| {"event": "start", "run": "emactl_genreg_s2", "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": 1.3} |
| {"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": 1.3} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10933.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.628225, "cps": 85.414062, "mean_p": 0.513249, "boundary_entropy": 0.686295, "bpb": 8.157895}, "de": {"chunks": 16287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.793211, "cps": 127.242188, "mean_p": 0.524631, "boundary_entropy": 0.686391, "bpb": 8.150604}, "vi": {"chunks": 5099.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276593, "cps": 39.835938, "mean_p": 0.481643, "boundary_entropy": 0.68728, "bpb": 7.885576}, "tr": {"chunks": 8187.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.452245, "cps": 63.960938, "mean_p": 0.498002, "boundary_entropy": 0.686826, "bpb": 8.139464}, "fi": {"chunks": 13540.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.719792, "cps": 105.78125, "mean_p": 0.517169, "boundary_entropy": 0.686661, "bpb": 8.097447}, "ru": {"chunks": 33477.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.766596, "cps": 261.539062, "mean_p": 0.530376, "boundary_entropy": 0.688543, "bpb": 8.304526}, "el": {"chunks": 20696.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.00843, "cps": 161.6875, "mean_p": 0.503648, "boundary_entropy": 0.690112, "bpb": 8.265181}, "hi": {"chunks": 875.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.052006, "cps": 6.835938, "mean_p": 0.461978, "boundary_entropy": 0.68743, "bpb": 8.053074}, "ta": {"chunks": 1261.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.067451, "cps": 9.851562, "mean_p": 0.477518, "boundary_entropy": 0.690053, "bpb": 8.127442}, "th": {"chunks": 3630.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.225676, "cps": 28.359375, "mean_p": 0.484845, "boundary_entropy": 0.690398, "bpb": 7.710377}, "zh": {"chunks": 11722.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.984761, "cps": 91.578125, "mean_p": 0.520316, "boundary_entropy": 0.68562, "bpb": 8.124258}, "my": {"chunks": 1628.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085919, "cps": 12.71875, "mean_p": 0.474628, "boundary_entropy": 0.689685, "bpb": 7.688162}}, "bpb": {"en": 8.15789, "de": 8.1506, "vi": 7.88558, "tr": 8.13946, "fi": 8.09745, "ru": 8.30453, "el": 8.26518, "hi": 8.05307, "ta": 8.12744, "th": 7.71038, "zh": 8.12426, "my": 7.68816}, "gini_cpc": 0.489538, "gini_cps": 0.463551, "bpic": 3.013947, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2626223171148012, "vi": 0.440277652395235, "tr": 0.7198781817505053, "fi": 1.1457544508645294, "ru": 2.8120438593717987, "el": 1.6052044042347315, "hi": 0.08278235025738971, "ta": 0.10736788276924242, "th": 0.3592281253258146, "zh": 3.159315851384177, "my": 0.13676526035979295}, "t": 3.3} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.65891, "bpb": 8.16408, "chunk_loss": 0.065103, "F": 0.62439, "G": 0.5091, "chunks": 639.38, "bytes_per_chunk": 1.602, "boundary_entropy": 0.69066, "router_noise": 0.0, "grad_norm": 7.3916, "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": 10931.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.62811, "cps": 85.398438, "mean_p": 0.513237, "boundary_entropy": 0.6863, "bpb": 8.134186}, "de": {"chunks": 16277.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.792724, "cps": 127.164062, "mean_p": 0.524625, "boundary_entropy": 0.686395, "bpb": 8.129297}, "vi": {"chunks": 5092.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.276214, "cps": 39.78125, "mean_p": 0.481598, "boundary_entropy": 0.687276, "bpb": 7.872886}, "tr": {"chunks": 8217.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.453903, "cps": 64.195312, "mean_p": 0.498129, "boundary_entropy": 0.686829, "bpb": 8.124928}, "fi": {"chunks": 13531.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.719313, "cps": 105.710938, "mean_p": 0.517125, "boundary_entropy": 0.686667, "bpb": 8.076826}, "ru": {"chunks": 33476.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.766544, "cps": 261.53125, "mean_p": 0.53043, "boundary_entropy": 0.688536, "bpb": 8.26764}, "el": {"chunks": 20697.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.008478, "cps": 161.695312, "mean_p": 0.503649, "boundary_entropy": 0.690112, "bpb": 8.259751}, "hi": {"chunks": 876.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.052065, "cps": 6.84375, "mean_p": 0.462018, "boundary_entropy": 0.687436, "bpb": 8.046612}, "ta": {"chunks": 1279.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.068414, "cps": 9.992188, "mean_p": 0.477595, "boundary_entropy": 0.690061, "bpb": 8.121091}, "th": {"chunks": 3633.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.225863, "cps": 28.382812, "mean_p": 0.484847, "boundary_entropy": 0.690397, "bpb": 7.707639}, "zh": {"chunks": 11710.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.982729, "cps": 91.484375, "mean_p": 0.520278, "boundary_entropy": 0.685623, "bpb": 8.120828}, "my": {"chunks": 1617.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085339, "cps": 12.632812, "mean_p": 0.474602, "boundary_entropy": 0.689682, "bpb": 7.686636}}, "bpb": {"en": 8.13419, "de": 8.1293, "vi": 7.87289, "tr": 8.12493, "fi": 8.07683, "ru": 8.26764, "el": 8.25975, "hi": 8.04661, "ta": 8.12109, "th": 7.70764, "zh": 8.12083, "my": 7.68664}, "gini_cpc": 0.489268, "gini_cps": 0.463367, "bpic": 3.013924, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2620779588703526, "vi": 0.4397536764194128, "tr": 0.7226482602098392, "fi": 1.1452023657212618, "ru": 2.812474352794417, "el": 1.6055756771512673, "hi": 0.08289212231309648, "ta": 0.10892041838262422, "th": 0.3595907888823745, "zh": 3.156659064154633, "my": 0.13586602506169015}, "t": 4.9} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10927.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.62788, "cps": 85.367188, "mean_p": 0.513223, "boundary_entropy": 0.68631, "bpb": 8.088293}, "de": {"chunks": 16286.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.793162, "cps": 127.234375, "mean_p": 0.524617, "boundary_entropy": 0.686403, "bpb": 8.087217}, "vi": {"chunks": 5073.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.275183, "cps": 39.632812, "mean_p": 0.481502, "boundary_entropy": 0.687266, "bpb": 7.851395}, "tr": {"chunks": 8277.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.457217, "cps": 64.664062, "mean_p": 0.498371, "boundary_entropy": 0.686836, "bpb": 8.095007}, "fi": {"chunks": 13519.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.718675, "cps": 105.617188, "mean_p": 0.517052, "boundary_entropy": 0.686677, "bpb": 8.035252}, "ru": {"chunks": 33485.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.767018, "cps": 261.601562, "mean_p": 0.530579, "boundary_entropy": 0.688517, "bpb": 8.193549}, "el": {"chunks": 20681.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.007699, "cps": 161.570312, "mean_p": 0.503626, "boundary_entropy": 0.690113, "bpb": 8.254527}, "hi": {"chunks": 897.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.053314, "cps": 7.007812, "mean_p": 0.462179, "boundary_entropy": 0.687462, "bpb": 8.02111}, "ta": {"chunks": 1306.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.069858, "cps": 10.203125, "mean_p": 0.477805, "boundary_entropy": 0.690081, "bpb": 8.107852}, "th": {"chunks": 3678.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.22866, "cps": 28.734375, "mean_p": 0.484935, "boundary_entropy": 0.690402, "bpb": 7.694271}, "zh": {"chunks": 11694.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.98002, "cps": 91.359375, "mean_p": 0.520225, "boundary_entropy": 0.685626, "bpb": 8.113904}, "my": {"chunks": 1622.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085603, "cps": 12.671875, "mean_p": 0.474601, "boundary_entropy": 0.689682, "bpb": 7.68659}}, "bpb": {"en": 8.08829, "de": 8.08722, "vi": 7.85139, "tr": 8.09501, "fi": 8.03525, "ru": 8.19355, "el": 8.25453, "hi": 8.02111, "ta": 8.10785, "th": 7.69427, "zh": 8.1139, "my": 7.68659}, "gini_cpc": 0.488228, "gini_cps": 0.462428, "bpic": 3.011346, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.263238055326777, "vi": 0.4382731825667834, "tr": 0.7281914592207113, "fi": 1.1446055877699337, "ru": 2.814260311836793, "el": 1.6049217644244091, "hi": 0.0849103336188519, "ta": 0.11126046855284821, "th": 0.36417810923078725, "zh": 3.153499917610423, "my": 0.13633603232413533}, "t": 6.4} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10921.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.627535, "cps": 85.320312, "mean_p": 0.513077, "boundary_entropy": 0.686344, "bpb": 7.941284}, "de": {"chunks": 16311.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.79438, "cps": 127.429688, "mean_p": 0.524501, "boundary_entropy": 0.686437, "bpb": 7.955596}, "vi": {"chunks": 5036.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.273176, "cps": 39.34375, "mean_p": 0.481293, "boundary_entropy": 0.687244, "bpb": 7.781651}, "tr": {"chunks": 8480.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.468431, "cps": 66.25, "mean_p": 0.499112, "boundary_entropy": 0.686855, "bpb": 7.997115}, "fi": {"chunks": 13476.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.716389, "cps": 105.28125, "mean_p": 0.516729, "boundary_entropy": 0.686715, "bpb": 7.903054}, "ru": {"chunks": 33528.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.769288, "cps": 261.9375, "mean_p": 0.531281, "boundary_entropy": 0.688427, "bpb": 7.949979}, "el": {"chunks": 20463.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.997076, "cps": 159.867188, "mean_p": 0.503363, "boundary_entropy": 0.690116, "bpb": 8.206512}, "hi": {"chunks": 938.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.05575, "cps": 7.328125, "mean_p": 0.462569, "boundary_entropy": 0.687524, "bpb": 7.939445}, "ta": {"chunks": 1401.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.07494, "cps": 10.945312, "mean_p": 0.478287, "boundary_entropy": 0.690124, "bpb": 8.032558}, "th": {"chunks": 3735.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.232204, "cps": 29.179688, "mean_p": 0.484902, "boundary_entropy": 0.690395, "bpb": 7.621806}, "zh": {"chunks": 11641.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.971046, "cps": 90.945312, "mean_p": 0.519956, "boundary_entropy": 0.685644, "bpb": 8.086994}, "my": {"chunks": 1588.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.083808, "cps": 12.40625, "mean_p": 0.474427, "boundary_entropy": 0.689662, "bpb": 7.683362}}, "bpb": {"en": 7.94128, "de": 7.9556, "vi": 7.78165, "tr": 7.99712, "fi": 7.90305, "ru": 7.94998, "el": 8.20651, "hi": 7.93945, "ta": 8.03256, "th": 7.62181, "zh": 8.08699, "my": 7.68336}, "gini_cpc": 0.485798, "gini_cps": 0.460112, "bpic": 3.009622, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.265872291188209, "vi": 0.43531566187790427, "tr": 0.7464608137718268, "fi": 1.1415917773010082, "ru": 2.819422404947598, "el": 1.588876611433824, "hi": 0.08884018997438446, "ta": 0.11941926087913002, "th": 0.37002515906975625, "zh": 3.14093218441975, "my": 0.13355151995803358}, "t": 8.0} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 10611.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.609722, "cps": 82.898438, "mean_p": 0.511398, "boundary_entropy": 0.686513, "bpb": 7.483408}, "de": {"chunks": 16138.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.785954, "cps": 126.078125, "mean_p": 0.522866, "boundary_entropy": 0.686672, "bpb": 7.541433}, "vi": {"chunks": 4839.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.26249, "cps": 37.804688, "mean_p": 0.48052, "boundary_entropy": 0.687161, "bpb": 7.56082}, "tr": {"chunks": 9127.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.504171, "cps": 71.304688, "mean_p": 0.501698, "boundary_entropy": 0.686913, "bpb": 7.662294}, "fi": {"chunks": 13005.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.691351, "cps": 101.601562, "mean_p": 0.51441, "boundary_entropy": 0.686914, "bpb": 7.489155}, "ru": {"chunks": 33816.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.784485, "cps": 264.1875, "mean_p": 0.535772, "boundary_entropy": 0.687825, "bpb": 7.383502}, "el": {"chunks": 19653.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.957609, "cps": 153.539062, "mean_p": 0.502313, "boundary_entropy": 0.690126, "bpb": 7.973775}, "hi": {"chunks": 1042.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.061932, "cps": 8.140625, "mean_p": 0.463591, "boundary_entropy": 0.687684, "bpb": 7.685043}, "ta": {"chunks": 1658.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.088687, "cps": 12.953125, "mean_p": 0.479374, "boundary_entropy": 0.690218, "bpb": 7.810084}, "th": {"chunks": 3771.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.234442, "cps": 29.460938, "mean_p": 0.484366, "boundary_entropy": 0.690344, "bpb": 7.382868}, "zh": {"chunks": 11527.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.951744, "cps": 90.054688, "mean_p": 0.519229, "boundary_entropy": 0.685688, "bpb": 8.009413}, "my": {"chunks": 1617.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.085339, "cps": 12.632812, "mean_p": 0.474683, "boundary_entropy": 0.689693, "bpb": 7.572966}}, "bpb": {"en": 7.48341, "de": 7.54143, "vi": 7.56082, "tr": 7.66229, "fi": 7.48916, "ru": 7.3835, "el": 7.97377, "hi": 7.68504, "ta": 7.81008, "th": 7.38287, "zh": 8.00941, "my": 7.57297}, "gini_cpc": 0.481342, "gini_cps": 0.454715, "bpic": 3.026569, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2890361875800695, "vi": 0.43050706779177145, "tr": 0.8268853637146945, "fi": 1.133877868817755, "ru": 2.926717646769209, "el": 1.57056463719701, "hi": 0.10157350799298101, "ta": 0.14545439972330862, "th": 0.38450613431665903, "zh": 3.201036720706199, "my": 0.13996338893123506}, "t": 9.7} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6656.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.382463, "cps": 52.0, "mean_p": 0.496497, "boundary_entropy": 0.686792, "bpb": 6.904403}, "de": {"chunks": 11891.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.579117, "cps": 92.898438, "mean_p": 0.506573, "boundary_entropy": 0.687732, "bpb": 6.995068}, "vi": {"chunks": 3716.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.201573, "cps": 29.03125, "mean_p": 0.47566, "boundary_entropy": 0.686649, "bpb": 7.200353}, "tr": {"chunks": 9638.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.532398, "cps": 75.296875, "mean_p": 0.503094, "boundary_entropy": 0.687052, "bpb": 7.195272}, "fi": {"chunks": 8341.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.443411, "cps": 65.164062, "mean_p": 0.498603, "boundary_entropy": 0.687378, "bpb": 6.973773}, "ru": {"chunks": 34322.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.811187, "cps": 268.140625, "mean_p": 0.556767, "boundary_entropy": 0.683954, "bpb": 6.514413}, "el": {"chunks": 18483.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.900599, "cps": 144.398438, "mean_p": 0.500433, "boundary_entropy": 0.690138, "bpb": 7.390376}, "hi": {"chunks": 872.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.051828, "cps": 6.8125, "mean_p": 0.4625, "boundary_entropy": 0.687544, "bpb": 6.861709}, "ta": {"chunks": 2294.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.122707, "cps": 17.921875, "mean_p": 0.480716, "boundary_entropy": 0.690311, "bpb": 6.970221}, "th": {"chunks": 2532.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.157414, "cps": 19.78125, "mean_p": 0.479476, "boundary_entropy": 0.689918, "bpb": 6.641495}, "zh": {"chunks": 10888.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.843549, "cps": 85.0625, "mean_p": 0.516182, "boundary_entropy": 0.68585, "bpb": 7.789761}, "my": {"chunks": 1802.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.095102, "cps": 14.078125, "mean_p": 0.476161, "boundary_entropy": 0.689873, "bpb": 7.158891}}, "bpb": {"en": 6.9044, "de": 6.99507, "vi": 7.20035, "tr": 7.19527, "fi": 6.97377, "ru": 6.51441, "el": 7.39038, "hi": 6.86171, "ta": 6.97022, "th": 6.64149, "zh": 7.78976, "my": 7.15889}, "gini_cpc": 0.512086, "gini_cps": 0.483457, "bpic": 3.44399, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.5141774664915577, "vi": 0.5270397485447831, "tr": 1.392025456495736, "fi": 1.1593566158993511, "ru": 4.735590924370814, "el": 2.3547370940413006, "hi": 0.13551027260258316, "ta": 0.32083279221614175, "th": 0.41157922305779393, "zh": 4.820204640326656, "my": 0.2486578747655527}, "t": 11.7} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2427.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.139459, "cps": 18.960938, "mean_p": 0.482491, "boundary_entropy": 0.686398, "bpb": 6.389212}, "de": {"chunks": 3274.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159451, "cps": 25.578125, "mean_p": 0.484916, "boundary_entropy": 0.687531, "bpb": 6.478528}, "vi": {"chunks": 3923.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.212802, "cps": 30.648438, "mean_p": 0.480806, "boundary_entropy": 0.687459, "bpb": 6.792049}, "tr": {"chunks": 7869.