| {"event": "start", "run": "emactl_genreg_s0", "condition": "A", "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": "global", "p3_objective": "global", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_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": 1.5} |
| {"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": 8.125775}, "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": 8.121816}, "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": 8.087584}, "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": 8.14748}, "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": 8.208588}, "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": 8.082947}, "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": 8.196862}, "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": 8.272031}, "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": 8.008816}, "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": 8.270247}, "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": 8.131836}, "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": 8.106955}}, "bpb": {"en": 8.12577, "de": 8.12182, "vi": 8.08758, "tr": 8.14748, "fi": 8.20859, "ru": 8.08295, "el": 8.19686, "hi": 8.27203, "ta": 8.00882, "th": 8.27025, "zh": 8.13184, "my": 8.10696}, "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": 5.63738, "bpb": 8.13302, "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": 7.4901, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 3.6} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6343.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.364477, "cps": 49.554688, "mean_p": 0.49658, "boundary_entropy": 0.68699, "bpb": 8.10565}, "de": {"chunks": 8607.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.419179, "cps": 67.242188, "mean_p": 0.498815, "boundary_entropy": 0.68782, "bpb": 8.104411}, "vi": {"chunks": 13053.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.708055, "cps": 101.976562, "mean_p": 0.505612, "boundary_entropy": 0.688386, "bpb": 8.073819}, "tr": {"chunks": 5260.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.29056, "cps": 41.09375, "mean_p": 0.489476, "boundary_entropy": 0.687283, "bpb": 8.133896}, "fi": {"chunks": 8850.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.470469, "cps": 69.140625, "mean_p": 0.500541, "boundary_entropy": 0.68751, "bpb": 8.191987}, "ru": {"chunks": 10636.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.561266, "cps": 83.09375, "mean_p": 0.493511, "boundary_entropy": 0.689889, "bpb": 8.047595}, "el": {"chunks": 14967.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.729279, "cps": 116.929688, "mean_p": 0.497406, "boundary_entropy": 0.690233, "bpb": 8.183471}, "hi": {"chunks": 2508.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.149064, "cps": 19.59375, "mean_p": 0.484458, "boundary_entropy": 0.690244, "bpb": 8.262066}, "ta": {"chunks": 37839.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.024017, "cps": 295.617188, "mean_p": 0.512671, "boundary_entropy": 0.690392, "bpb": 7.98748}, "th": {"chunks": 33875.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.105999, "cps": 264.648438, "mean_p": 0.511786, "boundary_entropy": 0.690558, "bpb": 8.25513}, "zh": {"chunks": 11820.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.001355, "cps": 92.34375, "mean_p": 0.520016, "boundary_entropy": 0.685676, "bpb": 8.129447}, "my": {"chunks": 48280.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.548026, "cps": 377.1875, "mean_p": 0.52173, "boundary_entropy": 0.690191, "bpb": 8.087111}}, "bpb": {"en": 8.10565, "de": 8.10441, "vi": 8.07382, "tr": 8.1339, "fi": 8.19199, "ru": 8.04759, "el": 8.18347, "hi": 8.26207, "ta": 7.98748, "th": 8.25513, "zh": 8.12945, "my": 8.08711}, "gini_cpc": 0.425556, "gini_cps": 0.426908, "bpic": 1.899549, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1500819954149706, "vi": 1.9426591360426666, "tr": 0.7971950649483452, "fi": 1.2908054668252422, "ru": 1.5399212894192464, "el": 2.0008905002695974, "hi": 0.40897980926429883, "ta": 5.55320351332114, "th": 5.778134507425989, "zh": 5.491025274361123, "my": 6.990903288141485}, "t": 5.0} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6282.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.360972, "cps": 49.078125, "mean_p": 0.496334, "boundary_entropy": 0.686981, "bpb": 8.060672}, "de": {"chunks": 8563.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.417036, "cps": 66.898438, "mean_p": 0.498717, "boundary_entropy": 0.687817, "bpb": 8.065101}, "vi": {"chunks": 12903.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.699919, "cps": 100.804688, "mean_p": 0.505287, "boundary_entropy": 0.688389, "bpb": 8.042861}, "tr": {"chunks": 5221.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.288405, "cps": 40.789062, "mean_p": 0.489332, "boundary_entropy": 0.687276, "bpb": 8.101938}, "fi": {"chunks": 8783.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.466908, "cps": 68.617188, "mean_p": 0.500294, "boundary_entropy": 0.687505, "bpb": 8.15397}, "ru": {"chunks": 10491.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.553615, "cps": 81.960938, "mean_p": 0.493302, "boundary_entropy": 0.689883, "bpb": 7.989111}, "el": {"chunks": 14970.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.729426, "cps": 116.953125, "mean_p": 0.497401, "boundary_entropy": 0.690233, "bpb": 8.168061}, "hi": {"chunks": 2423.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.144012, "cps": 18.929688, "mean_p": 0.484204, "boundary_entropy": 0.690226, "bpb": 8.235381}, "ta": {"chunks": 37435.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.002407, "cps": 292.460938, "mean_p": 0.512171, "boundary_entropy": 0.690416, "bpb": 7.950557}, "th": {"chunks": 33648.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.091887, "cps": 262.875, "mean_p": 0.511603, "boundary_entropy": 0.690565, "bpb": 8.233799}, "zh": {"chunks": 11802.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.998307, "cps": 92.203125, "mean_p": 0.519959, "boundary_entropy": 0.68568, "bpb": 8.123234}, "my": {"chunks": 48413.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.555045, "cps": 378.226562, "mean_p": 0.521973, "boundary_entropy": 0.690169, "bpb": 8.054418}}, "bpb": {"en": 8.06067, "de": 8.0651, "vi": 8.04286, "tr": 8.10194, "fi": 8.15397, "ru": 7.98911, "el": 8.16806, "hi": 8.23538, "ta": 7.95056, "th": 8.2338, "zh": 8.12323, "my": 8.05442}, "gini_cpc": 0.427116, "gini_cps": 0.428451, "bpic": 1.909985, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1553131723429255, "vi": 1.9389818482502217, "tr": 0.7989678978524722, "fi": 1.2934724632506627, "ru": 1.5336768452646463, "el": 2.020724668821493, "hi": 0.3989555668144036, "ta": 5.547260438962248, "th": 5.795145951887047, "zh": 5.535901521157155, "my": 7.078232231216829}, "t": 6.5} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6012.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.345458, "cps": 46.96875, "mean_p": 0.495304, "boundary_entropy": 0.686943, "bpb": 7.908784}, "de": {"chunks": 8396.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.408903, "cps": 65.59375, "mean_p": 0.49818, "boundary_entropy": 0.687805, "bpb": 7.930173}, "vi": {"chunks": 12515.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.678872, "cps": 97.773438, "mean_p": 0.504182, "boundary_entropy": 0.688395, "bpb": 7.941824}, "tr": {"chunks": 5042.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.278517, "cps": 39.390625, "mean_p": 0.488689, "boundary_entropy": 0.687243, "bpb": 7.994073}, "fi": {"chunks": 8463.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.449896, "cps": 66.117188, "mean_p": 0.49919, "boundary_entropy": 0.687484, "bpb": 8.023524}, "ru": {"chunks": 9650.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.509235, "cps": 75.390625, "mean_p": 0.492206, "boundary_entropy": 0.689848, "bpb": 7.791313}, "el": {"chunks": 14757.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.719047, "cps": 115.289062, "mean_p": 0.497202, "boundary_entropy": 0.690228, "bpb": 8.098125}, "hi": {"chunks": 2130.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.126597, "cps": 16.640625, "mean_p": 0.483037, "boundary_entropy": 0.690144, "bpb": 8.136898}, "ta": {"chunks": 35788.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.914309, "cps": 279.59375, "mean_p": 0.51029, "boundary_entropy": 0.690497, "bpb": 7.818991}, "th": {"chunks": 32664.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.030712, "cps": 255.1875, "mean_p": 0.510803, "boundary_entropy": 0.690594, "bpb": 8.143656}, "zh": {"chunks": 11758.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.990857, "cps": 91.859375, "mean_p": 0.519677, "boundary_entropy": 0.685698, "bpb": 8.097827}, "my": {"chunks": 48594.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.564598, "cps": 379.640625, "mean_p": 0.522454, "boundary_entropy": 0.690123, "bpb": 7.97698}}, "bpb": {"en": 7.90878, "de": 7.93017, "vi": 7.94182, "tr": 7.99407, "fi": 8.02352, "ru": 7.79131, "el": 8.09813, "hi": 8.1369, "ta": 7.81899, "th": 8.14366, "zh": 8.09783, "my": 7.97698}, "gini_cpc": 0.434174, "gini_cps": 0.434377, "bpic": 1.960377, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.183655092775536, "vi": 1.965137124719912, "tr": 0.8062271854677652, "fi": 1.3023196868184848, "ru": 1.4740874451624455, "el": 2.0814327345892925, "hi": 0.36646261711346995, "ta": 5.541369467537244, "th": 5.878323055474379, "zh": 5.762954112094132, "my": 7.423768517429991}, "t": 8.1} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5168.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.29696, "cps": 40.375, "mean_p": 0.492001, "boundary_entropy": 0.68679, "bpb": 7.479578}, "de": {"chunks": 7856.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.382604, "cps": 61.375, "mean_p": 0.49642, "boundary_entropy": 0.687763, "bpb": 7.540147}, "vi": {"chunks": 11071.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.600542, "cps": 86.492188, "mean_p": 0.500662, "boundary_entropy": 0.688379, "bpb": 7.633096}, "tr": {"chunks": 4418.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.244048, "cps": 34.515625, "mean_p": 0.486563, "boundary_entropy": 0.687118, "bpb": 7.665818}, "fi": {"chunks": 7409.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.393865, "cps": 57.882812, "mean_p": 0.495667, "boundary_entropy": 0.687385, "bpb": 7.637518}, "ru": {"chunks": 6524.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.344274, "cps": 50.96875, "mean_p": 0.48768, "boundary_entropy": 0.689644, "bpb": 7.249202}, "el": {"chunks": 13164.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.641427, "cps": 102.84375, "mean_p": 0.495758, "boundary_entropy": 0.690195, "bpb": 7.831362}, "hi": {"chunks": 1175.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.069837, "cps": 9.179688, "mean_p": 0.477956, "boundary_entropy": 0.689721, "bpb": 7.801163}, "ta": {"chunks": 28557.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.527521, "cps": 223.101562, "mean_p": 0.503462, "boundary_entropy": 0.690677, "bpb": 7.392576}, "th": {"chunks": 28028.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.742493, "cps": 218.96875, "mean_p": 0.506755, "boundary_entropy": 0.69071, "bpb": 7.702866}, "zh": {"chunks": 11562.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.95767, "cps": 90.328125, "mean_p": 0.518679, "boundary_entropy": 0.68576, "bpb": 8.013352}, "my": {"chunks": 49636.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.61959, "cps": 387.78125, "mean_p": 0.525217, "boundary_entropy": 0.689848, "bpb": 7.651208}}, "bpb": {"en": 7.47958, "de": 7.54015, "vi": 7.6331, "tr": 7.66582, "fi": 7.63752, "ru": 7.2492, "el": 7.83136, "hi": 7.80116, "ta": 7.39258, "th": 7.70287, "zh": 8.01335, "my": 7.65121}, "gini_cpc": 0.46583, "gini_cps": 0.4579, "bpic": 2.198461, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.288399901148329, "vi": 2.0222987993964234, "tr": 0.8218201309021358, "fi": 1.326323078453701, "ru": 1.159328075349012, "el": 2.1599745977065075, "hi": 0.23517134819830804, "ta": 5.143855112250838, "th": 5.867764276661044, "zh": 6.59236337030401, "my": 8.821349214830095}, "t": 9.8} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2566.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.147446, "cps": 20.046875, "mean_p": 0.480144, "boundary_entropy": 0.685969, "bpb": 6.886495}, "de": {"chunks": 4870.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.237179, "cps": 38.046875, "mean_p": 0.487306, "boundary_entropy": 0.68744, "bpb": 6.972675}, "vi": {"chunks": 7625.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.413615, "cps": 59.570312, "mean_p": 0.491771, "boundary_entropy": 0.68815, "bpb": 7.197799}, "tr": {"chunks": 2862.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.158095, "cps": 22.359375, "mean_p": 0.480206, "boundary_entropy": 0.686638, "bpb": 7.184362}, "fi": {"chunks": 4154.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.220828, "cps": 32.453125, "mean_p": 0.484394, "boundary_entropy": 0.686813, "bpb": 7.07877}, "ru": {"chunks": 709.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.037414, "cps": 5.539062, "mean_p": 0.468102, "boundary_entropy": 0.687786, "bpb": 6.530374}, "el": {"chunks": 4389.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.213858, "cps": 34.289062, "mean_p": 0.48332, "boundary_entropy": 0.689589, "bpb": 7.204364}, "hi": {"chunks": 370.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.021991, "cps": 2.890625, "mean_p": 0.465352, "boundary_entropy": 0.688227, "bpb": 7.149443}, "ta": {"chunks": 14381.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.769243, "cps": 112.351562, "mean_p": 0.489906, "boundary_entropy": 0.690513, "bpb": 6.707955}, "th": {"chunks": 15762.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.979919, "cps": 123.140625, "mean_p": 0.493951, "boundary_entropy": 0.690659, "bpb": 6.555155}, "zh": {"chunks": 10987.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.860312, "cps": 85.835938, "mean_p": 0.515842, "boundary_entropy": 0.685925, "bpb": 7.796888}, "my": {"chunks": 52115.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.750422, "cps": 407.148438, "mean_p": 0.533874, "boundary_entropy": 0.688776, "bpb": 6.828328}}, "bpb": {"en": 6.88649, "de": 6.97267, "vi": 7.1978, "tr": 7.18436, "fi": 7.07877, "ru": 6.53037, "el": 7.20436, "hi": 7.14944, "ta": 6.70795, "th": 6.55515, "zh": 7.79689, "my": 6.82833}, "gini_cpc": 0.59662, "gini_cps": 0.578757, "bpic": 3.177258, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.608585028173001, "vi": 2.80520222195868, "tr": 1.0722265236479798, "fi": 1.4976905065372565, "ru": 0.25374908741673646, "el": 1.4504147372400351, "hi": 0.1491468616344262, "ta": 5.2171231090544214, "th": 6.6459600461403046, "zh": 12.616914209864653, "my": 18.65377930179672}, "t": 11.6} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 1674.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.09619, "cps": 13.078125, "mean_p": 0.483814, "boundary_entropy": 0.68678, "bpb": 6.388466}, "de": {"chunks": 4620.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.225004, "cps": 36.09375, "mean_p": 0.4909, "boundary_entropy": 0.688086, "bpb": 6.470807}, "vi": {"chunks": 5503.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.298508, "cps": 42.992188, "mean_p": 0.488566, "boundary_entropy": 0.688296, "bpb": 6.93321}, "tr": {"chunks": 4586.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.253328, "cps": 35.828125, "mean_p": 0.491357, "boundary_entropy": 0.687791, "bpb": 6.739348}, "fi": {"chunks": 3387.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180054, "cps": 26.460938, "mean_p": 0.487719, "boundary_entropy": 0.687571, "bpb": 6.544507}, "ru": {"chunks": 3791.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.200053, "cps": 29.617188, "mean_p": 0.485792, "boundary_entropy": 0.689708, "bpb": 5.838241}, "el": {"chunks": 222.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.010817, "cps": 1.734375, "mean_p": 0.446249, "boundary_entropy": 0.684197, "bpb": 6.771603}, "hi": {"chunks": 476.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.028291, "cps": 3.71875, "mean_p": 0.477441, "boundary_entropy": 0.689765, "bpb": 6.161915}, "ta": {"chunks": 3169.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.169511, "cps": 24.757812, "mean_p": 0.477819, "boundary_entropy": 0.68983, "bpb": 5.972316}, "th": {"chunks": 3887.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.241654, "cps": 30.367188, "mean_p": 0.487341, "boundary_entropy": 0.690574, "bpb": 5.603778}, "zh": {"chunks": 8906.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.507958, "cps": 69.578125, "mean_p": 0.506225, "boundary_entropy": 0.686322, "bpb": 7.386018}, "my": {"chunks": 53213.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.80837, "cps": 415.726562, "mean_p": 0.541247, "boundary_entropy": 0.687627, "bpb": 5.834638}}, "bpb": {"en": 6.38847, "de": 6.47081, "vi": 6.93321, "tr": 6.73935, "fi": 6.54451, "ru": 5.83824, "el": 6.7716, "hi": 6.16191, "ta": 5.97232, "th": 5.60378, "zh": 7.38602, "my": 5.83464}, "gini_cpc": 0.638665, "gini_cps": 0.621225, "bpic": 4.107509, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 2.3391508764543203, "vi": 3.1033091176690744, "tr": 2.6336142626313745, "fi": 1.8718540341586911, "ru": 2.079760074143426, "el": 0.11245552459108296, "hi": 0.2941172836547423, "ta": 1.7622415477275755, "th": 2.512245875685065, "zh": 15.67681793741132, "my": 29.195978448468285}, "t": 14.3} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.23264, "bpb": 6.1064, "chunk_loss": 0.040058, "F": 0.30142, "G": 0.49253, "chunks": 308.66, "bytes_per_chunk": 3.318, "boundary_entropy": 0.69175, "router_noise": 0.0, "grad_norm": 1.5887, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.247, "alpha": 0.03, "t": 16.0} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2381.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.136815, "cps": 18.601562, "mean_p": 0.489314, "boundary_entropy": 0.687282, "bpb": 5.81171}, "de": {"chunks": 3909.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.190376, "cps": 30.539062, "mean_p": 0.490798, "boundary_entropy": 0.688189, "bpb": 5.928519}, "vi": {"chunks": 3581.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.19425, "cps": 27.976562, "mean_p": 0.483287, "boundary_entropy": 0.688035, "bpb": 6.283885}, "tr": {"chunks": 3805.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.210186, "cps": 29.726562, "mean_p": 0.490979, "boundary_entropy": 0.68791, "bpb": 6.223243}, "fi": {"chunks": 2979.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158365, "cps": 23.273438, "mean_p": 0.490169, "boundary_entropy": 0.68791, "bpb": 5.977182}, "ru": {"chunks": 7767.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.409868, "cps": 60.679688, "mean_p": 0.49455, "boundary_entropy": 0.690266, "bpb": 5.383619}, "el": {"chunks": 648.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.031574, "cps": 5.0625, "mean_p": 0.479762, "boundary_entropy": 0.689615, "bpb": 5.745531}, "hi": {"chunks": 9611.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.571233, "cps": 75.085938, "mean_p": 0.492127, "boundary_entropy": 0.690663, "bpb": 5.167397}, "ta": {"chunks": 3023.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.161701, "cps": 23.617188, "mean_p": 0.478871, "boundary_entropy": 0.689833, "bpb": 5.137685}, "th": {"chunks": 12220.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.759714, "cps": 95.46875, "mean_p": 0.49116, "boundary_entropy": 0.690555, "bpb": 4.876469}, "zh": {"chunks": 3520.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.596004, "cps": 27.5, "mean_p": 0.488524, "boundary_entropy": 0.686319, "bpb": 6.995088}, "my": {"chunks": 37957.