}, "zh": {"chunks": 3020.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.511344, "cps": 23.59375, "mean_p": 0.299934, "boundary_entropy": 0.466813, "bpb": 3.248021}, "my": {"chunks": 2854.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.150623, "cps": 22.296875, "mean_p": 0.27358, "boundary_entropy": 0.548369, "bpb": 1.151322}}, "bpb": {"en": 2.12067, "de": 2.33322, "vi": 1.99347, "tr": 2.32829, "fi": 2.52074, "ru": 1.5066, "el": 1.62057, "hi": 1.28687, "ta": 1.04345, "th": 1.21111, "zh": 3.24802, "my": 1.15132}, "gini_cpc": 0.18187, "gini_cps": 0.067323, "bpic": 10.353152, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.949974191312351, "vi": 0.8653052022961767, "tr": 0.9444570269461592, "fi": 0.948698471093022, "ru": 0.8522676934620532, "el": 0.6700941103998206, "hi": 0.8073889738531349, "ta": 0.7062022697826634, "th": 0.8913815629483001, "zh": 2.6035478394991403, "my": 0.7669069164439204}, "t": 191.1} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.12908, "de": 2.34221, "vi": 1.9948, "tr": 2.33582, "fi": 2.53315, "ru": 1.48492, "el": 1.61372, "hi": 1.28354, "ta": 1.04886, "th": 1.22823, "zh": 3.24997, "my": 1.16038}, "bytes_per_chunk": {"en": 5.196, "de": 5.342, "vi": 7.917, "tr": 5.518, "fi": 5.191, "ru": 10.551, "el": 13.19, "hi": 15.73, "ta": 19.912, "th": 16.063, "zh": 5.401, "my": 18.884}, "bpb_mean": 1.86706, "t": 192.3} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "parity_a", "lm_loss": 1.39788, "bpb": 2.01671, "chunk_loss": 0.030472, "F": 0.14221, "G": 0.22184, "chunks": 145.62, "bytes_per_chunk": 7.032, "boundary_entropy": 0.33971, "router_noise": 0.0, "grad_norm": 0.6605, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.6778, "alpha": 0.03, "t": 192.4} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "parity_a", "lm_loss": 1.26409, "bpb": 1.8237, "chunk_loss": 0.030075, "F": 0.13354, "G": 0.20609, "chunks": 136.75, "bytes_per_chunk": 7.488, "boundary_entropy": 0.30933, "router_noise": 0.0, "grad_norm": 0.6457, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.8297, "alpha": 0.03, "t": 194.4} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "parity_a", "lm_loss": 1.45038, "bpb": 2.09246, "chunk_loss": 0.030397, "F": 0.15652, "G": 0.21778, "chunks": 160.28, "bytes_per_chunk": 6.389, "boundary_entropy": 0.29941, "router_noise": 0.0, "grad_norm": 0.7968, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.7543, "alpha": 0.03, "t": 196.3} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "parity_a", "lm_loss": 1.37693, "bpb": 1.98649, "chunk_loss": 0.030294, "F": 0.13812, "G": 0.21588, "chunks": 141.44, "bytes_per_chunk": 7.24, "boundary_entropy": 0.32339, "router_noise": 0.0, "grad_norm": 0.664, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.6795, "alpha": 0.03, "t": 198.3} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3415.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196231, "cps": 26.679688, "mean_p": 0.225044, "boundary_entropy": 0.235534, "bpb": 2.104013}, "de": {"chunks": 3887.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189305, "cps": 30.367188, "mean_p": 0.220842, "boundary_entropy": 0.24416, "bpb": 2.318509}, "vi": {"chunks": 3261.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.176892, "cps": 25.476562, "mean_p": 0.209496, "boundary_entropy": 0.317627, "bpb": 1.991738}, "tr": {"chunks": 3636.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.200851, "cps": 28.40625, "mean_p": 0.218024, "boundary_entropy": 0.246757, "bpb": 2.309156}, "fi": {"chunks": 3511.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.186646, "cps": 27.429688, "mean_p": 0.22134, "boundary_entropy": 0.259714, "bpb": 2.481392}, "ru": {"chunks": 3175.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.167546, "cps": 24.804688, "mean_p": 0.162061, "boundary_entropy": 0.287549, "bpb": 1.490732}, "el": {"chunks": 2786.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.13575, "cps": 21.765625, "mean_p": 0.19594, "boundary_entropy": 0.394332, "bpb": 1.609242}, "hi": {"chunks": 2362.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.140386, "cps": 18.453125, "mean_p": 0.226699, "boundary_entropy": 0.463897, "bpb": 1.283149}, "ta": {"chunks": 2725.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.145761, "cps": 21.289062, "mean_p": 0.216332, "boundary_entropy": 0.462319, "bpb": 1.040034}, "th": {"chunks": 3054.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.189866, "cps": 23.859375, "mean_p": 0.178341, "boundary_entropy": 0.385252, "bpb": 1.195226}, "zh": {"chunks": 2755.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.466475, "cps": 21.523438, "mean_p": 0.293242, "boundary_entropy": 0.468644, "bpb": 3.249555}, "my": {"chunks": 2910.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.153578, "cps": 22.734375, "mean_p": 0.237488, "boundary_entropy": 0.505512, "bpb": 1.149535}}, "bpb": {"en": 2.10401, "de": 2.31851, "vi": 1.99174, "tr": 2.30916, "fi": 2.48139, "ru": 1.49073, "el": 1.60924, "hi": 1.28315, "ta": 1.04003, "th": 1.19523, "zh": 3.24955, "my": 1.14954}, "gini_cpc": 0.169977, "gini_cps": 0.076903, "bpic": 10.240441, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9647072573029781, "vi": 0.9014488085040987, "tr": 1.023544514959906, "fi": 0.9511573280423055, "ru": 0.8538231533240361, "el": 0.6917890430256124, "hi": 0.7154153143961066, "ta": 0.7428043686566, "th": 0.9675677392840287, "zh": 2.377177290484055, "my": 0.782641774312745}, "t": 201.2} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.11444, "de": 2.32781, "vi": 1.99378, "tr": 2.31474, "fi": 2.49531, "ru": 1.46759, "el": 1.60148, "hi": 1.27945, "ta": 1.04498, "th": 1.21177, "zh": 3.24959, "my": 1.15875}, "bytes_per_chunk": {"en": 5.156, "de": 5.213, "vi": 7.547, "tr": 5.137, "fi": 5.179, "ru": 10.473, "el": 12.56, "hi": 18.242, "ta": 18.44, "th": 14.814, "zh": 5.935, "my": 17.781}, "bpb_mean": 1.85497, "t": 202.5} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "parity_a", "lm_loss": 1.63609, "bpb": 2.36039, "chunk_loss": 0.030602, "F": 0.16797, "G": 0.23916, "chunks": 172.0, "bytes_per_chunk": 5.953, "boundary_entropy": 0.33418, "router_noise": 0.0, "grad_norm": 0.7499, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 7.8431, "alpha": 0.03, "t": 202.5} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "parity_a", "lm_loss": 1.43843, "bpb": 2.07522, "chunk_loss": 0.030468, "F": 0.15311, "G": 0.21761, "chunks": 156.78, "bytes_per_chunk": 6.531, "boundary_entropy": 0.30769, "router_noise": 0.0, "grad_norm": 0.7562, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.856, "alpha": 0.03, "t": 204.5} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "parity_a", "lm_loss": 1.36427, "bpb": 1.96823, "chunk_loss": 0.030669, "F": 0.13763, "G": 0.21334, "chunks": 140.94, "bytes_per_chunk": 7.266, "boundary_entropy": 0.32511, "router_noise": 0.0, "grad_norm": 0.7019, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 10.295, "alpha": 0.03, "t": 206.5} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "parity_a", "lm_loss": 1.39945, "bpb": 2.01897, "chunk_loss": 0.030477, "F": 0.15335, "G": 0.21951, "chunks": 157.03, "bytes_per_chunk": 6.521, "boundary_entropy": 0.31545, "router_noise": 0.0, "grad_norm": 0.6903, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.0563, "alpha": 0.03, "t": 208.5} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3455.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198529, "cps": 26.992188, "mean_p": 0.225087, "boundary_entropy": 0.229039, "bpb": 2.092278}, "de": {"chunks": 3847.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187357, "cps": 30.054688, "mean_p": 0.215589, "boundary_entropy": 0.235809, "bpb": 2.296659}, "vi": {"chunks": 3108.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.168592, "cps": 24.28125, "mean_p": 0.205619, "boundary_entropy": 0.310724, "bpb": 1.97644}, "tr": {"chunks": 3680.