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.434679, "cps": 61.476562, "mean_p": 0.497171, "boundary_entropy": 0.687415, "bpb": 6.743942}, "fi": {"chunks": 2271.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.120727, "cps": 17.742188, "mean_p": 0.481547, "boundary_entropy": 0.686967, "bpb": 6.485234}, "ru": {"chunks": 19370.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.022164, "cps": 151.328125, "mean_p": 0.504061, "boundary_entropy": 0.690078, "bpb": 5.909203}, "el": {"chunks": 19211.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.936072, "cps": 150.085938, "mean_p": 0.501024, "boundary_entropy": 0.690219, "bpb": 6.298318}, "hi": {"chunks": 2186.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.129926, "cps": 17.078125, "mean_p": 0.48091, "boundary_entropy": 0.689978, "bpb": 5.887968}, "ta": {"chunks": 1088.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.058197, "cps": 8.5, "mean_p": 0.478542, "boundary_entropy": 0.690143, "bpb": 6.059309}, "th": {"chunks": 2708.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.168356, "cps": 21.15625, "mean_p": 0.480468, "boundary_entropy": 0.69004, "bpb": 5.7338}, "zh": {"chunks": 4178.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.707416, "cps": 32.640625, "mean_p": 0.486723, "boundary_entropy": 0.685654, "bpb": 7.360496}, "my": {"chunks": 3206.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.1692, "cps": 25.046875, "mean_p": 0.484672, "boundary_entropy": 0.69071, "bpb": 5.884437}}, "bpb": {"en": 6.38921, "de": 6.47853, "vi": 6.79205, "tr": 6.74394, "fi": 6.48523, "ru": 5.9092, "el": 6.29832, "hi": 5.88797, "ta": 6.05931, "th": 5.7338, "zh": 7.3605, "my": 5.88444}, "gini_cpc": 0.462241, "gini_cps": 0.471779, "bpic": 5.351773, "worst_lang": "ru", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1433537270071505, "vi": 1.5259120760712392, "tr": 3.1169033644656756, "fi": 0.865684406442618, "ru": 7.329506769163532, "el": 6.712178088113267, "hi": 0.931642792727433, "ta": 0.41730901786567337, "th": 1.2072075381365912, "zh": 5.072585043723736, "my": 1.2132616936391913}, "t": 14.3} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.28018, "bpb": 6.17499, "chunk_loss": 0.037038, "F": 0.25903, "G": 0.4867, "chunks": 265.25, "bytes_per_chunk": 3.861, "boundary_entropy": 0.69114, "router_noise": 0.0, "grad_norm": 1.7944, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.243, "alpha": 0.03, "t": 15.8} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5177.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.297477, "cps": 40.445312, "mean_p": 0.495828, "boundary_entropy": 0.687454, "bpb": 5.771576}, "de": {"chunks": 6965.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.33921, "cps": 54.414062, "mean_p": 0.496245, "boundary_entropy": 0.688307, "bpb": 5.907643}, "vi": {"chunks": 6196.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.3361, "cps": 48.40625, "mean_p": 0.488794, "boundary_entropy": 0.688242, "bpb": 6.209645}, "tr": {"chunks": 5950.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.328675, "cps": 46.484375, "mean_p": 0.49336, "boundary_entropy": 0.687733, "bpb": 6.209298}, "fi": {"chunks": 2794.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.14853, "cps": 21.828125, "mean_p": 0.488487, "boundary_entropy": 0.687747, "bpb": 5.898366}, "ru": {"chunks": 13513.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.713087, "cps": 105.570312, "mean_p": 0.49729, "boundary_entropy": 0.690172, "bpb": 5.428046}, "el": {"chunks": 10613.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.517127, "cps": 82.914062, "mean_p": 0.494819, "boundary_entropy": 0.690374, "bpb": 5.669029}, "hi": {"chunks": 22297.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.32523, "cps": 174.195312, "mean_p": 0.503087, "boundary_entropy": 0.690852, "bpb": 5.147269}, "ta": {"chunks": 12711.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.679914, "cps": 99.304688, "mean_p": 0.49345, "boundary_entropy": 0.691056, "bpb": 5.238127}, "th": {"chunks": 3952.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.245695, "cps": 30.875, "mean_p": 0.474476, "boundary_entropy": 0.689162, "bpb": 5.013473}, "zh": {"chunks": 6399.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.083474, "cps": 49.992188, "mean_p": 0.498855, "boundary_entropy": 0.68693, "bpb": 6.949892}, "my": {"chunks": 16845.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.889012, "cps": 131.601562, "mean_p": 0.498224, "boundary_entropy": 0.691338, "bpb": 5.104834}}, "bpb": {"en": 5.77158, "de": 5.90764, "vi": 6.20965, "tr": 6.2093, "fi": 5.89837, "ru": 5.42805, "el": 5.66903, "hi": 5.14727, "ta": 5.23813, "th": 5.01347, "zh": 6.94989, "my": 5.10483}, "gini_cpc": 0.335044, "gini_cps": 0.320766, "bpic": 3.383954, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1402883015050864, "vi": 1.1298329140296797, "tr": 1.1048730222722378, "fi": 0.499298743978344, "ru": 2.3971130579083555, "el": 1.7383742259858346, "hi": 4.454893397789572, "ta": 2.285599879277586, "th": 0.8259273084983082, "zh": 3.642206983518298, "my": 2.988502300390129}, "t": 17.7} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.31481, "bpb": 4.78226, "chunk_loss": 0.036876, "F": 0.26648, "G": 0.47724, "chunks": 272.88, "bytes_per_chunk": 3.753, "boundary_entropy": 0.6883, "router_noise": 0.0, "grad_norm": 1.8091, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.085, "alpha": 0.03, "t": 20.0} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4036.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.231914, "cps": 31.53125, "mean_p": 0.414389, "boundary_entropy": 0.651562, "bpb": 4.377855}, "de": {"chunks": 5375.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.261774, "cps": 41.992188, "mean_p": 0.414982, "boundary_entropy": 0.653758, "bpb": 4.597076}, "vi": {"chunks": 5116.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.277516, "cps": 39.96875, "mean_p": 0.462705, "boundary_entropy": 0.680616, "bpb": 4.80328}, "tr": {"chunks": 7225.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.399105, "cps": 56.445312, "mean_p": 0.447874, "boundary_entropy": 0.666784, "bpb": 4.866145}, "fi": {"chunks": 6254.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.332465, "cps": 48.859375, "mean_p": 0.438212, "boundary_entropy": 0.666008, "bpb": 4.636178}, "ru": {"chunks": 5803.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306227, "cps": 45.335938, "mean_p": 0.487002, "boundary_entropy": 0.688152, "bpb": 3.662839}, "el": {"chunks": 9568.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.466209, "cps": 74.75, "mean_p": 0.488364, "boundary_entropy": 0.689709, "bpb": 4.208704}, "hi": {"chunks": 14709.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.874235, "cps": 114.914062, "mean_p": 0.494285, "boundary_entropy": 0.689983, "bpb": 3.4439}, "ta": {"chunks": 13202.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.706178, "cps": 103.140625, "mean_p": 0.483988, "boundary_entropy": 0.689256, "bpb": 3.516294}, "th": {"chunks": 12986.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.807336, "cps": 101.453125, "mean_p": 0.481666, "boundary_entropy": 0.689018, "bpb": 3.280643}, "zh": {"chunks": 5457.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.923976, "cps": 42.632812, "mean_p": 0.495051, "boundary_entropy": 0.686612, "bpb": 6.269553}, "my": {"chunks": 45292.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.390331, "cps": 353.84375, "mean_p": 0.514319, "boundary_entropy": 0.690858, "bpb": 3.673614}}, "bpb": {"en": 4.37786, "de": 4.59708, "vi": 4.80328, "tr": 4.86615, "fi": 4.63618, "ru": 3.66284, "el": 4.2087, "hi": 3.4439, "ta": 3.51629, "th": 3.28064, "zh": 6.26955, "my": 3.67361}, "gini_cpc": 0.392503, "gini_cps": 0.39799, "bpic": 2.842338, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.128753276711805, "vi": 1.1966312947508226, "tr": 1.720918338828355, "fi": 1.433570171892445, "ru": 1.3204328458124899, "el": 2.010264873846245, "hi": 3.7696500735578895, "ta": 3.0449994711941284, "th": 3.481186542134159, "zh": 3.984129752310479, "my": 10.306971738206679}, "t": 22.7} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.61744, "bpb": 3.77617, "chunk_loss": 0.040042, "F": 0.31235, "G": 0.45901, "chunks": 319.84, "bytes_per_chunk": 3.202, "boundary_entropy": 0.67619, "router_noise": 0.0, "grad_norm": 1.9941, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.082, "alpha": 0.03, "t": 24.1} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3713.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.213354, "cps": 29.007812, "mean_p": 0.404996, "boundary_entropy": 0.643248, "bpb": 3.77829}, "de": {"chunks": 4096.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.199484, "cps": 32.0, "mean_p": 0.403003, "boundary_entropy": 0.646167, "bpb": 4.034409}, "vi": {"chunks": 5727.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.310659, "cps": 44.742188, "mean_p": 0.441363, "boundary_entropy": 0.670347, "bpb": 3.52239}, "tr": {"chunks": 5228.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.288792, "cps": 40.84375, "mean_p": 0.431378, "boundary_entropy": 0.660851, "bpb": 4.118573}, "fi": {"chunks": 4013.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.213333, "cps": 31.351562, "mean_p": 0.412786, "boundary_entropy": 0.654625, "bpb": 4.056047}, "ru": {"chunks": 7512.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.396412, "cps": 58.6875, "mean_p": 0.486227, "boundary_entropy": 0.688909, "bpb": 2.82077}, "el": {"chunks": 15115.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.736491, "cps": 118.085938, "mean_p": 0.493376, "boundary_entropy": 0.68956, "bpb": 2.975094}, "hi": {"chunks": 16453.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.97789, "cps": 128.539062, "mean_p": 0.476083, "boundary_entropy": 0.683105, "bpb": 2.414888}, "ta": {"chunks": 23538.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.259053, "cps": 183.890625, "mean_p": 0.506017, "boundary_entropy": 0.686341, "bpb": 2.246813}, "th": {"chunks": 12131.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.754181, "cps": 94.773438, "mean_p": 0.480337, "boundary_entropy": 0.689225, "bpb": 2.32904}, "zh": {"chunks": 4003.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.677785, "cps": 31.273438, "mean_p": 0.399791, "boundary_entropy": 0.615008, "bpb": 4.950198}, "my": {"chunks": 10773.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.568556, "cps": 84.164062, "mean_p": 0.481163, "boundary_entropy": 0.689201, "bpb": 2.436363}}, "bpb": {"en": 3.77829, "de": 4.03441, "vi": 3.52239, "tr": 4.11857, "fi": 4.05605, "ru": 2.82077, "el": 2.97509, "hi": 2.41489, "ta": 2.24681, "th": 2.32904, "zh": 4.9502, "my": 2.43636}, "gini_cpc": 0.326386, "gini_cps": 0.346116, "bpic": 3.417401, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9349894527119437, "vi": 1.4560732122989684, "tr": 1.353580840544082, "fi": 0.9998997211592492, "ru": 1.8579992566941939, "el": 3.4519657443182616, "hi": 4.583415148334294, "ta": 5.901239761960517, "th": 3.5348802705395963, "zh": 3.17681056545362, "my": 2.6648480758514625}, "t": 28.0} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.78203, "de": 4.03431, "vi": 3.5264, "tr": 4.12695, "fi": 4.05741, "ru": 2.80825, "el": 2.96869, "hi": 2.41079, "ta": 2.26378, "th": 2.36237, "zh": 4.9547, "my": 2.45346}, "bytes_per_chunk": {"en": 4.556, "de": 4.915, "vi": 4.262, "tr": 3.474, "fi": 4.549, "ru": 4.253, "el": 2.63, "hi": 2.699, "ta": 2.214, "th": 4.129, "zh": 4.541, "my": 5.518}, "bpb_mean": 3.31243, "t": 29.5} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.45356, "bpb": 3.53974, "chunk_loss": 0.035282, "F": 0.25247, "G": 0.43333, "chunks": 258.53, "bytes_per_chunk": 3.961, "boundary_entropy": 0.65917, "router_noise": 0.0, "grad_norm": 1.3468, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.108, "alpha": 0.03, "t": 29.6} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.38397, "bpb": 3.43935, "chunk_loss": 0.033082, "F": 0.21634, "G": 0.42891, "chunks": 221.53, "bytes_per_chunk": 4.622, "boundary_entropy": 0.65475, "router_noise": 0.0, "grad_norm": 0.9107, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.053, "alpha": 0.03, "t": 32.2} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4009.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.230363, "cps": 31.320312, "mean_p": 0.416313, "boundary_entropy": 0.649485, "bpb": 3.580175}, "de": {"chunks": 3667.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178591, "cps": 28.648438, "mean_p": 0.406683, "boundary_entropy": 0.650073, "bpb": 3.818075}, "vi": {"chunks": 6521.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.353729, "cps": 50.945312, "mean_p": 0.451899, "boundary_entropy": 0.673529, "bpb": 3.217886}, "tr": {"chunks": 5671.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.313263, "cps": 44.304688, "mean_p": 0.441631, "boundary_entropy": 0.66463, "bpb": 3.821494}, "fi": {"chunks": 5038.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.267822, "cps": 39.359375, "mean_p": 0.441146, "boundary_entropy": 0.667135, "bpb": 3.878048}, "ru": {"chunks": 7226.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.381319, "cps": 56.453125, "mean_p": 0.46031, "boundary_entropy": 0.683861, "bpb": 2.647827}, "el": {"chunks": 15599.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.760074, "cps": 121.867188, "mean_p": 0.497006, "boundary_entropy": 0.685713, "bpb": 2.732207}, "hi": {"chunks": 17942.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.066389, "cps": 140.171875, "mean_p": 0.468665, "boundary_entropy": 0.680375, "bpb": 2.097105}, "ta": {"chunks": 21045.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.125702, "cps": 164.414062, "mean_p": 0.49497, "boundary_entropy": 0.680207, "bpb": 1.918371}, "th": {"chunks": 4033.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.25073, "cps": 31.507812, "mean_p": 0.468257, "boundary_entropy": 0.686457, "bpb": 2.103462}, "zh": {"chunks": 9893.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.675076, "cps": 77.289062, "mean_p": 0.414725, "boundary_entropy": 0.604543, "bpb": 4.502987}, "my": {"chunks": 29252.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.543804, "cps": 228.53125, "mean_p": 0.496133, "boundary_entropy": 0.689279, "bpb": 2.210921}}, "bpb": {"en": 3.58018, "de": 3.81807, "vi": 3.21789, "tr": 3.82149, "fi": 3.87805, "ru": 2.64783, "el": 2.73221, "hi": 2.0971, "ta": 1.91837, "th": 2.10346, "zh": 4.50299, "my": 2.21092}, "gini_cpc": 0.406782, "gini_cps": 0.383831, "bpic": 2.954525, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.7752585538526237, "vi": 1.535532886643782, "tr": 1.359869270002848, "fi": 1.1626107750570625, "ru": 1.6553003499382326, "el": 3.299468426707938, "hi": 4.62917760447611, "ta": 4.886653264975949, "th": 1.0884167601448187, "zh": 7.271476926661034, "my": 6.701627007377795}, "t": 33.6} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.2798, "bpb": 3.28906, "chunk_loss": 0.034928, "F": 0.25269, "G": 0.43147, "chunks": 258.75, "bytes_per_chunk": 3.957, "boundary_entropy": 0.66375, "router_noise": 0.0, "grad_norm": 1.5405, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.077, "alpha": 0.03, "t": 36.1} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.04992, "bpb": 2.95741, "chunk_loss": 0.036056, "F": 0.27362, "G": 0.42839, "chunks": 280.19, "bytes_per_chunk": 3.655, "boundary_entropy": 0.65672, "router_noise": 0.0, "grad_norm": 1.0006, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.052, "alpha": 0.03, "t": 38.7} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4908.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.28202, "cps": 38.34375, "mean_p": 0.433006, "boundary_entropy": 0.656702, "bpb": 3.181321}, "de": {"chunks": 5248.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.255589, "cps": 41.0, "mean_p": 0.437449, "boundary_entropy": 0.664353, "bpb": 3.508162}, "vi": {"chunks": 7022.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.380906, "cps": 54.859375, "mean_p": 0.454163, "boundary_entropy": 0.675238, "bpb": 2.75746}, "tr": {"chunks": 5521.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.304977, "cps": 43.132812, "mean_p": 0.447689, "boundary_entropy": 0.66936, "bpb": 3.433751}, "fi": {"chunks": 4745.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.252246, "cps": 37.070312, "mean_p": 0.445424, "boundary_entropy": 0.672262, "bpb": 3.547089}, "ru": {"chunks": 8396.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.443061, "cps": 65.59375, "mean_p": 0.411233, "boundary_entropy": 0.632381, "bpb": 2.35259}, "el": {"chunks": 3450.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.168104, "cps": 26.953125, "mean_p": 0.440715, "boundary_entropy": 0.674852, "bpb": 2.50235}, "hi": {"chunks": 11161.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.663358, "cps": 87.195312, "mean_p": 0.434003, "boundary_entropy": 0.659257, "bpb": 1.807542}, "ta": {"chunks": 7616.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.407382, "cps": 59.5, "mean_p": 0.411712, "boundary_entropy": 0.642471, "bpb": 1.547733}, "th": {"chunks": 17032.