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.003219, "cps": 296.539062, "mean_p": 0.509427, "boundary_entropy": 0.690902, "bpb": 4.931557}}, "bpb": {"en": 5.81171, "de": 5.92852, "vi": 6.28389, "tr": 6.22324, "fi": 5.97718, "ru": 5.38362, "el": 5.74553, "hi": 5.1674, "ta": 5.13768, "th": 4.87647, "zh": 6.99509, "my": 4.93156}, "gini_cpc": 0.508296, "gini_cps": 0.52013, "bpic": 4.198871, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.3914832666186983, "vi": 1.4197958546931584, "tr": 1.5362745443538899, "fi": 1.1575062496942607, "ru": 2.9957724022300556, "el": 0.23078038073794874, "hi": 4.17520908779447, "ta": 1.1818909372013149, "th": 5.552836235783833, "zh": 4.3562615371465006, "my": 14.641758136971523}, "t": 17.9} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.26097, "bpb": 4.70459, "chunk_loss": 0.038242, "F": 0.28479, "G": 0.47398, "chunks": 291.62, "bytes_per_chunk": 3.511, "boundary_entropy": 0.68638, "router_noise": 0.0, "grad_norm": 1.9085, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 20.1} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5347.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.307246, "cps": 41.773438, "mean_p": 0.444609, "boundary_entropy": 0.667595, "bpb": 4.449651}, "de": {"chunks": 5780.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.281498, "cps": 45.15625, "mean_p": 0.441821, "boundary_entropy": 0.668226, "bpb": 4.668533}, "vi": {"chunks": 6926.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.375698, "cps": 54.109375, "mean_p": 0.472904, "boundary_entropy": 0.683169, "bpb": 4.822987}, "tr": {"chunks": 8650.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.477821, "cps": 67.578125, "mean_p": 0.466775, "boundary_entropy": 0.674185, "bpb": 4.915791}, "fi": {"chunks": 7470.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.397108, "cps": 58.359375, "mean_p": 0.465324, "boundary_entropy": 0.675411, "bpb": 4.679762}, "ru": {"chunks": 10993.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.580106, "cps": 85.882812, "mean_p": 0.494107, "boundary_entropy": 0.690017, "bpb": 3.602344}, "el": {"chunks": 18131.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.883448, "cps": 141.648438, "mean_p": 0.485999, "boundary_entropy": 0.687546, "bpb": 4.334593}, "hi": {"chunks": 17860.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.061516, "cps": 139.53125, "mean_p": 0.410601, "boundary_entropy": 0.624804, "bpb": 3.361567}, "ta": {"chunks": 4232.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.226371, "cps": 33.0625, "mean_p": 0.401967, "boundary_entropy": 0.637726, "bpb": 3.13428}, "th": {"chunks": 4951.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.307802, "cps": 38.679688, "mean_p": 0.435444, "boundary_entropy": 0.667942, "bpb": 3.036038}, "zh": {"chunks": 3663.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.620217, "cps": 28.617188, "mean_p": 0.468553, "boundary_entropy": 0.680344, "bpb": 6.132763}, "my": {"chunks": 23702.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.250897, "cps": 185.171875, "mean_p": 0.496003, "boundary_entropy": 0.691022, "bpb": 3.565708}}, "bpb": {"en": 4.44965, "de": 4.66853, "vi": 4.82299, "tr": 4.91579, "fi": 4.67976, "ru": 3.60234, "el": 4.33459, "hi": 3.36157, "ta": 3.13428, "th": 3.03604, "zh": 6.13276, "my": 3.56571}, "gini_cpc": 0.302565, "gini_cps": 0.335066, "bpic": 3.260533, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.916198058964431, "vi": 1.2227939501447393, "tr": 1.5551758090698529, "fi": 1.2924764971595382, "ru": 1.888082425328631, "el": 2.875377359694995, "hi": 3.4549384735048827, "ta": 0.7367737183363324, "th": 1.0018110027620588, "zh": 2.0186332018783646, "my": 4.07132295453003}, "t": 22.6} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.81416, "bpb": 4.05997, "chunk_loss": 0.037505, "F": 0.29526, "G": 0.43571, "chunks": 302.34, "bytes_per_chunk": 3.387, "boundary_entropy": 0.66024, "router_noise": 0.0, "grad_norm": 1.5166, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.078, "alpha": 0.03, "t": 24.0} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4516.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.259495, "cps": 35.28125, "mean_p": 0.418759, "boundary_entropy": 0.647149, "bpb": 3.824149}, "de": {"chunks": 6191.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.301515, "cps": 48.367188, "mean_p": 0.416285, "boundary_entropy": 0.644517, "bpb": 4.073353}, "vi": {"chunks": 5968.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.323732, "cps": 46.625, "mean_p": 0.449861, "boundary_entropy": 0.666269, "bpb": 3.567412}, "tr": {"chunks": 7558.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.4175, "cps": 59.046875, "mean_p": 0.445841, "boundary_entropy": 0.65512, "bpb": 4.094817}, "fi": {"chunks": 6659.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.353995, "cps": 52.023438, "mean_p": 0.444179, "boundary_entropy": 0.656415, "bpb": 4.094987}, "ru": {"chunks": 5092.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.268707, "cps": 39.78125, "mean_p": 0.460604, "boundary_entropy": 0.684695, "bpb": 2.798237}, "el": {"chunks": 15012.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.731472, "cps": 117.28125, "mean_p": 0.478195, "boundary_entropy": 0.68601, "bpb": 2.996152}, "hi": {"chunks": 17070.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.014562, "cps": 133.359375, "mean_p": 0.387912, "boundary_entropy": 0.598844, "bpb": 2.346133}, "ta": {"chunks": 8618.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.460979, "cps": 67.328125, "mean_p": 0.388575, "boundary_entropy": 0.616136, "bpb": 2.18471}, "th": {"chunks": 5004.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.311097, "cps": 39.09375, "mean_p": 0.409166, "boundary_entropy": 0.647919, "bpb": 2.287766}, "zh": {"chunks": 3708.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.627836, "cps": 28.96875, "mean_p": 0.390499, "boundary_entropy": 0.60152, "bpb": 4.822386}, "my": {"chunks": 11450.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.604285, "cps": 89.453125, "mean_p": 0.481863, "boundary_entropy": 0.689475, "bpb": 2.381137}}, "bpb": {"en": 3.82415, "de": 4.07335, "vi": 3.56741, "tr": 4.09482, "fi": 4.09499, "ru": 2.79824, "el": 2.99615, "hi": 2.34613, "ta": 2.18471, "th": 2.28777, "zh": 4.82239, "my": 2.38114}, "gini_cpc": 0.245162, "gini_cps": 0.266261, "bpic": 3.962797, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.161926304809424, "vi": 1.2475439644666353, "tr": 1.6088906325662222, "fi": 1.364166305554309, "ru": 1.035498246048643, "el": 2.8188235912551955, "hi": 3.909746820616016, "ta": 1.7764427143950567, "th": 1.1988543480122347, "zh": 2.419448987788311, "my": 2.328693320940084}, "t": 27.7} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.82948, "de": 4.07247, "vi": 3.57138, "tr": 4.10293, "fi": 4.09787, "ru": 2.7831, "el": 2.98887, "hi": 2.34217, "ta": 2.20081, "th": 2.32089, "zh": 4.83059, "my": 2.39916}, "bytes_per_chunk": {"en": 3.662, "de": 3.365, "vi": 4.158, "tr": 2.525, "fi": 2.824, "ru": 6.26, "el": 2.407, "hi": 2.666, "ta": 5.934, "th": 9.374, "zh": 3.91, "my": 6.329}, "bpb_mean": 3.29498, "t": 29.2} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.43728, "bpb": 3.51625, "chunk_loss": 0.035222, "F": 0.25955, "G": 0.42743, "chunks": 265.78, "bytes_per_chunk": 3.853, "boundary_entropy": 0.65331, "router_noise": 0.0, "grad_norm": 2.0802, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.104, "alpha": 0.03, "t": 29.2} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.50798, "bpb": 3.61825, "chunk_loss": 0.036635, "F": 0.28543, "G": 0.43104, "chunks": 292.28, "bytes_per_chunk": 3.503, "boundary_entropy": 0.65462, "router_noise": 0.0, "grad_norm": 1.5165, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.052, "alpha": 0.03, "t": 31.8} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4761.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.273574, "cps": 37.195312, "mean_p": 0.43595, "boundary_entropy": 0.65433, "bpb": 3.578364}, "de": {"chunks": 5862.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.285492, "cps": 45.796875, "mean_p": 0.429654, "boundary_entropy": 0.654419, "bpb": 3.868259}, "vi": {"chunks": 5835.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.316517, "cps": 45.585938, "mean_p": 0.452506, "boundary_entropy": 0.668044, "bpb": 3.174673}, "tr": {"chunks": 7512.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.414959, "cps": 58.6875, "mean_p": 0.445915, "boundary_entropy": 0.657718, "bpb": 3.772206}, "fi": {"chunks": 6212.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.330232, "cps": 48.53125, "mean_p": 0.437414, "boundary_entropy": 0.656919, "bpb": 3.869624}, "ru": {"chunks": 4933.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.260317, "cps": 38.539062, "mean_p": 0.478585, "boundary_entropy": 0.688109, "bpb": 2.642471}, "el": {"chunks": 7874.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.383667, "cps": 61.515625, "mean_p": 0.470765, "boundary_entropy": 0.686054, "bpb": 2.753917}, "hi": {"chunks": 16457.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.978128, "cps": 128.570312, "mean_p": 0.399856, "boundary_entropy": 0.603264, "bpb": 2.074262}, "ta": {"chunks": 11484.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.614282, "cps": 89.71875, "mean_p": 0.38576, "boundary_entropy": 0.622444, "bpb": 1.892182}, "th": {"chunks": 5420.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.33696, "cps": 42.34375, "mean_p": 0.403754, "boundary_entropy": 0.635154, "bpb": 2.041388}, "zh": {"chunks": 5099.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.863359, "cps": 39.835938, "mean_p": 0.397143, "boundary_entropy": 0.607638, "bpb": 4.425177}, "my": {"chunks": 8427.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.444744, "cps": 65.835938, "mean_p": 0.472718, "boundary_entropy": 0.688356, "bpb": 2.164708}}, "bpb": {"en": 3.57836, "de": 3.86826, "vi": 3.17467, "tr": 3.77221, "fi": 3.86962, "ru": 2.64247, "el": 2.75392, "hi": 2.07426, "ta": 1.89218, "th": 2.04139, "zh": 4.42518, "my": 2.16471}, "gini_cpc": 0.249104, "gini_cps": 0.211923, "bpic": 4.270117, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0435646173152775, "vi": 1.156974153811763, "tr": 1.5168092433060265, "fi": 1.2071062605079272, "ru": 0.9515417317153055, "el": 1.4024277739506363, "hi": 3.5753744718182934, "ta": 2.2453996625767028, "th": 1.2316977838471934, "zh": 3.1558583956891533, "my": 1.6256818482022035}, "t": 33.2} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.29045, "bpb": 3.30442, "chunk_loss": 0.037073, "F": 0.2933, "G": 0.42788, "chunks": 300.34, "bytes_per_chunk": 3.409, "boundary_entropy": 0.65161, "router_noise": 0.0, "grad_norm": 1.279, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.073, "alpha": 0.03, "t": 35.5} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.02305, "bpb": 2.91864, "chunk_loss": 0.036811, "F": 0.28073, "G": 0.42796, "chunks": 287.47, "bytes_per_chunk": 3.562, "boundary_entropy": 0.65138, "router_noise": 0.0, "grad_norm": 0.7249, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.048, "alpha": 0.03, "t": 37.9} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4640.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.266621, "cps": 36.25, "mean_p": 0.439605, "boundary_entropy": 0.6617, "bpb": 3.222174}, "de": {"chunks": 6036.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.293966, "cps": 47.15625, "mean_p": 0.440557, "boundary_entropy": 0.663547, "bpb": 3.493676}, "vi": {"chunks": 9419.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.51093, "cps": 73.585938, "mean_p": 0.462751, "boundary_entropy": 0.672315, "bpb": 2.768488}, "tr": {"chunks": 7441.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.411037, "cps": 58.132812, "mean_p": 0.451038, "boundary_entropy": 0.661783, "bpb": 3.410093}, "fi": {"chunks": 5841.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.31051, "cps": 45.632812, "mean_p": 0.443317, "boundary_entropy": 0.661312, "bpb": 3.531081}, "ru": {"chunks": 4811.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.253879, "cps": 37.585938, "mean_p": 0.397981, "boundary_entropy": 0.638101, "bpb": 2.359052}, "el": {"chunks": 18366.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.894898, "cps": 143.484375, "mean_p": 0.487357, "boundary_entropy": 0.678794, "bpb": 2.584824}, "hi": {"chunks": 19356.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.150431, "cps": 151.21875, "mean_p": 0.414444, "boundary_entropy": 0.627901, "bpb": 1.889939}, "ta": {"chunks": 15146.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.810163, "cps": 118.328125, "mean_p": 0.385076, "boundary_entropy": 0.619294, "bpb": 1.582095}, "th": {"chunks": 11439.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.711159, "cps": 89.367188, "mean_p": 0.400379, "boundary_entropy": 0.625406, "bpb": 1.931147}, "zh": {"chunks": 2152.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.364375, "cps": 16.8125, "mean_p": 0.404992, "boundary_entropy": 0.64058, "bpb": 4.151042}, "my": {"chunks": 2897.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.152892, "cps": 22.632812, "mean_p": 0.386205, "boundary_entropy": 0.627163, "bpb": 1.953972}}, "bpb": {"en": 3.22217, "de": 3.49368, "vi": 2.76849, "tr": 3.41009, "fi": 3.53108, "ru": 2.35905, "el": 2.58482, "hi": 1.88994, "ta": 1.5821, "th": 1.93115, "zh": 4.15104, "my": 1.95397}, "gini_cpc": 0.316443, "gini_cps": 0.346591, "bpic": 3.568595, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.102561855851892, "vi": 1.9163189514884542, "tr": 1.541653924287649, "fi": 1.1646125409930728, "ru": 0.9522090005459011, "el": 3.3564476231041036, "hi": 4.314859712045909, "ta": 3.038635606975865, "th": 2.667307796404875, "zh": 1.3666426311057138, "my": 0.5734443244851609}, "t": 41.3} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.23158, "de": 3.48972, "vi": 2.77376, "tr": 3.41862, "fi": 3.52324, "ru": 2.34157, "el": 2.58023, "hi": 1.88697, "ta": 1.59797, "th": 1.95977, "zh": 4.15751, "my": 1.96968}, "bytes_per_chunk": {"en": 3.74, "de": 3.307, "vi": 2.55, "tr": 2.591, "fi": 3.313, "ru": 7.191, "el": 2.011, "hi": 2.202, "ta": 3.327, "th": 3.817, "zh": 6.224, "my": 15.761}, "bpb_mean": 2.74422, "t": 42.8} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.02711, "bpb": 2.9245, "chunk_loss": 0.036554, "F": 0.27737, "G": 0.42884, "chunks": 284.03, "bytes_per_chunk": 3.605, "boundary_entropy": 0.65473, "router_noise": 0.0, "grad_norm": 0.909, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.099, "alpha": 0.03, "t": 42.8} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.11435, "bpb": 3.05036, "chunk_loss": 0.033924, "F": 0.23495, "G": 0.41607, "chunks": 240.59, "bytes_per_chunk": 4.256, "boundary_entropy": 0.6524, "router_noise": 0.0, "grad_norm": 0.7492, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.049, "alpha": 0.03, "t": 45.3} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.95957, "bpb": 2.82707, "chunk_loss": 0.036005, "F": 0.2691, "G": 0.42317, "chunks": 275.56, "bytes_per_chunk": 3.716, "boundary_entropy": 0.65237, "router_noise": 0.0, "grad_norm": 1.0349, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 47.7} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.91427, "bpb": 2.76171, "chunk_loss": 0.03591, "F": 0.26401, "G": 0.43047, "chunks": 270.34, "bytes_per_chunk": 3.788, "boundary_entropy": 0.65864, "router_noise": 0.0, "grad_norm": 0.8965, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 50.0} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5292.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.304086, "cps": 41.34375, "mean_p": 0.447306, "boundary_entropy": 0.664083, "bpb": 2.933472}, "de": {"chunks": 5630.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.274193, "cps": 43.984375, "mean_p": 0.441298, "boundary_entropy": 0.666255, "bpb": 3.246896}, "vi": {"chunks": 6754.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.366368, "cps": 52.765625, "mean_p": 0.454207, "boundary_entropy": 0.671345, "bpb": 2.573275}, "tr": {"chunks": 6645.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.367066, "cps": 51.914062, "mean_p": 0.445071, "boundary_entropy": 0.660255, "bpb": 3.154749}, "fi": {"chunks": 5504.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.292595, "cps": 43.0, "mean_p": 0.442748, "boundary_entropy": 0.663324, "bpb": 3.361077}, "ru": {"chunks": 5850.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.308707, "cps": 45.703125, "mean_p": 0.426399, "boundary_entropy": 0.658734, "bpb": 2.149622}, "el": {"chunks": 6207.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.302441, "cps": 48.492188, "mean_p": 0.44264, "boundary_entropy": 0.663042, "bpb": 2.28076}, "hi": {"chunks": 6032.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.358514, "cps": 47.125, "mean_p": 0.408854, "boundary_entropy": 0.656657, "bpb": 1.643797}, "ta": {"chunks": 12817.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.685584, "cps": 100.132812, "mean_p": 0.382823, "boundary_entropy": 0.636904, "bpb": 1.415817}, "th": {"chunks": 13514.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.840162, "cps": 105.578125, "mean_p": 0.426765, "boundary_entropy": 0.649368, "bpb": 1.757858}, "zh": {"chunks": 4798.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.812394, "cps": 37.484375, "mean_p": 0.430317, "boundary_entropy": 0.651918, "bpb": 3.947883}, "my": {"chunks": 11341.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.598533, "cps": 88.601562, "mean_p": 0.415602, "boundary_entropy": 0.659775, "bpb": 1.721533}}, "bpb": {"en": 2.93347, "de": 3.2469, "vi": 2.57327, "tr": 3.15475, "fi": 3.36108, "ru": 2.14962, "el": 2.28076, "hi": 1.6438, "ta": 1.41582, "th": 1.75786, "zh": 3.94788, "my": 1.72153}, "gini_cpc": 0.230479, "gini_cps": 0.196052, "bpic": 4.246117, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9016962683708201, "vi": 1.204820085524491, "tr": 1.2071151998334548, "fi": 0.9622121277763492, "ru": 1.0151984276560229, "el": 0.9945925112724034, "hi": 1.178991148713076, "ta": 2.254577660608685, "th": 2.7629125176129925, "zh": 2.6715978523704322, "my": 1.968304380720693}, "t": 53.3} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.94301, "de": 3.24463, "vi": 2.57804, "tr": 3.16813, "fi": 3.34993, "ru": 2.13077, "el": 2.27313, "hi": 1.6399, "ta": 1.42888, "th": 1.78821, "zh": 3.95947, "my": 1.73545}, "bytes_per_chunk": {"en": 3.17, "de": 3.607, "vi": 3.717, "tr": 2.752, "fi": 3.539, "ru": 6.232, "el": 6.078, "hi": 6.733, "ta": 4.255, "th": 3.259, "zh": 2.701, "my": 4.522}, "bpb_mean": 2.51996, "t": 54.7} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.94005, "bpb": 2.79889, "chunk_loss": 0.036349, "F": 0.2742, "G": 0.43528, "chunks": 280.78, "bytes_per_chunk": 3.647, "boundary_entropy": 0.66264, "router_noise": 0.0, "grad_norm": 0.8538, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.095, "alpha": 0.03, "t": 54.8} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.72258, "bpb": 2.48516, "chunk_loss": 0.034997, "F": 0.24927, "G": 0.43212, "chunks": 255.25, "bytes_per_chunk": 4.012, "boundary_entropy": 0.66343, "router_noise": 0.0, "grad_norm": 0.8387, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 57.0} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.75311, "bpb": 2.5292, "chunk_loss": 0.036699, "F": 0.27844, "G": 0.43653, "chunks": 285.12, "bytes_per_chunk": 3.591, "boundary_entropy": 0.66501, "router_noise": 0.0, "grad_norm": 0.7142, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 59.3} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.68902, "bpb": 2.43674, "chunk_loss": 0.034693, "F": 0.2453, "G": 0.42887, "chunks": 251.19, "bytes_per_chunk": 4.077, "boundary_entropy": 0.66285, "router_noise": 0.0, "grad_norm": 0.7364, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 61.5} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5194.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.298454, "cps": 40.578125, "mean_p": 0.444528, "boundary_entropy": 0.664804, "bpb": 2.72969}, "de": {"chunks": 5779.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.281449, "cps": 45.148438, "mean_p": 0.445685, "boundary_entropy": 0.668368, "bpb": 3.053589}, "vi": {"chunks": 6401.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.34722, "cps": 50.007812, "mean_p": 0.451409, "boundary_entropy": 0.67155, "bpb": 2.445015}, "tr": {"chunks": 6617.