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203281, "cps": 28.75, "mean_p": 0.218484, "boundary_entropy": 0.241486, "bpb": 2.287514}, "fi": {"chunks": 3629.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.192919, "cps": 28.351562, "mean_p": 0.224466, "boundary_entropy": 0.255823, "bpb": 2.464252}, "ru": {"chunks": 3085.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.162797, "cps": 24.101562, "mean_p": 0.152509, "boundary_entropy": 0.275253, "bpb": 1.478806}, "el": {"chunks": 3009.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.146616, "cps": 23.507812, "mean_p": 0.198222, "boundary_entropy": 0.394629, "bpb": 1.575679}, "hi": {"chunks": 2699.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.160416, "cps": 21.085938, "mean_p": 0.231446, "boundary_entropy": 0.46901, "bpb": 1.268922}, "ta": {"chunks": 2919.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.156138, "cps": 22.804688, "mean_p": 0.218032, "boundary_entropy": 0.463138, "bpb": 1.032507}, "th": {"chunks": 2978.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.185141, "cps": 23.265625, "mean_p": 0.178561, "boundary_entropy": 0.386178, "bpb": 1.184786}, "zh": {"chunks": 3013.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.510159, "cps": 23.539062, "mean_p": 0.299181, "boundary_entropy": 0.463512, "bpb": 3.215394}, "my": {"chunks": 2656.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.140173, "cps": 20.75, "mean_p": 0.279074, "boundary_entropy": 0.552803, "bpb": 1.140923}}, "bpb": {"en": 2.09228, "de": 2.29666, "vi": 1.97644, "tr": 2.28751, "fi": 2.46425, "ru": 1.47881, "el": 1.57568, "hi": 1.26892, "ta": 1.03251, "th": 1.18479, "zh": 3.21539, "my": 1.14092}, "gini_cpc": 0.177206, "gini_cps": 0.064897, "bpic": 10.078812, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9437258423364655, "vi": 0.8492077259758442, "tr": 1.0239372339002746, "fi": 0.9717424012384293, "ru": 0.8200154263829332, "el": 0.738511752688893, "hi": 0.8080232924697286, "ta": 0.7864745869117671, "th": 0.9325662555157636, "zh": 2.569696053413757, "my": 0.7060586258316233}, "t": 211.4} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.10214, "de": 2.30478, "vi": 1.97618, "tr": 2.29303, "fi": 2.47744, "ru": 1.45642, "el": 1.56764, "hi": 1.26426, "ta": 1.03733, "th": 1.1997, "zh": 3.21498, "my": 1.14926}, "bytes_per_chunk": {"en": 5.115, "de": 5.298, "vi": 7.928, "tr": 5.112, "fi": 5.066, "ru": 10.727, "el": 11.797, "hi": 16.634, "ta": 17.088, "th": 15.077, "zh": 5.491, "my": 21.11}, "bpb_mean": 1.83693, "t": 212.6} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "parity_a", "lm_loss": 1.47158, "bpb": 2.12304, "chunk_loss": 0.030478, "F": 0.14868, "G": 0.22726, "chunks": 152.25, "bytes_per_chunk": 6.726, "boundary_entropy": 0.33882, "router_noise": 0.0, "grad_norm": 0.6841, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 9.1031, "alpha": 0.03, "t": 212.6} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "parity_a", "lm_loss": 1.46945, "bpb": 2.11997, "chunk_loss": 0.030705, "F": 0.15375, "G": 0.22341, "chunks": 157.44, "bytes_per_chunk": 6.504, "boundary_entropy": 0.32543, "router_noise": 0.0, "grad_norm": 0.6873, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.1437, "alpha": 0.03, "t": 214.7} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "parity_a", "lm_loss": 1.46902, "bpb": 2.11935, "chunk_loss": 0.030559, "F": 0.16412, "G": 0.22265, "chunks": 168.06, "bytes_per_chunk": 6.093, "boundary_entropy": 0.28887, "router_noise": 0.0, "grad_norm": 0.7515, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.4141, "alpha": 0.03, "t": 216.7} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "parity_a", "lm_loss": 1.54655, "bpb": 2.23119, "chunk_loss": 0.031235, "F": 0.16226, "G": 0.23443, "chunks": 166.16, "bytes_per_chunk": 6.163, "boundary_entropy": 0.33746, "router_noise": 0.0, "grad_norm": 0.6803, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 9.0542, "alpha": 0.03, "t": 218.7} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3434.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197322, "cps": 26.828125, "mean_p": 0.224593, "boundary_entropy": 0.228291, "bpb": 2.075675}, "de": {"chunks": 3892.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189549, "cps": 30.40625, "mean_p": 0.216971, "boundary_entropy": 0.2351, "bpb": 2.287905}, "vi": {"chunks": 3105.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.16843, "cps": 24.257812, "mean_p": 0.201001, "boundary_entropy": 0.310138, "bpb": 1.966545}, "tr": {"chunks": 3641.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201127, "cps": 28.445312, "mean_p": 0.215181, "boundary_entropy": 0.235939, "bpb": 2.273646}, "fi": {"chunks": 3686.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.195949, "cps": 28.796875, "mean_p": 0.22875, "boundary_entropy": 0.257898, "bpb": 2.434871}, "ru": {"chunks": 3094.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.163272, "cps": 24.171875, "mean_p": 0.148856, "boundary_entropy": 0.262887, "bpb": 1.469341}, "el": {"chunks": 2875.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.140087, "cps": 22.460938, "mean_p": 0.189616, "boundary_entropy": 0.381549, "bpb": 1.559942}, "hi": {"chunks": 2562.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.152273, "cps": 20.015625, "mean_p": 0.215587, "boundary_entropy": 0.44713, "bpb": 1.264384}, "ta": {"chunks": 2310.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.123562, "cps": 18.046875, "mean_p": 0.212073, "boundary_entropy": 0.457753, "bpb": 1.031161}, "th": {"chunks": 2738.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170221, "cps": 21.390625, "mean_p": 0.172977, "boundary_entropy": 0.373416, "bpb": 1.173102}, "zh": {"chunks": 2980.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.504572, "cps": 23.28125, "mean_p": 0.293425, "boundary_entropy": 0.454461, "bpb": 3.196568}, "my": {"chunks": 4095.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.216118, "cps": 31.992188, "mean_p": 0.305599, "boundary_entropy": 0.569117, "bpb": 1.136624}}, "bpb": {"en": 2.07567, "de": 2.2879, "vi": 1.96654, "tr": 2.27365, "fi": 2.43487, "ru": 1.46934, "el": 1.55994, "hi": 1.26438, "ta": 1.03116, "th": 1.1731, "zh": 3.19657, "my": 1.13662}, "gini_cpc": 0.188839, "gini_cps": 0.09425, "bpic": 9.991175, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9606036971438026, "vi": 0.8535761894765076, "tr": 1.0192810659485514, "fi": 0.9930412220238055, "ru": 0.8274369758875624, "el": 0.7099386710212849, "hi": 0.7716989012073132, "ta": 0.626196073031465, "th": 0.8626531409197896, "zh": 2.557093750449924, "my": 1.0952527679285082}, "t": 221.6} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.08609, "de": 2.2946, "vi": 1.96837, "tr": 2.27939, "fi": 2.44778, "ru": 1.44631, "el": 1.55355, "hi": 1.26081, "ta": 1.03678, "th": 1.18849, "zh": 3.19531, "my": 1.14483}, "bytes_per_chunk": {"en": 5.158, "de": 5.279, "vi": 7.861, "tr": 5.181, "fi": 4.926, "ru": 10.709, "el": 12.444, "hi": 16.887, "ta": 21.699, "th": 16.412, "zh": 5.587, "my": 13.885}, "bpb_mean": 1.82519, "t": 222.8} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "parity_a", "lm_loss": 1.34368, "bpb": 1.93852, "chunk_loss": 0.030251, "F": 0.15421, "G": 0.21715, "chunks": 157.91, "bytes_per_chunk": 6.485, "boundary_entropy": 0.29782, "router_noise": 0.0, "grad_norm": 0.7006, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.083, "n_target_mean": 8.7169, "alpha": 0.03, "t": 222.8} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "parity_a", "lm_loss": 1.53561, "bpb": 2.21541, "chunk_loss": 0.030536, "F": 0.16516, "G": 0.23039, "chunks": 169.12, "bytes_per_chunk": 6.055, "boundary_entropy": 0.31491, "router_noise": 0.0, "grad_norm": 0.6958, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.3032, "alpha": 0.03, "t": 224.8} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "parity_a", "lm_loss": 1.40526, "bpb": 2.02736, "chunk_loss": 0.030508, "F": 0.15622, "G": 0.21908, "chunks": 159.97, "bytes_per_chunk": 6.401, "boundary_entropy": 0.3001, "router_noise": 0.0, "grad_norm": 0.7028, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.04, "n_target_mean": 8.5769, "alpha": 0.03, "t": 226.9} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "parity_a", "lm_loss": 1.38987, "bpb": 2.00515, "chunk_loss": 0.030474, "F": 0.16339, "G": 0.21544, "chunks": 167.31, "bytes_per_chunk": 6.12, "boundary_entropy": 0.27016, "router_noise": 0.0, "grad_norm": 0.7556, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.041, "n_target_mean": 8.3506, "alpha": 0.03, "t": 228.9} |