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.058875, "cps": 133.0625, "mean_p": 0.458255, "boundary_entropy": 0.676032, "bpb": 1.908995}, "zh": {"chunks": 3683.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.623603, "cps": 28.773438, "mean_p": 0.412755, "boundary_entropy": 0.638775, "bpb": 4.134757}, "my": {"chunks": 2660.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.140384, "cps": 20.78125, "mean_p": 0.449524, "boundary_entropy": 0.678214, "bpb": 1.93299}}, "bpb": {"en": 3.18132, "de": 3.50816, "vi": 2.75746, "tr": 3.43375, "fi": 3.54709, "ru": 2.35259, "el": 2.50235, "hi": 1.80754, "ta": 1.54773, "th": 1.90899, "zh": 4.13476, "my": 1.93299}, "gini_cpc": 0.310369, "gini_cps": 0.287988, "bpic": 4.712323, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9062770563028303, "vi": 1.3506326662833732, "tr": 1.0814009875843567, "fi": 0.8944249378614889, "ru": 1.5710238628226385, "el": 0.596070757038409, "hi": 2.3521640139941944, "ta": 1.4445115942379636, "th": 3.7546040936254754, "zh": 2.2111990665495367, "my": 0.49778043937519384}, "t": 42.0} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.19198, "de": 3.50728, "vi": 2.76233, "tr": 3.4438, "fi": 3.53901, "ru": 2.33676, "el": 2.49648, "hi": 1.8011, "ta": 1.56535, "th": 1.9403, "zh": 4.14188, "my": 1.94826}, "bytes_per_chunk": {"en": 3.454, "de": 3.81, "vi": 3.481, "tr": 3.361, "fi": 3.758, "ru": 3.99, "el": 10.452, "hi": 4.216, "ta": 6.751, "th": 2.673, "zh": 3.997, "my": 18.99}, "bpb_mean": 2.72288, "t": 43.4} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.19891, "bpb": 3.17236, "chunk_loss": 0.036194, "F": 0.27484, "G": 0.42756, "chunks": 281.44, "bytes_per_chunk": 3.638, "boundary_entropy": 0.65523, "router_noise": 0.0, "grad_norm": 0.8667, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.095, "alpha": 0.03, "t": 43.4} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.16124, "bpb": 3.11801, "chunk_loss": 0.034177, "F": 0.23923, "G": 0.42795, "chunks": 244.97, "bytes_per_chunk": 4.18, "boundary_entropy": 0.66037, "router_noise": 0.0, "grad_norm": 0.868, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 45.8} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.06545, "bpb": 2.97982, "chunk_loss": 0.033314, "F": 0.22516, "G": 0.42798, "chunks": 230.56, "bytes_per_chunk": 4.441, "boundary_entropy": 0.66276, "router_noise": 0.0, "grad_norm": 1.2817, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.048, "alpha": 0.03, "t": 48.2} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.92418, "bpb": 2.77601, "chunk_loss": 0.032979, "F": 0.21661, "G": 0.43149, "chunks": 221.81, "bytes_per_chunk": 4.617, "boundary_entropy": 0.66202, "router_noise": 0.0, "grad_norm": 0.718, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 50.5} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4833.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.277711, "cps": 37.757812, "mean_p": 0.443862, "boundary_entropy": 0.663228, "bpb": 2.889279}, "de": {"chunks": 4869.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.23713, "cps": 38.039062, "mean_p": 0.442375, "boundary_entropy": 0.668774, "bpb": 3.205895}, "vi": {"chunks": 5062.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.274586, "cps": 39.546875, "mean_p": 0.450311, "boundary_entropy": 0.674874, "bpb": 2.572995}, "tr": {"chunks": 5612.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.310004, "cps": 43.84375, "mean_p": 0.453004, "boundary_entropy": 0.670507, "bpb": 3.111351}, "fi": {"chunks": 3927.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.208761, "cps": 30.679688, "mean_p": 0.439253, "boundary_entropy": 0.67053, "bpb": 3.352982}, "ru": {"chunks": 6060.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319789, "cps": 47.34375, "mean_p": 0.423163, "boundary_entropy": 0.651834, "bpb": 2.149439}, "el": {"chunks": 3913.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.190664, "cps": 30.570312, "mean_p": 0.437697, "boundary_entropy": 0.673015, "bpb": 2.246887}, "hi": {"chunks": 7426.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.441367, "cps": 58.015625, "mean_p": 0.423879, "boundary_entropy": 0.657151, "bpb": 1.6571}, "ta": {"chunks": 13715.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.733619, "cps": 107.148438, "mean_p": 0.434022, "boundary_entropy": 0.653744, "bpb": 1.388821}, "th": {"chunks": 2746.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170718, "cps": 21.453125, "mean_p": 0.432237, "boundary_entropy": 0.67172, "bpb": 1.768305}, "zh": {"chunks": 3198.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.541483, "cps": 24.984375, "mean_p": 0.407128, "boundary_entropy": 0.636656, "bpb": 3.967482}, "my": {"chunks": 6645.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.350697, "cps": 51.914062, "mean_p": 0.471629, "boundary_entropy": 0.686024, "bpb": 1.664026}}, "bpb": {"en": 2.88928, "de": 3.2059, "vi": 2.573, "tr": 3.11135, "fi": 3.35298, "ru": 2.14944, "el": 2.24689, "hi": 1.6571, "ta": 1.38882, "th": 1.76831, "zh": 3.96748, "my": 1.66403}, "gini_cpc": 0.240626, "gini_cps": 0.234085, "bpic": 5.64334, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.85387577484616, "vi": 0.9887496070967213, "tr": 1.116283321594519, "fi": 0.7517204116177914, "ru": 1.151518010249456, "el": 0.6865565705104504, "hi": 1.5893048073100704, "ta": 2.6416645458169126, "th": 0.6147333754358587, "zh": 1.9498102173026077, "my": 1.262812680711181}, "t": 53.8} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.89929, "de": 3.20539, "vi": 2.57798, "tr": 3.12387, "fi": 3.34498, "ru": 2.13359, "el": 2.23573, "hi": 1.65087, "ta": 1.40253, "th": 1.79572, "zh": 3.97818, "my": 1.6772}, "bytes_per_chunk": {"en": 3.618, "de": 4.494, "vi": 5.308, "tr": 3.512, "fi": 4.955, "ru": 5.468, "el": 10.525, "hi": 8.2, "ta": 3.383, "th": 13.766, "zh": 4.302, "my": 7.376}, "bpb_mean": 2.50211, "t": 55.1} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.91463, "bpb": 2.76222, "chunk_loss": 0.03267, "F": 0.21243, "G": 0.43334, "chunks": 217.53, "bytes_per_chunk": 4.707, "boundary_entropy": 0.66474, "router_noise": 0.0, "grad_norm": 1.0423, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 55.2} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.94596, "bpb": 2.80742, "chunk_loss": 0.034185, "F": 0.23929, "G": 0.43519, "chunks": 245.03, "bytes_per_chunk": 4.179, "boundary_entropy": 0.6662, "router_noise": 0.0, "grad_norm": 0.798, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 57.5} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.7802, "bpb": 2.56828, "chunk_loss": 0.035255, "F": 0.25626, "G": 0.43691, "chunks": 262.41, "bytes_per_chunk": 3.902, "boundary_entropy": 0.66515, "router_noise": 0.0, "grad_norm": 0.9498, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 59.8} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.86793, "bpb": 2.69486, "chunk_loss": 0.033593, "F": 0.22714, "G": 0.43564, "chunks": 232.59, "bytes_per_chunk": 4.403, "boundary_entropy": 0.66648, "router_noise": 0.0, "grad_norm": 0.7469, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 62.1} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4593.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.26392, "cps": 35.882812, "mean_p": 0.443795, "boundary_entropy": 0.664237, "bpb": 2.683231}, "de": {"chunks": 4819.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.234695, "cps": 37.648438, "mean_p": 0.446001, "boundary_entropy": 0.670945, "bpb": 3.024648}, "vi": {"chunks": 4558.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.247247, "cps": 35.609375, "mean_p": 0.444911, "boundary_entropy": 0.673443, "bpb": 2.39719}, "tr": {"chunks": 4852.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.268022, "cps": 37.90625, "mean_p": 0.449383, "boundary_entropy": 0.67177, "bpb": 3.005853}, "fi": {"chunks": 3427.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.182181, "cps": 26.773438, "mean_p": 0.441428, "boundary_entropy": 0.672531, "bpb": 3.22877}, "ru": {"chunks": 5297.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.279525, "cps": 41.382812, "mean_p": 0.42687, "boundary_entropy": 0.655704, "bpb": 2.010338}, "el": {"chunks": 7718.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.376066, "cps": 60.296875, "mean_p": 0.45565, "boundary_entropy": 0.678082, "bpb": 2.127384}, "hi": {"chunks": 6115.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.363447, "cps": 47.773438, "mean_p": 0.435399, "boundary_entropy": 0.663066, "bpb": 1.570924}, "ta": {"chunks": 10865.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.581171, "cps": 84.882812, "mean_p": 0.429283, "boundary_entropy": 0.653268, "bpb": 1.352395}, "th": {"chunks": 1888.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.117376, "cps": 14.75, "mean_p": 0.41981, "boundary_entropy": 0.668003, "bpb": 1.636759}, "zh": {"chunks": 2842.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.481206, "cps": 22.203125, "mean_p": 0.431712, "boundary_entropy": 0.664894, "bpb": 3.828787}, "my": {"chunks": 20197.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.065917, "cps": 157.789062, "mean_p": 0.466983, "boundary_entropy": 0.681873, "bpb": 1.480508}}, "bpb": {"en": 2.68323, "de": 3.02465, "vi": 2.39719, "tr": 3.00585, "fi": 3.22877, "ru": 2.01034, "el": 2.12738, "hi": 1.57092, "ta": 1.3524, "th": 1.63676, "zh": 3.82879, "my": 1.48051}, "gini_cpc": 0.312065, "gini_cps": 0.338613, "bpic": 4.973125, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8892670361005908, "vi": 0.936825823851593, "tr": 1.0155420613152268, "fi": 0.6902873199009669, "ru": 1.0591279605824397, "el": 1.4249236801214538, "hi": 1.3771113675471656, "ta": 2.202074135513378, "th": 0.44474246567721654, "zh": 1.823300729499808, "my": 4.038788994790239}, "t": 65.3} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.69245, "de": 3.02793, "vi": 2.40259, "tr": 3.01706, "fi": 3.22555, "ru": 1.99537, "el": 2.11383, "hi": 1.56602, "ta": 1.36537, "th": 1.6637, "zh": 3.83907, "my": 1.49065}, "bytes_per_chunk": {"en": 3.81, "de": 4.194, "vi": 5.695, "tr": 3.998, "fi": 5.773, "ru": 6.528, "el": 4.979, "hi": 9.38, "ta": 4.999, "th": 21.069, "zh": 5.266, "my": 2.697}, "bpb_mean": 2.36663, "t": 66.6} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.85132, "bpb": 2.67089, "chunk_loss": 0.032643, "F": 0.20972, "G": 0.43591, "chunks": 214.75, "bytes_per_chunk": 4.768, "boundary_entropy": 0.66823, "router_noise": 0.0, "grad_norm": 0.8404, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 66.7} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.64696, "bpb": 2.37606, "chunk_loss": 0.033479, "F": 0.22321, "G": 0.4402, "chunks": 228.56, "bytes_per_chunk": 4.48, "boundary_entropy": 0.66893, "router_noise": 0.0, "grad_norm": 0.73, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 68.9} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.73248, "bpb": 2.49944, "chunk_loss": 0.033933, "F": 0.23276, "G": 0.44009, "chunks": 238.34, "bytes_per_chunk": 4.296, "boundary_entropy": 0.66994, "router_noise": 0.0, "grad_norm": 0.8231, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.046, "alpha": 0.03, "t": 71.2} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.88816, "bpb": 2.72404, "chunk_loss": 0.033608, "F": 0.2276, "G": 0.4378, "chunks": 233.06, "bytes_per_chunk": 4.394, "boundary_entropy": 0.66884, "router_noise": 0.0, "grad_norm": 0.8259, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 73.5} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5020.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.288456, "cps": 39.21875, "mean_p": 0.444522, "boundary_entropy": 0.662912, "bpb": 2.553715}, "de": {"chunks": 4932.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.240199, "cps": 38.53125, "mean_p": 0.445545, "boundary_entropy": 0.670517, "bpb": 2.889454}, "vi": {"chunks": 7812.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.423759, "cps": 61.03125, "mean_p": 0.460069, "boundary_entropy": 0.675381, "bpb": 2.268263}, "tr": {"chunks": 5743.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.31724, "cps": 44.867188, "mean_p": 0.452155, "boundary_entropy": 0.669313, "bpb": 2.872238}, "fi": {"chunks": 4266.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.226782, "cps": 33.328125, "mean_p": 0.446217, "boundary_entropy": 0.672673, "bpb": 3.124532}, "ru": {"chunks": 8301.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.438047, "cps": 64.851562, "mean_p": 0.435514, "boundary_entropy": 0.65678, "bpb": 1.917575}, "el": {"chunks": 11917.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.580666, "cps": 93.101562, "mean_p": 0.464824, "boundary_entropy": 0.679874, "bpb": 2.02048}, "hi": {"chunks": 14892.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.885111, "cps": 116.34375, "mean_p": 0.439486, "boundary_entropy": 0.66203, "bpb": 1.524343}, "ta": {"chunks": 8630.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.461621, "cps": 67.421875, "mean_p": 0.423412, "boundary_entropy": 0.64946, "bpb": 1.426152}, "th": {"chunks": 3721.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.231334, "cps": 29.070312, "mean_p": 0.428591, "boundary_entropy": 0.670406, "bpb": 1.587326}, "zh": {"chunks": 2535.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.429225, "cps": 19.804688, "mean_p": 0.435668, "boundary_entropy": 0.668489, "bpb": 3.701011}, "my": {"chunks": 6385.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.336975, "cps": 49.882812, "mean_p": 0.464116, "boundary_entropy": 0.684114, "bpb": 1.409843}}, "bpb": {"en": 2.55371, "de": 2.88945, "vi": 2.26826, "tr": 2.87224, "fi": 3.12453, "ru": 1.91757, "el": 2.02048, "hi": 1.52434, "ta": 1.42615, "th": 1.58733, "zh": 3.70101, "my": 1.40984}, "gini_cpc": 0.224757, "gini_cps": 0.263507, "bpic": 4.560461, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8327047918005085, "vi": 1.469059871174375, "tr": 1.0997872180284, "fi": 0.7861933585100351, "ru": 1.5185937306184234, "el": 2.013012618434317, "hi": 3.0684451022063306, "ta": 1.6003159333780153, "th": 0.8019716347541138, "zh": 1.4880068280889311, "my": 1.1682019546891598}, "t": 76.7} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.56321, "de": 2.89242, "vi": 2.2739, "tr": 2.88334, "fi": 3.12126, "ru": 1.89917, "el": 2.01018, "hi": 1.51939, "ta": 1.43427, "th": 1.6133, "zh": 3.71182, "my": 1.41913}, "bytes_per_chunk": {"en": 3.457, "de": 4.025, "vi": 3.081, "tr": 3.257, "fi": 4.356, "ru": 4.128, "el": 2.933, "hi": 3.013, "ta": 6.513, "th": 11.848, "zh": 6.666, "my": 8.79}, "bpb_mean": 2.27845, "t": 78.1} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.56527, "bpb": 2.2582, "chunk_loss": 0.034638, "F": 0.24313, "G": 0.4386, "chunks": 248.97, "bytes_per_chunk": 4.113, "boundary_entropy": 0.66803, "router_noise": 0.0, "grad_norm": 0.7829, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.093, "alpha": 0.03, "t": 78.2} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.77616, "bpb": 2.56246, "chunk_loss": 0.034832, "F": 0.24643, "G": 0.44339, "chunks": 252.34, "bytes_per_chunk": 4.058, "boundary_entropy": 0.67151, "router_noise": 0.0, "grad_norm": 0.9656, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.045, "alpha": 0.03, "t": 80.4} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.60053, "bpb": 2.30908, "chunk_loss": 0.03358, "F": 0.22632, "G": 0.43953, "chunks": 231.75, "bytes_per_chunk": 4.419, "boundary_entropy": 0.66943, "router_noise": 0.0, "grad_norm": 0.8141, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 82.6} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.62029, "bpb": 2.33758, "chunk_loss": 0.032003, "F": 0.19992, "G": 0.43849, "chunks": 204.72, "bytes_per_chunk": 5.002, "boundary_entropy": 0.67014, "router_noise": 0.0, "grad_norm": 0.85, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 84.8} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4111.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.236224, "cps": 32.117188, "mean_p": 0.442584, "boundary_entropy": 0.664246, "bpb": 2.479104}, "de": {"chunks": 4573.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.222715, "cps": 35.726562, "mean_p": 0.450257, "boundary_entropy": 0.673031, "bpb": 2.806344}, "vi": {"chunks": 5192.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.281638, "cps": 40.5625, "mean_p": 0.45074, "boundary_entropy": 0.674875, "bpb": 2.233787}, "tr": {"chunks": 5075.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.28034, "cps": 39.648438, "mean_p": 0.456529, "boundary_entropy": 0.672798, "bpb": 2.789823}, "fi": {"chunks": 3540.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188188, "cps": 27.65625, "mean_p": 0.447523, "boundary_entropy": 0.67477, "bpb": 3.048851}, "ru": {"chunks": 6713.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.354248, "cps": 52.445312, "mean_p": 0.430318, "boundary_entropy": 0.658971, "bpb": 1.836656}, "el": {"chunks": 7692.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.374799, "cps": 60.09375, "mean_p": 0.447164, "boundary_entropy": 0.676533, "bpb": 1.947232}, "hi": {"chunks": 10485.