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.36552, "cps": 51.695312, "mean_p": 0.450159, "boundary_entropy": 0.662552, "bpb": 2.995826}, "fi": {"chunks": 6106.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.324597, "cps": 47.703125, "mean_p": 0.455946, "boundary_entropy": 0.668381, "bpb": 3.227648}, "ru": {"chunks": 6414.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.33847, "cps": 50.109375, "mean_p": 0.422687, "boundary_entropy": 0.661056, "bpb": 2.024086}, "el": {"chunks": 4179.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.203625, "cps": 32.648438, "mean_p": 0.430898, "boundary_entropy": 0.659868, "bpb": 2.198443}, "hi": {"chunks": 10247.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.609034, "cps": 80.054688, "mean_p": 0.42496, "boundary_entropy": 0.663803, "bpb": 1.577099}, "ta": {"chunks": 9959.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.532709, "cps": 77.804688, "mean_p": 0.390879, "boundary_entropy": 0.645684, "bpb": 1.314888}, "th": {"chunks": 10894.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.677277, "cps": 85.109375, "mean_p": 0.440124, "boundary_entropy": 0.667706, "bpb": 1.649394}, "zh": {"chunks": 2423.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.410261, "cps": 18.929688, "mean_p": 0.421839, "boundary_entropy": 0.655712, "bpb": 3.838435}, "my": {"chunks": 15949.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.841725, "cps": 124.601562, "mean_p": 0.432043, "boundary_entropy": 0.670235, "bpb": 1.490845}}, "bpb": {"en": 2.72969, "de": 3.05359, "vi": 2.44501, "tr": 2.99583, "fi": 3.22765, "ru": 2.02409, "el": 2.19844, "hi": 1.5771, "ta": 1.31489, "th": 1.64939, "zh": 3.83843, "my": 1.49085}, "gini_cpc": 0.223533, "gini_cps": 0.246813, "bpic": 4.256572, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9430233844478982, "vi": 1.1633941084301753, "tr": 1.2247085735454106, "fi": 1.0875947222566402, "ru": 1.13407536400332, "el": 0.6822659555061029, "hi": 2.040627984856226, "ta": 1.7848940334080212, "th": 2.2692820641784492, "zh": 1.374618379512895, "my": 2.8202801900570713}, "t": 64.7} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.74056, "de": 3.05424, "vi": 2.45302, "tr": 3.00799, "fi": 3.22434, "ru": 2.00641, "el": 2.18584, "hi": 1.57282, "ta": 1.32587, "th": 1.67601, "zh": 3.85103, "my": 1.50243}, "bytes_per_chunk": {"en": 3.349, "de": 3.487, "vi": 4.103, "tr": 2.789, "fi": 3.164, "ru": 5.466, "el": 8.592, "hi": 4.132, "ta": 5.071, "th": 3.961, "zh": 5.178, "my": 3.317}, "bpb_mean": 2.38338, "t": 66.0} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.83052, "bpb": 2.64088, "chunk_loss": 0.03523, "F": 0.25473, "G": 0.43676, "chunks": 260.84, "bytes_per_chunk": 3.926, "boundary_entropy": 0.6666, "router_noise": 0.0, "grad_norm": 0.9178, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 66.1} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.86603, "bpb": 2.69211, "chunk_loss": 0.036174, "F": 0.26913, "G": 0.43903, "chunks": 275.59, "bytes_per_chunk": 3.716, "boundary_entropy": 0.66697, "router_noise": 0.0, "grad_norm": 0.8425, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 68.3} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.67932, "bpb": 2.42275, "chunk_loss": 0.034624, "F": 0.2453, "G": 0.4336, "chunks": 251.19, "bytes_per_chunk": 4.077, "boundary_entropy": 0.6664, "router_noise": 0.0, "grad_norm": 0.9503, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 70.5} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.81753, "bpb": 2.62214, "chunk_loss": 0.034614, "F": 0.24518, "G": 0.43545, "chunks": 251.06, "bytes_per_chunk": 4.079, "boundary_entropy": 0.66647, "router_noise": 0.0, "grad_norm": 0.8371, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 72.7} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5076.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.291674, "cps": 39.65625, "mean_p": 0.447919, "boundary_entropy": 0.665886, "bpb": 2.592808}, "de": {"chunks": 5948.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.28968, "cps": 46.46875, "mean_p": 0.449919, "boundary_entropy": 0.66993, "bpb": 2.923982}, "vi": {"chunks": 5248.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.284676, "cps": 41.0, "mean_p": 0.443701, "boundary_entropy": 0.669822, "bpb": 2.322893}, "tr": {"chunks": 6123.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.338231, "cps": 47.835938, "mean_p": 0.448591, "boundary_entropy": 0.662607, "bpb": 2.897807}, "fi": {"chunks": 5248.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.278986, "cps": 41.0, "mean_p": 0.446464, "boundary_entropy": 0.66687, "bpb": 3.145786}, "ru": {"chunks": 8576.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.452559, "cps": 67.0, "mean_p": 0.428259, "boundary_entropy": 0.664344, "bpb": 1.934508}, "el": {"chunks": 9688.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.472056, "cps": 75.6875, "mean_p": 0.453773, "boundary_entropy": 0.669371, "bpb": 2.075743}, "hi": {"chunks": 3620.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.215156, "cps": 28.28125, "mean_p": 0.404032, "boundary_entropy": 0.656898, "bpb": 1.539729}, "ta": {"chunks": 6020.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.322011, "cps": 47.03125, "mean_p": 0.384889, "boundary_entropy": 0.643442, "bpb": 1.278172}, "th": {"chunks": 9430.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.58626, "cps": 73.671875, "mean_p": 0.436426, "boundary_entropy": 0.667082, "bpb": 1.587469}, "zh": {"chunks": 2214.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.374873, "cps": 17.296875, "mean_p": 0.422562, "boundary_entropy": 0.657874, "bpb": 3.685815}, "my": {"chunks": 4201.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.221712, "cps": 32.820312, "mean_p": 0.424572, "boundary_entropy": 0.675673, "bpb": 1.520271}}, "bpb": {"en": 2.59281, "de": 2.92398, "vi": 2.32289, "tr": 2.89781, "fi": 3.14579, "ru": 1.93451, "el": 2.07574, "hi": 1.53973, "ta": 1.27817, "th": 1.58747, "zh": 3.68582, "my": 1.52027}, "gini_cpc": 0.165082, "gini_cps": 0.20238, "bpic": 5.375686, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9931642069808859, "vi": 0.9760075814262751, "tr": 1.1596213824517936, "fi": 0.9564988444842581, "ru": 1.5515939045765574, "el": 1.618436974764748, "hi": 0.7376596095953688, "ta": 1.104011325261745, "th": 2.009986471103939, "zh": 1.2852472422563421, "my": 0.7601368945043571}, "t": 75.8} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.59924, "de": 2.92562, "vi": 2.32823, "tr": 2.9109, "fi": 3.145, "ru": 1.91666, "el": 2.06458, "hi": 1.5359, "ta": 1.28891, "th": 1.6119, "zh": 3.69565, "my": 1.53087}, "bytes_per_chunk": {"en": 3.365, "de": 3.437, "vi": 5.253, "tr": 3.126, "fi": 3.837, "ru": 3.839, "el": 3.782, "hi": 11.555, "ta": 8.156, "th": 4.647, "zh": 5.695, "my": 12.232}, "bpb_mean": 2.29612, "t": 77.1} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.75337, "bpb": 2.52958, "chunk_loss": 0.034479, "F": 0.24146, "G": 0.43491, "chunks": 247.25, "bytes_per_chunk": 4.142, "boundary_entropy": 0.66637, "router_noise": 0.0, "grad_norm": 0.8415, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 77.2} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.6138, "bpb": 2.32822, "chunk_loss": 0.033414, "F": 0.22305, "G": 0.43374, "chunks": 228.41, "bytes_per_chunk": 4.483, "boundary_entropy": 0.66769, "router_noise": 0.0, "grad_norm": 0.8016, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 79.3} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.81264, "bpb": 2.61509, "chunk_loss": 0.035108, "F": 0.25198, "G": 0.43904, "chunks": 258.03, "bytes_per_chunk": 3.969, "boundary_entropy": 0.66868, "router_noise": 0.0, "grad_norm": 0.7801, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 81.5} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.66301, "bpb": 2.39922, "chunk_loss": 0.03472, "F": 0.24606, "G": 0.43444, "chunks": 251.97, "bytes_per_chunk": 4.064, "boundary_entropy": 0.66725, "router_noise": 0.0, "grad_norm": 0.7652, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 83.7} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4862.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.279377, "cps": 37.984375, "mean_p": 0.445516, "boundary_entropy": 0.665279, "bpb": 2.485828}, "de": {"chunks": 4936.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.240394, "cps": 38.5625, "mean_p": 0.444999, "boundary_entropy": 0.669762, "bpb": 2.815693}, "vi": {"chunks": 5762.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.312558, "cps": 45.015625, "mean_p": 0.45099, "boundary_entropy": 0.672834, "bpb": 2.280081}, "tr": {"chunks": 6366.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.351654, "cps": 49.734375, "mean_p": 0.450269, "boundary_entropy": 0.66343, "bpb": 2.783331}, "fi": {"chunks": 5037.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.267769, "cps": 39.351562, "mean_p": 0.446892, "boundary_entropy": 0.668712, "bpb": 3.055167}, "ru": {"chunks": 6389.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.33715, "cps": 49.914062, "mean_p": 0.424734, "boundary_entropy": 0.665532, "bpb": 1.866286}, "el": {"chunks": 10253.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.499586, "cps": 80.101562, "mean_p": 0.451916, "boundary_entropy": 0.670226, "bpb": 1.975765}, "hi": {"chunks": 7505.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.446062, "cps": 58.632812, "mean_p": 0.409406, "boundary_entropy": 0.655931, "bpb": 1.51738}, "ta": {"chunks": 7501.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.40123, "cps": 58.601562, "mean_p": 0.395562, "boundary_entropy": 0.651623, "bpb": 1.295072}, "th": {"chunks": 11989.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.745353, "cps": 93.664062, "mean_p": 0.454, "boundary_entropy": 0.671267, "bpb": 1.518875}, "zh": {"chunks": 3123.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.528784, "cps": 24.398438, "mean_p": 0.430413, "boundary_entropy": 0.661503, "bpb": 3.625837}, "my": {"chunks": 7755.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.409278, "cps": 60.585938, "mean_p": 0.430365, "boundary_entropy": 0.674158, "bpb": 1.407899}}, "bpb": {"en": 2.48583, "de": 2.81569, "vi": 2.28008, "tr": 2.78333, "fi": 3.05517, "ru": 1.86629, "el": 1.97577, "hi": 1.51738, "ta": 1.29507, "th": 1.51887, "zh": 3.62584, "my": 1.4079}, "gini_cpc": 0.179467, "gini_cps": 0.187533, "bpic": 4.710241, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.860462423833869, "vi": 1.1187660470482468, "tr": 1.258708742216513, "fi": 0.9584494954803942, "ru": 1.2067931587270406, "el": 1.7882131239698504, "hi": 1.5966318534188477, "ta": 1.4361601160194681, "th": 2.667909298184125, "zh": 1.8927258226126775, "my": 1.464966156482567}, "t": 86.9} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.497, "de": 2.82173, "vi": 2.2865, "tr": 2.79577, "fi": 3.05757, "ru": 1.84834, "el": 1.96544, "hi": 1.5144, "ta": 1.3023, "th": 1.54294, "zh": 3.63615, "my": 1.41796}, "bytes_per_chunk": {"en": 3.423, "de": 4.094, "vi": 4.457, "tr": 3.136, "fi": 4.193, "ru": 5.277, "el": 3.496, "hi": 5.779, "ta": 6.911, "th": 3.862, "zh": 4.247, "my": 6.51}, "bpb_mean": 2.22384, "t": 88.2} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.72451, "bpb": 2.48794, "chunk_loss": 0.035638, "F": 0.26086, "G": 0.43914, "chunks": 267.12, "bytes_per_chunk": 3.833, "boundary_entropy": 0.66795, "router_noise": 0.0, "grad_norm": 0.7831, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 88.3} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.84373, "bpb": 2.65994, "chunk_loss": 0.035306, "F": 0.25455, "G": 0.43905, "chunks": 260.66, "bytes_per_chunk": 3.929, "boundary_entropy": 0.66845, "router_noise": 0.0, "grad_norm": 0.8789, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 90.5} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.52479, "bpb": 2.19981, "chunk_loss": 0.035402, "F": 0.25629, "G": 0.43724, "chunks": 262.44, "bytes_per_chunk": 3.902, "boundary_entropy": 0.66806, "router_noise": 0.0, "grad_norm": 0.7869, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 92.6} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.52172, "bpb": 2.19538, "chunk_loss": 0.034792, "F": 0.24649, "G": 0.43678, "chunks": 252.41, "bytes_per_chunk": 4.057, "boundary_entropy": 0.66777, "router_noise": 0.0, "grad_norm": 0.8247, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 94.8} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4589.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.26369, "cps": 35.851562, "mean_p": 0.446036, "boundary_entropy": 0.665329, "bpb": 2.428424}, "de": {"chunks": 5087.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.247748, "cps": 39.742188, "mean_p": 0.447395, "boundary_entropy": 0.670224, "bpb": 2.749546}, "vi": {"chunks": 4769.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.258693, "cps": 37.257812, "mean_p": 0.445534, "boundary_entropy": 0.672144, "bpb": 2.202745}, "tr": {"chunks": 5812.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.321052, "cps": 45.40625, "mean_p": 0.450108, "boundary_entropy": 0.663403, "bpb": 2.720576}, "fi": {"chunks": 5230.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.278029, "cps": 40.859375, "mean_p": 0.445858, "boundary_entropy": 0.667162, "bpb": 2.983633}, "ru": {"chunks": 7897.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.416728, "cps": 61.695312, "mean_p": 0.432486, "boundary_entropy": 0.668947, "bpb": 1.803798}, "el": {"chunks": 7862.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.383082, "cps": 61.421875, "mean_p": 0.447596, "boundary_entropy": 0.670742, "bpb": 1.904173}, "hi": {"chunks": 4768.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.283388, "cps": 37.25, "mean_p": 0.403933, "boundary_entropy": 0.655711, "bpb": 1.465246}, "ta": {"chunks": 4704.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.251618, "cps": 36.75, "mean_p": 0.384765, "boundary_entropy": 0.647768, "bpb": 1.19331}, "th": {"chunks": 12240.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.760957, "cps": 95.625, "mean_p": 0.452985, "boundary_entropy": 0.672155, "bpb": 1.457388}, "zh": {"chunks": 2823.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.477988, "cps": 22.054688, "mean_p": 0.427003, "boundary_entropy": 0.661205, "bpb": 3.576883}, "my": {"chunks": 14583.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.769633, "cps": 113.929688, "mean_p": 0.44331, "boundary_entropy": 0.676461, "bpb": 1.339753}}, "bpb": {"en": 2.42842, "de": 2.74955, "vi": 2.20275, "tr": 2.72058, "fi": 2.98363, "ru": 1.8038, "el": 1.90417, "hi": 1.46525, "ta": 1.19331, "th": 1.45739, "zh": 3.57688, "my": 1.33975}, "gini_cpc": 0.229489, "gini_cps": 0.248154, "bpic": 4.775534, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9395402563096824, "vi": 0.9810479689199115, "tr": 1.2175340528298475, "fi": 1.054376864544895, "ru": 1.5803707615607128, "el": 1.452774662556709, "hi": 1.0746999689482926, "ta": 0.9542186622416676, "th": 2.8858012358484793, "zh": 1.8126898293354068, "my": 2.9187006997347926}, "t": 97.9} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.43725, "de": 2.75561, "vi": 2.20834, "tr": 2.73298, "fi": 2.9824, "ru": 1.78403, "el": 1.89404, "hi": 1.46167, "ta": 1.20187, "th": 1.4776, "zh": 3.58453, "my": 1.34975}, "bytes_per_chunk": {"en": 3.753, "de": 4.125, "vi": 5.239, "tr": 3.426, "fi": 3.791, "ru": 4.256, "el": 4.741, "hi": 9.642, "ta": 11.076, "th": 3.68, "zh": 4.82, "my": 3.649}, "bpb_mean": 2.15584, "t": 99.3} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.58031, "bpb": 2.2799, "chunk_loss": 0.035088, "F": 0.25128, "G": 0.44102, "chunks": 257.31, "bytes_per_chunk": 3.98, "boundary_entropy": 0.66909, "router_noise": 0.0, "grad_norm": 0.7446, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 99.3} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.67318, "bpb": 2.41389, "chunk_loss": 0.034474, "F": 0.24139, "G": 0.4356, "chunks": 247.19, "bytes_per_chunk": 4.143, "boundary_entropy": 0.66801, "router_noise": 0.0, "grad_norm": 0.8141, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 101.4} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.74684, "bpb": 2.52015, "chunk_loss": 0.034588, "F": 0.24332, "G": 0.43963, "chunks": 249.16, "bytes_per_chunk": 4.11, "boundary_entropy": 0.67072, "router_noise": 0.0, "grad_norm": 0.7395, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 103.6} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.61323, "bpb": 2.32739, "chunk_loss": 0.035353, "F": 0.25586, "G": 0.43917, "chunks": 262.0, "bytes_per_chunk": 3.908, "boundary_entropy": 0.66865, "router_noise": 0.0, "grad_norm": 0.8573, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 105.8} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4339.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.249325, "cps": 33.898438, "mean_p": 0.444739, "boundary_entropy": 0.665411, "bpb": 2.381796}, "de": {"chunks": 5387.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.262358, "cps": 42.085938, "mean_p": 0.449642, "boundary_entropy": 0.671074, "bpb": 2.674554}, "vi": {"chunks": 5425.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.294277, "cps": 42.382812, "mean_p": 0.451235, "boundary_entropy": 0.672761, "bpb": 2.170455}, "tr": {"chunks": 6769.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.373916, "cps": 52.882812, "mean_p": 0.452702, "boundary_entropy": 0.664681, "bpb": 2.63905}, "fi": {"chunks": 6005.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.319228, "cps": 46.914062, "mean_p": 0.458616, "boundary_entropy": 0.670489, "bpb": 2.919571}, "ru": {"chunks": 6698.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.353456, "cps": 52.328125, "mean_p": 0.431251, "boundary_entropy": 0.66854, "bpb": 1.742442}, "el": {"chunks": 6581.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.320665, "cps": 51.414062, "mean_p": 0.441282, "boundary_entropy": 0.670608, "bpb": 1.867485}, "hi": {"chunks": 6853.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.407311, "cps": 53.539062, "mean_p": 0.414193, "boundary_entropy": 0.661523, "bpb": 1.440437}, "ta": {"chunks": 3673.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.19647, "cps": 28.695312, "mean_p": 0.39593, "boundary_entropy": 0.652784, "bpb": 1.202091}, "th": {"chunks": 9297.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.577992, "cps": 72.632812, "mean_p": 0.441042, "boundary_entropy": 0.671119, "bpb": 1.392265}, "zh": {"chunks": 2947.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.498984, "cps": 23.023438, "mean_p": 0.428618, "boundary_entropy": 0.662314, "bpb": 3.561711}, "my": {"chunks": 15973.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.842991, "cps": 124.789062, "mean_p": 0.44485, "boundary_entropy": 0.675414, "bpb": 1.299951}}, "bpb": {"en": 2.3818, "de": 2.67455, "vi": 2.17046, "tr": 2.63905, "fi": 2.91957, "ru": 1.74244, "el": 1.86748, "hi": 1.44044, "ta": 1.20209, "th": 1.39227, "zh": 3.56171, "my": 1.29995}, "gini_cpc": 0.221564, "gini_cps": 0.232242, "bpic": 4.800443, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0522744812938463, "vi": 1.1802963677779317, "tr": 1.4997139553178052, "fi": 1.2803703202152967, "ru": 1.4176544940318556, "el": 1.2861319163474136, "hi": 1.6336541824778, "ta": 0.7880067341731798, "th": 2.318228473804932, "zh": 2.0013413260072705, "my": 3.3810966518524053}, "t": 108.9} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.39239, "de": 2.67992, "vi": 2.1765, "tr": 2.64935, "fi": 2.92137, "ru": 1.72169, "el": 1.85762, "hi": 1.43822, "ta": 1.20884, "th": 1.40883, "zh": 3.56666, "my": 1.31267}, "bytes_per_chunk": {"en": 3.878, "de": 3.73, "vi": 4.723, "tr": 2.881, "fi": 3.281, "ru": 5.038, "el": 5.671, "hi": 6.296, "ta": 15.299, "th": 4.775, "zh": 5.11, "my": 3.362}, "bpb_mean": 2.11117, "t": 110.2} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.6078, "bpb": 2.31957, "chunk_loss": 0.034302, "F": 0.23746, "G": 0.4386, "chunks": 243.16, "bytes_per_chunk": 4.211, "boundary_entropy": 0.66929, "router_noise": 0.0, "grad_norm": 0.8761, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 110.3} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.58783, "bpb": 2.29075, "chunk_loss": 0.034253, "F": 0.23669, "G": 0.43483, "chunks": 242.38, "bytes_per_chunk": 4.225, "boundary_entropy": 0.66765, "router_noise": 0.0, "grad_norm": 0.8661, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 112.4} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.55593, "bpb": 2.24473, "chunk_loss": 0.033216, "F": 0.21964, "G": 0.43813, "chunks": 224.91, "bytes_per_chunk": 4.553, "boundary_entropy": 0.6705, "router_noise": 0.0, "grad_norm": 0.7545, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 114.6} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.56918, "bpb": 2.26385, "chunk_loss": 0.034453, "F": 0.242, "G": 0.43844, "chunks": 247.81, "bytes_per_chunk": 4.132, "boundary_entropy": 0.66971, "router_noise": 0.0, "grad_norm": 0.7958, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 116.7} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4940.