| {"event": "probe", "step": 4200, "phase": 3, "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": 0.222645, "boundary_entropy": 0.21876, "bpb": 2.066132}, "de": {"chunks": 3881.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.189013, "cps": 30.320312, "mean_p": 0.215124, "boundary_entropy": 0.230354, "bpb": 2.264092}, "vi": {"chunks": 3078.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.166965, "cps": 24.046875, "mean_p": 0.202664, "boundary_entropy": 0.30497, "bpb": 1.956701}, "tr": {"chunks": 3455.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190852, "cps": 26.992188, "mean_p": 0.208184, "boundary_entropy": 0.227914, "bpb": 2.254656}, "fi": {"chunks": 3651.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.194089, "cps": 28.523438, "mean_p": 0.224934, "boundary_entropy": 0.248528, "bpb": 2.425108}, "ru": {"chunks": 3069.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.161953, "cps": 23.976562, "mean_p": 0.149556, "boundary_entropy": 0.258063, "bpb": 1.455991}, "el": {"chunks": 2344.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.114213, "cps": 18.3125, "mean_p": 0.171356, "boundary_entropy": 0.359173, "bpb": 1.556094}, "hi": {"chunks": 2627.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.156137, "cps": 20.523438, "mean_p": 0.226097, "boundary_entropy": 0.45963, "bpb": 1.252913}, "ta": {"chunks": 2413.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.129072, "cps": 18.851562, "mean_p": 0.205584, "boundary_entropy": 0.446618, "bpb": 1.015352}, "th": {"chunks": 2744.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170594, "cps": 21.4375, "mean_p": 0.169309, "boundary_entropy": 0.369033, "bpb": 1.157965}, "zh": {"chunks": 3167.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.536234, "cps": 24.742188, "mean_p": 0.29519, "boundary_entropy": 0.449512, "bpb": 3.190023}, "my": {"chunks": 2108.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.111252, "cps": 16.46875, "mean_p": 0.262951, "boundary_entropy": 0.535712, "bpb": 1.138483}}, "bpb": {"en": 2.06613, "de": 2.26409, "vi": 1.9567, "tr": 2.25466, "fi": 2.42511, "ru": 1.45599, "el": 1.55609, "hi": 1.25291, "ta": 1.01535, "th": 1.15796, "zh": 3.19002, "my": 1.13848}, "gini_cpc": 0.220776, "gini_cps": 0.101736, "bpic": 10.665916, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9548301622429879, "vi": 0.8434519905382857, "tr": 0.9641228907284048, "fi": 0.9804712047015076, "ru": 0.8181304709130666, "el": 0.5769679060916963, "hi": 0.7887509472780717, "ta": 0.6520287511970285, "th": 0.8617830258608906, "zh": 2.708878427733574, "my": 0.5620075170208405}, "t": 231.9} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.07728, "de": 2.2731, "vi": 1.95796, "tr": 2.26218, "fi": 2.43472, "ru": 1.43491, "el": 1.54853, "hi": 1.24778, "ta": 1.02157, "th": 1.17371, "zh": 3.18706, "my": 1.14743}, "bytes_per_chunk": {"en": 5.155, "de": 5.256, "vi": 8.001, "tr": 5.487, "fi": 5.022, "ru": 10.771, "el": 14.828, "hi": 16.569, "ta": 19.906, "th": 16.756, "zh": 5.258, "my": 26.152}, "bpb_mean": 1.81385, "t": 233.1} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 233.1} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.20543, "bpb": 1.73907, "chunk_loss": 0.029866, "F": 0.12662, "G": 0.21671, "chunks": 129.66, "bytes_per_chunk": 7.898, "boundary_entropy": 0.33698, "router_noise": 0.0, "grad_norm": 0.6517, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 233.2} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.25077, "bpb": 1.80448, "chunk_loss": 0.029973, "F": 0.16229, "G": 0.23629, "chunks": 166.19, "bytes_per_chunk": 6.162, "boundary_entropy": 0.32635, "router_noise": 0.0, "grad_norm": 0.6946, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 235.2} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.47869, "bpb": 2.13329, "chunk_loss": 0.03032, "F": 0.18439, "G": 0.24616, "chunks": 188.81, "bytes_per_chunk": 5.423, "boundary_entropy": 0.31425, "router_noise": 0.0, "grad_norm": 0.7454, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 237.2} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.46876, "bpb": 2.11897, "chunk_loss": 0.030254, "F": 0.17792, "G": 0.24506, "chunks": 182.19, "bytes_per_chunk": 5.621, "boundary_entropy": 0.32098, "router_noise": 0.0, "grad_norm": 0.7579, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 239.3} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3491.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200598, "cps": 27.273438, "mean_p": 0.229334, "boundary_entropy": 0.233822, "bpb": 2.05099}, "de": {"chunks": 3938.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.191789, "cps": 30.765625, "mean_p": 0.223558, "boundary_entropy": 0.245472, "bpb": 2.254809}, "vi": {"chunks": 3595.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.195009, "cps": 28.085938, "mean_p": 0.230837, "boundary_entropy": 0.344662, "bpb": 1.944175}, "tr": {"chunks": 3685.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203557, "cps": 28.789062, "mean_p": 0.222559, "boundary_entropy": 0.251467, "bpb": 2.229285}, "fi": {"chunks": 3759.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19983, "cps": 29.367188, "mean_p": 0.236298, "boundary_entropy": 0.270941, "bpb": 2.404433}, "ru": {"chunks": 4246.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.224063, "cps": 33.171875, "mean_p": 0.24522, "boundary_entropy": 0.404064, "bpb": 1.434581}, "el": {"chunks": 5506.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.268284, "cps": 43.015625, "mean_p": 0.283391, "boundary_entropy": 0.487481, "bpb": 1.533402}, "hi": {"chunks": 7171.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.426211, "cps": 56.023438, "mean_p": 0.331152, "boundary_entropy": 0.568957, "bpb": 1.230696}, "ta": {"chunks": 8353.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.446804, "cps": 65.257812, "mean_p": 0.340787, "boundary_entropy": 0.581874, "bpb": 1.004902}, "th": {"chunks": 6363.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.395586, "cps": 49.710938, "mean_p": 0.286855, "boundary_entropy": 0.515308, "bpb": 1.148268}, "zh": {"chunks": 2850.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.48256, "cps": 22.265625, "mean_p": 0.285831, "boundary_entropy": 0.444481, "bpb": 3.165651}, "my": {"chunks": 10173.