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.62318, "cps": 81.914062, "mean_p": 0.442425, "boundary_entropy": 0.665041, "bpb": 1.476939}, "ta": {"chunks": 6445.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.344745, "cps": 50.351562, "mean_p": 0.418731, "boundary_entropy": 0.652964, "bpb": 1.217323}, "th": {"chunks": 4773.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.296736, "cps": 37.289062, "mean_p": 0.421237, "boundary_entropy": 0.664566, "bpb": 1.471741}, "zh": {"chunks": 3386.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.573315, "cps": 26.453125, "mean_p": 0.448141, "boundary_entropy": 0.673641, "bpb": 3.645204}, "my": {"chunks": 4540.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.239603, "cps": 35.46875, "mean_p": 0.46, "boundary_entropy": 0.684444, "bpb": 1.370316}}, "bpb": {"en": 2.4791, "de": 2.80634, "vi": 2.23379, "tr": 2.78982, "fi": 3.04885, "ru": 1.83666, "el": 1.94723, "hi": 1.47694, "ta": 1.21732, "th": 1.47174, "zh": 3.6452, "my": 1.37032}, "gini_cpc": 0.201123, "gini_cps": 0.182081, "bpic": 5.768974, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9428127296386142, "vi": 1.1922523448668845, "tr": 1.1867579195901805, "fi": 0.7966508466137203, "ru": 1.4996298495124414, "el": 1.5866278524205737, "hi": 2.6380924154128578, "ta": 1.4593991723000457, "th": 1.256165936753384, "zh": 2.427002113629214, "my": 1.0143062935768028}, "t": 88.0} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.48704, "de": 2.80733, "vi": 2.23861, "tr": 2.8008, "fi": 3.05141, "ru": 1.81784, "el": 1.93669, "hi": 1.47356, "ta": 1.22838, "th": 1.49547, "zh": 3.65179, "my": 1.37906}, "bytes_per_chunk": {"en": 4.384, "de": 4.434, "vi": 4.655, "tr": 3.796, "fi": 5.768, "ru": 5.04, "el": 4.59, "hi": 4.516, "ta": 7.782, "th": 9.388, "zh": 5.012, "my": 11.348}, "bpb_mean": 2.19733, "t": 89.3} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.70515, "bpb": 2.46002, "chunk_loss": 0.033076, "F": 0.21838, "G": 0.4403, "chunks": 223.62, "bytes_per_chunk": 4.579, "boundary_entropy": 0.67077, "router_noise": 0.0, "grad_norm": 0.8167, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 89.4} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.60795, "bpb": 2.31978, "chunk_loss": 0.033637, "F": 0.22717, "G": 0.44098, "chunks": 232.62, "bytes_per_chunk": 4.402, "boundary_entropy": 0.67135, "router_noise": 0.0, "grad_norm": 0.8592, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 91.6} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.7207, "bpb": 2.48245, "chunk_loss": 0.034328, "F": 0.23932, "G": 0.44165, "chunks": 245.06, "bytes_per_chunk": 4.179, "boundary_entropy": 0.67064, "router_noise": 0.0, "grad_norm": 0.8223, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 93.8} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.51035, "bpb": 2.17898, "chunk_loss": 0.033673, "F": 0.22864, "G": 0.43982, "chunks": 234.12, "bytes_per_chunk": 4.374, "boundary_entropy": 0.66954, "router_noise": 0.0, "grad_norm": 0.7734, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 96.0} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4795.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.275527, "cps": 37.460938, "mean_p": 0.446464, "boundary_entropy": 0.664413, "bpb": 2.409148}, "de": {"chunks": 5005.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.243754, "cps": 39.101562, "mean_p": 0.452245, "boundary_entropy": 0.672499, "bpb": 2.740646}, "vi": {"chunks": 4933.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.267589, "cps": 38.539062, "mean_p": 0.448117, "boundary_entropy": 0.674077, "bpb": 2.15999}, "tr": {"chunks": 5418.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.299287, "cps": 42.328125, "mean_p": 0.456935, "boundary_entropy": 0.672273, "bpb": 2.691935}, "fi": {"chunks": 3691.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.196215, "cps": 28.835938, "mean_p": 0.446159, "boundary_entropy": 0.673441, "bpb": 2.97946}, "ru": {"chunks": 5331.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.281319, "cps": 41.648438, "mean_p": 0.429686, "boundary_entropy": 0.663311, "bpb": 1.804787}, "el": {"chunks": 8922.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.434732, "cps": 69.703125, "mean_p": 0.451888, "boundary_entropy": 0.677354, "bpb": 1.856197}, "hi": {"chunks": 10638.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.632273, "cps": 83.109375, "mean_p": 0.446352, "boundary_entropy": 0.668711, "bpb": 1.430941}, "ta": {"chunks": 7843.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.419524, "cps": 61.273438, "mean_p": 0.42394, "boundary_entropy": 0.652221, "bpb": 1.163059}, "th": {"chunks": 10523.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.654212, "cps": 82.210938, "mean_p": 0.435204, "boundary_entropy": 0.670813, "bpb": 1.398333}, "zh": {"chunks": 3078.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.521165, "cps": 24.046875, "mean_p": 0.450153, "boundary_entropy": 0.67524, "bpb": 3.579008}, "my": {"chunks": 5931.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.313015, "cps": 46.335938, "mean_p": 0.4599, "boundary_entropy": 0.684262, "bpb": 1.348221}}, "bpb": {"en": 2.40915, "de": 2.74065, "vi": 2.15999, "tr": 2.69194, "fi": 2.97946, "ru": 1.80479, "el": 1.8562, "hi": 1.43094, "ta": 1.16306, "th": 1.39833, "zh": 3.57901, "my": 1.34822}, "gini_cpc": 0.211776, "gini_cps": 0.20908, "bpic": 5.042584, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8846819842439854, "vi": 0.9711883904019222, "tr": 1.0862354144633042, "fi": 0.7121437554369521, "ru": 1.0210217107199946, "el": 1.5778179134912955, "hi": 2.2947766479136424, "ta": 1.5226225388006418, "th": 2.3744007120669877, "zh": 1.8915188844910196, "my": 1.1360568290407003}, "t": 99.1} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.41617, "de": 2.74464, "vi": 2.16509, "tr": 2.7038, "fi": 2.9803, "ru": 1.78624, "el": 1.84626, "hi": 1.42649, "ta": 1.17066, "th": 1.41911, "zh": 3.58423, "my": 1.35826}, "bytes_per_chunk": {"en": 3.623, "de": 4.078, "vi": 5.041, "tr": 3.613, "fi": 5.705, "ru": 6.474, "el": 4.126, "hi": 4.363, "ta": 7.102, "th": 4.173, "zh": 5.626, "my": 9.1}, "bpb_mean": 2.13344, "t": 100.5} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.7086, "bpb": 2.46498, "chunk_loss": 0.033309, "F": 0.22168, "G": 0.44076, "chunks": 227.0, "bytes_per_chunk": 4.511, "boundary_entropy": 0.67187, "router_noise": 0.0, "grad_norm": 0.8173, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 100.5} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.71046, "bpb": 2.46767, "chunk_loss": 0.032195, "F": 0.20328, "G": 0.4408, "chunks": 208.16, "bytes_per_chunk": 4.919, "boundary_entropy": 0.67281, "router_noise": 0.0, "grad_norm": 0.7518, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 102.7} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.6758, "bpb": 2.41766, "chunk_loss": 0.035575, "F": 0.25891, "G": 0.44587, "chunks": 265.12, "bytes_per_chunk": 3.862, "boundary_entropy": 0.67109, "router_noise": 0.0, "grad_norm": 0.931, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 104.9} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.56813, "bpb": 2.26233, "chunk_loss": 0.035098, "F": 0.25128, "G": 0.4448, "chunks": 257.31, "bytes_per_chunk": 3.98, "boundary_entropy": 0.67072, "router_noise": 0.0, "grad_norm": 0.8549, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 107.1} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5277.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.303224, "cps": 41.226562, "mean_p": 0.448122, "boundary_entropy": 0.662525, "bpb": 2.345843}, "de": {"chunks": 5140.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.250329, "cps": 40.15625, "mean_p": 0.451386, "boundary_entropy": 0.671895, "bpb": 2.671138}, "vi": {"chunks": 6036.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.327421, "cps": 47.15625, "mean_p": 0.452277, "boundary_entropy": 0.674437, "bpb": 2.120237}, "tr": {"chunks": 5446.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.300834, "cps": 42.546875, "mean_p": 0.45535, "boundary_entropy": 0.670987, "bpb": 2.642573}, "fi": {"chunks": 4866.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.258678, "cps": 38.015625, "mean_p": 0.454996, "boundary_entropy": 0.674848, "bpb": 2.89988}, "ru": {"chunks": 7273.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.383799, "cps": 56.820312, "mean_p": 0.436948, "boundary_entropy": 0.663277, "bpb": 1.742867}, "el": {"chunks": 9679.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.471617, "cps": 75.617188, "mean_p": 0.457575, "boundary_entropy": 0.678257, "bpb": 1.806615}, "hi": {"chunks": 9414.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.559525, "cps": 73.546875, "mean_p": 0.45272, "boundary_entropy": 0.668837, "bpb": 1.401148}, "ta": {"chunks": 9985.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.5341, "cps": 78.007812, "mean_p": 0.442704, "boundary_entropy": 0.658616, "bpb": 1.135695}, "th": {"chunks": 6684.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.415542, "cps": 52.21875, "mean_p": 0.430634, "boundary_entropy": 0.669128, "bpb": 1.340991}, "zh": {"chunks": 3672.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.621741, "cps": 28.6875, "mean_p": 0.462145, "boundary_entropy": 0.678831, "bpb": 3.517681}, "my": {"chunks": 3485.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.183924, "cps": 27.226562, "mean_p": 0.451166, "boundary_entropy": 0.682049, "bpb": 1.297439}}, "bpb": {"en": 2.34584, "de": 2.67114, "vi": 2.12024, "tr": 2.64257, "fi": 2.89988, "ru": 1.74287, "el": 1.80662, "hi": 1.40115, "ta": 1.1357, "th": 1.34099, "zh": 3.51768, "my": 1.29744}, "gini_cpc": 0.194726, "gini_cps": 0.187199, "bpic": 4.986954, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8255582805775818, "vi": 1.0797994829335218, "tr": 0.992119787545068, "fi": 0.853094706570542, "ru": 1.2657309487130641, "el": 1.5553447612139815, "hi": 1.8452539645398276, "ta": 1.7614066260082064, "th": 1.3704159416035793, "zh": 2.0504361777762545, "my": 0.6065637225269169}, "t": 110.2} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.35823, "de": 2.67759, "vi": 2.12492, "tr": 2.65034, "fi": 2.90446, "ru": 1.72532, "el": 1.79679, "hi": 1.39662, "ta": 1.14402, "th": 1.35954, "zh": 3.52801, "my": 1.30708}, "bytes_per_chunk": {"en": 3.336, "de": 3.968, "vi": 4.078, "tr": 3.618, "fi": 3.844, "ru": 4.687, "el": 3.735, "hi": 4.929, "ta": 5.01, "th": 6.244, "zh": 4.428, "my": 13.352}, "bpb_mean": 2.08108, "t": 111.6} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.79312, "bpb": 2.58692, "chunk_loss": 0.035591, "F": 0.25833, "G": 0.44889, "chunks": 264.53, "bytes_per_chunk": 3.871, "boundary_entropy": 0.67378, "router_noise": 0.0, "grad_norm": 0.8756, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 111.7} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.57244, "bpb": 2.26855, "chunk_loss": 0.033116, "F": 0.21771, "G": 0.44091, "chunks": 222.94, "bytes_per_chunk": 4.593, "boundary_entropy": 0.67112, "router_noise": 0.0, "grad_norm": 0.8237, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 113.8} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.55543, "bpb": 2.24402, "chunk_loss": 0.032955, "F": 0.21622, "G": 0.44033, "chunks": 221.41, "bytes_per_chunk": 4.625, "boundary_entropy": 0.67058, "router_noise": 0.0, "grad_norm": 0.7413, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 116.0} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.51826, "bpb": 2.19039, "chunk_loss": 0.033343, "F": 0.22293, "G": 0.43969, "chunks": 228.28, "bytes_per_chunk": 4.486, "boundary_entropy": 0.66985, "router_noise": 0.0, "grad_norm": 0.801, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 118.2} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4310.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.247658, "cps": 33.671875, "mean_p": 0.440954, "boundary_entropy": 0.661758, "bpb": 2.314975}, "de": {"chunks": 4776.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.232601, "cps": 37.3125, "mean_p": 0.445797, "boundary_entropy": 0.670317, "bpb": 2.620434}, "vi": {"chunks": 6429.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.348739, "cps": 50.226562, "mean_p": 0.448476, "boundary_entropy": 0.672175, "bpb": 2.093713}, "tr": {"chunks": 5140.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.283931, "cps": 40.15625, "mean_p": 0.454069, "boundary_entropy": 0.671264, "bpb": 2.59728}, "fi": {"chunks": 3767.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.200255, "cps": 29.429688, "mean_p": 0.444772, "boundary_entropy": 0.672281, "bpb": 2.841635}, "ru": {"chunks": 6687.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.352876, "cps": 52.242188, "mean_p": 0.432937, "boundary_entropy": 0.66296, "bpb": 1.695728}, "el": {"chunks": 7883.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.384106, "cps": 61.585938, "mean_p": 0.449319, "boundary_entropy": 0.676509, "bpb": 1.770333}, "hi": {"chunks": 6791.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.403626, "cps": 53.054688, "mean_p": 0.440307, "boundary_entropy": 0.666919, "bpb": 1.381121}, "ta": {"chunks": 14600.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.780957, "cps": 114.0625, "mean_p": 0.440432, "boundary_entropy": 0.659957, "bpb": 1.123431}, "th": {"chunks": 12112.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.753, "cps": 94.625, "mean_p": 0.432783, "boundary_entropy": 0.668967, "bpb": 1.304682}, "zh": {"chunks": 3186.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.539451, "cps": 24.890625, "mean_p": 0.459375, "boundary_entropy": 0.678388, "bpb": 3.504794}, "my": {"chunks": 20076.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.059531, "cps": 156.84375, "mean_p": 0.465264, "boundary_entropy": 0.681947, "bpb": 1.279334}}, "bpb": {"en": 2.31498, "de": 2.62043, "vi": 2.09371, "tr": 2.59728, "fi": 2.84164, "ru": 1.69573, "el": 1.77033, "hi": 1.38112, "ta": 1.12343, "th": 1.30468, "zh": 3.50479, "my": 1.27933}, "gini_cpc": 0.289348, "gini_cps": 0.312116, "bpic": 4.007864, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9392014642718196, "vi": 1.408144210202398, "tr": 1.1464613484606583, "fi": 0.8085941347690995, "ru": 1.424849383834612, "el": 1.550949051220107, "hi": 1.6297669540754938, "ta": 3.1533649517315756, "th": 3.0404764710676386, "zh": 2.178207147868816, "my": 4.278195838947673}, "t": 121.3} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.32617, "de": 2.6253, "vi": 2.09909, "tr": 2.60084, "fi": 2.84774, "ru": 1.67682, "el": 1.76109, "hi": 1.378, "ta": 1.13011, "th": 1.32329, "zh": 3.51098, "my": 1.28854}, "bytes_per_chunk": {"en": 4.083, "de": 4.235, "vi": 3.759, "tr": 3.799, "fi": 5.232, "ru": 5.062, "el": 4.516, "hi": 7.005, "ta": 3.399, "th": 3.605, "zh": 4.951, "my": 2.691}, "bpb_mean": 2.04733, "t": 122.7} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.45785, "bpb": 2.10323, "chunk_loss": 0.034807, "F": 0.24625, "G": 0.44305, "chunks": 252.16, "bytes_per_chunk": 4.061, "boundary_entropy": 0.67146, "router_noise": 0.0, "grad_norm": 0.6935, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 122.8} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.43855, "bpb": 2.07539, "chunk_loss": 0.03414, "F": 0.23572, "G": 0.44289, "chunks": 241.38, "bytes_per_chunk": 4.242, "boundary_entropy": 0.67109, "router_noise": 0.0, "grad_norm": 0.7549, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 124.9} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.54903, "bpb": 2.23478, "chunk_loss": 0.034599, "F": 0.24396, "G": 0.44112, "chunks": 249.81, "bytes_per_chunk": 4.099, "boundary_entropy": 0.66895, "router_noise": 0.0, "grad_norm": 0.7199, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 127.1} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.61695, "bpb": 2.33277, "chunk_loss": 0.033469, "F": 0.2244, "G": 0.44161, "chunks": 229.78, "bytes_per_chunk": 4.456, "boundary_entropy": 0.67055, "router_noise": 0.0, "grad_norm": 0.7796, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 129.3} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4726.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.271562, "cps": 36.921875, "mean_p": 0.440692, "boundary_entropy": 0.659826, "bpb": 2.271839}, "de": {"chunks": 4764.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.232017, "cps": 37.21875, "mean_p": 0.443365, "boundary_entropy": 0.668909, "bpb": 2.589004}, "vi": {"chunks": 6049.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.328126, "cps": 47.257812, "mean_p": 0.446056, "boundary_entropy": 0.671272, "bpb": 2.079886}, "tr": {"chunks": 5778.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.319174, "cps": 45.140625, "mean_p": 0.455776, "boundary_entropy": 0.670586, "bpb": 2.531904}, "fi": {"chunks": 4116.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.218808, "cps": 32.15625, "mean_p": 0.446451, "boundary_entropy": 0.672192, "bpb": 2.815654}, "ru": {"chunks": 6174.