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.283859, "cps": 38.59375, "mean_p": 0.447416, "boundary_entropy": 0.665632, "bpb": 2.326525}, "de": {"chunks": 5811.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.283008, "cps": 45.398438, "mean_p": 0.453145, "boundary_entropy": 0.671725, "bpb": 2.616113}, "vi": {"chunks": 6451.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.349932, "cps": 50.398438, "mean_p": 0.455527, "boundary_entropy": 0.673282, "bpb": 2.130254}, "tr": {"chunks": 6164.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.340496, "cps": 48.15625, "mean_p": 0.451189, "boundary_entropy": 0.6641, "bpb": 2.622419}, "fi": {"chunks": 5738.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.305034, "cps": 44.828125, "mean_p": 0.457301, "boundary_entropy": 0.669906, "bpb": 2.848928}, "ru": {"chunks": 6126.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.323272, "cps": 47.859375, "mean_p": 0.429607, "boundary_entropy": 0.668731, "bpb": 1.709531}, "el": {"chunks": 6837.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.333138, "cps": 53.414062, "mean_p": 0.442292, "boundary_entropy": 0.672719, "bpb": 1.798647}, "hi": {"chunks": 7209.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.42847, "cps": 56.320312, "mean_p": 0.420647, "boundary_entropy": 0.664406, "bpb": 1.398472}, "ta": {"chunks": 8923.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.477293, "cps": 69.710938, "mean_p": 0.408621, "boundary_entropy": 0.657347, "bpb": 1.137035}, "th": {"chunks": 10413.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.647373, "cps": 81.351562, "mean_p": 0.441643, "boundary_entropy": 0.670121, "bpb": 1.361849}, "zh": {"chunks": 2830.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.479174, "cps": 22.109375, "mean_p": 0.42979, "boundary_entropy": 0.663449, "bpb": 3.506014}, "my": {"chunks": 11928.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.629512, "cps": 93.1875, "mean_p": 0.432911, "boundary_entropy": 0.672594, "bpb": 1.267024}}, "bpb": {"en": 2.32653, "de": 2.61611, "vi": 2.13025, "tr": 2.62242, "fi": 2.84893, "ru": 1.70953, "el": 1.79865, "hi": 1.39847, "ta": 1.13703, "th": 1.36185, "zh": 3.50601, "my": 1.26702}, "gini_cpc": 0.160559, "gini_cps": 0.18096, "bpic": 4.603347, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9970011047684472, "vi": 1.2327672015364193, "tr": 1.1995248511609569, "fi": 1.074597514547544, "ru": 1.138845865424674, "el": 1.1736048781844641, "hi": 1.5094444116921635, "ta": 1.6814447236861056, "th": 2.2806149894475074, "zh": 1.688068882122209, "my": 2.217692999976924}, "t": 119.8} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.33559, "de": 2.61984, "vi": 2.13866, "tr": 2.63155, "fi": 2.85021, "ru": 1.69001, "el": 1.7901, "hi": 1.39605, "ta": 1.14374, "th": 1.38269, "zh": 3.5104, "my": 1.27639}, "bytes_per_chunk": {"en": 3.537, "de": 3.523, "vi": 3.581, "tr": 3.19, "fi": 3.368, "ru": 5.382, "el": 5.532, "hi": 5.862, "ta": 5.975, "th": 4.274, "zh": 5.481, "my": 4.373}, "bpb_mean": 2.06377, "t": 121.2} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.53816, "bpb": 2.2191, "chunk_loss": 0.034335, "F": 0.23849, "G": 0.43984, "chunks": 244.22, "bytes_per_chunk": 4.193, "boundary_entropy": 0.6705, "router_noise": 0.0, "grad_norm": 0.7458, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 121.2} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.58571, "bpb": 2.2877, "chunk_loss": 0.035351, "F": 0.25424, "G": 0.44324, "chunks": 260.34, "bytes_per_chunk": 3.933, "boundary_entropy": 0.6702, "router_noise": 0.0, "grad_norm": 0.7438, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 123.4} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.63986, "bpb": 2.36582, "chunk_loss": 0.033149, "F": 0.21823, "G": 0.43282, "chunks": 223.47, "bytes_per_chunk": 4.582, "boundary_entropy": 0.667, "router_noise": 0.0, "grad_norm": 0.7538, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 125.5} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.48558, "bpb": 2.14324, "chunk_loss": 0.03331, "F": 0.22235, "G": 0.43685, "chunks": 227.69, "bytes_per_chunk": 4.497, "boundary_entropy": 0.67083, "router_noise": 0.0, "grad_norm": 0.7538, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 127.7} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4304.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.247314, "cps": 33.625, "mean_p": 0.444974, "boundary_entropy": 0.665565, "bpb": 2.294995}, "de": {"chunks": 5063.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.246579, "cps": 39.554688, "mean_p": 0.446197, "boundary_entropy": 0.670083, "bpb": 2.583736}, "vi": {"chunks": 5851.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.317385, "cps": 45.710938, "mean_p": 0.451722, "boundary_entropy": 0.67238, "bpb": 2.105065}, "tr": {"chunks": 6043.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.333812, "cps": 47.210938, "mean_p": 0.448913, "boundary_entropy": 0.662988, "bpb": 2.553711}, "fi": {"chunks": 5543.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.294668, "cps": 43.304688, "mean_p": 0.454499, "boundary_entropy": 0.668729, "bpb": 2.787109}, "ru": {"chunks": 6744.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.355884, "cps": 52.6875, "mean_p": 0.434104, "boundary_entropy": 0.67056, "bpb": 1.670316}, "el": {"chunks": 7984.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.389027, "cps": 62.375, "mean_p": 0.448515, "boundary_entropy": 0.673684, "bpb": 1.743817}, "hi": {"chunks": 9243.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.549361, "cps": 72.210938, "mean_p": 0.419504, "boundary_entropy": 0.661447, "bpb": 1.375586}, "ta": {"chunks": 10291.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.550468, "cps": 80.398438, "mean_p": 0.412066, "boundary_entropy": 0.658258, "bpb": 1.105642}, "th": {"chunks": 10666.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.663102, "cps": 83.328125, "mean_p": 0.444694, "boundary_entropy": 0.672416, "bpb": 1.322191}, "zh": {"chunks": 3096.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.524213, "cps": 24.1875, "mean_p": 0.436951, "boundary_entropy": 0.667234, "bpb": 3.462063}, "my": {"chunks": 12282.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.648195, "cps": 95.953125, "mean_p": 0.435941, "boundary_entropy": 0.674128, "bpb": 1.248105}}, "bpb": {"en": 2.29499, "de": 2.58374, "vi": 2.10507, "tr": 2.55371, "fi": 2.78711, "ru": 1.67032, "el": 1.74382, "hi": 1.37559, "ta": 1.10564, "th": 1.32219, "zh": 3.46206, "my": 1.24811}, "gini_cpc": 0.191154, "gini_cps": 0.21094, "bpic": 4.405705, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9970280523129237, "vi": 1.2833313798203876, "tr": 1.349751753544926, "fi": 1.191474776286674, "ru": 1.4389980480819218, "el": 1.5730102068818608, "hi": 2.221312894601656, "ta": 2.225788869158702, "th": 2.681219514590818, "zh": 2.1196266288331316, "my": 2.620942990716914}, "t": 130.8} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.30294, "de": 2.5922, "vi": 2.11276, "tr": 2.5608, "fi": 2.79337, "ru": 1.64981, "el": 1.73553, "hi": 1.37401, "ta": 1.11234, "th": 1.33957, "zh": 3.46668, "my": 1.25757}, "bytes_per_chunk": {"en": 3.866, "de": 4.08, "vi": 4.056, "tr": 3.198, "fi": 3.529, "ru": 5.029, "el": 4.832, "hi": 4.452, "ta": 5.058, "th": 4.17, "zh": 5.056, "my": 4.316}, "bpb_mean": 2.0248, "t": 132.2} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.60206, "bpb": 2.31128, "chunk_loss": 0.034499, "F": 0.24185, "G": 0.4418, "chunks": 247.66, "bytes_per_chunk": 4.135, "boundary_entropy": 0.67091, "router_noise": 0.0, "grad_norm": 0.7845, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 132.2} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.61653, "bpb": 2.33216, "chunk_loss": 0.033698, "F": 0.22791, "G": 0.43966, "chunks": 233.38, "bytes_per_chunk": 4.388, "boundary_entropy": 0.67116, "router_noise": 0.0, "grad_norm": 0.7278, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 134.4} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.55665, "bpb": 2.24578, "chunk_loss": 0.03505, "F": 0.25076, "G": 0.44008, "chunks": 256.78, "bytes_per_chunk": 3.988, "boundary_entropy": 0.66968, "router_noise": 0.0, "grad_norm": 0.7678, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 136.5} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.50961, "bpb": 2.17791, "chunk_loss": 0.034931, "F": 0.24847, "G": 0.44071, "chunks": 254.44, "bytes_per_chunk": 4.025, "boundary_entropy": 0.67002, "router_noise": 0.0, "grad_norm": 0.7584, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 138.6} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4426.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.254324, "cps": 34.578125, "mean_p": 0.445592, "boundary_entropy": 0.665718, "bpb": 2.259136}, "de": {"chunks": 4889.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.238105, "cps": 38.195312, "mean_p": 0.447783, "boundary_entropy": 0.671165, "bpb": 2.535651}, "vi": {"chunks": 5862.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.317982, "cps": 45.796875, "mean_p": 0.452766, "boundary_entropy": 0.673137, "bpb": 2.091615}, "tr": {"chunks": 5572.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.307794, "cps": 43.53125, "mean_p": 0.452314, "boundary_entropy": 0.666593, "bpb": 2.518236}, "fi": {"chunks": 5197.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.276275, "cps": 40.601562, "mean_p": 0.455885, "boundary_entropy": 0.670736, "bpb": 2.750367}, "ru": {"chunks": 6618.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.349235, "cps": 51.703125, "mean_p": 0.433634, "boundary_entropy": 0.670741, "bpb": 1.640507}, "el": {"chunks": 7690.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.374702, "cps": 60.078125, "mean_p": 0.446865, "boundary_entropy": 0.673684, "bpb": 1.711938}, "hi": {"chunks": 5351.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.318039, "cps": 41.804688, "mean_p": 0.416336, "boundary_entropy": 0.66469, "bpb": 1.352716}, "ta": {"chunks": 8726.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.466756, "cps": 68.171875, "mean_p": 0.403651, "boundary_entropy": 0.654602, "bpb": 1.101038}, "th": {"chunks": 10030.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.623562, "cps": 78.359375, "mean_p": 0.443272, "boundary_entropy": 0.671632, "bpb": 1.287471}, "zh": {"chunks": 2675.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.452929, "cps": 20.898438, "mean_p": 0.434581, "boundary_entropy": 0.667244, "bpb": 3.475415}, "my": {"chunks": 7648.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.403631, "cps": 59.75, "mean_p": 0.42533, "boundary_entropy": 0.672651, "bpb": 1.229195}}, "bpb": {"en": 2.25914, "de": 2.53565, "vi": 2.09161, "tr": 2.51824, "fi": 2.75037, "ru": 1.64051, "el": 1.71194, "hi": 1.35272, "ta": 1.10104, "th": 1.28747, "zh": 3.47541, "my": 1.22919}, "gini_cpc": 0.153404, "gini_cps": 0.173574, "bpic": 5.138731, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9362252302395124, "vi": 1.2503033153085714, "tr": 1.2102449160036912, "fi": 1.0863094160686044, "ru": 1.37318882067705, "el": 1.4733238026246285, "hi": 1.250525605482664, "ta": 1.8352804980465671, "th": 2.4518425541221434, "zh": 1.7809144361222338, "my": 1.5870741347332404}, "t": 141.7} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.26737, "de": 2.54434, "vi": 2.09596, "tr": 2.52318, "fi": 2.76001, "ru": 1.61955, "el": 1.7027, "hi": 1.35001, "ta": 1.10768, "th": 1.30485, "zh": 3.48204, "my": 1.23913}, "bytes_per_chunk": {"en": 3.825, "de": 4.186, "vi": 3.976, "tr": 3.533, "fi": 3.76, "ru": 5.129, "el": 5.021, "hi": 8.258, "ta": 6.326, "th": 4.426, "zh": 5.804, "my": 6.657}, "bpb_mean": 1.99973, "t": 143.0} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.45617, "bpb": 2.10081, "chunk_loss": 0.033291, "F": 0.22095, "G": 0.43831, "chunks": 226.25, "bytes_per_chunk": 4.526, "boundary_entropy": 0.67083, "router_noise": 0.0, "grad_norm": 0.7628, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 143.1} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.54985, "bpb": 2.23596, "chunk_loss": 0.03388, "F": 0.23169, "G": 0.43731, "chunks": 237.25, "bytes_per_chunk": 4.316, "boundary_entropy": 0.66956, "router_noise": 0.0, "grad_norm": 0.7667, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 145.2} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.48722, "bpb": 2.1456, "chunk_loss": 0.033595, "F": 0.2265, "G": 0.43797, "chunks": 231.94, "bytes_per_chunk": 4.415, "boundary_entropy": 0.67001, "router_noise": 0.0, "grad_norm": 0.8075, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 147.3} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.49603, "bpb": 2.15832, "chunk_loss": 0.033475, "F": 0.22452, "G": 0.43835, "chunks": 229.91, "bytes_per_chunk": 4.454, "boundary_entropy": 0.66995, "router_noise": 0.0, "grad_norm": 0.7541, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 149.4} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4081.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.2345, "cps": 31.882812, "mean_p": 0.441853, "boundary_entropy": 0.664612, "bpb": 2.236968}, "de": {"chunks": 5404.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.263186, "cps": 42.21875, "mean_p": 0.450003, "boundary_entropy": 0.67083, "bpb": 2.492977}, "vi": {"chunks": 6114.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.331652, "cps": 47.765625, "mean_p": 0.451748, "boundary_entropy": 0.672901, "bpb": 2.07681}, "tr": {"chunks": 6260.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.345799, "cps": 48.90625, "mean_p": 0.451708, "boundary_entropy": 0.663903, "bpb": 2.477169}, "fi": {"chunks": 5610.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.29823, "cps": 43.828125, "mean_p": 0.456722, "boundary_entropy": 0.669544, "bpb": 2.701549}, "ru": {"chunks": 6108.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.322322, "cps": 47.71875, "mean_p": 0.434355, "boundary_entropy": 0.67157, "bpb": 1.608047}, "el": {"chunks": 9221.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.449301, "cps": 72.039062, "mean_p": 0.454247, "boundary_entropy": 0.675634, "bpb": 1.698548}, "hi": {"chunks": 8522.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.506508, "cps": 66.578125, "mean_p": 0.423871, "boundary_entropy": 0.664281, "bpb": 1.326886}, "ta": {"chunks": 7440.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.397967, "cps": 58.125, "mean_p": 0.398095, "boundary_entropy": 0.653215, "bpb": 1.076827}, "th": {"chunks": 10388.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.645819, "cps": 81.15625, "mean_p": 0.439964, "boundary_entropy": 0.669921, "bpb": 1.272202}, "zh": {"chunks": 3074.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.520488, "cps": 24.015625, "mean_p": 0.435711, "boundary_entropy": 0.666515, "bpb": 3.410781}, "my": {"chunks": 11041.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.5827, "cps": 86.257812, "mean_p": 0.436053, "boundary_entropy": 0.674758, "bpb": 1.194853}}, "bpb": {"en": 2.23697, "de": 2.49298, "vi": 2.07681, "tr": 2.47717, "fi": 2.70155, "ru": 1.60805, "el": 1.69855, "hi": 1.32689, "ta": 1.07683, "th": 1.2722, "zh": 3.41078, "my": 1.19485}, "gini_cpc": 0.174918, "gini_cps": 0.189981, "bpic": 4.609262, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1223297074641758, "vi": 1.41429438731433, "tr": 1.4746240315649746, "fi": 1.2717697865889963, "ru": 1.3745082139446656, "el": 1.9159964597903993, "hi": 2.159951414757549, "ta": 1.6970904574188672, "th": 2.754028315483087, "zh": 2.219565399043896, "my": 2.484863628371581}, "t": 152.5} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.24541, "de": 2.50021, "vi": 2.08264, "tr": 2.48511, "fi": 2.7093, "ru": 1.5867, "el": 1.69088, "hi": 1.32507, "ta": 1.08317, "th": 1.29019, "zh": 3.41565, "my": 1.20488}, "bytes_per_chunk": {"en": 4.168, "de": 3.785, "vi": 3.794, "tr": 3.108, "fi": 3.466, "ru": 5.38, "el": 4.137, "hi": 4.963, "ta": 7.35, "th": 4.269, "zh": 5.116, "my": 4.861}, "bpb_mean": 1.96827, "t": 153.9} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 153.9} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "global", "lm_loss": 1.37567, "bpb": 1.98467, "chunk_loss": 0.034238, "F": 0.23633, "G": 0.4397, "chunks": 242.0, "bytes_per_chunk": 4.231, "boundary_entropy": 0.67006, "router_noise": 0.0, "grad_norm": 0.7335, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 153.9} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "global", "lm_loss": 1.46682, "bpb": 2.11617, "chunk_loss": 0.034207, "F": 0.23615, "G": 0.4428, "chunks": 241.81, "bytes_per_chunk": 4.235, "boundary_entropy": 0.67217, "router_noise": 0.0, "grad_norm": 0.7134, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 156.1} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "global", "lm_loss": 1.50253, "bpb": 2.16769, "chunk_loss": 0.034139, "F": 0.23627, "G": 0.43881, "chunks": 241.94, "bytes_per_chunk": 4.232, "boundary_entropy": 0.66985, "router_noise": 0.0, "grad_norm": 0.7561, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 158.2} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "global", "lm_loss": 1.57128, "bpb": 2.26687, "chunk_loss": 0.034458, "F": 0.24106, "G": 0.44142, "chunks": 246.84, "bytes_per_chunk": 4.148, "boundary_entropy": 0.67038, "router_noise": 0.0, "grad_norm": 0.7827, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 160.3} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4015.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.230707, "cps": 31.367188, "mean_p": 0.439853, "boundary_entropy": 0.664036, "bpb": 2.206066}, "de": {"chunks": 5226.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.254517, "cps": 40.828125, "mean_p": 0.448382, "boundary_entropy": 0.670016, "bpb": 2.451903}, "vi": {"chunks": 5169.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.280391, "cps": 40.382812, "mean_p": 0.449773, "boundary_entropy": 0.671642, "bpb": 2.05755}, "tr": {"chunks": 5907.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.3263, "cps": 46.148438, "mean_p": 0.452894, "boundary_entropy": 0.665547, "bpb": 2.45364}, "fi": {"chunks": 5210.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.276966, "cps": 40.703125, "mean_p": 0.454539, "boundary_entropy": 0.668782, "bpb": 2.674304}, "ru": {"chunks": 5081.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.268127, "cps": 39.695312, "mean_p": 0.434607, "boundary_entropy": 0.672199, "bpb": 1.595921}, "el": {"chunks": 7960.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.387858, "cps": 62.1875, "mean_p": 0.446491, "boundary_entropy": 0.674297, "bpb": 1.680516}, "hi": {"chunks": 8861.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.526657, "cps": 69.226562, "mean_p": 0.420849, "boundary_entropy": 0.66242, "bpb": 1.318485}, "ta": {"chunks": 9281.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.496443, "cps": 72.507812, "mean_p": 0.403163, "boundary_entropy": 0.653793, "bpb": 1.075958}, "th": {"chunks": 9915.