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.53689, "cps": 79.476562, "mean_p": 0.39213, "boundary_entropy": 0.629365, "bpb": 1.117892}}, "bpb": {"en": 2.05099, "de": 2.25481, "vi": 1.94418, "tr": 2.22928, "fi": 2.40443, "ru": 1.43458, "el": 1.5334, "hi": 1.2307, "ta": 1.0049, "th": 1.14827, "zh": 3.16565, "my": 1.11789}, "gini_cpc": 0.217908, "gini_cps": 0.222137, "bpic": 6.079217, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9560873586856423, "vi": 0.9721427108041858, "tr": 1.0147550286464324, "fi": 0.996172878668038, "ru": 1.1169790996675206, "el": 1.3374256043198516, "hi": 2.124706375085499, "ta": 2.227364447409144, "th": 1.972037319286407, "zh": 2.4056125940605044, "my": 2.6764549627205687}, "t": 242.2} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.06307, "de": 2.26196, "vi": 1.9453, "tr": 2.23652, "fi": 2.41506, "ru": 1.41194, "el": 1.52559, "hi": 1.2261, "ta": 1.01113, "th": 1.16487, "zh": 3.16589, "my": 1.12656}, "bytes_per_chunk": {"en": 5.058, "de": 5.16, "vi": 6.654, "tr": 5.185, "fi": 4.87, "ru": 8.008, "el": 6.743, "hi": 5.829, "ta": 6.461, "th": 6.999, "zh": 5.669, "my": 5.986}, "bpb_mean": 1.79617, "t": 243.4} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.48395, "bpb": 2.14089, "chunk_loss": 0.030279, "F": 0.18341, "G": 0.25368, "chunks": 187.81, "bytes_per_chunk": 5.452, "boundary_entropy": 0.3393, "router_noise": 0.0, "grad_norm": 0.7414, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 243.5} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.48792, "bpb": 2.14661, "chunk_loss": 0.030077, "F": 0.17371, "G": 0.26334, "chunks": 177.88, "bytes_per_chunk": 5.757, "boundary_entropy": 0.37269, "router_noise": 0.0, "grad_norm": 0.6955, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 245.5} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.41582, "bpb": 2.0426, "chunk_loss": 0.03018, "F": 0.18265, "G": 0.26056, "chunks": 187.03, "bytes_per_chunk": 5.475, "boundary_entropy": 0.33536, "router_noise": 0.0, "grad_norm": 0.7547, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 247.5} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.22603, "bpb": 1.76879, "chunk_loss": 0.030077, "F": 0.17664, "G": 0.27712, "chunks": 180.88, "bytes_per_chunk": 5.661, "boundary_entropy": 0.3823, "router_noise": 0.0, "grad_norm": 0.7092, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 249.5} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3476.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199736, "cps": 27.15625, "mean_p": 0.229844, "boundary_entropy": 0.234696, "bpb": 2.042524}, "de": {"chunks": 3986.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.194127, "cps": 31.140625, "mean_p": 0.226357, "boundary_entropy": 0.245929, "bpb": 2.241472}, "vi": {"chunks": 3757.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.203797, "cps": 29.351562, "mean_p": 0.249476, "boundary_entropy": 0.365158, "bpb": 1.939317}, "tr": {"chunks": 3686.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.203613, "cps": 28.796875, "mean_p": 0.226783, "boundary_entropy": 0.257859, "bpb": 2.218391}, "fi": {"chunks": 3759.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19983, "cps": 29.367188, "mean_p": 0.236762, "boundary_entropy": 0.271842, "bpb": 2.388775}, "ru": {"chunks": 4818.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.254248, "cps": 37.640625, "mean_p": 0.314609, "boundary_entropy": 0.498333, "bpb": 1.423439}, "el": {"chunks": 5497.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.267846, "cps": 42.945312, "mean_p": 0.309263, "boundary_entropy": 0.529338, "bpb": 1.520748}, "hi": {"chunks": 7586.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.450877, "cps": 59.265625, "mean_p": 0.376426, "boundary_entropy": 0.613941, "bpb": 1.223962}, "ta": {"chunks": 10213.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.546296, "cps": 79.789062, "mean_p": 0.383918, "boundary_entropy": 0.613907, "bpb": 1.003053}, "th": {"chunks": 7891.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.490581, "cps": 61.648438, "mean_p": 0.341134, "boundary_entropy": 0.577112, "bpb": 1.138732}, "zh": {"chunks": 2921.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.494582, "cps": 22.820312, "mean_p": 0.287246, "boundary_entropy": 0.445535, "bpb": 3.148932}, "my": {"chunks": 9042.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.477201, "cps": 70.640625, "mean_p": 0.413378, "boundary_entropy": 0.648039, "bpb": 1.111156}}, "bpb": {"en": 2.04252, "de": 2.24147, "vi": 1.93932, "tr": 2.21839, "fi": 2.38877, "ru": 1.42344, "el": 1.52075, "hi": 1.22396, "ta": 1.00305, "th": 1.13873, "zh": 3.14893, "my": 1.11116}, "gini_cpc": 0.22358, "gini_cps": 0.231285, "bpic": 5.75971, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9719171367296306, "vi": 1.0203341101740542, "tr": 1.01941056886667, "fi": 1.0004716684206332, "ru": 1.272922414081026, "el": 1.3410014400750634, "hi": 2.257366719273917, "ta": 2.7350937356406564, "th": 2.4561525132772424, "zh": 2.4761814553825037, "my": 2.389161342301248}, "t": 252.4} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.05279, "de": 2.24811, "vi": 1.941, "tr": 2.22507, "fi": 2.39989, "ru": 1.40025, "el": 1.51237, "hi": 1.21971, "ta": 1.00926, "th": 1.15445, "zh": 3.14671, "my": 1.12016}, "bytes_per_chunk": {"en": 5.066, "de": 5.105, "vi": 6.531, "tr": 5.134, "fi": 4.877, "ru": 7.083, "el": 6.618, "hi": 5.577, "ta": 5.156, "th": 5.795, "zh": 5.513, "my": 6.63}, "bpb_mean": 1.78581, "t": 253.7} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.30487, "bpb": 1.88253, "chunk_loss": 0.030121, "F": 0.17551, "G": 0.2891, "chunks": 179.72, "bytes_per_chunk": 5.698, "boundary_entropy": 0.41514, "router_noise": 0.0, "grad_norm": 0.7143, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 253.7} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.40667, "bpb": 2.02939, "chunk_loss": 0.030258, "F": 0.17999, "G": 0.28493, "chunks": 184.31, "bytes_per_chunk": 5.556, "boundary_entropy": 0.39987, "router_noise": 0.0, "grad_norm": 0.7053, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 255.7} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.32335, "bpb": 1.90919, "chunk_loss": 0.030245, "F": 0.17981, "G": 0.27002, "chunks": 184.12, "bytes_per_chunk": 5.561, "boundary_entropy": 0.36178, "router_noise": 0.0, "grad_norm": 0.7554, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 257.7} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.33674, "bpb": 1.92851, "chunk_loss": 0.030299, "F": 0.18405, "G": 0.27569, "chunks": 188.47, "bytes_per_chunk": 5.433, "boundary_entropy": 0.35738, "router_noise": 0.0, "grad_norm": 0.7368, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 259.7} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3469.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.199333, "cps": 27.101562, "mean_p": 0.228613, "boundary_entropy": 0.233968, "bpb": 2.035312}, "de": {"chunks": 4003.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.194954, "cps": 31.273438, "mean_p": 0.227872, "boundary_entropy": 0.246666, "bpb": 2.232124}, "vi": {"chunks": 3835.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.208028, "cps": 29.960938, "mean_p": 0.253948, "boundary_entropy": 0.372317, "bpb": 1.933403}, "tr": {"chunks": 3609.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.199359, "cps": 28.195312, "mean_p": 0.223938, "boundary_entropy": 0.255473, "bpb": 2.208296}, "fi": {"chunks": 3649.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.193982, "cps": 28.507812, "mean_p": 0.234107, "boundary_entropy": 0.270667, "bpb": 2.372313}, "ru": {"chunks": 4980.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.262797, "cps": 38.90625, "mean_p": 0.346743, "boundary_entropy": 0.532994, "bpb": 1.411961}, "el": {"chunks": 6466.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.315061, "cps": 50.515625, "mean_p": 0.334456, "boundary_entropy": 0.549199, "bpb": 1.510533}, "hi": {"chunks": 6993.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.415632, "cps": 54.632812, "mean_p": 0.395682, "boundary_entropy": 0.630641, "bpb": 1.216163}, "ta": {"chunks": 8993.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.481038, "cps": 70.257812, "mean_p": 0.38484, "boundary_entropy": 0.62045, "bpb": 0.994942}, "th": {"chunks": 8652.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.537892, "cps": 67.59375, "mean_p": 0.363269, "boundary_entropy": 0.598022, "bpb": 1.138311}, "zh": {"chunks": 2809.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.475618, "cps": 21.945312, "mean_p": 0.28048, "boundary_entropy": 0.436119, "bpb": 3.1431}, "my": {"chunks": 10556.