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.325805, "cps": 48.234375, "mean_p": 0.433916, "boundary_entropy": 0.664333, "bpb": 1.673766}, "el": {"chunks": 7824.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.381231, "cps": 61.125, "mean_p": 0.446443, "boundary_entropy": 0.675856, "bpb": 1.748338}, "hi": {"chunks": 11788.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.700624, "cps": 92.09375, "mean_p": 0.446656, "boundary_entropy": 0.66776, "bpb": 1.350393}, "ta": {"chunks": 11116.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.594597, "cps": 86.84375, "mean_p": 0.434788, "boundary_entropy": 0.658087, "bpb": 1.102294}, "th": {"chunks": 11115.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.691016, "cps": 86.835938, "mean_p": 0.437887, "boundary_entropy": 0.671753, "bpb": 1.291769}, "zh": {"chunks": 3219.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545039, "cps": 25.148438, "mean_p": 0.459188, "boundary_entropy": 0.678069, "bpb": 3.467648}, "my": {"chunks": 6368.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.336078, "cps": 49.75, "mean_p": 0.457264, "boundary_entropy": 0.683101, "bpb": 1.253504}}, "bpb": {"en": 2.27184, "de": 2.589, "vi": 2.07989, "tr": 2.5319, "fi": 2.81565, "ru": 1.67377, "el": 1.74834, "hi": 1.35039, "ta": 1.10229, "th": 1.29177, "zh": 3.46765, "my": 1.2535}, "gini_cpc": 0.219046, "gini_cps": 0.219959, "bpic": 4.621807, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8543773934588208, "vi": 1.2082890658707417, "tr": 1.1753234167704811, "fi": 0.8057380730019642, "ru": 1.1997418647419484, "el": 1.4038425433071602, "hi": 2.5799747594634717, "ta": 2.1895429640586572, "th": 2.544595792217366, "zh": 2.0070488220089056, "my": 1.2375708791086566}, "t": 132.4} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.28026, "de": 2.59462, "vi": 2.08411, "tr": 2.539, "fi": 2.8177, "ru": 1.65437, "el": 1.74168, "hi": 1.34697, "ta": 1.11028, "th": 1.31056, "zh": 3.47255, "my": 1.26261}, "bytes_per_chunk": {"en": 3.742, "de": 4.177, "vi": 3.867, "tr": 3.369, "fi": 4.564, "ru": 5.414, "el": 4.546, "hi": 3.97, "ta": 4.357, "th": 3.938, "zh": 4.982, "my": 8.823}, "bpb_mean": 2.01789, "t": 133.8} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.41971, "bpb": 2.04821, "chunk_loss": 0.033358, "F": 0.2229, "G": 0.4425, "chunks": 228.25, "bytes_per_chunk": 4.486, "boundary_entropy": 0.66994, "router_noise": 0.0, "grad_norm": 1.2268, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.091, "alpha": 0.03, "t": 133.8} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.59092, "bpb": 2.29522, "chunk_loss": 0.033687, "F": 0.22894, "G": 0.44187, "chunks": 234.44, "bytes_per_chunk": 4.368, "boundary_entropy": 0.67054, "router_noise": 0.0, "grad_norm": 0.7565, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 136.0} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.64791, "bpb": 2.37743, "chunk_loss": 0.032665, "F": 0.21094, "G": 0.44362, "chunks": 216.0, "bytes_per_chunk": 4.741, "boundary_entropy": 0.67251, "router_noise": 0.0, "grad_norm": 0.8611, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 138.2} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.37925, "bpb": 1.98983, "chunk_loss": 0.032414, "F": 0.2081, "G": 0.43694, "chunks": 213.09, "bytes_per_chunk": 4.805, "boundary_entropy": 0.66799, "router_noise": 0.0, "grad_norm": 0.7695, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 140.4} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4611.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.264954, "cps": 36.023438, "mean_p": 0.439952, "boundary_entropy": 0.659559, "bpb": 2.238612}, "de": {"chunks": 4527.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.220474, "cps": 35.367188, "mean_p": 0.442304, "boundary_entropy": 0.66874, "bpb": 2.533827}, "vi": {"chunks": 5803.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.314782, "cps": 45.335938, "mean_p": 0.44509, "boundary_entropy": 0.671321, "bpb": 2.058878}, "tr": {"chunks": 5621.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.310501, "cps": 43.914062, "mean_p": 0.454763, "boundary_entropy": 0.670628, "bpb": 2.502269}, "fi": {"chunks": 4125.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.219287, "cps": 32.226562, "mean_p": 0.445623, "boundary_entropy": 0.671686, "bpb": 2.736412}, "ru": {"chunks": 5968.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.314934, "cps": 46.625, "mean_p": 0.430993, "boundary_entropy": 0.662926, "bpb": 1.638494}, "el": {"chunks": 8074.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.393412, "cps": 63.078125, "mean_p": 0.450871, "boundary_entropy": 0.677144, "bpb": 1.716493}, "hi": {"chunks": 8505.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.505498, "cps": 66.445312, "mean_p": 0.446002, "boundary_entropy": 0.668249, "bpb": 1.338283}, "ta": {"chunks": 13344.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.713774, "cps": 104.25, "mean_p": 0.441476, "boundary_entropy": 0.66203, "bpb": 1.084208}, "th": {"chunks": 13443.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.835748, "cps": 105.023438, "mean_p": 0.429867, "boundary_entropy": 0.664855, "bpb": 1.261974}, "zh": {"chunks": 3213.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.544023, "cps": 25.101562, "mean_p": 0.46131, "boundary_entropy": 0.67863, "bpb": 3.460354}, "my": {"chunks": 18514.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.977095, "cps": 144.640625, "mean_p": 0.468823, "boundary_entropy": 0.684258, "bpb": 1.230412}}, "bpb": {"en": 2.23861, "de": 2.53383, "vi": 2.05888, "tr": 2.50227, "fi": 2.73641, "ru": 1.63849, "el": 1.71649, "hi": 1.33828, "ta": 1.08421, "th": 1.26197, "zh": 3.46035, "my": 1.23041}, "gini_cpc": 0.279535, "gini_cps": 0.296687, "bpic": 4.00824, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8321221553599185, "vi": 1.188060143439495, "tr": 1.1719039871283787, "fi": 0.8276392289330012, "ru": 1.1886352298535854, "el": 1.4848305617472008, "hi": 1.9078676451395542, "ta": 2.6939501914355413, "th": 3.1543082465488625, "zh": 2.05327103586199, "my": 3.6877874438812714}, "t": 143.5} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.24798, "de": 2.54004, "vi": 2.06319, "tr": 2.50669, "fi": 2.74369, "ru": 1.62047, "el": 1.71065, "hi": 1.33446, "ta": 1.09216, "th": 1.27894, "zh": 3.4638, "my": 1.23979}, "bytes_per_chunk": {"en": 3.752, "de": 4.417, "vi": 4.11, "tr": 3.446, "fi": 4.49, "ru": 5.575, "el": 4.407, "hi": 5.309, "ta": 3.679, "th": 3.24, "zh": 5.144, "my": 2.909}, "bpb_mean": 1.98682, "t": 144.9} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.28949, "bpb": 1.86034, "chunk_loss": 0.034277, "F": 0.23865, "G": 0.4412, "chunks": 244.38, "bytes_per_chunk": 4.19, "boundary_entropy": 0.66884, "router_noise": 0.0, "grad_norm": 0.7411, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.092, "alpha": 0.03, "t": 145.0} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.53349, "bpb": 2.21237, "chunk_loss": 0.033566, "F": 0.22665, "G": 0.44097, "chunks": 232.09, "bytes_per_chunk": 4.412, "boundary_entropy": 0.66983, "router_noise": 0.0, "grad_norm": 0.7607, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 147.1} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.48511, "bpb": 2.14256, "chunk_loss": 0.033471, "F": 0.2247, "G": 0.43981, "chunks": 230.09, "bytes_per_chunk": 4.45, "boundary_entropy": 0.66882, "router_noise": 0.0, "grad_norm": 0.7734, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 149.3} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.46234, "bpb": 2.10971, "chunk_loss": 0.032817, "F": 0.21414, "G": 0.43954, "chunks": 219.28, "bytes_per_chunk": 4.67, "boundary_entropy": 0.66969, "router_noise": 0.0, "grad_norm": 0.7748, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 151.4} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4631.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.266104, "cps": 36.179688, "mean_p": 0.440131, "boundary_entropy": 0.659156, "bpb": 2.214993}, "de": {"chunks": 4992.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.243121, "cps": 39.0, "mean_p": 0.444985, "boundary_entropy": 0.668437, "bpb": 2.488543}, "vi": {"chunks": 4830.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.262002, "cps": 37.734375, "mean_p": 0.442322, "boundary_entropy": 0.670441, "bpb": 2.047687}, "tr": {"chunks": 5404.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.298514, "cps": 42.21875, "mean_p": 0.454918, "boundary_entropy": 0.670071, "bpb": 2.467964}, "fi": {"chunks": 4061.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.215884, "cps": 31.726562, "mean_p": 0.447305, "boundary_entropy": 0.672061, "bpb": 2.699616}, "ru": {"chunks": 6867.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.362375, "cps": 53.648438, "mean_p": 0.434582, "boundary_entropy": 0.664467, "bpb": 1.608071}, "el": {"chunks": 7333.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.357306, "cps": 57.289062, "mean_p": 0.442486, "boundary_entropy": 0.674396, "bpb": 1.696965}, "hi": {"chunks": 9273.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.551144, "cps": 72.445312, "mean_p": 0.446925, "boundary_entropy": 0.669706, "bpb": 1.304085}, "ta": {"chunks": 10221.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.546724, "cps": 79.851562, "mean_p": 0.438681, "boundary_entropy": 0.66195, "bpb": 1.062953}, "th": {"chunks": 4646.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.288841, "cps": 36.296875, "mean_p": 0.430017, "boundary_entropy": 0.67056, "bpb": 1.23384}, "zh": {"chunks": 3386.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.573315, "cps": 26.453125, "mean_p": 0.460813, "boundary_entropy": 0.67811, "bpb": 3.412689}, "my": {"chunks": 6125.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.323253, "cps": 47.851562, "mean_p": 0.460939, "boundary_entropy": 0.683753, "bpb": 1.218942}}, "bpb": {"en": 2.21499, "de": 2.48854, "vi": 2.04769, "tr": 2.46796, "fi": 2.69962, "ru": 1.60807, "el": 1.69697, "hi": 1.30409, "ta": 1.06295, "th": 1.23384, "zh": 3.41269, "my": 1.21894}, "gini_cpc": 0.182821, "gini_cps": 0.181612, "bpic": 5.347448, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9136324341317913, "vi": 0.9845852560107627, "tr": 1.1217966225630864, "fi": 0.8112793939132978, "ru": 1.3617806921235747, "el": 1.3427345967514388, "hi": 2.0711641778402092, "ta": 2.054552569636499, "th": 1.0854441426649493, "zh": 2.154481902208961, "my": 1.2147644006050586}, "t": 154.5} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.22612, "de": 2.49479, "vi": 2.05026, "tr": 2.47459, "fi": 2.71008, "ru": 1.58984, "el": 1.6901, "hi": 1.30257, "ta": 1.07181, "th": 1.2527, "zh": 3.41501, "my": 1.22667}, "bytes_per_chunk": {"en": 3.781, "de": 4.013, "vi": 4.961, "tr": 3.574, "fi": 4.759, "ru": 4.938, "el": 4.863, "hi": 4.905, "ta": 4.859, "th": 9.765, "zh": 4.92, "my": 8.926}, "bpb_mean": 1.95871, "t": 155.9} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 155.9} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "global", "lm_loss": 1.31937, "bpb": 1.90345, "chunk_loss": 0.032936, "F": 0.21619, "G": 0.44325, "chunks": 221.38, "bytes_per_chunk": 4.626, "boundary_entropy": 0.67134, "router_noise": 0.0, "grad_norm": 0.7659, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 155.9} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "global", "lm_loss": 1.47532, "bpb": 2.12844, "chunk_loss": 0.033177, "F": 0.21967, "G": 0.44012, "chunks": 224.94, "bytes_per_chunk": 4.552, "boundary_entropy": 0.66923, "router_noise": 0.0, "grad_norm": 0.83, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 158.0} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "global", "lm_loss": 1.4678, "bpb": 2.11759, "chunk_loss": 0.032813, "F": 0.21439, "G": 0.43998, "chunks": 219.53, "bytes_per_chunk": 4.664, "boundary_entropy": 0.66963, "router_noise": 0.0, "grad_norm": 0.7332, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 160.2} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "global", "lm_loss": 1.53652, "bpb": 2.21674, "chunk_loss": 0.034338, "F": 0.23981, "G": 0.43986, "chunks": 245.56, "bytes_per_chunk": 4.17, "boundary_entropy": 0.66757, "router_noise": 0.0, "grad_norm": 0.8135, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 162.4} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4171.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.239671, "cps": 32.585938, "mean_p": 0.43566, "boundary_entropy": 0.657621, "bpb": 2.195731}, "de": {"chunks": 4604.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.224224, "cps": 35.96875, "mean_p": 0.443399, "boundary_entropy": 0.66871, "bpb": 2.467904}, "vi": {"chunks": 5231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.283754, "cps": 40.867188, "mean_p": 0.44151, "boundary_entropy": 0.669971, "bpb": 2.036578}, "tr": {"chunks": 5615.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.31017, "cps": 43.867188, "mean_p": 0.455438, "boundary_entropy": 0.669964, "bpb": 2.420103}, "fi": {"chunks": 4198.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.223167, "cps": 32.796875, "mean_p": 0.446902, "boundary_entropy": 0.671758, "bpb": 2.658882}, "ru": {"chunks": 6369.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.336095, "cps": 49.757812, "mean_p": 0.433998, "boundary_entropy": 0.664304, "bpb": 1.586232}, "el": {"chunks": 8149.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.397067, "cps": 63.664062, "mean_p": 0.447681, "boundary_entropy": 0.675893, "bpb": 1.658312}, "hi": {"chunks": 5864.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.348529, "cps": 45.8125, "mean_p": 0.429355, "boundary_entropy": 0.664265, "bpb": 1.308429}, "ta": {"chunks": 8908.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.476491, "cps": 69.59375, "mean_p": 0.425612, "boundary_entropy": 0.657975, "bpb": 1.05797}, "th": {"chunks": 10288.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.639602, "cps": 80.375, "mean_p": 0.43147, "boundary_entropy": 0.668014, "bpb": 1.234063}, "zh": {"chunks": 2953.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.5, "cps": 23.070312, "mean_p": 0.4598, "boundary_entropy": 0.678354, "bpb": 3.402875}, "my": {"chunks": 15322.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.808634, "cps": 119.703125, "mean_p": 0.470683, "boundary_entropy": 0.685114, "bpb": 1.19779}}, "bpb": {"en": 2.19573, "de": 2.4679, "vi": 2.03658, "tr": 2.4201, "fi": 2.65888, "ru": 1.58623, "el": 1.65831, "hi": 1.30843, "ta": 1.05797, "th": 1.23406, "zh": 3.40287, "my": 1.19779}, "gini_cpc": 0.230425, "gini_cps": 0.250463, "bpic": 4.699052, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9355496446458067, "vi": 1.1839285905880017, "tr": 1.2941455982723227, "fi": 0.9311389229466412, "ru": 1.402316244403973, "el": 1.656713468803729, "hi": 1.4541955759769127, "ta": 1.9881020320840117, "th": 2.668663530556136, "zh": 2.0861903620234954, "my": 3.37392956796749}, "t": 165.5} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.20451, "de": 2.47438, "vi": 2.03939, "tr": 2.42817, "fi": 2.66866, "ru": 1.56518, "el": 1.64906, "hi": 1.3066, "ta": 1.0667, "th": 1.2517, "zh": 3.40524, "my": 1.20682}, "bytes_per_chunk": {"en": 4.161, "de": 4.422, "vi": 4.562, "tr": 3.41, "fi": 4.597, "ru": 5.346, "el": 4.441, "hi": 7.742, "ta": 5.537, "th": 4.231, "zh": 5.99, "my": 3.502}, "bpb_mean": 1.93887, "t": 166.8} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "global", "lm_loss": 1.50331, "bpb": 2.16882, "chunk_loss": 0.033423, "F": 0.22424, "G": 0.43885, "chunks": 229.62, "bytes_per_chunk": 4.459, "boundary_entropy": 0.66881, "router_noise": 0.0, "grad_norm": 0.7818, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 166.9} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "global", "lm_loss": 1.44572, "bpb": 2.08573, "chunk_loss": 0.032789, "F": 0.21365, "G": 0.43792, "chunks": 218.78, "bytes_per_chunk": 4.68, "boundary_entropy": 0.66858, "router_noise": 0.0, "grad_norm": 0.7723, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 169.0} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "global", "lm_loss": 1.5031, "bpb": 2.16851, "chunk_loss": 0.033026, "F": 0.21814, "G": 0.43801, "chunks": 223.38, "bytes_per_chunk": 4.584, "boundary_entropy": 0.66839, "router_noise": 0.0, "grad_norm": 0.8624, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 171.2} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "global", "lm_loss": 1.5349, "bpb": 2.21439, "chunk_loss": 0.034003, "F": 0.23419, "G": 0.44116, "chunks": 239.81, "bytes_per_chunk": 4.27, "boundary_entropy": 0.66977, "router_noise": 0.0, "grad_norm": 0.8255, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 173.3} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4167.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.239441, "cps": 32.554688, "mean_p": 0.435634, "boundary_entropy": 0.657848, "bpb": 2.177883}, "de": {"chunks": 4651.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.226513, "cps": 36.335938, "mean_p": 0.442605, "boundary_entropy": 0.668064, "bpb": 2.431318}, "vi": {"chunks": 5489.