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.616413, "cps": 77.460938, "mean_p": 0.441, "boundary_entropy": 0.672382, "bpb": 1.237039}, "zh": {"chunks": 3610.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.611243, "cps": 28.203125, "mean_p": 0.438946, "boundary_entropy": 0.666837, "bpb": 3.379368}, "my": {"chunks": 9305.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.491081, "cps": 72.695312, "mean_p": 0.430848, "boundary_entropy": 0.672935, "bpb": 1.187005}}, "bpb": {"en": 2.20607, "de": 2.4519, "vi": 2.05755, "tr": 2.45364, "fi": 2.6743, "ru": 1.59592, "el": 1.68052, "hi": 1.31849, "ta": 1.07596, "th": 1.23704, "zh": 3.37937, "my": 1.187}, "gini_cpc": 0.193251, "gini_cps": 0.182043, "bpic": 4.825006, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1032033420186405, "vi": 1.2153516664015231, "tr": 1.414343796797473, "fi": 1.2005062086761586, "ru": 1.1621937917505922, "el": 1.681166754621929, "hi": 2.282791434053947, "ta": 2.1518295826329252, "th": 2.6718386250503725, "zh": 2.64942926943029, "my": 2.1285878162716285}, "t": 163.4} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.21374, "de": 2.46146, "vi": 2.06023, "tr": 2.45684, "fi": 2.68123, "ru": 1.57462, "el": 1.67134, "hi": 1.31445, "ta": 1.08177, "th": 1.25334, "zh": 3.38319, "my": 1.1953}, "bytes_per_chunk": {"en": 4.233, "de": 3.91, "vi": 4.757, "tr": 3.271, "fi": 3.906, "ru": 6.54, "el": 4.852, "hi": 4.91, "ta": 5.837, "th": 4.409, "zh": 4.485, "my": 5.504}, "bpb_mean": 1.94563, "t": 164.7} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "global", "lm_loss": 1.60775, "bpb": 2.31949, "chunk_loss": 0.033563, "F": 0.22635, "G": 0.43973, "chunks": 231.78, "bytes_per_chunk": 4.418, "boundary_entropy": 0.67043, "router_noise": 0.0, "grad_norm": 0.8595, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 164.8} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "global", "lm_loss": 1.50377, "bpb": 2.16948, "chunk_loss": 0.033944, "F": 0.23254, "G": 0.43953, "chunks": 238.12, "bytes_per_chunk": 4.3, "boundary_entropy": 0.67071, "router_noise": 0.0, "grad_norm": 0.7599, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 166.9} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "global", "lm_loss": 1.51437, "bpb": 2.18477, "chunk_loss": 0.034235, "F": 0.23724, "G": 0.44065, "chunks": 242.94, "bytes_per_chunk": 4.215, "boundary_entropy": 0.67025, "router_noise": 0.0, "grad_norm": 0.7625, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 169.0} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "global", "lm_loss": 1.40909, "bpb": 2.03288, "chunk_loss": 0.033438, "F": 0.22491, "G": 0.43721, "chunks": 230.31, "bytes_per_chunk": 4.446, "boundary_entropy": 0.6689, "router_noise": 0.0, "grad_norm": 0.7811, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 171.2} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4004.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.230075, "cps": 31.28125, "mean_p": 0.441423, "boundary_entropy": 0.664153, "bpb": 2.188389}, "de": {"chunks": 4874.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.237374, "cps": 38.078125, "mean_p": 0.447351, "boundary_entropy": 0.670484, "bpb": 2.431747}, "vi": {"chunks": 5943.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.322376, "cps": 46.429688, "mean_p": 0.451168, "boundary_entropy": 0.67216, "bpb": 2.047828}, "tr": {"chunks": 6050.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.334199, "cps": 47.265625, "mean_p": 0.455319, "boundary_entropy": 0.66646, "bpb": 2.425716}, "fi": {"chunks": 4833.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.256924, "cps": 37.757812, "mean_p": 0.453135, "boundary_entropy": 0.66958, "bpb": 2.647304}, "ru": {"chunks": 6336.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.334354, "cps": 49.5, "mean_p": 0.436995, "boundary_entropy": 0.672252, "bpb": 1.557021}, "el": {"chunks": 8164.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.397798, "cps": 63.78125, "mean_p": 0.450081, "boundary_entropy": 0.675383, "bpb": 1.633745}, "hi": {"chunks": 9137.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.543061, "cps": 71.382812, "mean_p": 0.422863, "boundary_entropy": 0.663666, "bpb": 1.289}, "ta": {"chunks": 8869.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.474405, "cps": 69.289062, "mean_p": 0.404513, "boundary_entropy": 0.656026, "bpb": 1.064057}, "th": {"chunks": 10578.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.657631, "cps": 82.640625, "mean_p": 0.441436, "boundary_entropy": 0.672912, "bpb": 1.22732}, "zh": {"chunks": 3264.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.552658, "cps": 25.5, "mean_p": 0.436555, "boundary_entropy": 0.666783, "bpb": 3.367425}, "my": {"chunks": 10152.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.535782, "cps": 79.3125, "mean_p": 0.427247, "boundary_entropy": 0.670532, "bpb": 1.177793}}, "bpb": {"en": 2.18839, "de": 2.43175, "vi": 2.04783, "tr": 2.42572, "fi": 2.6473, "ru": 1.55702, "el": 1.63374, "hi": 1.289, "ta": 1.06406, "th": 1.22732, "zh": 3.36743, "my": 1.17779}, "gini_cpc": 0.18959, "gini_cps": 0.195487, "bpic": 4.668641, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0317231348985112, "vi": 1.4011758379943897, "tr": 1.45256265581626, "fi": 1.1166960066673, "ru": 1.453235524370089, "el": 1.7289888886146743, "hi": 2.3603619411791774, "ta": 2.061955255228858, "th": 2.858331193208408, "zh": 2.4020760817238975, "my": 2.328725422772288}, "t": 174.3} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.19783, "de": 2.44159, "vi": 2.05046, "tr": 2.42897, "fi": 2.65336, "ru": 1.53572, "el": 1.62505, "hi": 1.28459, "ta": 1.07075, "th": 1.24371, "zh": 3.36777, "my": 1.18545}, "bytes_per_chunk": {"en": 4.204, "de": 4.228, "vi": 3.883, "tr": 3.229, "fi": 4.253, "ru": 5.267, "el": 4.642, "hi": 4.722, "ta": 6.176, "th": 4.174, "zh": 4.941, "my": 5.165}, "bpb_mean": 1.92377, "t": 175.6} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "global", "lm_loss": 1.47848, "bpb": 2.133, "chunk_loss": 0.033216, "F": 0.22079, "G": 0.44018, "chunks": 226.09, "bytes_per_chunk": 4.529, "boundary_entropy": 0.67179, "router_noise": 0.0, "grad_norm": 0.858, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 175.7} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "global", "lm_loss": 1.43385, "bpb": 2.06861, "chunk_loss": 0.0336, "F": 0.22675, "G": 0.43894, "chunks": 232.19, "bytes_per_chunk": 4.41, "boundary_entropy": 0.67026, "router_noise": 0.0, "grad_norm": 0.7221, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 177.8} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "global", "lm_loss": 1.45814, "bpb": 2.10365, "chunk_loss": 0.03481, "F": 0.24643, "G": 0.44175, "chunks": 252.34, "bytes_per_chunk": 4.058, "boundary_entropy": 0.67063, "router_noise": 0.0, "grad_norm": 0.7826, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 179.9} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "global", "lm_loss": 1.35968, "bpb": 1.9616, "chunk_loss": 0.0337, "F": 0.22946, "G": 0.43598, "chunks": 234.97, "bytes_per_chunk": 4.358, "boundary_entropy": 0.669, "router_noise": 0.0, "grad_norm": 0.7191, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 182.0} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4134.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.237545, "cps": 32.296875, "mean_p": 0.441944, "boundary_entropy": 0.663855, "bpb": 2.158752}, "de": {"chunks": 5484.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.267082, "cps": 42.84375, "mean_p": 0.450322, "boundary_entropy": 0.670483, "bpb": 2.393987}, "vi": {"chunks": 6277.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.340494, "cps": 49.039062, "mean_p": 0.454037, "boundary_entropy": 0.67238, "bpb": 2.020409}, "tr": {"chunks": 6114.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.337734, "cps": 47.765625, "mean_p": 0.454104, "boundary_entropy": 0.664985, "bpb": 2.387751}, "fi": {"chunks": 5047.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.2683, "cps": 39.429688, "mean_p": 0.452815, "boundary_entropy": 0.668655, "bpb": 2.59409}, "ru": {"chunks": 6095.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.321636, "cps": 47.617188, "mean_p": 0.434413, "boundary_entropy": 0.67145, "bpb": 1.542587}, "el": {"chunks": 7736.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.376943, "cps": 60.4375, "mean_p": 0.449389, "boundary_entropy": 0.675048, "bpb": 1.604118}, "hi": {"chunks": 8199.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.487311, "cps": 64.054688, "mean_p": 0.418691, "boundary_entropy": 0.662381, "bpb": 1.275341}, "ta": {"chunks": 8149.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.435892, "cps": 63.664062, "mean_p": 0.401179, "boundary_entropy": 0.654709, "bpb": 1.037605}, "th": {"chunks": 9598.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.596705, "cps": 74.984375, "mean_p": 0.439766, "boundary_entropy": 0.671065, "bpb": 1.221501}, "zh": {"chunks": 3053.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.516932, "cps": 23.851562, "mean_p": 0.436271, "boundary_entropy": 0.666622, "bpb": 3.337255}, "my": {"chunks": 8115.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.428277, "cps": 63.398438, "mean_p": 0.422716, "boundary_entropy": 0.668643, "bpb": 1.171404}}, "bpb": {"en": 2.15875, "de": 2.39399, "vi": 2.02041, "tr": 2.38775, "fi": 2.59409, "ru": 1.54259, "el": 1.60412, "hi": 1.27534, "ta": 1.0376, "th": 1.2215, "zh": 3.33726, "my": 1.1714}, "gini_cpc": 0.155509, "gini_cps": 0.158689, "bpic": 4.920206, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1243426543288983, "vi": 1.4333842713365874, "tr": 1.4217673309004988, "fi": 1.1294709600402877, "ru": 1.3539983762938907, "el": 1.586825839563909, "hi": 2.0514430328425677, "ta": 1.8349849058932814, "th": 2.511963521080183, "zh": 2.1761408903985373, "my": 1.8029297124541905}, "t": 185.1} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.16978, "de": 2.40159, "vi": 2.02042, "tr": 2.39338, "fi": 2.60064, "ru": 1.52112, "el": 1.59605, "hi": 1.2724, "ta": 1.04443, "th": 1.23875, "zh": 3.33902, "my": 1.18123}, "bytes_per_chunk": {"en": 4.097, "de": 3.73, "vi": 3.716, "tr": 3.193, "fi": 3.956, "ru": 5.458, "el": 4.958, "hi": 5.272, "ta": 6.57, "th": 4.573, "zh": 5.192, "my": 6.304}, "bpb_mean": 1.89824, "t": 186.4} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "global", "lm_loss": 1.43768, "bpb": 2.07414, "chunk_loss": 0.033391, "F": 0.22382, "G": 0.43567, "chunks": 229.19, "bytes_per_chunk": 4.468, "boundary_entropy": 0.66892, "router_noise": 0.0, "grad_norm": 0.7343, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 186.4} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "global", "lm_loss": 1.39474, "bpb": 2.01218, "chunk_loss": 0.034289, "F": 0.23825, "G": 0.44023, "chunks": 243.97, "bytes_per_chunk": 4.197, "boundary_entropy": 0.6708, "router_noise": 0.0, "grad_norm": 0.7035, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 188.6} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "global", "lm_loss": 1.40937, "bpb": 2.03329, "chunk_loss": 0.034073, "F": 0.23523, "G": 0.43921, "chunks": 240.88, "bytes_per_chunk": 4.251, "boundary_entropy": 0.66968, "router_noise": 0.0, "grad_norm": 0.7805, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 190.7} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "global", "lm_loss": 1.35584, "bpb": 1.95606, "chunk_loss": 0.033614, "F": 0.2276, "G": 0.43897, "chunks": 233.06, "bytes_per_chunk": 4.394, "boundary_entropy": 0.67048, "router_noise": 0.0, "grad_norm": 0.719, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 192.9} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4174.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.239844, "cps": 32.609375, "mean_p": 0.441067, "boundary_entropy": 0.663028, "bpb": 2.147656}, "de": {"chunks": 5159.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.251254, "cps": 40.304688, "mean_p": 0.447575, "boundary_entropy": 0.66952, "bpb": 2.367112}, "vi": {"chunks": 4768.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.258638, "cps": 37.25, "mean_p": 0.445495, "boundary_entropy": 0.67099, "bpb": 2.008286}, "tr": {"chunks": 5865.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.323979, "cps": 45.820312, "mean_p": 0.451806, "boundary_entropy": 0.66449, "bpb": 2.370826}, "fi": {"chunks": 4949.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.263091, "cps": 38.664062, "mean_p": 0.454628, "boundary_entropy": 0.669752, "bpb": 2.576198}, "ru": {"chunks": 7009.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.369868, "cps": 54.757812, "mean_p": 0.436819, "boundary_entropy": 0.671228, "bpb": 1.513804}, "el": {"chunks": 7808.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.380451, "cps": 61.0, "mean_p": 0.448187, "boundary_entropy": 0.67481, "bpb": 1.58111}, "hi": {"chunks": 8123.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.482793, "cps": 63.460938, "mean_p": 0.414345, "boundary_entropy": 0.658787, "bpb": 1.281017}, "ta": {"chunks": 11841.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.633378, "cps": 92.507812, "mean_p": 0.410811, "boundary_entropy": 0.657108, "bpb": 1.02}, "th": {"chunks": 10030.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.623562, "cps": 78.359375, "mean_p": 0.437326, "boundary_entropy": 0.670748, "bpb": 1.218399}, "zh": {"chunks": 3169.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.536573, "cps": 24.757812, "mean_p": 0.435621, "boundary_entropy": 0.666153, "bpb": 3.294454}, "my": {"chunks": 10798.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.569875, "cps": 84.359375, "mean_p": 0.428534, "boundary_entropy": 0.670637, "bpb": 1.151917}}, "bpb": {"en": 2.14766, "de": 2.36711, "vi": 2.00829, "tr": 2.37083, "fi": 2.5762, "ru": 1.5138, "el": 1.58111, "hi": 1.28102, "ta": 1.02, "th": 1.2184, "zh": 3.29445, "my": 1.15192}, "gini_cpc": 0.197851, "gini_cps": 0.215871, "bpic": 4.585581, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0475742054039436, "vi": 1.0783625083482038, "tr": 1.3507940547091961, "fi": 1.09692578614135, "ru": 1.542121248133628, "el": 1.5862463470519046, "hi": 2.012950316506841, "ta": 2.6407943800136686, "th": 2.599869464433303, "zh": 2.2371776503019434, "my": 2.3760283737227925}, "t": 195.9} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.15847, "de": 2.37811, "vi": 2.01003, "tr": 2.37559, "fi": 2.58506, "ru": 1.49281, "el": 1.57194, "hi": 1.27607, "ta": 1.02843, "th": 1.2355, "zh": 3.29927, "my": 1.16018}, "bytes_per_chunk": {"en": 4.104, "de": 3.973, "vi": 5.114, "tr": 3.321, "fi": 4.062, "ru": 4.74, "el": 4.894, "hi": 5.417, "ta": 4.464, "th": 4.449, "zh": 5.001, "my": 4.876}, "bpb_mean": 1.88095, "t": 197.3} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "global", "lm_loss": 1.38224, "bpb": 1.99415, "chunk_loss": 0.033243, "F": 0.22198, "G": 0.43656, "chunks": 227.31, "bytes_per_chunk": 4.505, "boundary_entropy": 0.66962, "router_noise": 0.0, "grad_norm": 0.7084, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 197.3} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "global", "lm_loss": 1.44468, "bpb": 2.08423, "chunk_loss": 0.033298, "F": 0.22205, "G": 0.43535, "chunks": 227.38, "bytes_per_chunk": 4.504, "boundary_entropy": 0.66844, "router_noise": 0.0, "grad_norm": 0.7104, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 199.4} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "global", "lm_loss": 1.41258, "bpb": 2.03792, "chunk_loss": 0.033543, "F": 0.22659, "G": 0.43766, "chunks": 232.03, "bytes_per_chunk": 4.413, "boundary_entropy": 0.66926, "router_noise": 0.0, "grad_norm": 0.8032, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 201.6} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "global", "lm_loss": 1.47712, "bpb": 2.13103, "chunk_loss": 0.033786, "F": 0.23062, "G": 0.43771, "chunks": 236.16, "bytes_per_chunk": 4.336, "boundary_entropy": 0.66941, "router_noise": 0.0, "grad_norm": 0.7743, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 203.7} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4354.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.250187, "cps": 34.015625, "mean_p": 0.442299, "boundary_entropy": 0.663391, "bpb": 2.126243}, "de": {"chunks": 5202.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.253348, "cps": 40.640625, "mean_p": 0.448068, "boundary_entropy": 0.669561, "bpb": 2.339362}, "vi": {"chunks": 5362.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.29086, "cps": 41.890625, "mean_p": 0.45058, "boundary_entropy": 0.672404, "bpb": 1.990719}, "tr": {"chunks": 5856.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.323482, "cps": 45.75, "mean_p": 0.452919, "boundary_entropy": 0.664952, "bpb": 2.341268}, "fi": {"chunks": 5001.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.265855, "cps": 39.070312, "mean_p": 0.454871, "boundary_entropy": 0.67012, "bpb": 2.564109}, "ru": {"chunks": 6539.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.345066, "cps": 51.085938, "mean_p": 0.43535, "boundary_entropy": 0.671153, "bpb": 1.496525}, "el": {"chunks": 7619.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.371242, "cps": 59.523438, "mean_p": 0.448754, "boundary_entropy": 0.675238, "bpb": 1.583952}, "hi": {"chunks": 9025.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.536404, "cps": 70.507812, "mean_p": 0.419072, "boundary_entropy": 0.660853, "bpb": 1.256162}, "ta": {"chunks": 8694.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.465044, "cps": 67.921875, "mean_p": 0.406732, "boundary_entropy": 0.656911, "bpb": 1.026109}, "th": {"chunks": 9448.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.58738, "cps": 73.8125, "mean_p": 0.43848, "boundary_entropy": 0.672062, "bpb": 1.176361}, "zh": {"chunks": 3220.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545208, "cps": 25.15625, "mean_p": 0.435497, "boundary_entropy": 0.666216, "bpb": 3.274797}, "my": {"chunks": 10303.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.543751, "cps": 80.492188, "mean_p": 0.428185, "boundary_entropy": 0.670847, "bpb": 1.140793}}, "bpb": {"en": 2.12624, "de": 2.33936, "vi": 1.99072, "tr": 2.34127, "fi": 2.56411, "ru": 1.49652, "el": 1.58395, "hi": 1.25616, "ta": 1.02611, "th": 1.17636, "zh": 3.2748, "my": 1.14079}, "gini_cpc": 0.173714, "gini_cps": 0.182786, "bpic": 4.760193, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.012636637380462, "vi": 1.1625706728341605, "tr": 1.292963342437653, "fi": 1.0626265596266729, "ru": 1.3792335680167935, "el": 1.4838596449291093, "hi": 2.14401506769748, "ta": 1.858788007523032, "th": 2.3477644101608233, "zh": 2.1792051899915394, "my": 2.1733817665425543}, "t": 206.8} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.13526, "de": 2.35269, "vi": 1.9907, "tr": 2.34562, "fi": 2.57146, "ru": 1.47389, "el": 1.57507, "hi": 1.25344, "ta": 1.03398, "th": 1.19568, "zh": 3.27905, "my": 1.14994}, "bytes_per_chunk": {"en": 3.905, "de": 3.934, "vi": 4.529, "tr": 3.308, "fi": 3.985, "ru": 5.038, "el": 4.977, "hi": 4.841, "ta": 6.162, "th": 4.647, "zh": 5.065, "my": 5.087}, "bpb_mean": 1.86306, "t": 208.1} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "global", "lm_loss": 1.40243, "bpb": 2.02328, "chunk_loss": 0.033309, "F": 0.2226, "G": 0.43743, "chunks": 227.94, "bytes_per_chunk": 4.492, "boundary_entropy": 0.67041, "router_noise": 0.0, "grad_norm": 0.732, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 208.1} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "global", "lm_loss": 1.26211, "bpb": 1.82084, "chunk_loss": 0.033188, "F": 0.2204, "G": 0.43704, "chunks": 225.69, "bytes_per_chunk": 4.537, "boundary_entropy": 0.66952, "router_noise": 0.0, "grad_norm": 0.7364, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 210.3} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "global", "lm_loss": 1.45222, "bpb": 2.0951, "chunk_loss": 0.033906, "F": 0.23257, "G": 0.43722, "chunks": 238.16, "bytes_per_chunk": 4.3, "boundary_entropy": 0.66903, "router_noise": 0.0, "grad_norm": 0.8498, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 212.4} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "global", "lm_loss": 1.37913, "bpb": 1.98967, "chunk_loss": 0.034336, "F": 0.23929, "G": 0.44005, "chunks": 245.03, "bytes_per_chunk": 4.179, "boundary_entropy": 0.67034, "router_noise": 0.0, "grad_norm": 0.7227, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 214.5} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3952.