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.557104, "cps": 82.46875, "mean_p": 0.433145, "boundary_entropy": 0.658167, "bpb": 1.108314}}, "bpb": {"en": 2.03531, "de": 2.23212, "vi": 1.9334, "tr": 2.2083, "fi": 2.37231, "ru": 1.41196, "el": 1.51053, "hi": 1.21616, "ta": 0.99494, "th": 1.13831, "zh": 3.1431, "my": 1.10831}, "gini_cpc": 0.229102, "gini_cps": 0.239931, "bpic": 5.642677, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9780318619485437, "vi": 1.043619166951455, "tr": 1.0001293007898957, "fi": 0.9731545113405934, "ru": 1.3183780046563949, "el": 1.58057342438165, "hi": 2.0851066611211935, "ta": 2.413231270649117, "th": 2.6984555331008973, "zh": 2.386042180770575, "my": 2.7948327637661428}, "t": 262.7} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.04603, "de": 2.23927, "vi": 1.93414, "tr": 2.21401, "fi": 2.38387, "ru": 1.38941, "el": 1.50218, "hi": 1.2106, "ta": 1.00103, "th": 1.15336, "zh": 3.1392, "my": 1.11659}, "bytes_per_chunk": {"en": 5.091, "de": 5.07, "vi": 6.37, "tr": 5.235, "fi": 4.986, "ru": 6.788, "el": 5.673, "hi": 6.226, "ta": 5.71, "th": 5.321, "zh": 5.757, "my": 5.743}, "bpb_mean": 1.77748, "t": 263.9} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.2597, "bpb": 1.81736, "chunk_loss": 0.030257, "F": 0.18146, "G": 0.28697, "chunks": 185.81, "bytes_per_chunk": 5.511, "boundary_entropy": 0.39707, "router_noise": 0.0, "grad_norm": 0.6846, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 264.0} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.49065, "bpb": 2.15055, "chunk_loss": 0.030259, "F": 0.18451, "G": 0.26091, "chunks": 188.94, "bytes_per_chunk": 5.42, "boundary_entropy": 0.34048, "router_noise": 0.0, "grad_norm": 0.7609, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 266.0} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.40581, "bpb": 2.02815, "chunk_loss": 0.030224, "F": 0.17847, "G": 0.28969, "chunks": 182.75, "bytes_per_chunk": 5.603, "boundary_entropy": 0.41122, "router_noise": 0.0, "grad_norm": 0.7033, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 268.0} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.52051, "bpb": 2.19364, "chunk_loss": 0.030358, "F": 0.18811, "G": 0.27111, "chunks": 192.62, "bytes_per_chunk": 5.316, "boundary_entropy": 0.3473, "router_noise": 0.0, "grad_norm": 0.7726, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 270.0} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3490.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20054, "cps": 27.265625, "mean_p": 0.228061, "boundary_entropy": 0.229208, "bpb": 2.02658}, "de": {"chunks": 3952.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.192471, "cps": 30.875, "mean_p": 0.223451, "boundary_entropy": 0.240744, "bpb": 2.217497}, "vi": {"chunks": 3868.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.209818, "cps": 30.21875, "mean_p": 0.25954, "boundary_entropy": 0.377598, "bpb": 1.927111}, "tr": {"chunks": 3649.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.201569, "cps": 28.507812, "mean_p": 0.22252, "boundary_entropy": 0.24988, "bpb": 2.201409}, "fi": {"chunks": 3736.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.198607, "cps": 29.1875, "mean_p": 0.235076, "boundary_entropy": 0.265086, "bpb": 2.357827}, "ru": {"chunks": 5105.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.269393, "cps": 39.882812, "mean_p": 0.356524, "boundary_entropy": 0.540834, "bpb": 1.405452}, "el": {"chunks": 5495.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.267748, "cps": 42.929688, "mean_p": 0.32975, "boundary_entropy": 0.550105, "bpb": 1.503434}, "hi": {"chunks": 6544.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.388945, "cps": 51.125, "mean_p": 0.399581, "boundary_entropy": 0.635094, "bpb": 1.209299}, "ta": {"chunks": 7887.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.421878, "cps": 61.617188, "mean_p": 0.37542, "boundary_entropy": 0.61624, "bpb": 0.983367}, "th": {"chunks": 7478.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.464905, "cps": 58.421875, "mean_p": 0.35789, "boundary_entropy": 0.599826, "bpb": 1.124977}, "zh": {"chunks": 2833.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.479682, "cps": 22.132812, "mean_p": 0.277434, "boundary_entropy": 0.430032, "bpb": 3.1354}, "my": {"chunks": 10083.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.532141, "cps": 78.773438, "mean_p": 0.433653, "boundary_entropy": 0.657655, "bpb": 1.101261}}, "bpb": {"en": 2.02658, "de": 2.2175, "vi": 1.92711, "tr": 2.20141, "fi": 2.35783, "ru": 1.40545, "el": 1.50343, "hi": 1.2093, "ta": 0.98337, "th": 1.12498, "zh": 3.1354, "my": 1.10126}, "gini_cpc": 0.212425, "gini_cps": 0.216001, "bpic": 5.985356, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9597612732428629, "vi": 1.0462657692209054, "tr": 1.0051294668980286, "fi": 0.9903613367719521, "ru": 1.3433377686718933, "el": 1.3351361188586095, "hi": 1.9394871698804053, "ta": 2.1037060918330828, "th": 2.3182650550041726, "zh": 2.3919485016936783, "my": 2.6535366132494724}, "t": 273.0} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.03689, "de": 2.22474, "vi": 1.92778, "tr": 2.20809, "fi": 2.36942, "ru": 1.38158, "el": 1.49467, "hi": 1.20374, "ta": 0.98976, "th": 1.14151, "zh": 3.13259, "my": 1.11062}, "bytes_per_chunk": {"en": 5.048, "de": 5.133, "vi": 6.431, "tr": 5.222, "fi": 4.897, "ru": 6.597, "el": 6.636, "hi": 6.566, "ta": 6.571, "th": 6.197, "zh": 5.725, "my": 6.577}, "bpb_mean": 1.76845, "t": 274.2} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.28664, "bpb": 1.85623, "chunk_loss": 0.030108, "F": 0.17944, "G": 0.29249, "chunks": 183.75, "bytes_per_chunk": 5.573, "boundary_entropy": 0.41031, "router_noise": 0.0, "grad_norm": 0.7351, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 274.3} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.26562, "bpb": 1.8259, "chunk_loss": 0.030227, "F": 0.18222, "G": 0.28459, "chunks": 186.59, "bytes_per_chunk": 5.488, "boundary_entropy": 0.3801, "router_noise": 0.0, "grad_norm": 0.7237, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 276.2} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.30648, "bpb": 1.88485, "chunk_loss": 0.030088, "F": 0.1785, "G": 0.27668, "chunks": 182.78, "bytes_per_chunk": 5.602, "boundary_entropy": 0.36958, "router_noise": 0.0, "grad_norm": 0.7345, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 278.2} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.33308, "bpb": 1.92323, "chunk_loss": 0.03002, "F": 0.17813, "G": 0.28366, "chunks": 182.41, "bytes_per_chunk": 5.614, "boundary_entropy": 0.37766, "router_noise": 0.0, "grad_norm": 0.7354, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 280.2} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3482.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20008, "cps": 27.203125, "mean_p": 0.228174, "boundary_entropy": 0.225235, "bpb": 2.021305}, "de": {"chunks": 4024.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.195977, "cps": 31.4375, "mean_p": 0.225563, "boundary_entropy": 0.23973, "bpb": 2.211554}, "vi": {"chunks": 3871.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.209981, "cps": 30.242188, "mean_p": 0.257807, "boundary_entropy": 0.372135, "bpb": 1.920478}, "tr": {"chunks": 3670.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.202729, "cps": 28.671875, "mean_p": 0.22477, "boundary_entropy": 0.251598, "bpb": 2.188136}, "fi": {"chunks": 3712.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.197331, "cps": 29.0, "mean_p": 0.233592, "boundary_entropy": 0.261886, "bpb": 2.345119}, "ru": {"chunks": 5052.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.266596, "cps": 39.46875, "mean_p": 0.35944, "boundary_entropy": 0.54404, "bpb": 1.394979}, "el": {"chunks": 6134.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.298884, "cps": 47.921875, "mean_p": 0.340274, "boundary_entropy": 0.557941, "bpb": 1.499669}, "hi": {"chunks": 7495.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.445468, "cps": 58.554688, "mean_p": 0.409453, "boundary_entropy": 0.638381, "bpb": 1.194367}, "ta": {"chunks": 8076.