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.297749, "cps": 42.882812, "mean_p": 0.442722, "boundary_entropy": 0.669918, "bpb": 2.016126}, "tr": {"chunks": 5289.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.292162, "cps": 41.320312, "mean_p": 0.454351, "boundary_entropy": 0.669819, "bpb": 2.398482}, "fi": {"chunks": 4034.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.214449, "cps": 31.515625, "mean_p": 0.448795, "boundary_entropy": 0.672425, "bpb": 2.632621}, "ru": {"chunks": 6512.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.343641, "cps": 50.875, "mean_p": 0.43183, "boundary_entropy": 0.66332, "bpb": 1.568201}, "el": {"chunks": 7254.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.353457, "cps": 56.671875, "mean_p": 0.446283, "boundary_entropy": 0.675427, "bpb": 1.640655}, "hi": {"chunks": 9612.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.571293, "cps": 75.09375, "mean_p": 0.441998, "boundary_entropy": 0.668225, "bpb": 1.276698}, "ta": {"chunks": 9489.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.507569, "cps": 74.132812, "mean_p": 0.432172, "boundary_entropy": 0.658858, "bpb": 1.044428}, "th": {"chunks": 5560.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.345664, "cps": 43.4375, "mean_p": 0.424697, "boundary_entropy": 0.666785, "bpb": 1.198165}, "zh": {"chunks": 3181.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.538605, "cps": 24.851562, "mean_p": 0.460838, "boundary_entropy": 0.678023, "bpb": 3.351185}, "my": {"chunks": 4889.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.258022, "cps": 38.195312, "mean_p": 0.459306, "boundary_entropy": 0.684004, "bpb": 1.186683}}, "bpb": {"en": 2.17788, "de": 2.43132, "vi": 2.01613, "tr": 2.39848, "fi": 2.63262, "ru": 1.5682, "el": 1.64066, "hi": 1.2767, "ta": 1.04443, "th": 1.19816, "zh": 3.35119, "my": 1.18668}, "gini_cpc": 0.18492, "gini_cps": 0.181769, "bpic": 5.472657, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9460074402346649, "vi": 1.243514084373657, "tr": 1.2201792502680544, "fi": 0.8956217466772558, "ru": 1.4351780955148918, "el": 1.4761732319709995, "hi": 2.3859388508949, "ta": 2.1198034609223773, "th": 1.4436248395123727, "zh": 2.249421522742793, "my": 1.0775992512112647}, "t": 176.4} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.19076, "de": 2.43684, "vi": 2.01852, "tr": 2.40253, "fi": 2.64039, "ru": 1.54758, "el": 1.63272, "hi": 1.27256, "ta": 1.05224, "th": 1.21707, "zh": 3.35503, "my": 1.19556}, "bytes_per_chunk": {"en": 4.201, "de": 4.388, "vi": 4.309, "tr": 3.657, "fi": 4.928, "ru": 5.174, "el": 4.828, "hi": 4.626, "ta": 5.075, "th": 7.92, "zh": 5.187, "my": 10.395}, "bpb_mean": 1.91348, "t": 177.7} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "global", "lm_loss": 1.41524, "bpb": 2.04176, "chunk_loss": 0.032838, "F": 0.21466, "G": 0.43659, "chunks": 219.81, "bytes_per_chunk": 4.659, "boundary_entropy": 0.66752, "router_noise": 0.0, "grad_norm": 0.8045, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 177.7} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "global", "lm_loss": 1.33052, "bpb": 1.91954, "chunk_loss": 0.033674, "F": 0.2272, "G": 0.44072, "chunks": 232.66, "bytes_per_chunk": 4.401, "boundary_entropy": 0.67014, "router_noise": 0.0, "grad_norm": 0.7438, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 179.9} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "global", "lm_loss": 1.41279, "bpb": 2.03822, "chunk_loss": 0.032524, "F": 0.21024, "G": 0.43521, "chunks": 215.28, "bytes_per_chunk": 4.757, "boundary_entropy": 0.66708, "router_noise": 0.0, "grad_norm": 0.7521, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 182.0} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "global", "lm_loss": 1.3092, "bpb": 1.88878, "chunk_loss": 0.032285, "F": 0.20541, "G": 0.43458, "chunks": 210.34, "bytes_per_chunk": 4.868, "boundary_entropy": 0.66707, "router_noise": 0.0, "grad_norm": 0.7185, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 184.1} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4136.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.23766, "cps": 32.3125, "mean_p": 0.435331, "boundary_entropy": 0.65764, "bpb": 2.16032}, "de": {"chunks": 4439.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.216189, "cps": 34.679688, "mean_p": 0.441626, "boundary_entropy": 0.668213, "bpb": 2.422721}, "vi": {"chunks": 4603.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.249688, "cps": 35.960938, "mean_p": 0.438366, "boundary_entropy": 0.668346, "bpb": 2.007446}, "tr": {"chunks": 5326.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.294205, "cps": 41.609375, "mean_p": 0.45314, "boundary_entropy": 0.669289, "bpb": 2.377943}, "fi": {"chunks": 3777.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.200787, "cps": 29.507812, "mean_p": 0.442637, "boundary_entropy": 0.670421, "bpb": 2.617984}, "ru": {"chunks": 6453.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.340528, "cps": 50.414062, "mean_p": 0.430395, "boundary_entropy": 0.662248, "bpb": 1.543459}, "el": {"chunks": 7470.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.363982, "cps": 58.359375, "mean_p": 0.446782, "boundary_entropy": 0.675347, "bpb": 1.607397}, "hi": {"chunks": 8987.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.534146, "cps": 70.210938, "mean_p": 0.444353, "boundary_entropy": 0.669117, "bpb": 1.258481}, "ta": {"chunks": 10241.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.547794, "cps": 80.007812, "mean_p": 0.432267, "boundary_entropy": 0.659335, "bpb": 1.033108}, "th": {"chunks": 9168.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.569972, "cps": 71.625, "mean_p": 0.432288, "boundary_entropy": 0.668745, "bpb": 1.178925}, "zh": {"chunks": 3110.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.526583, "cps": 24.296875, "mean_p": 0.460296, "boundary_entropy": 0.677958, "bpb": 3.348485}, "my": {"chunks": 7952.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.419675, "cps": 62.125, "mean_p": 0.460954, "boundary_entropy": 0.683685, "bpb": 1.166705}}, "bpb": {"en": 2.16032, "de": 2.42272, "vi": 2.00745, "tr": 2.37794, "fi": 2.61798, "ru": 1.54346, "el": 1.6074, "hi": 1.25848, "ta": 1.03311, "th": 1.17893, "zh": 3.34848, "my": 1.16671}, "gini_cpc": 0.201894, "gini_cps": 0.207293, "bpic": 5.072308, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9096541977385839, "vi": 1.0506097407431305, "tr": 1.2379246213185222, "fi": 0.8448482163045677, "ru": 1.4328345360640595, "el": 1.5315223774505577, "hi": 2.2475184154784866, "ta": 2.304944575008678, "th": 2.3982647792884633, "zh": 2.215697851118195, "my": 1.76586128624855}, "t": 187.2} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.1712, "de": 2.42792, "vi": 2.0107, "tr": 2.38321, "fi": 2.62806, "ru": 1.52028, "el": 1.5999, "hi": 1.25585, "ta": 1.04038, "th": 1.19867, "zh": 3.34669, "my": 1.17683}, "bytes_per_chunk": {"en": 4.225, "de": 4.532, "vi": 5.0, "tr": 3.639, "fi": 5.168, "ru": 5.257, "el": 4.706, "hi": 5.008, "ta": 4.785, "th": 4.774, "zh": 5.303, "my": 7.412}, "bpb_mean": 1.89664, "t": 188.6} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "global", "lm_loss": 1.52435, "bpb": 2.19917, "chunk_loss": 0.032385, "F": 0.20694, "G": 0.44164, "chunks": 211.91, "bytes_per_chunk": 4.832, "boundary_entropy": 0.67181, "router_noise": 0.0, "grad_norm": 0.7555, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.09, "alpha": 0.03, "t": 188.6} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "global", "lm_loss": 1.48535, "bpb": 2.1429, "chunk_loss": 0.032904, "F": 0.21634, "G": 0.43932, "chunks": 221.53, "bytes_per_chunk": 4.622, "boundary_entropy": 0.66924, "router_noise": 0.0, "grad_norm": 0.8087, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 190.7} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "global", "lm_loss": 1.42245, "bpb": 2.05216, "chunk_loss": 0.03249, "F": 0.2088, "G": 0.44221, "chunks": 213.81, "bytes_per_chunk": 4.789, "boundary_entropy": 0.67205, "router_noise": 0.0, "grad_norm": 0.842, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 192.8} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "global", "lm_loss": 1.32551, "bpb": 1.91231, "chunk_loss": 0.033182, "F": 0.22046, "G": 0.43906, "chunks": 225.75, "bytes_per_chunk": 4.536, "boundary_entropy": 0.66959, "router_noise": 0.0, "grad_norm": 0.7974, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 195.0} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4372.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.251221, "cps": 34.15625, "mean_p": 0.437321, "boundary_entropy": 0.65814, "bpb": 2.134516}, "de": {"chunks": 4928.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.240004, "cps": 38.5, "mean_p": 0.445105, "boundary_entropy": 0.668167, "bpb": 2.38482}, "vi": {"chunks": 5503.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.298508, "cps": 42.992188, "mean_p": 0.441784, "boundary_entropy": 0.669065, "bpb": 1.989316}, "tr": {"chunks": 5619.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.310391, "cps": 43.898438, "mean_p": 0.455575, "boundary_entropy": 0.669363, "bpb": 2.332958}, "fi": {"chunks": 4371.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.232364, "cps": 34.148438, "mean_p": 0.447583, "boundary_entropy": 0.671391, "bpb": 2.563385}, "ru": {"chunks": 6133.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.323641, "cps": 47.914062, "mean_p": 0.430993, "boundary_entropy": 0.66274, "bpb": 1.525327}, "el": {"chunks": 8359.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.407299, "cps": 65.304688, "mean_p": 0.445315, "boundary_entropy": 0.673935, "bpb": 1.590082}, "hi": {"chunks": 9427.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.560297, "cps": 73.648438, "mean_p": 0.445057, "boundary_entropy": 0.669139, "bpb": 1.254495}, "ta": {"chunks": 8445.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.451725, "cps": 65.976562, "mean_p": 0.426599, "boundary_entropy": 0.658198, "bpb": 1.034542}, "th": {"chunks": 8866.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.551197, "cps": 69.265625, "mean_p": 0.435789, "boundary_entropy": 0.670021, "bpb": 1.170326}, "zh": {"chunks": 3295.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.557907, "cps": 25.742188, "mean_p": 0.461706, "boundary_entropy": 0.677961, "bpb": 3.296002}, "my": {"chunks": 8973.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.473559, "cps": 70.101562, "mean_p": 0.462759, "boundary_entropy": 0.684263, "bpb": 1.154256}}, "bpb": {"en": 2.13452, "de": 2.38482, "vi": 1.98932, "tr": 2.33296, "fi": 2.56339, "ru": 1.52533, "el": 1.59008, "hi": 1.2545, "ta": 1.03454, "th": 1.17033, "zh": 3.296, "my": 1.15426}, "gini_cpc": 0.178025, "gini_cps": 0.17775, "bpic": 4.901981, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9553494521950673, "vi": 1.1882295203517912, "tr": 1.2355275891805249, "fi": 0.924938570086812, "ru": 1.2882724437352593, "el": 1.6212778410886515, "hi": 2.230295463907155, "ta": 1.798117845894771, "th": 2.1940707546056357, "zh": 2.220782074085962, "my": 1.8850299664454613}, "t": 198.1} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.14853, "de": 2.38935, "vi": 1.99138, "tr": 2.34159, "fi": 2.57632, "ru": 1.50476, "el": 1.58161, "hi": 1.25067, "ta": 1.04163, "th": 1.18902, "zh": 3.29784, "my": 1.16337}, "bytes_per_chunk": {"en": 4.03, "de": 4.149, "vi": 4.19, "tr": 3.43, "fi": 4.38, "ru": 5.448, "el": 4.292, "hi": 4.787, "ta": 5.748, "th": 5.015, "zh": 4.96, "my": 6.195}, "bpb_mean": 1.873, "t": 199.4} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "global", "lm_loss": 1.39855, "bpb": 2.01768, "chunk_loss": 0.033318, "F": 0.22278, "G": 0.43895, "chunks": 228.12, "bytes_per_chunk": 4.489, "boundary_entropy": 0.66916, "router_noise": 0.0, "grad_norm": 0.743, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 199.5} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "global", "lm_loss": 1.38329, "bpb": 1.99566, "chunk_loss": 0.034148, "F": 0.23605, "G": 0.44257, "chunks": 241.72, "bytes_per_chunk": 4.236, "boundary_entropy": 0.66899, "router_noise": 0.0, "grad_norm": 0.7637, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 201.6} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "global", "lm_loss": 1.28695, "bpb": 1.85667, "chunk_loss": 0.033028, "F": 0.21796, "G": 0.43867, "chunks": 223.19, "bytes_per_chunk": 4.588, "boundary_entropy": 0.66924, "router_noise": 0.0, "grad_norm": 0.7787, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 203.7} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "global", "lm_loss": 1.36488, "bpb": 1.96911, "chunk_loss": 0.032412, "F": 0.20773, "G": 0.43715, "chunks": 212.72, "bytes_per_chunk": 4.814, "boundary_entropy": 0.66973, "router_noise": 0.0, "grad_norm": 0.7197, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 205.8} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4238.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.243521, "cps": 33.109375, "mean_p": 0.436327, "boundary_entropy": 0.657818, "bpb": 2.123321}, "de": {"chunks": 4481.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.218234, "cps": 35.007812, "mean_p": 0.442593, "boundary_entropy": 0.668145, "bpb": 2.370498}, "vi": {"chunks": 4967.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.269433, "cps": 38.804688, "mean_p": 0.441198, "boundary_entropy": 0.66954, "bpb": 1.98411}, "tr": {"chunks": 5012.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.27686, "cps": 39.15625, "mean_p": 0.450475, "boundary_entropy": 0.668152, "bpb": 2.328532}, "fi": {"chunks": 4008.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.213067, "cps": 31.3125, "mean_p": 0.44478, "boundary_entropy": 0.669997, "bpb": 2.552775}, "ru": {"chunks": 6616.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.349129, "cps": 51.6875, "mean_p": 0.432571, "boundary_entropy": 0.663288, "bpb": 1.518505}, "el": {"chunks": 7830.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.381523, "cps": 61.171875, "mean_p": 0.447448, "boundary_entropy": 0.675278, "bpb": 1.567827}, "hi": {"chunks": 9136.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.543001, "cps": 71.375, "mean_p": 0.442779, "boundary_entropy": 0.667039, "bpb": 1.252137}, "ta": {"chunks": 9482.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.507194, "cps": 74.078125, "mean_p": 0.431868, "boundary_entropy": 0.65919, "bpb": 1.018084}, "th": {"chunks": 9192.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.571464, "cps": 71.8125, "mean_p": 0.432292, "boundary_entropy": 0.66872, "bpb": 1.156081}, "zh": {"chunks": 3170.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.536742, "cps": 24.765625, "mean_p": 0.459243, "boundary_entropy": 0.67704, "bpb": 3.274972}, "my": {"chunks": 5194.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.274119, "cps": 40.578125, "mean_p": 0.457578, "boundary_entropy": 0.68354, "bpb": 1.164019}}, "bpb": {"en": 2.12332, "de": 2.3705, "vi": 1.98411, "tr": 2.32853, "fi": 2.55278, "ru": 1.5185, "el": 1.56783, "hi": 1.25214, "ta": 1.01808, "th": 1.15608, "zh": 3.27497, "my": 1.16402}, "gini_cpc": 0.199597, "gini_cps": 0.195756, "bpic": 5.233901, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8961603079718793, "vi": 1.10640514297585, "tr": 1.1369036969694963, "fi": 0.8749414614851004, "ru": 1.4336708334319095, "el": 1.566693655574035, "hi": 2.22979114177466, "ta": 2.082752427413002, "th": 2.3466705237909213, "zh": 2.204088290884707, "my": 1.1256457528181498}, "t": 208.9} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.13391, "de": 2.37561, "vi": 1.98549, "tr": 2.3374, "fi": 2.56301, "ru": 1.49685, "el": 1.56068, "hi": 1.2495, "ta": 1.02606, "th": 1.17292, "zh": 3.27612, "my": 1.17345}, "bytes_per_chunk": {"en": 4.122, "de": 4.5, "vi": 4.529, "tr": 3.877, "fi": 4.854, "ru": 4.966, "el": 4.588, "hi": 5.015, "ta": 5.225, "th": 4.765, "zh": 5.192, "my": 9.809}, "bpb_mean": 1.86258, "t": 210.2} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "global", "lm_loss": 1.31407, "bpb": 1.8958, "chunk_loss": 0.032431, "F": 0.20856, "G": 0.43789, "chunks": 213.56, "bytes_per_chunk": 4.795, "boundary_entropy": 0.66891, "router_noise": 0.0, "grad_norm": 0.802, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 210.3} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "global", "lm_loss": 1.21755, "bpb": 1.75656, "chunk_loss": 0.032073, "F": 0.20245, "G": 0.43594, "chunks": 207.31, "bytes_per_chunk": 4.939, "boundary_entropy": 0.66704, "router_noise": 0.0, "grad_norm": 0.7044, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 212.4} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "global", "lm_loss": 1.32798, "bpb": 1.91587, "chunk_loss": 0.034176, "F": 0.23648, "G": 0.44193, "chunks": 242.16, "bytes_per_chunk": 4.229, "boundary_entropy": 0.66909, "router_noise": 0.0, "grad_norm": 0.79, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 214.5} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "global", "lm_loss": 1.37542, "bpb": 1.98431, "chunk_loss": 0.03262, "F": 0.21103, "G": 0.44005, "chunks": 216.09, "bytes_per_chunk": 4.739, "boundary_entropy": 0.67075, "router_noise": 0.0, "grad_norm": 0.7814, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 216.6} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4347.