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.227087, "cps": 30.875, "mean_p": 0.43695, "boundary_entropy": 0.662512, "bpb": 2.115548}, "de": {"chunks": 5048.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.245848, "cps": 39.4375, "mean_p": 0.445956, "boundary_entropy": 0.669101, "bpb": 2.333534}, "vi": {"chunks": 4928.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.267318, "cps": 38.5, "mean_p": 0.443677, "boundary_entropy": 0.670039, "bpb": 1.999978}, "tr": {"chunks": 5633.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.311164, "cps": 44.007812, "mean_p": 0.449511, "boundary_entropy": 0.664006, "bpb": 2.31897}, "fi": {"chunks": 4908.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.260911, "cps": 38.34375, "mean_p": 0.453213, "boundary_entropy": 0.669058, "bpb": 2.517176}, "ru": {"chunks": 5982.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.315673, "cps": 46.734375, "mean_p": 0.43282, "boundary_entropy": 0.67063, "bpb": 1.491157}, "el": {"chunks": 7943.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.387029, "cps": 62.054688, "mean_p": 0.448393, "boundary_entropy": 0.67514, "bpb": 1.558972}, "hi": {"chunks": 7859.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.467103, "cps": 61.398438, "mean_p": 0.410878, "boundary_entropy": 0.657194, "bpb": 1.236646}, "ta": {"chunks": 8679.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.464242, "cps": 67.804688, "mean_p": 0.395222, "boundary_entropy": 0.649851, "bpb": 1.013718}, "th": {"chunks": 9691.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.602487, "cps": 75.710938, "mean_p": 0.433972, "boundary_entropy": 0.66844, "bpb": 1.167375}, "zh": {"chunks": 3196.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.541145, "cps": 24.96875, "mean_p": 0.434271, "boundary_entropy": 0.665929, "bpb": 3.283061}, "my": {"chunks": 9206.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.485856, "cps": 71.921875, "mean_p": 0.425707, "boundary_entropy": 0.671362, "bpb": 1.135085}}, "bpb": {"en": 2.11555, "de": 2.33353, "vi": 1.99998, "tr": 2.31897, "fi": 2.51718, "ru": 1.49116, "el": 1.55897, "hi": 1.23665, "ta": 1.01372, "th": 1.16737, "zh": 3.28306, "my": 1.13509}, "gini_cpc": 0.180133, "gini_cps": 0.182825, "bpic": 4.982551, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.082615207552728, "vi": 1.1771579540326063, "tr": 1.3702391885524152, "fi": 1.1489466280497005, "ru": 1.3900946716802154, "el": 1.7043190630794205, "hi": 2.0569294686310013, "ta": 2.044331889060813, "th": 2.653106677078627, "zh": 2.3829806304163332, "my": 2.1395122566427274}, "t": 217.6} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.12727, "de": 2.34557, "vi": 1.99992, "tr": 2.32523, "fi": 2.52758, "ru": 1.46804, "el": 1.54936, "hi": 1.2324, "ta": 1.0203, "th": 1.18522, "zh": 3.28399, "my": 1.14501}, "bytes_per_chunk": {"en": 4.357, "de": 4.096, "vi": 4.734, "tr": 3.429, "fi": 4.102, "ru": 5.613, "el": 4.881, "hi": 5.731, "ta": 6.375, "th": 4.588, "zh": 5.274, "my": 5.721}, "bpb_mean": 1.85082, "t": 218.9} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "global", "lm_loss": 1.64643, "bpb": 2.3753, "chunk_loss": 0.033648, "F": 0.22733, "G": 0.43768, "chunks": 232.78, "bytes_per_chunk": 4.399, "boundary_entropy": 0.66948, "router_noise": 0.0, "grad_norm": 0.812, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 218.9} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "global", "lm_loss": 1.44173, "bpb": 2.07997, "chunk_loss": 0.034405, "F": 0.24088, "G": 0.43896, "chunks": 246.66, "bytes_per_chunk": 4.152, "boundary_entropy": 0.66977, "router_noise": 0.0, "grad_norm": 0.7925, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 221.1} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "global", "lm_loss": 1.35344, "bpb": 1.9526, "chunk_loss": 0.034002, "F": 0.23425, "G": 0.43855, "chunks": 239.88, "bytes_per_chunk": 4.269, "boundary_entropy": 0.67006, "router_noise": 0.0, "grad_norm": 0.743, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 223.2} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "global", "lm_loss": 1.39832, "bpb": 2.01735, "chunk_loss": 0.033876, "F": 0.23193, "G": 0.43808, "chunks": 237.5, "bytes_per_chunk": 4.312, "boundary_entropy": 0.66967, "router_noise": 0.0, "grad_norm": 0.7616, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 225.3} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4413.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.253577, "cps": 34.476562, "mean_p": 0.441414, "boundary_entropy": 0.662567, "bpb": 2.102022}, "de": {"chunks": 5757.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.280378, "cps": 44.976562, "mean_p": 0.451387, "boundary_entropy": 0.670006, "bpb": 2.304005}, "vi": {"chunks": 5362.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.29086, "cps": 41.890625, "mean_p": 0.44836, "boundary_entropy": 0.671491, "bpb": 1.974118}, "tr": {"chunks": 6023.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.332707, "cps": 47.054688, "mean_p": 0.454662, "boundary_entropy": 0.664788, "bpb": 2.292421}, "fi": {"chunks": 5116.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.271969, "cps": 39.96875, "mean_p": 0.455037, "boundary_entropy": 0.669696, "bpb": 2.501048}, "ru": {"chunks": 7174.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.378575, "cps": 56.046875, "mean_p": 0.437661, "boundary_entropy": 0.671314, "bpb": 1.466671}, "el": {"chunks": 8258.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.402378, "cps": 64.515625, "mean_p": 0.451355, "boundary_entropy": 0.675537, "bpb": 1.536703}, "hi": {"chunks": 9262.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.55049, "cps": 72.359375, "mean_p": 0.41711, "boundary_entropy": 0.659289, "bpb": 1.224773}, "ta": {"chunks": 10476.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.560364, "cps": 81.84375, "mean_p": 0.40483, "boundary_entropy": 0.654824, "bpb": 1.003711}, "th": {"chunks": 10185.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.633199, "cps": 79.570312, "mean_p": 0.440949, "boundary_entropy": 0.671641, "bpb": 1.15371}, "zh": {"chunks": 3203.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.54233, "cps": 25.023438, "mean_p": 0.436922, "boundary_entropy": 0.666927, "bpb": 3.245203}, "my": {"chunks": 11014.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.581275, "cps": 86.046875, "mean_p": 0.424886, "boundary_entropy": 0.669296, "bpb": 1.13449}}, "bpb": {"en": 2.10202, "de": 2.30401, "vi": 1.97412, "tr": 2.29242, "fi": 2.50105, "ru": 1.46667, "el": 1.5367, "hi": 1.22477, "ta": 1.00371, "th": 1.15371, "zh": 3.2452, "my": 1.13449}, "gini_cpc": 0.178386, "gini_cps": 0.197238, "bpic": 4.449996, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.105691612212312, "vi": 1.1470275797688498, "tr": 1.3120564014820697, "fi": 1.0725285091988608, "ru": 1.4929399884127428, "el": 1.5868074329432846, "hi": 2.1709003891986023, "ta": 2.2098368584103216, "th": 2.497066802022819, "zh": 2.13871880858121, "my": 2.292302292997817}, "t": 228.4} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.11374, "de": 2.31498, "vi": 1.97311, "tr": 2.2972, "fi": 2.5098, "ru": 1.44525, "el": 1.52697, "hi": 1.2192, "ta": 1.01095, "th": 1.1701, "zh": 3.2456, "my": 1.14193}, "bytes_per_chunk": {"en": 3.952, "de": 3.583, "vi": 4.428, "tr": 3.256, "fi": 3.904, "ru": 4.659, "el": 4.655, "hi": 4.725, "ta": 5.17, "th": 4.33, "zh": 5.153, "my": 4.856}, "bpb_mean": 1.83074, "t": 229.7} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "global", "lm_loss": 1.47598, "bpb": 2.12938, "chunk_loss": 0.033607, "F": 0.22729, "G": 0.4384, "chunks": 232.75, "bytes_per_chunk": 4.4, "boundary_entropy": 0.67028, "router_noise": 0.0, "grad_norm": 0.7475, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 229.8} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "global", "lm_loss": 1.48055, "bpb": 2.13598, "chunk_loss": 0.033575, "F": 0.22714, "G": 0.43604, "chunks": 232.59, "bytes_per_chunk": 4.403, "boundary_entropy": 0.66966, "router_noise": 0.0, "grad_norm": 0.7811, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 231.9} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "global", "lm_loss": 1.48292, "bpb": 2.1394, "chunk_loss": 0.03371, "F": 0.22897, "G": 0.43882, "chunks": 234.47, "bytes_per_chunk": 4.367, "boundary_entropy": 0.66948, "router_noise": 0.0, "grad_norm": 0.835, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 234.1} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "global", "lm_loss": 1.55636, "bpb": 2.24535, "chunk_loss": 0.033264, "F": 0.22192, "G": 0.43635, "chunks": 227.25, "bytes_per_chunk": 4.506, "boundary_entropy": 0.66999, "router_noise": 0.0, "grad_norm": 0.7713, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 236.2} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4289.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.246452, "cps": 33.507812, "mean_p": 0.440086, "boundary_entropy": 0.661786, "bpb": 2.08436}, "de": {"chunks": 5271.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.256709, "cps": 41.179688, "mean_p": 0.448271, "boundary_entropy": 0.668904, "bpb": 2.29506}, "vi": {"chunks": 5435.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.29482, "cps": 42.460938, "mean_p": 0.448724, "boundary_entropy": 0.671289, "bpb": 1.966236}, "tr": {"chunks": 5972.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.32989, "cps": 46.65625, "mean_p": 0.452848, "boundary_entropy": 0.66405, "bpb": 2.280048}, "fi": {"chunks": 5007.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.266174, "cps": 39.117188, "mean_p": 0.453822, "boundary_entropy": 0.668872, "bpb": 2.478011}, "ru": {"chunks": 6848.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.361372, "cps": 53.5, "mean_p": 0.435263, "boundary_entropy": 0.670906, "bpb": 1.457467}, "el": {"chunks": 8075.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.393461, "cps": 63.085938, "mean_p": 0.450498, "boundary_entropy": 0.675187, "bpb": 1.524018}, "hi": {"chunks": 7805.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.463893, "cps": 60.976562, "mean_p": 0.412485, "boundary_entropy": 0.658768, "bpb": 1.215491}, "ta": {"chunks": 9483.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.507248, "cps": 74.085938, "mean_p": 0.402714, "boundary_entropy": 0.653753, "bpb": 0.994144}, "th": {"chunks": 9476.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.58912, "cps": 74.03125, "mean_p": 0.434287, "boundary_entropy": 0.669186, "bpb": 1.142875}, "zh": {"chunks": 3308.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.560108, "cps": 25.84375, "mean_p": 0.434754, "boundary_entropy": 0.665637, "bpb": 3.218977}, "my": {"chunks": 9606.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.506966, "cps": 75.046875, "mean_p": 0.426144, "boundary_entropy": 0.670862, "bpb": 1.122517}}, "bpb": {"en": 2.08436, "de": 2.29506, "vi": 1.96624, "tr": 2.28005, "fi": 2.47801, "ru": 1.45747, "el": 1.52402, "hi": 1.21549, "ta": 0.99414, "th": 1.14287, "zh": 3.21898, "my": 1.12252}, "gini_cpc": 0.168933, "gini_cps": 0.175109, "bpic": 4.763028, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0416184959021149, "vi": 1.196256967840793, "tr": 1.3385583932320027, "fi": 1.080024933943357, "ru": 1.4662992462155442, "el": 1.5965030766282418, "hi": 1.882287284552868, "ta": 2.0582037020258257, "th": 2.39040814952449, "zh": 2.272689639827887, "my": 2.0570615198814113}, "t": 239.3} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.09497, "de": 2.30643, "vi": 1.9666, "tr": 2.28509, "fi": 2.488, "ru": 1.43583, "el": 1.51498, "hi": 1.2115, "ta": 1.00196, "th": 1.15855, "zh": 3.21941, "my": 1.13031}, "bytes_per_chunk": {"en": 3.989, "de": 3.893, "vi": 4.335, "tr": 3.272, "fi": 3.955, "ru": 4.9, "el": 4.746, "hi": 5.62, "ta": 5.783, "th": 4.7, "zh": 5.116, "my": 5.47}, "bpb_mean": 1.8178, "t": 240.6} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "global", "lm_loss": 1.34008, "bpb": 1.93333, "chunk_loss": 0.033368, "F": 0.22421, "G": 0.43506, "chunks": 229.59, "bytes_per_chunk": 4.46, "boundary_entropy": 0.66735, "router_noise": 0.0, "grad_norm": 0.7222, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 240.6} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "global", "lm_loss": 1.54669, "bpb": 2.2314, "chunk_loss": 0.033906, "F": 0.23248, "G": 0.43779, "chunks": 238.06, "bytes_per_chunk": 4.301, "boundary_entropy": 0.66828, "router_noise": 0.0, "grad_norm": 0.7535, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 242.7} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "global", "lm_loss": 1.40197, "bpb": 2.02262, "chunk_loss": 0.033852, "F": 0.23163, "G": 0.43893, "chunks": 237.19, "bytes_per_chunk": 4.317, "boundary_entropy": 0.66943, "router_noise": 0.0, "grad_norm": 0.7242, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 244.9} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "global", "lm_loss": 1.3852, "bpb": 1.99841, "chunk_loss": 0.03432, "F": 0.23959, "G": 0.43828, "chunks": 245.34, "bytes_per_chunk": 4.174, "boundary_entropy": 0.66776, "router_noise": 0.0, "grad_norm": 0.7885, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 247.0} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4251.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.244268, "cps": 33.210938, "mean_p": 0.43941, "boundary_entropy": 0.661757, "bpb": 2.075037}, "de": {"chunks": 5302.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.258218, "cps": 41.421875, "mean_p": 0.44764, "boundary_entropy": 0.669036, "bpb": 2.276964}, "vi": {"chunks": 5413.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.293626, "cps": 42.289062, "mean_p": 0.449975, "boundary_entropy": 0.672311, "bpb": 1.956491}, "tr": {"chunks": 5946.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.328454, "cps": 46.453125, "mean_p": 0.452499, "boundary_entropy": 0.663351, "bpb": 2.257099}, "fi": {"chunks": 5161.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.274361, "cps": 40.320312, "mean_p": 0.455413, "boundary_entropy": 0.669017, "bpb": 2.459205}, "ru": {"chunks": 6547.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.345488, "cps": 51.148438, "mean_p": 0.434412, "boundary_entropy": 0.670665, "bpb": 1.444824}, "el": {"chunks": 7885.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.384203, "cps": 61.601562, "mean_p": 0.448647, "boundary_entropy": 0.674944, "bpb": 1.511947}, "hi": {"chunks": 8182.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.4863, "cps": 63.921875, "mean_p": 0.414003, "boundary_entropy": 0.658704, "bpb": 1.204297}, "ta": {"chunks": 10970.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.586788, "cps": 85.703125, "mean_p": 0.405988, "boundary_entropy": 0.654, "bpb": 0.973197}, "th": {"chunks": 9950.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.618589, "cps": 77.734375, "mean_p": 0.434681, "boundary_entropy": 0.669422, "bpb": 1.12842}, "zh": {"chunks": 3488.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.590586, "cps": 27.25, "mean_p": 0.436796, "boundary_entropy": 0.6662, "bpb": 3.214653}, "my": {"chunks": 10173.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.53689, "cps": 79.476562, "mean_p": 0.42756, "boundary_entropy": 0.670923, "bpb": 1.113793}}, "bpb": {"en": 2.07504, "de": 2.27696, "vi": 1.95649, "tr": 2.2571, "fi": 2.45921, "ru": 1.44482, "el": 1.51195, "hi": 1.2043, "ta": 0.9732, "th": 1.12842, "zh": 3.21465, "my": 1.11379}, "gini_cpc": 0.185679, "gini_cps": 0.19169, "bpic": 4.608985, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0571103658427397, "vi": 1.2020648566105825, "tr": 1.3446441559162499, "fi": 1.123194558717596, "ru": 1.4143801148138975, "el": 1.5728737618543953, "hi": 1.9908448567123267, "ta": 2.402227715530836, "th": 2.532415894828308, "zh": 2.4177759253952953, "my": 2.1979544283362746}, "t": 250.1} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.08546, "de": 2.28716, "vi": 1.95644, "tr": 2.2608, "fi": 2.46847, "ru": 1.42353, "el": 1.50271, "hi": 1.19904, "ta": 0.9816, "th": 1.14555, "zh": 3.21352, "my": 1.12149}, "bytes_per_chunk": {"en": 4.024, "de": 3.886, "vi": 4.314, "tr": 3.261, "fi": 3.86, "ru": 5.062, "el": 4.858, "hi": 5.328, "ta": 4.867, "th": 4.487, "zh": 4.725, "my": 5.171}, "bpb_mean": 1.80381, "t": 251.4} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 251.4} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.19827, "bpb": 1.72874, "chunk_loss": 0.033307, "F": 0.2229, "G": 0.43445, "chunks": 228.25, "bytes_per_chunk": 4.486, "boundary_entropy": 0.66879, "router_noise": 0.0, "grad_norm": 0.7026, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 251.5} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.25219, "bpb": 1.80653, "chunk_loss": 0.033639, "F": 0.22842, "G": 0.43521, "chunks": 233.91, "bytes_per_chunk": 4.378, "boundary_entropy": 0.66792, "router_noise": 0.0, "grad_norm": 0.7564, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 253.6} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.47879, "bpb": 2.13345, "chunk_loss": 0.034486, "F": 0.24139, "G": 0.43888, "chunks": 247.19, "bytes_per_chunk": 4.143, "boundary_entropy": 0.66792, "router_noise": 0.0, "grad_norm": 0.7898, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 255.7} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.47327, "bpb": 2.12549, "chunk_loss": 0.033739, "F": 0.22977, "G": 0.43922, "chunks": 235.28, "bytes_per_chunk": 4.352, "boundary_entropy": 0.67048, "router_noise": 0.0, "grad_norm": 0.8153, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 257.8} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4237.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.243464, "cps": 33.101562, "mean_p": 0.437542, "boundary_entropy": 0.660747, "bpb": 2.067329}, "de": {"chunks": 5275.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.256904, "cps": 41.210938, "mean_p": 0.447925, "boundary_entropy": 0.668785, "bpb": 2.270221}, "vi": {"chunks": 5119.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.277678, "cps": 39.992188, "mean_p": 0.446185, "boundary_entropy": 0.67058, "bpb": 1.952365}, "tr": {"chunks": 5909.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.32641, "cps": 46.164062, "mean_p": 0.45123, "boundary_entropy": 0.662832, "bpb": 2.237616}, "fi": {"chunks": 5091.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.27064, "cps": 39.773438, "mean_p": 0.453421, "boundary_entropy": 0.668282, "bpb": 2.443772}, "ru": {"chunks": 6722.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.354723, "cps": 52.515625, "mean_p": 0.435835, "boundary_entropy": 0.67106, "bpb": 1.42999}, "el": {"chunks": 7822.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.381133, "cps": 61.109375, "mean_p": 0.448487, "boundary_entropy": 0.675442, "bpb": 1.500575}, "hi": {"chunks": 8140.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.483804, "cps": 63.59375, "mean_p": 0.413465, "boundary_entropy": 0.658901, "bpb": 1.193617}, "ta": {"chunks": 8373.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.447874, "cps": 65.414062, "mean_p": 0.401228, "boundary_entropy": 0.653274, "bpb": 0.972771}, "th": {"chunks": 9854.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.61262, "cps": 76.984375, "mean_p": 0.435002, "boundary_entropy": 0.668964, "bpb": 1.120968}, "zh": {"chunks": 3219.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545039, "cps": 25.148438, "mean_p": 0.435631, "boundary_entropy": 0.666151, "bpb": 3.19402}, "my": {"chunks": 7487.