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.431987, "cps": 63.09375, "mean_p": 0.377249, "boundary_entropy": 0.615666, "bpb": 0.982024}, "th": {"chunks": 7578.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.471122, "cps": 59.203125, "mean_p": 0.365188, "boundary_entropy": 0.606013, "bpb": 1.116628}, "zh": {"chunks": 2868.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.485608, "cps": 22.40625, "mean_p": 0.280857, "boundary_entropy": 0.433098, "bpb": 3.118979}, "my": {"chunks": 9085.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.47947, "cps": 70.976562, "mean_p": 0.43007, "boundary_entropy": 0.657037, "bpb": 1.099783}}, "bpb": {"en": 2.02131, "de": 2.21155, "vi": 1.92048, "tr": 2.18814, "fi": 2.34512, "ru": 1.39498, "el": 1.49967, "hi": 1.19437, "ta": 0.98202, "th": 1.11663, "zh": 3.11898, "my": 1.09978}, "gini_cpc": 0.204502, "gini_cps": 0.209875, "bpic": 5.900057, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9794920565102007, "vi": 1.0494829388756965, "tr": 1.0132365937297403, "fi": 0.9862600405112056, "ru": 1.3324455814221348, "el": 1.493820035144722, "hi": 2.2264447257088675, "ta": 2.159067371123782, "th": 2.354663702326651, "zh": 2.4270630457290583, "my": 2.3963867464230226}, "t": 283.2} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.03188, "de": 2.21782, "vi": 1.92211, "tr": 2.19433, "fi": 2.35753, "ru": 1.37243, "el": 1.49154, "hi": 1.18984, "ta": 0.98889, "th": 1.13314, "zh": 3.11686, "my": 1.10864}, "bytes_per_chunk": {"en": 5.043, "de": 5.076, "vi": 6.4, "tr": 5.16, "fi": 4.953, "ru": 6.725, "el": 6.018, "hi": 5.826, "ta": 6.6, "th": 6.133, "zh": 5.623, "my": 6.899}, "bpb_mean": 1.76042, "t": 284.4} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.2856, "bpb": 1.85474, "chunk_loss": 0.030175, "F": 0.1813, "G": 0.29643, "chunks": 185.66, "bytes_per_chunk": 5.516, "boundary_entropy": 0.40204, "router_noise": 0.0, "grad_norm": 0.7682, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.084, "alpha": 0.03, "t": 284.4} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.37606, "bpb": 1.98524, "chunk_loss": 0.030363, "F": 0.18539, "G": 0.29175, "chunks": 189.84, "bytes_per_chunk": 5.394, "boundary_entropy": 0.38738, "router_noise": 0.0, "grad_norm": 0.7293, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 286.4} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.37311, "bpb": 1.98098, "chunk_loss": 0.030244, "F": 0.18552, "G": 0.2893, "chunks": 189.97, "bytes_per_chunk": 5.39, "boundary_entropy": 0.38139, "router_noise": 0.0, "grad_norm": 0.7394, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 288.4} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.42301, "bpb": 2.05297, "chunk_loss": 0.030304, "F": 0.18582, "G": 0.28737, "chunks": 190.28, "bytes_per_chunk": 5.382, "boundary_entropy": 0.38152, "router_noise": 0.0, "grad_norm": 0.7774, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 290.4} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3487.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200368, "cps": 27.242188, "mean_p": 0.227291, "boundary_entropy": 0.225132, "bpb": 2.01323}, "de": {"chunks": 4030.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.196269, "cps": 31.484375, "mean_p": 0.227372, "boundary_entropy": 0.240263, "bpb": 2.206513}, "vi": {"chunks": 3802.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.206238, "cps": 29.703125, "mean_p": 0.259987, "boundary_entropy": 0.378415, "bpb": 1.914659}, "tr": {"chunks": 3723.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205657, "cps": 29.085938, "mean_p": 0.226131, "boundary_entropy": 0.251421, "bpb": 2.184323}, "fi": {"chunks": 3735.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.198554, "cps": 29.179688, "mean_p": 0.234399, "boundary_entropy": 0.262935, "bpb": 2.339437}, "ru": {"chunks": 5030.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.265435, "cps": 39.296875, "mean_p": 0.364011, "boundary_entropy": 0.548434, "bpb": 1.389078}, "el": {"chunks": 6080.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.296253, "cps": 47.5, "mean_p": 0.338067, "boundary_entropy": 0.555564, "bpb": 1.486861}, "hi": {"chunks": 6981.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.414918, "cps": 54.539062, "mean_p": 0.403811, "boundary_entropy": 0.636543, "bpb": 1.191777}, "ta": {"chunks": 7999.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.427868, "cps": 62.492188, "mean_p": 0.378462, "boundary_entropy": 0.617109, "bpb": 0.975417}, "th": {"chunks": 7420.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.461299, "cps": 57.96875, "mean_p": 0.369326, "boundary_entropy": 0.610613, "bpb": 1.110164}, "zh": {"chunks": 2896.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.490349, "cps": 22.625, "mean_p": 0.278213, "boundary_entropy": 0.425367, "bpb": 3.117906}, "my": {"chunks": 8534.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.450391, "cps": 66.671875, "mean_p": 0.42545, "boundary_entropy": 0.655372, "bpb": 1.096976}}, "bpb": {"en": 2.01323, "de": 2.20651, "vi": 1.91466, "tr": 2.18432, "fi": 2.33944, "ru": 1.38908, "el": 1.48686, "hi": 1.19178, "ta": 0.97542, "th": 1.11016, "zh": 3.11791, "my": 1.09698}, "gini_cpc": 0.199236, "gini_cps": 0.19954, "bpic": 6.023212, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9795459465245472, "vi": 1.029298034335599, "tr": 1.026395306936326, "fi": 0.9909480685850374, "ru": 1.3247408958869469, "el": 1.4785462255508557, "hi": 2.070783700528944, "ta": 2.1354155460548725, "th": 2.3022634183060506, "zh": 2.447244086229623, "my": 2.2478195043818774}, "t": 293.3} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.02539, "de": 2.21124, "vi": 1.91533, "tr": 2.18912, "fi": 2.3514, "ru": 1.36556, "el": 1.47861, "hi": 1.18634, "ta": 0.98166, "th": 1.1273, "zh": 3.11388, "my": 1.10629}, "bytes_per_chunk": {"en": 5.068, "de": 5.04, "vi": 6.557, "tr": 5.127, "fi": 4.91, "ru": 6.726, "el": 6.093, "hi": 6.288, "ta": 6.532, "th": 6.226, "zh": 5.607, "my": 6.956}, "bpb_mean": 1.75434, "t": 294.6} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.13932, "bpb": 1.6437, "chunk_loss": 0.02994, "F": 0.17361, "G": 0.31962, "chunks": 177.78, "bytes_per_chunk": 5.76, "boundary_entropy": 0.45146, "router_noise": 0.0, "grad_norm": 0.7169, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 294.7} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.19548, "bpb": 1.72472, "chunk_loss": 0.030004, "F": 0.17606, "G": 0.30808, "chunks": 180.28, "bytes_per_chunk": 5.68, "boundary_entropy": 0.42956, "router_noise": 0.0, "grad_norm": 0.7416, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 296.7} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.39068, "bpb": 2.00632, "chunk_loss": 0.030527, "F": 0.19049, "G": 0.27785, "chunks": 195.06, "bytes_per_chunk": 5.25, "boundary_entropy": 0.34778, "router_noise": 0.0, "grad_norm": 0.7781, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 298.7} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.40792, "bpb": 2.0312, "chunk_loss": 0.030232, "F": 0.18457, "G": 0.27574, "chunks": 189.0, "bytes_per_chunk": 5.418, "boundary_entropy": 0.35941, "router_noise": 0.0, "grad_norm": 0.8011, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 300.7} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3513.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201862, "cps": 27.445312, "mean_p": 0.228664, "boundary_entropy": 0.222917, "bpb": 2.007767}, "de": {"chunks": 4052.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.197341, "cps": 31.65625, "mean_p": 0.226852, "boundary_entropy": 0.237165, "bpb": 2.1984}, "vi": {"chunks": 3955.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.214538, "cps": 30.898438, "mean_p": 0.263063, "boundary_entropy": 0.376928, "bpb": 1.912688}, "tr": {"chunks": 3713.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.205104, "cps": 29.007812, "mean_p": 0.224997, "boundary_entropy": 0.249844, "bpb": 2.171518}, "fi": {"chunks": 3721.