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.249785, "cps": 33.960938, "mean_p": 0.436922, "boundary_entropy": 0.657661, "bpb": 2.110015}, "de": {"chunks": 4836.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.235523, "cps": 37.78125, "mean_p": 0.444737, "boundary_entropy": 0.668531, "bpb": 2.340216}, "vi": {"chunks": 4689.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.254353, "cps": 36.632812, "mean_p": 0.437029, "boundary_entropy": 0.667866, "bpb": 1.970607}, "tr": {"chunks": 5474.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.302381, "cps": 42.765625, "mean_p": 0.454402, "boundary_entropy": 0.668642, "bpb": 2.299376}, "fi": {"chunks": 3964.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.210728, "cps": 30.96875, "mean_p": 0.444571, "boundary_entropy": 0.670376, "bpb": 2.525473}, "ru": {"chunks": 7088.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.374037, "cps": 55.375, "mean_p": 0.434825, "boundary_entropy": 0.664359, "bpb": 1.501109}, "el": {"chunks": 8048.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.392145, "cps": 62.875, "mean_p": 0.447438, "boundary_entropy": 0.675013, "bpb": 1.552361}, "hi": {"chunks": 6459.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.383893, "cps": 50.460938, "mean_p": 0.435742, "boundary_entropy": 0.667487, "bpb": 1.240611}, "ta": {"chunks": 10018.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.535865, "cps": 78.265625, "mean_p": 0.4339, "boundary_entropy": 0.660912, "bpb": 1.015398}, "th": {"chunks": 8601.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.534722, "cps": 67.195312, "mean_p": 0.435789, "boundary_entropy": 0.670592, "bpb": 1.148421}, "zh": {"chunks": 3015.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.510498, "cps": 23.554688, "mean_p": 0.458885, "boundary_entropy": 0.677204, "bpb": 3.258569}, "my": {"chunks": 13430.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.708782, "cps": 104.921875, "mean_p": 0.471581, "boundary_entropy": 0.68515, "bpb": 1.144454}}, "bpb": {"en": 2.11001, "de": 2.34022, "vi": 1.97061, "tr": 2.29938, "fi": 2.52547, "ru": 1.50111, "el": 1.55236, "hi": 1.24061, "ta": 1.0154, "th": 1.14842, "zh": 3.25857, "my": 1.14445}, "gini_cpc": 0.20841, "gini_cps": 0.232762, "bpic": 4.799122, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.942905925612449, "vi": 1.018290215580342, "tr": 1.210566695513942, "fi": 0.8436382116259524, "ru": 1.497438429924377, "el": 1.569934750108002, "hi": 1.536896747974983, "ta": 2.1453099046829065, "th": 2.1407323027060565, "zh": 2.043752747495441, "my": 2.8375734803165487}, "t": 219.7} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.12122, "de": 2.35029, "vi": 1.97278, "tr": 2.30667, "fi": 2.53516, "ru": 1.4803, "el": 1.54607, "hi": 1.23585, "ta": 1.02363, "th": 1.16514, "zh": 3.25902, "my": 1.15332}, "bytes_per_chunk": {"en": 4.052, "de": 4.213, "vi": 5.0, "tr": 3.46, "fi": 4.988, "ru": 4.792, "el": 4.37, "hi": 7.058, "ta": 4.878, "th": 5.125, "zh": 5.561, "my": 4.067}, "bpb_mean": 1.84579, "t": 221.0} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "global", "lm_loss": 1.31985, "bpb": 1.90415, "chunk_loss": 0.032962, "F": 0.21661, "G": 0.43915, "chunks": 221.81, "bytes_per_chunk": 4.617, "boundary_entropy": 0.6697, "router_noise": 0.0, "grad_norm": 0.8033, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 221.1} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "global", "lm_loss": 1.31602, "bpb": 1.89862, "chunk_loss": 0.032694, "F": 0.21252, "G": 0.43578, "chunks": 217.62, "bytes_per_chunk": 4.705, "boundary_entropy": 0.66703, "router_noise": 0.0, "grad_norm": 0.7597, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 223.2} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "global", "lm_loss": 1.53085, "bpb": 2.20855, "chunk_loss": 0.034063, "F": 0.23431, "G": 0.44249, "chunks": 239.94, "bytes_per_chunk": 4.268, "boundary_entropy": 0.67046, "router_noise": 0.0, "grad_norm": 0.8123, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 225.3} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "global", "lm_loss": 1.43102, "bpb": 2.06453, "chunk_loss": 0.033996, "F": 0.23431, "G": 0.43751, "chunks": 239.94, "bytes_per_chunk": 4.268, "boundary_entropy": 0.66528, "router_noise": 0.0, "grad_norm": 0.7624, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 227.4} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4161.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.239097, "cps": 32.507812, "mean_p": 0.435536, "boundary_entropy": 0.657438, "bpb": 2.096493}, "de": {"chunks": 4521.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.220182, "cps": 35.320312, "mean_p": 0.441267, "boundary_entropy": 0.667326, "bpb": 2.329635}, "vi": {"chunks": 5217.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.282994, "cps": 40.757812, "mean_p": 0.441399, "boundary_entropy": 0.668841, "bpb": 1.958193}, "tr": {"chunks": 5274.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.291333, "cps": 41.203125, "mean_p": 0.452816, "boundary_entropy": 0.668681, "bpb": 2.272297}, "fi": {"chunks": 3504.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.186274, "cps": 27.375, "mean_p": 0.441417, "boundary_entropy": 0.66956, "bpb": 2.501888}, "ru": {"chunks": 6597.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.348127, "cps": 51.539062, "mean_p": 0.432057, "boundary_entropy": 0.663423, "bpb": 1.4805}, "el": {"chunks": 8065.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.392974, "cps": 63.007812, "mean_p": 0.447272, "boundary_entropy": 0.674898, "bpb": 1.530046}, "hi": {"chunks": 9637.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.572779, "cps": 75.289062, "mean_p": 0.446142, "boundary_entropy": 0.669447, "bpb": 1.216118}, "ta": {"chunks": 9464.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.506232, "cps": 73.9375, "mean_p": 0.433764, "boundary_entropy": 0.660661, "bpb": 0.999259}, "th": {"chunks": 8573.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.532981, "cps": 66.976562, "mean_p": 0.436438, "boundary_entropy": 0.670977, "bpb": 1.13268}, "zh": {"chunks": 3024.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.512022, "cps": 23.625, "mean_p": 0.460163, "boundary_entropy": 0.677436, "bpb": 3.248433}, "my": {"chunks": 6243.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.329481, "cps": 48.773438, "mean_p": 0.462541, "boundary_entropy": 0.684538, "bpb": 1.132801}}, "bpb": {"en": 2.09649, "de": 2.32963, "vi": 1.95819, "tr": 2.2723, "fi": 2.50189, "ru": 1.4805, "el": 1.53005, "hi": 1.21612, "ta": 0.99926, "th": 1.13268, "zh": 3.24843, "my": 1.1328}, "gini_cpc": 0.196644, "gini_cps": 0.202046, "bpic": 5.16668, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9208915846195642, "vi": 1.1835976635302938, "tr": 1.2184731915955884, "fi": 0.77907383417877, "ru": 1.4560077082140417, "el": 1.6435765299775682, "hi": 2.395593855488933, "ta": 2.11726718489593, "th": 2.229144198123747, "zh": 2.1414835791983564, "my": 1.3780226732168288}, "t": 230.5} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.10727, "de": 2.33788, "vi": 1.961, "tr": 2.28339, "fi": 2.51296, "ru": 1.45899, "el": 1.52313, "hi": 1.21362, "ta": 1.00663, "th": 1.14909, "zh": 3.24698, "my": 1.14016}, "bytes_per_chunk": {"en": 4.21, "de": 4.529, "vi": 4.373, "tr": 3.573, "fi": 5.644, "ru": 5.101, "el": 4.472, "hi": 4.69, "ta": 5.172, "th": 5.104, "zh": 5.524, "my": 9.141}, "bpb_mean": 1.82843, "t": 231.8} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "global", "lm_loss": 1.37546, "bpb": 1.98437, "chunk_loss": 0.033285, "F": 0.22186, "G": 0.43854, "chunks": 227.19, "bytes_per_chunk": 4.507, "boundary_entropy": 0.66741, "router_noise": 0.0, "grad_norm": 0.7544, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 231.8} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "global", "lm_loss": 1.53542, "bpb": 2.21515, "chunk_loss": 0.033027, "F": 0.21762, "G": 0.4414, "chunks": 222.84, "bytes_per_chunk": 4.595, "boundary_entropy": 0.66947, "router_noise": 0.0, "grad_norm": 0.7846, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 234.0} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "global", "lm_loss": 1.4735, "bpb": 2.12581, "chunk_loss": 0.032699, "F": 0.21289, "G": 0.43748, "chunks": 218.0, "bytes_per_chunk": 4.697, "boundary_entropy": 0.66782, "router_noise": 0.0, "grad_norm": 0.7681, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 236.0} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "global", "lm_loss": 1.4284, "bpb": 2.06075, "chunk_loss": 0.033026, "F": 0.21765, "G": 0.43825, "chunks": 222.88, "bytes_per_chunk": 4.595, "boundary_entropy": 0.66811, "router_noise": 0.0, "grad_norm": 0.7836, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 238.1} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4292.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.246624, "cps": 33.53125, "mean_p": 0.437916, "boundary_entropy": 0.657995, "bpb": 2.088511}, "de": {"chunks": 4793.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.233429, "cps": 37.445312, "mean_p": 0.443069, "boundary_entropy": 0.667483, "bpb": 2.306253}, "vi": {"chunks": 4503.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.244264, "cps": 35.179688, "mean_p": 0.438188, "boundary_entropy": 0.668255, "bpb": 1.953581}, "tr": {"chunks": 5353.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.295697, "cps": 41.820312, "mean_p": 0.451642, "boundary_entropy": 0.667579, "bpb": 2.260716}, "fi": {"chunks": 4373.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.23247, "cps": 34.164062, "mean_p": 0.448714, "boundary_entropy": 0.670759, "bpb": 2.47068}, "ru": {"chunks": 6788.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.358206, "cps": 53.03125, "mean_p": 0.434209, "boundary_entropy": 0.664186, "bpb": 1.468031}, "el": {"chunks": 8727.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.42523, "cps": 68.179688, "mean_p": 0.451772, "boundary_entropy": 0.675732, "bpb": 1.517042}, "hi": {"chunks": 9804.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.582704, "cps": 76.59375, "mean_p": 0.446391, "boundary_entropy": 0.669219, "bpb": 1.210131}, "ta": {"chunks": 8632.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.461728, "cps": 67.4375, "mean_p": 0.431412, "boundary_entropy": 0.660867, "bpb": 0.992581}, "th": {"chunks": 10187.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.633323, "cps": 79.585938, "mean_p": 0.436429, "boundary_entropy": 0.670117, "bpb": 1.124823}, "zh": {"chunks": 3102.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.525229, "cps": 24.234375, "mean_p": 0.4607, "boundary_entropy": 0.677478, "bpb": 3.208595}, "my": {"chunks": 8307.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.43841, "cps": 64.898438, "mean_p": 0.465288, "boundary_entropy": 0.684915, "bpb": 1.129706}}, "bpb": {"en": 2.08851, "de": 2.30625, "vi": 1.95358, "tr": 2.26072, "fi": 2.47068, "ru": 1.46803, "el": 1.51704, "hi": 1.21013, "ta": 0.99258, "th": 1.12482, "zh": 3.20859, "my": 1.12971}, "gini_cpc": 0.19759, "gini_cps": 0.201455, "bpic": 4.86655, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9464974074647821, "vi": 0.9904286894003793, "tr": 1.198977681249147, "fi": 0.9426099088480233, "ru": 1.4524360717737117, "el": 1.7242035610354922, "hi": 2.362722062124634, "ta": 1.87219193075083, "th": 2.5679682378186013, "zh": 2.1296721800913914, "my": 1.7776458418301246}, "t": 241.2} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.09939, "de": 2.31609, "vi": 1.95401, "tr": 2.26942, "fi": 2.48298, "ru": 1.44544, "el": 1.51102, "hi": 1.20782, "ta": 0.99983, "th": 1.14153, "zh": 3.20853, "my": 1.13825}, "bytes_per_chunk": {"en": 4.129, "de": 4.284, "vi": 5.082, "tr": 3.586, "fi": 4.51, "ru": 4.923, "el": 4.093, "hi": 4.482, "ta": 5.675, "th": 4.336, "zh": 5.248, "my": 6.734}, "bpb_mean": 1.81452, "t": 242.6} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "global", "lm_loss": 1.30965, "bpb": 1.88943, "chunk_loss": 0.034064, "F": 0.23495, "G": 0.44172, "chunks": 240.59, "bytes_per_chunk": 4.256, "boundary_entropy": 0.66914, "router_noise": 0.0, "grad_norm": 0.7698, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 242.6} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "global", "lm_loss": 1.46517, "bpb": 2.1138, "chunk_loss": 0.032006, "F": 0.20099, "G": 0.4362, "chunks": 205.81, "bytes_per_chunk": 4.975, "boundary_entropy": 0.66841, "router_noise": 0.0, "grad_norm": 0.7482, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 244.7} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "global", "lm_loss": 1.33033, "bpb": 1.91926, "chunk_loss": 0.032645, "F": 0.21097, "G": 0.43789, "chunks": 216.03, "bytes_per_chunk": 4.74, "boundary_entropy": 0.66763, "router_noise": 0.0, "grad_norm": 0.7884, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 246.8} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "global", "lm_loss": 1.48457, "bpb": 2.14178, "chunk_loss": 0.03293, "F": 0.21588, "G": 0.43917, "chunks": 221.06, "bytes_per_chunk": 4.632, "boundary_entropy": 0.66857, "router_noise": 0.0, "grad_norm": 0.7466, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 248.9} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4214.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.242142, "cps": 32.921875, "mean_p": 0.436437, "boundary_entropy": 0.6573, "bpb": 2.07458}, "de": {"chunks": 4839.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.235669, "cps": 37.804688, "mean_p": 0.443163, "boundary_entropy": 0.667307, "bpb": 2.293925}, "vi": {"chunks": 4794.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.260049, "cps": 37.453125, "mean_p": 0.440445, "boundary_entropy": 0.669035, "bpb": 1.943782}, "tr": {"chunks": 5344.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.2952, "cps": 41.75, "mean_p": 0.451861, "boundary_entropy": 0.667455, "bpb": 2.245183}, "fi": {"chunks": 4150.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.220616, "cps": 32.421875, "mean_p": 0.447066, "boundary_entropy": 0.670656, "bpb": 2.463868}, "ru": {"chunks": 6672.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.352084, "cps": 52.125, "mean_p": 0.433728, "boundary_entropy": 0.664437, "bpb": 1.455094}, "el": {"chunks": 8649.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.42143, "cps": 67.570312, "mean_p": 0.449229, "boundary_entropy": 0.674955, "bpb": 1.503724}, "hi": {"chunks": 6986.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.415215, "cps": 54.578125, "mean_p": 0.437557, "boundary_entropy": 0.66676, "bpb": 1.203821}, "ta": {"chunks": 8146.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.435731, "cps": 63.640625, "mean_p": 0.425576, "boundary_entropy": 0.659168, "bpb": 0.984901}, "th": {"chunks": 9868.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.613491, "cps": 77.09375, "mean_p": 0.433035, "boundary_entropy": 0.668672, "bpb": 1.118627}, "zh": {"chunks": 3118.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.527938, "cps": 24.359375, "mean_p": 0.460368, "boundary_entropy": 0.677176, "bpb": 3.195754}, "my": {"chunks": 11118.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.586764, "cps": 86.859375, "mean_p": 0.466285, "boundary_entropy": 0.684457, "bpb": 1.117874}}, "bpb": {"en": 2.07458, "de": 2.29393, "vi": 1.94378, "tr": 2.24518, "fi": 2.46387, "ru": 1.45509, "el": 1.50372, "hi": 1.20382, "ta": 0.9849, "th": 1.11863, "zh": 3.19575, "my": 1.11787}, "gini_cpc": 0.195425, "gini_cps": 0.208645, "bpic": 4.926712, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9732688054149511, "vi": 1.0739510245788477, "tr": 1.2191172796722105, "fi": 0.9110994872003443, "ru": 1.4540401952030735, "el": 1.7404223064721787, "hi": 1.7147590251415539, "ta": 1.7994862195361734, "th": 2.533597748907198, "zh": 2.1802799305879117, "my": 2.4232202108055025}, "t": 252.0} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.08578, "de": 2.30151, "vi": 1.94566, "tr": 2.25379, "fi": 2.47484, "ru": 1.43422, "el": 1.49703, "hi": 1.20014, "ta": 0.99416, "th": 1.13578, "zh": 3.19606, "my": 1.12683}, "bytes_per_chunk": {"en": 4.151, "de": 4.256, "vi": 4.752, "tr": 3.588, "fi": 4.717, "ru": 5.036, "el": 4.246, "hi": 6.454, "ta": 6.056, "th": 4.402, "zh": 5.206, "my": 4.781}, "bpb_mean": 1.80382, "t": 253.3} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 253.3} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "global", "lm_loss": 1.3005, "bpb": 1.87622, "chunk_loss": 0.033129, "F": 0.21967, "G": 0.43833, "chunks": 224.94, "bytes_per_chunk": 4.552, "boundary_entropy": 0.66898, "router_noise": 0.0, "grad_norm": 0.7485, "lr": 9.12e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 253.3} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "global", "lm_loss": 1.30899, "bpb": 1.88847, "chunk_loss": 0.033465, "F": 0.22443, "G": 0.43868, "chunks": 229.81, "bytes_per_chunk": 4.456, "boundary_entropy": 0.66801, "router_noise": 0.0, "grad_norm": 0.7499, "lr": 8.81e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 255.5} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "global", "lm_loss": 1.35395, "bpb": 1.95334, "chunk_loss": 0.032943, "F": 0.21725, "G": 0.43378, "chunks": 222.47, "bytes_per_chunk": 4.603, "boundary_entropy": 0.66517, "router_noise": 0.0, "grad_norm": 0.7556, "lr": 8.5e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 257.6} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "global", "lm_loss": 1.48681, "bpb": 2.14502, "chunk_loss": 0.03258, "F": 0.21039, "G": 0.43964, "chunks": 215.44, "bytes_per_chunk": 4.753, "boundary_entropy": 0.67034, "router_noise": 0.0, "grad_norm": 0.811, "lr": 8.21e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 259.7} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4106.