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.395134, "cps": 58.492188, "mean_p": 0.42081, "boundary_entropy": 0.669766, "bpb": 1.109286}}, "bpb": {"en": 2.06733, "de": 2.27022, "vi": 1.95236, "tr": 2.23762, "fi": 2.44377, "ru": 1.42999, "el": 1.50058, "hi": 1.19362, "ta": 0.97277, "th": 1.12097, "zh": 3.19402, "my": 1.10929}, "gini_cpc": 0.16737, "gini_cps": 0.164531, "bpic": 4.968167, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0552022607424802, "vi": 1.14053244375703, "tr": 1.3406922371710062, "fi": 1.111621325617746, "ru": 1.4569845618103614, "el": 1.5654623339120286, "hi": 1.9871698449342994, "ta": 1.8395910073529824, "th": 2.516269473249888, "zh": 2.238686035594545, "my": 1.6229685841563128}, "t": 260.9} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.07794, "de": 2.28144, "vi": 1.95235, "tr": 2.24237, "fi": 2.45159, "ru": 1.40793, "el": 1.49209, "hi": 1.18872, "ta": 0.98045, "th": 1.13818, "zh": 3.19433, "my": 1.11798}, "bytes_per_chunk": {"en": 4.037, "de": 3.92, "vi": 4.547, "tr": 3.265, "fi": 3.956, "ru": 4.919, "el": 4.94, "hi": 5.374, "ta": 6.537, "th": 4.501, "zh": 5.099, "my": 6.577}, "bpb_mean": 1.79378, "t": 262.2} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.48325, "bpb": 2.13988, "chunk_loss": 0.034581, "F": 0.24387, "G": 0.43864, "chunks": 249.72, "bytes_per_chunk": 4.101, "boundary_entropy": 0.66853, "router_noise": 0.0, "grad_norm": 0.8215, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 262.3} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.49808, "bpb": 2.16127, "chunk_loss": 0.033941, "F": 0.23352, "G": 0.43608, "chunks": 239.12, "bytes_per_chunk": 4.282, "boundary_entropy": 0.66895, "router_noise": 0.0, "grad_norm": 0.7577, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 264.4} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.418, "bpb": 2.04573, "chunk_loss": 0.034879, "F": 0.24799, "G": 0.44174, "chunks": 253.94, "bytes_per_chunk": 4.032, "boundary_entropy": 0.66944, "router_noise": 0.0, "grad_norm": 0.8015, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 266.5} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.2294, "bpb": 1.77365, "chunk_loss": 0.033162, "F": 0.22064, "G": 0.43312, "chunks": 225.94, "bytes_per_chunk": 4.532, "boundary_entropy": 0.66719, "router_noise": 0.0, "grad_norm": 0.7703, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 268.6} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4064.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.233523, "cps": 31.75, "mean_p": 0.436364, "boundary_entropy": 0.660621, "bpb": 2.056044}, "de": {"chunks": 5196.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.253056, "cps": 40.59375, "mean_p": 0.446946, "boundary_entropy": 0.668391, "bpb": 2.259934}, "vi": {"chunks": 5110.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.27719, "cps": 39.921875, "mean_p": 0.44632, "boundary_entropy": 0.670973, "bpb": 1.945151}, "tr": {"chunks": 5904.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.326134, "cps": 46.125, "mean_p": 0.451341, "boundary_entropy": 0.662449, "bpb": 2.225066}, "fi": {"chunks": 5014.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.266546, "cps": 39.171875, "mean_p": 0.453326, "boundary_entropy": 0.668186, "bpb": 2.425095}, "ru": {"chunks": 6907.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.364485, "cps": 53.960938, "mean_p": 0.43623, "boundary_entropy": 0.670881, "bpb": 1.425403}, "el": {"chunks": 8184.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.398772, "cps": 63.9375, "mean_p": 0.451266, "boundary_entropy": 0.675992, "bpb": 1.490503}, "hi": {"chunks": 9095.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.540565, "cps": 71.054688, "mean_p": 0.419909, "boundary_entropy": 0.661349, "bpb": 1.188978}, "ta": {"chunks": 8379.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.448195, "cps": 65.460938, "mean_p": 0.40344, "boundary_entropy": 0.654494, "bpb": 0.972466}, "th": {"chunks": 10041.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.624246, "cps": 78.445312, "mean_p": 0.434169, "boundary_entropy": 0.668875, "bpb": 1.113297}, "zh": {"chunks": 3164.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.535726, "cps": 24.71875, "mean_p": 0.435607, "boundary_entropy": 0.666358, "bpb": 3.178453}, "my": {"chunks": 8852.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.467173, "cps": 69.15625, "mean_p": 0.421581, "boundary_entropy": 0.668974, "bpb": 1.094684}}, "bpb": {"en": 2.05604, "de": 2.25993, "vi": 1.94515, "tr": 2.22507, "fi": 2.4251, "ru": 1.4254, "el": 1.4905, "hi": 1.18898, "ta": 0.97247, "th": 1.1133, "zh": 3.17845, "my": 1.09468}, "gini_cpc": 0.178086, "gini_cps": 0.181563, "bpic": 4.802665, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0836453111200677, "vi": 1.1869930581805834, "tr": 1.3965812861747804, "fi": 1.1414131380240329, "ru": 1.5608122416013963, "el": 1.707635585348645, "hi": 2.3148243995039253, "ta": 1.9192747150170686, "th": 2.6731684254905654, "zh": 2.294105853916076, "my": 2.0005455772865313}, "t": 271.7} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.06651, "de": 2.27076, "vi": 1.94343, "tr": 2.2313, "fi": 2.43392, "ru": 1.40247, "el": 1.48161, "hi": 1.18372, "ta": 0.97965, "th": 1.13108, "zh": 3.17806, "my": 1.10165}, "bytes_per_chunk": {"en": 4.239, "de": 3.99, "vi": 4.622, "tr": 3.308, "fi": 4.006, "ru": 4.787, "el": 4.709, "hi": 4.864, "ta": 6.459, "th": 4.447, "zh": 5.303, "my": 5.789}, "bpb_mean": 1.78368, "t": 273.0} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.30711, "bpb": 1.88576, "chunk_loss": 0.033614, "F": 0.22739, "G": 0.43768, "chunks": 232.84, "bytes_per_chunk": 4.398, "boundary_entropy": 0.66981, "router_noise": 0.0, "grad_norm": 0.7782, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 273.1} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.40743, "bpb": 2.03049, "chunk_loss": 0.033751, "F": 0.22992, "G": 0.43516, "chunks": 235.44, "bytes_per_chunk": 4.349, "boundary_entropy": 0.66818, "router_noise": 0.0, "grad_norm": 0.7171, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 275.2} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.32675, "bpb": 1.9141, "chunk_loss": 0.034717, "F": 0.24533, "G": 0.44114, "chunks": 251.22, "bytes_per_chunk": 4.076, "boundary_entropy": 0.66936, "router_noise": 0.0, "grad_norm": 0.806, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 277.3} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.34702, "bpb": 1.94334, "chunk_loss": 0.034341, "F": 0.23926, "G": 0.43837, "chunks": 245.0, "bytes_per_chunk": 4.18, "boundary_entropy": 0.66838, "router_noise": 0.0, "grad_norm": 0.783, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 279.4} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4189.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.240706, "cps": 32.726562, "mean_p": 0.437206, "boundary_entropy": 0.660657, "bpb": 2.048088}, "de": {"chunks": 5373.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.261676, "cps": 41.976562, "mean_p": 0.447351, "boundary_entropy": 0.668472, "bpb": 2.248046}, "vi": {"chunks": 5060.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.274478, "cps": 39.53125, "mean_p": 0.44611, "boundary_entropy": 0.671113, "bpb": 1.941165}, "tr": {"chunks": 5808.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.320831, "cps": 45.375, "mean_p": 0.450845, "boundary_entropy": 0.662753, "bpb": 2.215874}, "fi": {"chunks": 5014.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.266546, "cps": 39.171875, "mean_p": 0.453105, "boundary_entropy": 0.668824, "bpb": 2.413549}, "ru": {"chunks": 6799.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.358786, "cps": 53.117188, "mean_p": 0.435916, "boundary_entropy": 0.670999, "bpb": 1.411508}, "el": {"chunks": 7703.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.375335, "cps": 60.179688, "mean_p": 0.44825, "boundary_entropy": 0.675302, "bpb": 1.48084}, "hi": {"chunks": 7457.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.44321, "cps": 58.257812, "mean_p": 0.413037, "boundary_entropy": 0.659438, "bpb": 1.184007}, "ta": {"chunks": 8571.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.458465, "cps": 66.960938, "mean_p": 0.404004, "boundary_entropy": 0.654539, "bpb": 0.962134}, "th": {"chunks": 9797.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.609077, "cps": 76.539062, "mean_p": 0.431755, "boundary_entropy": 0.667707, "bpb": 1.111403}, "zh": {"chunks": 3475.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.588385, "cps": 27.148438, "mean_p": 0.437299, "boundary_entropy": 0.666518, "bpb": 3.171636}, "my": {"chunks": 9451.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.498786, "cps": 73.835938, "mean_p": 0.422424, "boundary_entropy": 0.668822, "bpb": 1.090888}}, "bpb": {"en": 2.04809, "de": 2.24805, "vi": 1.94116, "tr": 2.21587, "fi": 2.41355, "ru": 1.41151, "el": 1.48084, "hi": 1.18401, "ta": 0.96213, "th": 1.1114, "zh": 3.17164, "my": 1.09089}, "gini_cpc": 0.176057, "gini_cps": 0.171816, "bpic": 4.876692, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0871217712790606, "vi": 1.1403052785968752, "tr": 1.3328762088643529, "fi": 1.1073533045905155, "ru": 1.4905604261947518, "el": 1.5593112511733194, "hi": 1.8412926943381398, "ta": 1.9046701932256638, "th": 2.530380328365667, "zh": 2.4444160471562237, "my": 2.0721831930819143}, "t": 282.5} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.06036, "de": 2.25945, "vi": 1.9382, "tr": 2.21917, "fi": 2.42166, "ru": 1.38978, "el": 1.47013, "hi": 1.17929, "ta": 0.96923, "th": 1.12835, "zh": 3.16897, "my": 1.09786}, "bytes_per_chunk": {"en": 4.101, "de": 3.894, "vi": 4.695, "tr": 3.356, "fi": 4.043, "ru": 4.935, "el": 5.051, "hi": 5.911, "ta": 6.222, "th": 4.584, "zh": 4.81, "my": 5.503}, "bpb_mean": 1.7752, "t": 283.8} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.26285, "bpb": 1.8219, "chunk_loss": 0.033146, "F": 0.22009, "G": 0.43396, "chunks": 225.38, "bytes_per_chunk": 4.544, "boundary_entropy": 0.66754, "router_noise": 0.0, "grad_norm": 0.7341, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 283.9} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.49684, "bpb": 2.15948, "chunk_loss": 0.034584, "F": 0.24368, "G": 0.44093, "chunks": 249.53, "bytes_per_chunk": 4.104, "boundary_entropy": 0.66941, "router_noise": 0.0, "grad_norm": 0.7789, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 286.0} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.40874, "bpb": 2.03239, "chunk_loss": 0.033049, "F": 0.21826, "G": 0.43374, "chunks": 223.5, "bytes_per_chunk": 4.582, "boundary_entropy": 0.66863, "router_noise": 0.0, "grad_norm": 0.7496, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 288.1} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.52686, "bpb": 2.2028, "chunk_loss": 0.033733, "F": 0.23022, "G": 0.4362, "chunks": 235.75, "bytes_per_chunk": 4.344, "boundary_entropy": 0.66797, "router_noise": 0.0, "grad_norm": 0.8212, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 290.2} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4256.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.244556, "cps": 33.25, "mean_p": 0.436284, "boundary_entropy": 0.660007, "bpb": 2.040888}, "de": {"chunks": 5429.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.264404, "cps": 42.414062, "mean_p": 0.447134, "boundary_entropy": 0.667978, "bpb": 2.233375}, "vi": {"chunks": 5083.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.275726, "cps": 39.710938, "mean_p": 0.447139, "boundary_entropy": 0.670998, "bpb": 1.934195}, "tr": {"chunks": 5986.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.330663, "cps": 46.765625, "mean_p": 0.451702, "boundary_entropy": 0.661988, "bpb": 2.203314}, "fi": {"chunks": 5063.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.269151, "cps": 39.554688, "mean_p": 0.452085, "boundary_entropy": 0.667725, "bpb": 2.403028}, "ru": {"chunks": 6234.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.328971, "cps": 48.703125, "mean_p": 0.433496, "boundary_entropy": 0.67044, "bpb": 1.404555}, "el": {"chunks": 8469.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.412659, "cps": 66.164062, "mean_p": 0.451502, "boundary_entropy": 0.675844, "bpb": 1.467741}, "hi": {"chunks": 8256.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.490698, "cps": 64.5, "mean_p": 0.416467, "boundary_entropy": 0.660391, "bpb": 1.174376}, "ta": {"chunks": 8623.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.461246, "cps": 67.367188, "mean_p": 0.402745, "boundary_entropy": 0.654138, "bpb": 0.958219}, "th": {"chunks": 9736.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.605284, "cps": 76.0625, "mean_p": 0.430399, "boundary_entropy": 0.667506, "bpb": 1.103136}, "zh": {"chunks": 3316.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.561463, "cps": 25.90625, "mean_p": 0.435387, "boundary_entropy": 0.665999, "bpb": 3.172965}, "my": {"chunks": 9060.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.478151, "cps": 70.78125, "mean_p": 0.419906, "boundary_entropy": 0.667383, "bpb": 1.08353}}, "bpb": {"en": 2.04089, "de": 2.23337, "vi": 1.9342, "tr": 2.20331, "fi": 2.40303, "ru": 1.40455, "el": 1.46774, "hi": 1.17438, "ta": 0.95822, "th": 1.10314, "zh": 3.17297, "my": 1.08353}, "gini_cpc": 0.170855, "gini_cps": 0.17234, "bpic": 4.826766, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0811599148665507, "vi": 1.127455653474614, "tr": 1.3520995305060974, "fi": 1.100572216084574, "ru": 1.3451790177951475, "el": 1.6873834633177007, "hi": 2.006490520506318, "ta": 1.886059622009021, "th": 2.4750387392167084, "zh": 2.29585037135677, "my": 1.9551825951039727}, "t": 293.3} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.05117, "de": 2.244, "vi": 1.93309, "tr": 2.21, "fi": 2.4112, "ru": 1.38133, "el": 1.4589, "hi": 1.16886, "ta": 0.9654, "th": 1.12045, "zh": 3.1715, "my": 1.09091}, "bytes_per_chunk": {"en": 4.066, "de": 3.847, "vi": 4.786, "tr": 3.262, "fi": 4.033, "ru": 5.328, "el": 4.556, "hi": 5.35, "ta": 6.271, "th": 4.615, "zh": 5.057, "my": 5.651}, "bpb_mean": 1.76723, "t": 294.6} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.29614, "bpb": 1.86993, "chunk_loss": 0.033714, "F": 0.22961, "G": 0.43811, "chunks": 235.12, "bytes_per_chunk": 4.355, "boundary_entropy": 0.6697, "router_noise": 0.0, "grad_norm": 0.8207, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 294.6} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.26415, "bpb": 1.82379, "chunk_loss": 0.033748, "F": 0.23029, "G": 0.43589, "chunks": 235.81, "bytes_per_chunk": 4.342, "boundary_entropy": 0.66768, "router_noise": 0.0, "grad_norm": 0.7844, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 296.7} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.30512, "bpb": 1.88289, "chunk_loss": 0.033803, "F": 0.23102, "G": 0.43517, "chunks": 236.56, "bytes_per_chunk": 4.329, "boundary_entropy": 0.6667, "router_noise": 0.0, "grad_norm": 0.7801, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 298.8} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.34909, "bpb": 1.94632, "chunk_loss": 0.033381, "F": 0.22461, "G": 0.43481, "chunks": 230.0, "bytes_per_chunk": 4.452, "boundary_entropy": 0.66808, "router_noise": 0.0, "grad_norm": 0.7783, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 301.0} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4210.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.241912, "cps": 32.890625, "mean_p": 0.435706, "boundary_entropy": 0.659712, "bpb": 2.035109}, "de": {"chunks": 5332.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.25968, "cps": 41.65625, "mean_p": 0.446787, "boundary_entropy": 0.668192, "bpb": 2.225629}, "vi": {"chunks": 5253.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.284947, "cps": 41.039062, "mean_p": 0.447569, "boundary_entropy": 0.670978, "bpb": 1.927978}, "tr": {"chunks": 5880.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.324808, "cps": 45.9375, "mean_p": 0.450521, "boundary_entropy": 0.661827, "bpb": 2.195479}, "fi": {"chunks": 5040.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.267928, "cps": 39.375, "mean_p": 0.452007, "boundary_entropy": 0.66795, "bpb": 2.388217}, "ru": {"chunks": 6761.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.356781, "cps": 52.820312, "mean_p": 0.435857, "boundary_entropy": 0.67102, "bpb": 1.395275}, "el": {"chunks": 8102.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.394777, "cps": 63.296875, "mean_p": 0.450768, "boundary_entropy": 0.675616, "bpb": 1.464112}, "hi": {"chunks": 8346.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.496048, "cps": 65.203125, "mean_p": 0.414299, "boundary_entropy": 0.658594, "bpb": 1.164664}, "ta": {"chunks": 9290.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.496924, "cps": 72.578125, "mean_p": 0.405178, "boundary_entropy": 0.654716, "bpb": 0.951243}, "th": {"chunks": 9847.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.612185, "cps": 76.929688, "mean_p": 0.432162, "boundary_entropy": 0.667875, "bpb": 1.09401}, "zh": {"chunks": 3299.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.558584, "cps": 25.773438, "mean_p": 0.435663, "boundary_entropy": 0.666134, "bpb": 3.148092}, "my": {"chunks": 9639.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.508708, "cps": 75.304688, "mean_p": 0.422088, "boundary_entropy": 0.668008, "bpb": 1.080565}}, "bpb": {"en": 2.03511, "de": 2.22563, "vi": 1.92798, "tr": 2.19548, "fi": 2.38822, "ru": 1.39527, "el": 1.46411, "hi": 1.16466, "ta": 0.95124, "th": 1.09401, "zh": 3.14809, "my": 1.08057}, "gini_cpc": 0.174537, "gini_cps": 0.180347, "bpic": 4.738096, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0734449023777608, "vi": 1.177894199753, "tr": 1.3426684964356113, "fi": 1.107543206071244, "ru": 1.4748360543748706, "el": 1.631899532683509, "hi": 2.0505262429720506, "ta": 2.0541505438915353, "th": 2.5306081188342513, "zh": 2.309037027444211, "my": 2.1028613001102654}, "t": 304.1} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.04545, "de": 2.23641, "vi": 1.92781, "tr": 2.20045, "fi": 2.39716, "ru": 1.37239, "el": 1.45485, "hi": 1.16015, "ta": 0.95968, "th": 1.11112, "zh": 3.14673, "my": 1.08837}, "bytes_per_chunk": {"en": 4.078, "de": 3.865, "vi": 4.472, "tr": 3.321, "fi": 4.034, "ru": 4.911, "el": 4.748, "hi": 5.339, "ta": 5.765, "th": 4.565, "zh": 5.052, "my": 5.415}, "bpb_mean": 1.75838, "t": 305.4} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.29029, "bpb": 1.86149, "chunk_loss": 0.033399, "F": 0.22449, "G": 0.43608, "chunks": 229.88, "bytes_per_chunk": 4.455, "boundary_entropy": 0.66925, "router_noise": 0.0, "grad_norm": 0.8251, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 305.5} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.3758, "bpb": 1.98486, "chunk_loss": 0.034127, "F": 0.23618, "G": 0.43552, "chunks": 241.84, "bytes_per_chunk": 4.234, "boundary_entropy": 0.66705, "router_noise": 0.0, "grad_norm": 0.7362, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 307.6} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.38289, "bpb": 1.99509, "chunk_loss": 0.033933, "F": 0.23337, "G": 0.43629, "chunks": 238.97, "bytes_per_chunk": 4.285, "boundary_entropy": 0.66841, "router_noise": 0.0, "grad_norm": 0.8075, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 309.7} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.43018, "bpb": 2.06331, "chunk_loss": 0.033955, "F": 0.23346, "G": 0.43984, "chunks": 239.06, "bytes_per_chunk": 4.283, "boundary_entropy": 0.67037, "router_noise": 0.0, "grad_norm": 0.8287, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 311.9} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4203.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.24151, "cps": 32.835938, "mean_p": 0.4355, "boundary_entropy": 0.659763, "bpb": 2.027712}, "de": {"chunks": 5445.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.265183, "cps": 42.539062, "mean_p": 0.447289, "boundary_entropy": 0.668117, "bpb": 2.219632}, "vi": {"chunks": 5176.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.28077, "cps": 40.4375, "mean_p": 0.445511, "boundary_entropy": 0.670488, "bpb": 1.924193}, "tr": {"chunks": 5781.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.319339, "cps": 45.164062, "mean_p": 0.45044, "boundary_entropy": 0.662502, "bpb": 2.18883}, "fi": {"chunks": 5078.