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.19781, "cps": 29.070312, "mean_p": 0.233928, "boundary_entropy": 0.260676, "bpb": 2.328664}, "ru": {"chunks": 5214.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.275145, "cps": 40.734375, "mean_p": 0.370362, "boundary_entropy": 0.551974, "bpb": 1.385021}, "el": {"chunks": 6362.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.309994, "cps": 49.703125, "mean_p": 0.34426, "boundary_entropy": 0.559372, "bpb": 1.477896}, "hi": {"chunks": 7109.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.422526, "cps": 55.539062, "mean_p": 0.409119, "boundary_entropy": 0.63878, "bpb": 1.182522}, "ta": {"chunks": 8241.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.440813, "cps": 64.382812, "mean_p": 0.384101, "boundary_entropy": 0.621127, "bpb": 0.970711}, "th": {"chunks": 6807.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.423189, "cps": 53.179688, "mean_p": 0.368109, "boundary_entropy": 0.612521, "bpb": 1.108099}, "zh": {"chunks": 2851.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482729, "cps": 22.273438, "mean_p": 0.276474, "boundary_entropy": 0.423384, "bpb": 3.10914}, "my": {"chunks": 9671.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.510397, "cps": 75.554688, "mean_p": 0.435201, "boundary_entropy": 0.658646, "bpb": 1.093287}}, "bpb": {"en": 2.00777, "de": 2.1984, "vi": 1.91269, "tr": 2.17152, "fi": 2.32866, "ru": 1.38502, "el": 1.4779, "hi": 1.18252, "ta": 0.97071, "th": 1.1081, "zh": 3.10914, "my": 1.09329}, "gini_cpc": 0.202454, "gini_cps": 0.211061, "bpic": 5.885399, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9776040676517352, "vi": 1.0627945441284343, "tr": 1.016062371627804, "fi": 0.9799270744841456, "ru": 1.363037432769502, "el": 1.535673146368493, "hi": 2.093145468180382, "ta": 2.1837374145486597, "th": 2.09643139947976, "zh": 2.3913863450823505, "my": 2.528447716227089}, "t": 303.6} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.01902, "de": 2.20388, "vi": 1.91293, "tr": 2.17806, "fi": 2.34097, "ru": 1.36092, "el": 1.46937, "hi": 1.17863, "ta": 0.97782, "th": 1.12488, "zh": 3.10551, "my": 1.10187}, "bytes_per_chunk": {"en": 5.025, "de": 5.054, "vi": 6.294, "tr": 5.143, "fi": 4.918, "ru": 6.525, "el": 5.79, "hi": 6.233, "ta": 6.472, "th": 6.65, "zh": 5.623, "my": 6.296}, "bpb_mean": 1.74782, "t": 304.9} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.46065, "bpb": 2.10728, "chunk_loss": 0.030213, "F": 0.18359, "G": 0.28016, "chunks": 188.0, "bytes_per_chunk": 5.447, "boundary_entropy": 0.36452, "router_noise": 0.0, "grad_norm": 0.7646, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.086, "alpha": 0.03, "t": 305.0} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.27226, "bpb": 1.83549, "chunk_loss": 0.030157, "F": 0.18149, "G": 0.30869, "chunks": 185.84, "bytes_per_chunk": 5.51, "boundary_entropy": 0.42018, "router_noise": 0.0, "grad_norm": 0.7315, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.048, "alpha": 0.03, "t": 307.4} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.31607, "bpb": 1.89869, "chunk_loss": 0.030229, "F": 0.18442, "G": 0.28648, "chunks": 188.84, "bytes_per_chunk": 5.422, "boundary_entropy": 0.36971, "router_noise": 0.0, "grad_norm": 0.7491, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.05, "alpha": 0.03, "t": 309.9} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.20281, "bpb": 1.73529, "chunk_loss": 0.030198, "F": 0.18188, "G": 0.30375, "chunks": 186.25, "bytes_per_chunk": 5.498, "boundary_entropy": 0.40307, "router_noise": 0.0, "grad_norm": 0.7265, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.049, "alpha": 0.03, "t": 312.3} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3542.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203528, "cps": 27.671875, "mean_p": 0.229622, "boundary_entropy": 0.221763, "bpb": 2.00508}, "de": {"chunks": 4068.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.19812, "cps": 31.78125, "mean_p": 0.227779, "boundary_entropy": 0.236615, "bpb": 2.19568}, "vi": {"chunks": 3928.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.213073, "cps": 30.6875, "mean_p": 0.265174, "boundary_entropy": 0.379627, "bpb": 1.906869}, "tr": {"chunks": 3737.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.20643, "cps": 29.195312, "mean_p": 0.226538, "boundary_entropy": 0.248533, "bpb": 2.16243}, "fi": {"chunks": 3693.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.196321, "cps": 28.851562, "mean_p": 0.232147, "boundary_entropy": 0.257945, "bpb": 2.32305}, "ru": {"chunks": 5553.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.293034, "cps": 43.382812, "mean_p": 0.379441, "boundary_entropy": 0.55728, "bpb": 1.378978}, "el": {"chunks": 5600.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.272865, "cps": 43.75, "mean_p": 0.336816, "boundary_entropy": 0.556872, "bpb": 1.475486}, "hi": {"chunks": 7478.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.444458, "cps": 58.421875, "mean_p": 0.413771, "boundary_entropy": 0.641348, "bpb": 1.177615}, "ta": {"chunks": 8558.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.457769, "cps": 66.859375, "mean_p": 0.384797, "boundary_entropy": 0.621085, "bpb": 0.966659}, "th": {"chunks": 7671.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.476904, "cps": 59.929688, "mean_p": 0.379562, "boundary_entropy": 0.616515, "bpb": 1.104233}, "zh": {"chunks": 2852.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.482899, "cps": 22.28125, "mean_p": 0.275282, "boundary_entropy": 0.41932, "bpb": 3.104883}, "my": {"chunks": 11714.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.618218, "cps": 91.515625, "mean_p": 0.446237, "boundary_entropy": 0.661273, "bpb": 1.093418}}, "bpb": {"en": 2.00508, "de": 2.19568, "vi": 1.90687, "tr": 2.16243, "fi": 2.32305, "ru": 1.37898, "el": 1.47549, "hi": 1.17762, "ta": 0.96666, "th": 1.10423, "zh": 3.10488, "my": 1.09342}, "gini_cpc": 0.227394, "gini_cps": 0.23842, "bpic": 5.611326, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9734285965633594, "vi": 1.046896867946148, "tr": 1.0142572243220946, "fi": 0.9645905160012308, "ru": 1.4397729917209094, "el": 1.340672635868482, "hi": 2.1837652521262574, "ta": 2.2491704860136315, "th": 2.343184472944849, "zh": 2.372638874840026, "my": 3.037507835942453}, "t": 315.9} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.01503, "de": 2.20174, "vi": 1.90681, "tr": 2.17032, "fi": 2.33538, "ru": 1.35485, "el": 1.467, "hi": 1.17311, "ta": 0.97422, "th": 1.12039, "zh": 3.10123, "my": 1.10229}, "bytes_per_chunk": {"en": 4.977, "de": 5.027, "vi": 6.326, "tr": 5.083, "fi": 4.96, "ru": 6.156, "el": 6.496, "hi": 6.038, "ta": 6.256, "th": 5.926, "zh": 5.605, "my": 5.277}, "bpb_mean": 1.74353, "t": 317.4} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.25834, "bpb": 1.8154, "chunk_loss": 0.0303, "F": 0.18579, "G": 0.29675, "chunks": 190.25, "bytes_per_chunk": 5.382, "boundary_entropy": 0.3894, "router_noise": 0.0, "grad_norm": 0.7188, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.102, "alpha": 0.03, "t": 317.4} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.36354, "bpb": 1.96718, "chunk_loss": 0.030256, "F": 0.18414, "G": 0.29138, "chunks": 188.56, "bytes_per_chunk": 5.431, "boundary_entropy": 0.38608, "router_noise": 0.0, "grad_norm": 0.7209, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 319.6} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.32306, "bpb": 1.90877, "chunk_loss": 0.030129, "F": 0.18091, "G": 0.27805, "chunks": 185.25, "bytes_per_chunk": 5.528, "boundary_entropy": 0.36398, "router_noise": 0.0, "grad_norm": 0.7492, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.04, "alpha": 0.03, "t": 321.6} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.28281, "bpb": 1.8507, "chunk_loss": 0.030067, "F": 0.17786, "G": 0.2907, "chunks": 182.12, "bytes_per_chunk": 5.623, "boundary_entropy": 0.39166, "router_noise": 0.0, "grad_norm": 0.7307, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 323.6} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3554.