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.269948, "cps": 39.671875, "mean_p": 0.452898, "boundary_entropy": 0.668326, "bpb": 2.378427}, "ru": {"chunks": 6469.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.341372, "cps": 50.539062, "mean_p": 0.434302, "boundary_entropy": 0.670553, "bpb": 1.392296}, "el": {"chunks": 8049.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.392194, "cps": 62.882812, "mean_p": 0.449673, "boundary_entropy": 0.675309, "bpb": 1.454655}, "hi": {"chunks": 8184.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.486419, "cps": 63.9375, "mean_p": 0.413932, "boundary_entropy": 0.658791, "bpb": 1.160684}, "ta": {"chunks": 9034.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.483231, "cps": 70.578125, "mean_p": 0.402263, "boundary_entropy": 0.653876, "bpb": 0.947099}, "th": {"chunks": 9911.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.616164, "cps": 77.429688, "mean_p": 0.431448, "boundary_entropy": 0.667148, "bpb": 1.090799}, "zh": {"chunks": 3439.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.582289, "cps": 26.867188, "mean_p": 0.436127, "boundary_entropy": 0.66621, "bpb": 3.139889}, "my": {"chunks": 8571.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.452343, "cps": 66.960938, "mean_p": 0.419043, "boundary_entropy": 0.667165, "bpb": 1.076792}}, "bpb": {"en": 2.02771, "de": 2.21963, "vi": 1.92419, "tr": 2.18883, "fi": 2.37843, "ru": 1.3923, "el": 1.45465, "hi": 1.16068, "ta": 0.9471, "th": 1.0908, "zh": 3.13989, "my": 1.07679}, "gini_cpc": 0.175317, "gini_cps": 0.170849, "bpic": 4.837169, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0980198898498486, "vi": 1.1625612841538466, "tr": 1.3222608767035076, "fi": 1.1177522256310388, "ru": 1.4134897613653272, "el": 1.6239244192628295, "hi": 2.0140733269485924, "ta": 2.0008721848662336, "th": 2.5512977208650423, "zh": 2.411034713968761, "my": 1.8729787462949425}, "t": 314.9} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.03923, "de": 2.22941, "vi": 1.9223, "tr": 2.19395, "fi": 2.38811, "ru": 1.36903, "el": 1.44634, "hi": 1.15453, "ta": 0.9549, "th": 1.10846, "zh": 3.13878, "my": 1.08452}, "bytes_per_chunk": {"en": 4.115, "de": 3.807, "vi": 4.547, "tr": 3.366, "fi": 3.988, "ru": 5.111, "el": 4.777, "hi": 5.38, "ta": 5.95, "th": 4.546, "zh": 4.856, "my": 6.051}, "bpb_mean": 1.75246, "t": 316.2} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.13462, "bpb": 1.63692, "chunk_loss": 0.032668, "F": 0.21243, "G": 0.43107, "chunks": 217.53, "bytes_per_chunk": 4.707, "boundary_entropy": 0.66743, "router_noise": 0.0, "grad_norm": 0.7344, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 316.3} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.19748, "bpb": 1.72761, "chunk_loss": 0.033368, "F": 0.22394, "G": 0.43602, "chunks": 229.31, "bytes_per_chunk": 4.466, "boundary_entropy": 0.66897, "router_noise": 0.0, "grad_norm": 0.7945, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 318.4} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.39947, "bpb": 2.01901, "chunk_loss": 0.034467, "F": 0.24243, "G": 0.43625, "chunks": 248.25, "bytes_per_chunk": 4.125, "boundary_entropy": 0.66655, "router_noise": 0.0, "grad_norm": 0.8177, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 320.5} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.41486, "bpb": 2.04122, "chunk_loss": 0.034421, "F": 0.2413, "G": 0.43808, "chunks": 247.09, "bytes_per_chunk": 4.144, "boundary_entropy": 0.66817, "router_noise": 0.0, "grad_norm": 0.8484, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 322.6} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4225.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.242774, "cps": 33.007812, "mean_p": 0.435476, "boundary_entropy": 0.659644, "bpb": 2.022354}, "de": {"chunks": 5351.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.260605, "cps": 41.804688, "mean_p": 0.446479, "boundary_entropy": 0.667832, "bpb": 2.214877}, "vi": {"chunks": 5326.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.288907, "cps": 41.609375, "mean_p": 0.446307, "boundary_entropy": 0.670386, "bpb": 1.920134}, "tr": {"chunks": 5933.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.327736, "cps": 46.351562, "mean_p": 0.450734, "boundary_entropy": 0.661787, "bpb": 2.17523}, "fi": {"chunks": 5160.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.274308, "cps": 40.3125, "mean_p": 0.45277, "boundary_entropy": 0.667804, "bpb": 2.370455}, "ru": {"chunks": 6813.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.359525, "cps": 53.226562, "mean_p": 0.435661, "boundary_entropy": 0.670835, "bpb": 1.384324}, "el": {"chunks": 8105.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.394923, "cps": 63.320312, "mean_p": 0.450343, "boundary_entropy": 0.675573, "bpb": 1.449007}, "hi": {"chunks": 8020.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.476672, "cps": 62.65625, "mean_p": 0.415127, "boundary_entropy": 0.659835, "bpb": 1.157272}, "ta": {"chunks": 9518.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.50912, "cps": 74.359375, "mean_p": 0.405897, "boundary_entropy": 0.655514, "bpb": 0.942628}, "th": {"chunks": 9199.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.571899, "cps": 71.867188, "mean_p": 0.428205, "boundary_entropy": 0.667103, "bpb": 1.090591}, "zh": {"chunks": 3436.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.581781, "cps": 26.84375, "mean_p": 0.436769, "boundary_entropy": 0.666353, "bpb": 3.133687}, "my": {"chunks": 9183.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.484642, "cps": 71.742188, "mean_p": 0.422398, "boundary_entropy": 0.668548, "bpb": 1.073109}}, "bpb": {"en": 2.02235, "de": 2.21488, "vi": 1.92013, "tr": 2.17523, "fi": 2.37046, "ru": 1.38432, "el": 1.44901, "hi": 1.15727, "ta": 0.94263, "th": 1.09059, "zh": 3.13369, "my": 1.07311}, "gini_cpc": 0.16823, "gini_cps": 0.168819, "bpic": 4.781186, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0734453788777598, "vi": 1.1900231968069483, "tr": 1.3499609434115827, "fi": 1.129887554776429, "ru": 1.4809028930071348, "el": 1.6267079192719844, "hi": 1.9634357860678584, "ta": 2.0970927454260306, "th": 2.355683611288403, "zh": 2.3963879075937213, "my": 1.9962669429756679}, "t": 325.7} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.03424, "de": 2.22551, "vi": 1.91834, "tr": 2.18022, "fi": 2.38008, "ru": 1.36186, "el": 1.43938, "hi": 1.15304, "ta": 0.95064, "th": 1.10779, "zh": 3.1321, "my": 1.08012}, "bytes_per_chunk": {"en": 4.106, "de": 3.886, "vi": 4.401, "tr": 3.309, "fi": 3.995, "ru": 4.87, "el": 4.76, "hi": 5.497, "ta": 5.636, "th": 4.882, "zh": 4.868, "my": 5.706}, "bpb_mean": 1.74694, "t": 327.0} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.46978, "bpb": 2.12044, "chunk_loss": 0.033727, "F": 0.2301, "G": 0.43599, "chunks": 235.62, "bytes_per_chunk": 4.346, "boundary_entropy": 0.66839, "router_noise": 0.0, "grad_norm": 0.8211, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 327.0} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.27899, "bpb": 1.84519, "chunk_loss": 0.033654, "F": 0.22888, "G": 0.43499, "chunks": 234.38, "bytes_per_chunk": 4.369, "boundary_entropy": 0.6683, "router_noise": 0.0, "grad_norm": 0.771, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 329.2} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.33013, "bpb": 1.91897, "chunk_loss": 0.033821, "F": 0.23123, "G": 0.4372, "chunks": 236.78, "bytes_per_chunk": 4.325, "boundary_entropy": 0.66822, "router_noise": 0.0, "grad_norm": 0.8333, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 331.3} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.20849, "bpb": 1.74349, "chunk_loss": 0.033979, "F": 0.23434, "G": 0.43434, "chunks": 239.97, "bytes_per_chunk": 4.267, "boundary_entropy": 0.66684, "router_noise": 0.0, "grad_norm": 0.7809, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 333.4} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4219.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.242429, "cps": 32.960938, "mean_p": 0.435133, "boundary_entropy": 0.659364, "bpb": 2.018332}, "de": {"chunks": 5427.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.264306, "cps": 42.398438, "mean_p": 0.446859, "boundary_entropy": 0.667852, "bpb": 2.21134}, "vi": {"chunks": 5243.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.284405, "cps": 40.960938, "mean_p": 0.445873, "boundary_entropy": 0.670429, "bpb": 1.914808}, "tr": {"chunks": 5901.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.325968, "cps": 46.101562, "mean_p": 0.45, "boundary_entropy": 0.6616, "bpb": 2.169249}, "fi": {"chunks": 5098.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.271012, "cps": 39.828125, "mean_p": 0.451467, "boundary_entropy": 0.667459, "bpb": 2.364206}, "ru": {"chunks": 6720.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.354617, "cps": 52.5, "mean_p": 0.43462, "boundary_entropy": 0.670428, "bpb": 1.379928}, "el": {"chunks": 8300.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.404424, "cps": 64.84375, "mean_p": 0.450965, "boundary_entropy": 0.675804, "bpb": 1.443981}, "hi": {"chunks": 8373.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.497652, "cps": 65.414062, "mean_p": 0.41434, "boundary_entropy": 0.658337, "bpb": 1.15108}, "ta": {"chunks": 9155.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.489703, "cps": 71.523438, "mean_p": 0.402549, "boundary_entropy": 0.653496, "bpb": 0.940577}, "th": {"chunks": 9807.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.609698, "cps": 76.617188, "mean_p": 0.429504, "boundary_entropy": 0.666158, "bpb": 1.085579}, "zh": {"chunks": 3410.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.577379, "cps": 26.640625, "mean_p": 0.435788, "boundary_entropy": 0.666014, "bpb": 3.129267}, "my": {"chunks": 9224.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.486806, "cps": 72.0625, "mean_p": 0.420826, "boundary_entropy": 0.667355, "bpb": 1.06914}}, "bpb": {"en": 2.01833, "de": 2.21134, "vi": 1.91481, "tr": 2.16925, "fi": 2.36421, "ru": 1.37993, "el": 1.44398, "hi": 1.15108, "ta": 0.94058, "th": 1.08558, "zh": 3.12927, "my": 1.06914}, "gini_cpc": 0.173358, "gini_cps": 0.174029, "bpic": 4.745243, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0902397422970327, "vi": 1.173143964370565, "tr": 1.3445893220874043, "fi": 1.117898935266414, "ru": 1.4627653140929868, "el": 1.6682143091136674, "hi": 2.052771517232417, "ta": 2.0199818812753803, "th": 2.5149520247635606, "zh": 2.381636794251602, "my": 2.008031461762367}, "t": 336.5} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.02939, "de": 2.22295, "vi": 1.91287, "tr": 2.17497, "fi": 2.37462, "ru": 1.35779, "el": 1.43558, "hi": 1.14623, "ta": 0.94857, "th": 1.10257, "zh": 3.12798, "my": 1.07684}, "bytes_per_chunk": {"en": 4.071, "de": 3.848, "vi": 4.485, "tr": 3.296, "fi": 4.018, "ru": 4.964, "el": 4.662, "hi": 5.33, "ta": 5.851, "th": 4.617, "zh": 4.852, "my": 5.546}, "bpb_mean": 1.74253, "t": 337.8} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.25926, "bpb": 1.81673, "chunk_loss": 0.03348, "F": 0.22647, "G": 0.43004, "chunks": 231.91, "bytes_per_chunk": 4.416, "boundary_entropy": 0.66504, "router_noise": 0.0, "grad_norm": 0.7657, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 337.8} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.37595, "bpb": 1.98507, "chunk_loss": 0.033632, "F": 0.22815, "G": 0.43609, "chunks": 233.62, "bytes_per_chunk": 4.383, "boundary_entropy": 0.66817, "router_noise": 0.0, "grad_norm": 0.7576, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 340.0} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.3282, "bpb": 1.91619, "chunk_loss": 0.033685, "F": 0.22946, "G": 0.43751, "chunks": 234.97, "bytes_per_chunk": 4.358, "boundary_entropy": 0.66809, "router_noise": 0.0, "grad_norm": 0.7755, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 342.1} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.28482, "bpb": 1.8536, "chunk_loss": 0.033617, "F": 0.22839, "G": 0.43355, "chunks": 233.88, "bytes_per_chunk": 4.378, "boundary_entropy": 0.66718, "router_noise": 0.0, "grad_norm": 0.772, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 344.3} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4291.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.246567, "cps": 33.523438, "mean_p": 0.436294, "boundary_entropy": 0.659316, "bpb": 2.01389}, "de": {"chunks": 5377.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.261871, "cps": 42.007812, "mean_p": 0.446769, "boundary_entropy": 0.667575, "bpb": 2.206416}, "vi": {"chunks": 5070.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.27502, "cps": 39.609375, "mean_p": 0.44681, "boundary_entropy": 0.670793, "bpb": 1.913277}, "tr": {"chunks": 5988.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.330774, "cps": 46.78125, "mean_p": 0.451354, "boundary_entropy": 0.661655, "bpb": 2.166047}, "fi": {"chunks": 5181.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.275424, "cps": 40.476562, "mean_p": 0.453113, "boundary_entropy": 0.667917, "bpb": 2.362211}, "ru": {"chunks": 6941.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.36628, "cps": 54.226562, "mean_p": 0.434499, "boundary_entropy": 0.670094, "bpb": 1.37592}, "el": {"chunks": 8074.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.393412, "cps": 63.078125, "mean_p": 0.450347, "boundary_entropy": 0.675861, "bpb": 1.43868}, "hi": {"chunks": 8469.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.503358, "cps": 66.164062, "mean_p": 0.414627, "boundary_entropy": 0.658497, "bpb": 1.150185}, "ta": {"chunks": 9191.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.491629, "cps": 71.804688, "mean_p": 0.403639, "boundary_entropy": 0.654523, "bpb": 0.938217}, "th": {"chunks": 9666.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.600933, "cps": 75.515625, "mean_p": 0.429188, "boundary_entropy": 0.666521, "bpb": 1.080535}, "zh": {"chunks": 3445.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583305, "cps": 26.914062, "mean_p": 0.436332, "boundary_entropy": 0.665864, "bpb": 3.122795}, "my": {"chunks": 8535.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.450443, "cps": 66.679688, "mean_p": 0.418062, "boundary_entropy": 0.666439, "bpb": 1.072136}}, "bpb": {"en": 2.01389, "de": 2.20642, "vi": 1.91328, "tr": 2.16605, "fi": 2.36221, "ru": 1.37592, "el": 1.43868, "hi": 1.15019, "ta": 0.93822, "th": 1.08054, "zh": 3.1228, "my": 1.07214}, "gini_cpc": 0.17187, "gini_cps": 0.168551, "bpic": 4.783629, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0620702283220003, "vi": 1.1153994423964175, "tr": 1.3415190553683638, "fi": 1.1170363720974708, "ru": 1.4855197692415063, "el": 1.5955613423948596, "hi": 2.0414684177775593, "ta": 1.9938978371357505, "th": 2.437200907814152, "zh": 2.365709365943872, "my": 1.8268620537941986}, "t": 347.4} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.02616, "de": 2.21857, "vi": 1.91077, "tr": 2.17098, "fi": 2.3711, "ru": 1.3536, "el": 1.43041, "hi": 1.14548, "ta": 0.94665, "th": 1.09828, "zh": 3.12234, "my": 1.07921}, "bytes_per_chunk": {"en": 4.028, "de": 3.881, "vi": 4.683, "tr": 3.268, "fi": 3.942, "ru": 4.791, "el": 4.768, "hi": 5.213, "ta": 5.827, "th": 4.642, "zh": 4.798, "my": 6.039}, "bpb_mean": 1.73946, "t": 348.8} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4291.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.246567, "cps": 33.523438, "mean_p": 0.436294, "boundary_entropy": 0.659316, "bpb": 2.01389}, "de": {"chunks": 5377.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.261871, "cps": 42.007812, "mean_p": 0.446769, "boundary_entropy": 0.667575, "bpb": 2.206416}, "vi": {"chunks": 5070.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.27502, "cps": 39.609375, "mean_p": 0.44681, "boundary_entropy": 0.670793, "bpb": 1.913277}, "tr": {"chunks": 5988.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.330774, "cps": 46.78125, "mean_p": 0.451354, "boundary_entropy": 0.661655, "bpb": 2.166047}, "fi": {"chunks": 5181.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.275424, "cps": 40.476562, "mean_p": 0.453113, "boundary_entropy": 0.667917, "bpb": 2.362211}, "ru": {"chunks": 6941.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.36628, "cps": 54.226562, "mean_p": 0.434499, "boundary_entropy": 0.670094, "bpb": 1.37592}, "el": {"chunks": 8074.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.393412, "cps": 63.078125, "mean_p": 0.450347, "boundary_entropy": 0.675861, "bpb": 1.43868}, "hi": {"chunks": 8469.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.503358, "cps": 66.164062, "mean_p": 0.414627, "boundary_entropy": 0.658497, "bpb": 1.150185}, "ta": {"chunks": 9191.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.491629, "cps": 71.804688, "mean_p": 0.403639, "boundary_entropy": 0.654523, "bpb": 0.938217}, "th": {"chunks": 9666.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.600933, "cps": 75.515625, "mean_p": 0.429188, "boundary_entropy": 0.666521, "bpb": 1.080535}, "zh": {"chunks": 3445.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583305, "cps": 26.914062, "mean_p": 0.436332, "boundary_entropy": 0.665864, "bpb": 3.122795}, "my": {"chunks": 8535.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.450443, "cps": 66.679688, "mean_p": 0.418062, "boundary_entropy": 0.666439, "bpb": 1.072136}}, "bpb": {"en": 2.01389, "de": 2.20642, "vi": 1.91328, "tr": 2.16605, "fi": 2.36221, "ru": 1.37592, "el": 1.43868, "hi": 1.15019, "ta": 0.93822, "th": 1.08054, "zh": 3.1228, "my": 1.07214}, "gini_cpc": 0.17187, "gini_cps": 0.168551, "bpic": 4.783629, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0620702283220003, "vi": 1.1153994423964175, "tr": 1.3415190553683638, "fi": 1.1170363720974708, "ru": 1.4855197692415063, "el": 1.5955613423948596, "hi": 2.0414684177775593, "ta": 1.9938978371357505, "th": 2.437200907814152, "zh": 2.365709365943872, "my": 1.8268620537941986}, "t": 349.8} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s0/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s0/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s0/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s0/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s0/th0.70"], "t": 354.5} |
| {"event": "final", "run": "emactl_genreg_s0", "condition": "A", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.24656668390507383, "cps": 33.5234375, "bpb": 2.0362594194956314, "bytes_per_chunk": 4.029017582687815}, "de": {"cpc": 0.26187113427166026, "cps": 42.0078125, "bpb": 2.210904942235243, "bytes_per_chunk": 3.879821211852115}, "vi": {"cpc": 0.27502034174125306, "cps": 39.609375, "bpb": 1.914473000974849, "bytes_per_chunk": 4.687504470352621}, "tr": {"cpc": 0.3307739048776446, "cps": 46.78125, "bpb": 2.163528064484473, "bytes_per_chunk": 3.2757763199000314}, "fi": {"cpc": 0.2754239540694275, "cps": 40.4765625, "bpb": 2.387229897058361, "bytes_per_chunk": 3.928634050939282}, "ru": {"cpc": 0.3662796833773087, "cps": 54.2265625, "bpb": 1.354086676957211, "bytes_per_chunk": 4.78299502805273}, "el": {"cpc": 0.39341226916142863, "cps": 63.078125, "bpb": 1.4254445253235672, "bytes_per_chunk": 4.771372927322036}, "hi": {"cpc": 0.5033580980683506, "cps": 66.1640625, "bpb": 1.15441516864525, "bytes_per_chunk": 5.213268634157982}, "ta": {"cpc": 0.49162877774806096, "cps": 71.8046875, "bpb": 0.9478137860249648, "bytes_per_chunk": 5.832161608970366}, "th": {"cpc": 0.600932545850171, "cps": 75.515625, "bpb": 1.0961344833406215, "bytes_per_chunk": 4.624166519668372}, "zh": {"cpc": 0.5833051134439553, "cps": 26.9140625, "bpb": 3.1291131026693577, "bytes_per_chunk": 4.784129794048673}, "my": {"cpc": 0.45044331855604813, "cps": 66.6796875, "bpb": 1.0903263282513456, "bytes_per_chunk": 6.048406820331787}}, "gini_cpc": 0.1718704557282591, "gini_cps": 0.16855087999202278, "bpic": 4.783629156902827, "seed": 0, "mean_boundary_entropy": 0.6691246032714844, "dechunk": "ste", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 356.9} |
|
|