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.204218, "cps": 27.765625, "mean_p": 0.229418, "boundary_entropy": 0.219621, "bpb": 2.00097}, "de": {"chunks": 4074.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.198412, "cps": 31.828125, "mean_p": 0.226275, "boundary_entropy": 0.234372, "bpb": 2.190772}, "vi": {"chunks": 3910.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.212097, "cps": 30.546875, "mean_p": 0.262431, "boundary_entropy": 0.375861, "bpb": 1.903682}, "tr": {"chunks": 3737.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.20643, "cps": 29.195312, "mean_p": 0.225867, "boundary_entropy": 0.245853, "bpb": 2.163042}, "fi": {"chunks": 3735.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.198554, "cps": 29.179688, "mean_p": 0.233517, "boundary_entropy": 0.256381, "bpb": 2.323166}, "ru": {"chunks": 5416.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.285805, "cps": 42.3125, "mean_p": 0.375164, "boundary_entropy": 0.555337, "bpb": 1.374952}, "el": {"chunks": 6226.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.303367, "cps": 48.640625, "mean_p": 0.341666, "boundary_entropy": 0.558119, "bpb": 1.468961}, "hi": {"chunks": 7750.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.460624, "cps": 60.546875, "mean_p": 0.41527, "boundary_entropy": 0.6411, "bpb": 1.176798}, "ta": {"chunks": 8400.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449318, "cps": 65.625, "mean_p": 0.379954, "boundary_entropy": 0.618061, "bpb": 0.965125}, "th": {"chunks": 6746.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.419397, "cps": 52.703125, "mean_p": 0.372323, "boundary_entropy": 0.615118, "bpb": 1.101668}, "zh": {"chunks": 2871.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.486116, "cps": 22.429688, "mean_p": 0.273968, "boundary_entropy": 0.417905, "bpb": 3.099327}, "my": {"chunks": 10643.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.561695, "cps": 83.148438, "mean_p": 0.438119, "boundary_entropy": 0.658772, "bpb": 1.090673}}, "bpb": {"en": 2.00097, "de": 2.19077, "vi": 1.90368, "tr": 2.16304, "fi": 2.32317, "ru": 1.37495, "el": 1.46896, "hi": 1.1768, "ta": 0.96513, "th": 1.10167, "zh": 3.09933, "my": 1.09067}, "gini_cpc": 0.213933, "gini_cps": 0.224352, "bpic": 5.722779, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9715727247580181, "vi": 1.038580853899822, "tr": 1.0108326079203318, "fi": 0.9722667178266813, "ru": 1.3995104256529116, "el": 1.485507907443756, "hi": 2.255554505690593, "ta": 2.20019165579378, "th": 2.053676191602399, "zh": 2.380380843684808, "my": 2.7504729794506755}, "t": 326.6} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.01299, "de": 2.19767, "vi": 1.904, "tr": 2.16899, "fi": 2.33436, "ru": 1.3519, "el": 1.46079, "hi": 1.1729, "ta": 0.97211, "th": 1.11821, "zh": 3.0958, "my": 1.09919}, "bytes_per_chunk": {"en": 4.98, "de": 5.033, "vi": 6.337, "tr": 5.099, "fi": 4.907, "ru": 6.302, "el": 5.986, "hi": 5.887, "ta": 6.309, "th": 6.712, "zh": 5.615, "my": 5.706}, "bpb_mean": 1.74074, "t": 327.8} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3554.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.204218, "cps": 27.765625, "mean_p": 0.229418, "boundary_entropy": 0.219621, "bpb": 2.00097}, "de": {"chunks": 4074.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.198412, "cps": 31.828125, "mean_p": 0.226275, "boundary_entropy": 0.234372, "bpb": 2.190772}, "vi": {"chunks": 3910.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.212097, "cps": 30.546875, "mean_p": 0.262431, "boundary_entropy": 0.375861, "bpb": 1.903682}, "tr": {"chunks": 3737.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.20643, "cps": 29.195312, "mean_p": 0.225867, "boundary_entropy": 0.245853, "bpb": 2.163042}, "fi": {"chunks": 3735.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.198554, "cps": 29.179688, "mean_p": 0.233517, "boundary_entropy": 0.256381, "bpb": 2.323166}, "ru": {"chunks": 5416.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.285805, "cps": 42.3125, "mean_p": 0.375164, "boundary_entropy": 0.555337, "bpb": 1.374952}, "el": {"chunks": 6226.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.303367, "cps": 48.640625, "mean_p": 0.341666, "boundary_entropy": 0.558119, "bpb": 1.468961}, "hi": {"chunks": 7750.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.460624, "cps": 60.546875, "mean_p": 0.41527, "boundary_entropy": 0.6411, "bpb": 1.176798}, "ta": {"chunks": 8400.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.449318, "cps": 65.625, "mean_p": 0.379954, "boundary_entropy": 0.618061, "bpb": 0.965125}, "th": {"chunks": 6746.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.419397, "cps": 52.703125, "mean_p": 0.372323, "boundary_entropy": 0.615118, "bpb": 1.101668}, "zh": {"chunks": 2871.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.486116, "cps": 22.429688, "mean_p": 0.273968, "boundary_entropy": 0.417905, "bpb": 3.099327}, "my": {"chunks": 10643.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.561695, "cps": 83.148438, "mean_p": 0.438119, "boundary_entropy": 0.658772, "bpb": 1.090673}}, "bpb": {"en": 2.00097, "de": 2.19077, "vi": 1.90368, "tr": 2.16304, "fi": 2.32317, "ru": 1.37495, "el": 1.46896, "hi": 1.1768, "ta": 0.96513, "th": 1.10167, "zh": 3.09933, "my": 1.09067}, "gini_cpc": 0.213933, "gini_cps": 0.224352, "bpic": 5.722779, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9715727247580181, "vi": 1.038580853899822, "tr": 1.0108326079203318, "fi": 0.9722667178266813, "ru": 1.3995104256529116, "el": 1.485507907443756, "hi": 2.255554505690593, "ta": 2.20019165579378, "th": 2.053676191602399, "zh": 2.380380843684808, "my": 2.7504729794506755}, "t": 328.8} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s0/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s0/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s0/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s0/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_C_s0/th0.70"], "t": 332.3} |
| {"event": "final", "run": "pilot_C_s0", "condition": "C", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.20421766362121474, "cps": 27.765625, "bpb": 2.0218003220896423, "bytes_per_chunk": 4.931551174362496}, "de": {"cpc": 0.19841231188818, "cps": 31.828125, "bpb": 2.1884879495998555, "bytes_per_chunk": 5.053768001388059}, "vi": {"cpc": 0.21209655546514783, "cps": 30.546875, "bpb": 1.908820920141231, "bytes_per_chunk": 6.356469974903311}, "tr": {"cpc": 0.20642987350162956, "cps": 29.1953125, "bpb": 2.16043099015882, "bytes_per_chunk": 5.1186503558632}, "fi": {"cpc": 0.19855403753123171, "cps": 29.1796875, "bpb": 2.3519634356420727, "bytes_per_chunk": 4.898148321156972}, "ru": {"cpc": 0.28580474934036937, "cps": 42.3125, "bpb": 1.3522200363837824, "bytes_per_chunk": 6.302523230793273}, "el": {"cpc": 0.30336695414900355, "cps": 48.640625, "bpb": 1.4567373275916706, "bytes_per_chunk": 6.003458107659365}, "hi": {"cpc": 0.4606240713224368, "cps": 60.546875, "bpb": 1.1819406533249985, "bytes_per_chunk": 5.90753698633224}, "ta": {"cpc": 0.4493179994650976, "cps": 65.625, "bpb": 0.972932620730949, "bytes_per_chunk": 6.351387694283257}, "th": {"cpc": 0.4193969536835561, "cps": 52.703125, "bpb": 1.1166655464733943, "bytes_per_chunk": 6.694605120347315}, "zh": {"cpc": 0.48611581442600743, "cps": 22.4296875, "bpb": 3.1056148981036586, "bytes_per_chunk": 5.6012478365846885}, "my": {"cpc": 0.5616951657166983, "cps": 83.1484375, "bpb": 1.1093779590001915, "bytes_per_chunk": 5.724730572267779}}, "gini_cpc": 0.21393280088221256, "gini_cps": 0.2243520920938833, "bpic": 5.722778920998479, "seed": 0, "mean_boundary_entropy": 0.4017484784126282, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 334.7} |
|
|