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.235936, "cps": 32.078125, "mean_p": 0.43488, "boundary_entropy": 0.656932, "bpb": 2.064215}, "de": {"chunks": 4645.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.226221, "cps": 36.289062, "mean_p": 0.442074, "boundary_entropy": 0.667177, "bpb": 2.286148}, "vi": {"chunks": 4958.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.268945, "cps": 38.734375, "mean_p": 0.437821, "boundary_entropy": 0.667473, "bpb": 1.936193}, "tr": {"chunks": 5205.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.287521, "cps": 40.664062, "mean_p": 0.452181, "boundary_entropy": 0.667708, "bpb": 2.230834}, "fi": {"chunks": 4210.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.223805, "cps": 32.890625, "mean_p": 0.447717, "boundary_entropy": 0.670747, "bpb": 2.437586}, "ru": {"chunks": 6182.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.326227, "cps": 48.296875, "mean_p": 0.430016, "boundary_entropy": 0.66351, "bpb": 1.442808}, "el": {"chunks": 7337.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.357501, "cps": 57.320312, "mean_p": 0.443378, "boundary_entropy": 0.673907, "bpb": 1.497234}, "hi": {"chunks": 7975.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.473997, "cps": 62.304688, "mean_p": 0.440432, "boundary_entropy": 0.667798, "bpb": 1.187982}, "ta": {"chunks": 9055.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.484354, "cps": 70.742188, "mean_p": 0.431299, "boundary_entropy": 0.661703, "bpb": 0.977004}, "th": {"chunks": 4743.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.294871, "cps": 37.054688, "mean_p": 0.428041, "boundary_entropy": 0.66864, "bpb": 1.117966}, "zh": {"chunks": 2952.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.499831, "cps": 23.0625, "mean_p": 0.459666, "boundary_entropy": 0.677219, "bpb": 3.187477}, "my": {"chunks": 7577.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.399884, "cps": 59.195312, "mean_p": 0.464665, "boundary_entropy": 0.684729, "bpb": 1.113915}}, "bpb": {"en": 2.06421, "de": 2.28615, "vi": 1.93619, "tr": 2.23083, "fi": 2.43759, "ru": 1.44281, "el": 1.49723, "hi": 1.18798, "ta": 0.977, "th": 1.11797, "zh": 3.18748, "my": 1.11392}, "gini_cpc": 0.161839, "gini_cps": 0.173504, "bpic": 5.566481, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9588230951522823, "vi": 1.139904729707503, "tr": 1.2186397993900733, "fi": 0.9485831100213074, "ru": 1.3826904446417945, "el": 1.5152449633036191, "hi": 2.009003965493906, "ta": 2.0529017281671957, "th": 1.2497905443575525, "zh": 2.118498133363596, "my": 1.6948805128317448}, "t": 262.7} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.07491, "de": 2.29327, "vi": 1.93734, "tr": 2.23869, "fi": 2.44936, "ru": 1.42097, "el": 1.48983, "hi": 1.18469, "ta": 0.98456, "th": 1.13518, "zh": 3.18602, "my": 1.1231}, "bytes_per_chunk": {"en": 4.272, "de": 4.395, "vi": 4.572, "tr": 3.658, "fi": 4.601, "ru": 5.417, "el": 4.951, "hi": 5.685, "ta": 5.435, "th": 9.142, "zh": 5.525, "my": 7.314}, "bpb_mean": 1.79316, "t": 264.0} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "global", "lm_loss": 1.45101, "bpb": 2.09337, "chunk_loss": 0.032938, "F": 0.21649, "G": 0.43746, "chunks": 221.69, "bytes_per_chunk": 4.619, "boundary_entropy": 0.66797, "router_noise": 0.0, "grad_norm": 0.8246, "lr": 7.92e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 264.0} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "global", "lm_loss": 1.37865, "bpb": 1.98898, "chunk_loss": 0.033469, "F": 0.22522, "G": 0.44003, "chunks": 230.62, "bytes_per_chunk": 4.44, "boundary_entropy": 0.66883, "router_noise": 0.0, "grad_norm": 0.7886, "lr": 7.63e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 266.2} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "global", "lm_loss": 1.31058, "bpb": 1.89077, "chunk_loss": 0.033087, "F": 0.2197, "G": 0.43446, "chunks": 224.97, "bytes_per_chunk": 4.552, "boundary_entropy": 0.66538, "router_noise": 0.0, "grad_norm": 0.7412, "lr": 7.36e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 268.3} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "global", "lm_loss": 1.36908, "bpb": 1.97516, "chunk_loss": 0.032372, "F": 0.20706, "G": 0.43859, "chunks": 212.03, "bytes_per_chunk": 4.829, "boundary_entropy": 0.66953, "router_noise": 0.0, "grad_norm": 0.7905, "lr": 7.09e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 270.4} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3990.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.229271, "cps": 31.171875, "mean_p": 0.433686, "boundary_entropy": 0.656337, "bpb": 2.059577}, "de": {"chunks": 4739.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.230799, "cps": 37.023438, "mean_p": 0.441813, "boundary_entropy": 0.666699, "bpb": 2.269506}, "vi": {"chunks": 4064.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22045, "cps": 31.75, "mean_p": 0.436083, "boundary_entropy": 0.66803, "bpb": 1.932472}, "tr": {"chunks": 4989.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.27559, "cps": 38.976562, "mean_p": 0.44948, "boundary_entropy": 0.666751, "bpb": 2.222289}, "fi": {"chunks": 3884.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.206475, "cps": 30.34375, "mean_p": 0.442603, "boundary_entropy": 0.668921, "bpb": 2.431635}, "ru": {"chunks": 6116.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.322744, "cps": 47.78125, "mean_p": 0.430053, "boundary_entropy": 0.663374, "bpb": 1.434776}, "el": {"chunks": 8315.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.405155, "cps": 64.960938, "mean_p": 0.447784, "boundary_entropy": 0.674264, "bpb": 1.475828}, "hi": {"chunks": 8575.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.509658, "cps": 66.992188, "mean_p": 0.441427, "boundary_entropy": 0.668137, "bpb": 1.185538}, "ta": {"chunks": 9773.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.52276, "cps": 76.351562, "mean_p": 0.432535, "boundary_entropy": 0.66238, "bpb": 0.973484}, "th": {"chunks": 8756.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.544358, "cps": 68.40625, "mean_p": 0.431561, "boundary_entropy": 0.668278, "bpb": 1.104021}, "zh": {"chunks": 3276.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.55469, "cps": 25.59375, "mean_p": 0.462333, "boundary_entropy": 0.677468, "bpb": 3.171378}, "my": {"chunks": 6892.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.363732, "cps": 53.84375, "mean_p": 0.46274, "boundary_entropy": 0.68443, "bpb": 1.105876}}, "bpb": {"en": 2.05958, "de": 2.26951, "vi": 1.93247, "tr": 2.22229, "fi": 2.43164, "ru": 1.43478, "el": 1.47583, "hi": 1.18554, "ta": 0.97348, "th": 1.10402, "zh": 3.17138, "my": 1.10588}, "gini_cpc": 0.201868, "gini_cps": 0.200188, "bpic": 5.230833, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.006666290720714, "vi": 0.9615276596699234, "tr": 1.202026872960449, "fi": 0.9005722535566496, "ru": 1.4076979784553734, "el": 1.7671468174895653, "hi": 2.222952999139751, "ta": 2.280098735740126, "th": 2.374301747530156, "zh": 2.4193665674514766, "my": 1.5864745778315326}, "t": 273.4} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.07034, "de": 2.27757, "vi": 1.93387, "tr": 2.22906, "fi": 2.44107, "ru": 1.41379, "el": 1.46833, "hi": 1.18166, "ta": 0.98029, "th": 1.12069, "zh": 3.17065, "my": 1.11386}, "bytes_per_chunk": {"en": 4.333, "de": 4.303, "vi": 5.787, "tr": 3.801, "fi": 5.019, "ru": 5.563, "el": 4.354, "hi": 5.192, "ta": 4.996, "th": 5.153, "zh": 5.114, "my": 8.132}, "bpb_mean": 1.78343, "t": 274.8} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "global", "lm_loss": 1.41085, "bpb": 2.03542, "chunk_loss": 0.032894, "F": 0.21506, "G": 0.44092, "chunks": 220.22, "bytes_per_chunk": 4.65, "boundary_entropy": 0.6699, "router_noise": 0.0, "grad_norm": 0.7495, "lr": 6.82e-05, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 274.8} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "global", "lm_loss": 1.51282, "bpb": 2.18254, "chunk_loss": 0.032698, "F": 0.21179, "G": 0.44178, "chunks": 216.88, "bytes_per_chunk": 4.722, "boundary_entropy": 0.67127, "router_noise": 0.0, "grad_norm": 0.8283, "lr": 6.57e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 276.9} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "global", "lm_loss": 1.45473, "bpb": 2.09873, "chunk_loss": 0.033076, "F": 0.21835, "G": 0.44242, "chunks": 223.59, "bytes_per_chunk": 4.58, "boundary_entropy": 0.67114, "router_noise": 0.0, "grad_norm": 0.8387, "lr": 6.32e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 279.0} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "global", "lm_loss": 1.46197, "bpb": 2.10918, "chunk_loss": 0.033703, "F": 0.22873, "G": 0.43804, "chunks": 234.22, "bytes_per_chunk": 4.372, "boundary_entropy": 0.6664, "router_noise": 0.0, "grad_norm": 0.7668, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 281.1} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4070.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.233868, "cps": 31.796875, "mean_p": 0.433874, "boundary_entropy": 0.656533, "bpb": 2.046099}, "de": {"chunks": 4711.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.229436, "cps": 36.804688, "mean_p": 0.441707, "boundary_entropy": 0.666846, "bpb": 2.262583}, "vi": {"chunks": 4632.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.251261, "cps": 36.1875, "mean_p": 0.438451, "boundary_entropy": 0.668394, "bpb": 1.924717}, "tr": {"chunks": 5174.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.285809, "cps": 40.421875, "mean_p": 0.451078, "boundary_entropy": 0.666993, "bpb": 2.209569}, "fi": {"chunks": 4024.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.213917, "cps": 31.4375, "mean_p": 0.444582, "boundary_entropy": 0.669333, "bpb": 2.410367}, "ru": {"chunks": 6751.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.356253, "cps": 52.742188, "mean_p": 0.432785, "boundary_entropy": 0.664011, "bpb": 1.427603}, "el": {"chunks": 8037.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.391609, "cps": 62.789062, "mean_p": 0.444301, "boundary_entropy": 0.673372, "bpb": 1.464718}, "hi": {"chunks": 8885.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.528083, "cps": 69.414062, "mean_p": 0.441985, "boundary_entropy": 0.66848, "bpb": 1.174792}, "ta": {"chunks": 8740.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.467505, "cps": 68.28125, "mean_p": 0.428663, "boundary_entropy": 0.660246, "bpb": 0.966178}, "th": {"chunks": 8269.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.514081, "cps": 64.601562, "mean_p": 0.430602, "boundary_entropy": 0.667961, "bpb": 1.098769}, "zh": {"chunks": 3169.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.536573, "cps": 24.757812, "mean_p": 0.460294, "boundary_entropy": 0.677152, "bpb": 3.159633}, "my": {"chunks": 8080.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.42643, "cps": 63.125, "mean_p": 0.466606, "boundary_entropy": 0.68521, "bpb": 1.105964}}, "bpb": {"en": 2.0461, "de": 2.26258, "vi": 1.92472, "tr": 2.20957, "fi": 2.41037, "ru": 1.4276, "el": 1.46472, "hi": 1.17479, "ta": 0.96618, "th": 1.09877, "zh": 3.15963, "my": 1.10596}, "gini_cpc": 0.182964, "gini_cps": 0.182792, "bpic": 5.14852, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9810483417895876, "vi": 1.0743730845276818, "tr": 1.2220967352934429, "fi": 0.9146939352882684, "ru": 1.5233110928150506, "el": 1.6744910636008428, "hi": 2.258042283807514, "ta": 1.999013256247811, "th": 2.198171828255757, "zh": 2.294343860530789, "my": 1.8233821362383742}, "t": 284.1} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.0578, "de": 2.27005, "vi": 1.92563, "tr": 2.21688, "fi": 2.4219, "ru": 1.40545, "el": 1.45711, "hi": 1.17126, "ta": 0.97376, "th": 1.11495, "zh": 3.15741, "my": 1.11362}, "bytes_per_chunk": {"en": 4.326, "de": 4.35, "vi": 4.983, "tr": 3.661, "fi": 4.872, "ru": 5.009, "el": 4.521, "hi": 4.987, "ta": 5.571, "th": 5.327, "zh": 5.233, "my": 6.802}, "bpb_mean": 1.77382, "t": 285.5} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "global", "lm_loss": 1.35278, "bpb": 1.95165, "chunk_loss": 0.033504, "F": 0.22534, "G": 0.43838, "chunks": 230.75, "bytes_per_chunk": 4.438, "boundary_entropy": 0.66708, "router_noise": 0.0, "grad_norm": 0.7448, "lr": 5.85e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 285.5} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "global", "lm_loss": 1.28583, "bpb": 1.85507, "chunk_loss": 0.03346, "F": 0.22556, "G": 0.43714, "chunks": 230.97, "bytes_per_chunk": 4.434, "boundary_entropy": 0.66591, "router_noise": 0.0, "grad_norm": 0.7309, "lr": 5.62e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 287.6} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "global", "lm_loss": 1.38533, "bpb": 1.99861, "chunk_loss": 0.033935, "F": 0.23206, "G": 0.44306, "chunks": 237.62, "bytes_per_chunk": 4.309, "boundary_entropy": 0.67012, "router_noise": 0.0, "grad_norm": 0.8207, "lr": 5.41e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 289.7} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "global", "lm_loss": 1.2602, "bpb": 1.81808, "chunk_loss": 0.033029, "F": 0.21832, "G": 0.43494, "chunks": 223.56, "bytes_per_chunk": 4.58, "boundary_entropy": 0.66573, "router_noise": 0.0, "grad_norm": 0.7786, "lr": 5.2e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 291.8} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4276.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.245705, "cps": 33.40625, "mean_p": 0.435284, "boundary_entropy": 0.656573, "bpb": 2.037533}, "de": {"chunks": 4718.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.229776, "cps": 36.859375, "mean_p": 0.441194, "boundary_entropy": 0.666682, "bpb": 2.254899}, "vi": {"chunks": 5054.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.274152, "cps": 39.484375, "mean_p": 0.43907, "boundary_entropy": 0.667921, "bpb": 1.911937}, "tr": {"chunks": 5197.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.287079, "cps": 40.601562, "mean_p": 0.451152, "boundary_entropy": 0.66731, "bpb": 2.198032}, "fi": {"chunks": 4106.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.218277, "cps": 32.078125, "mean_p": 0.444394, "boundary_entropy": 0.669263, "bpb": 2.395081}, "ru": {"chunks": 6024.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.317889, "cps": 47.0625, "mean_p": 0.42901, "boundary_entropy": 0.663168, "bpb": 1.421461}, "el": {"chunks": 8038.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.391658, "cps": 62.796875, "mean_p": 0.445182, "boundary_entropy": 0.673589, "bpb": 1.458278}, "hi": {"chunks": 8809.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.523566, "cps": 68.820312, "mean_p": 0.441529, "boundary_entropy": 0.668428, "bpb": 1.1701}, "ta": {"chunks": 8560.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.457876, "cps": 66.875, "mean_p": 0.430004, "boundary_entropy": 0.661878, "bpb": 0.962418}, "th": {"chunks": 9481.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.589431, "cps": 74.070312, "mean_p": 0.431216, "boundary_entropy": 0.667975, "bpb": 1.100933}, "zh": {"chunks": 3068.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.519472, "cps": 23.96875, "mean_p": 0.458944, "boundary_entropy": 0.676824, "bpb": 3.148147}, "my": {"chunks": 8015.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.423, "cps": 62.617188, "mean_p": 0.466736, "boundary_entropy": 0.685281, "bpb": 1.098861}}, "bpb": {"en": 2.03753, "de": 2.2549, "vi": 1.91194, "tr": 2.19803, "fi": 2.39508, "ru": 1.42146, "el": 1.45828, "hi": 1.1701, "ta": 0.96242, "th": 1.10093, "zh": 3.14815, "my": 1.09886}, "gini_cpc": 0.186906, "gini_cps": 0.187101, "bpic": 5.093582, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9351729860399037, "vi": 1.1157798631604023, "tr": 1.1683920386550157, "fi": 0.8883691841034635, "ru": 1.2937851813274557, "el": 1.594019318123858, "hi": 2.130874934149898, "ta": 1.8635228350526685, "th": 2.3989406575742236, "zh": 2.1142110400008614, "my": 1.7215774850691756}, "t": 294.8} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.04839, "de": 2.26387, "vi": 1.91251, "tr": 2.20431, "fi": 2.40647, "ru": 1.40117, "el": 1.45092, "hi": 1.16641, "ta": 0.97045, "th": 1.11678, "zh": 3.14693, "my": 1.10703}, "bytes_per_chunk": {"en": 4.085, "de": 4.346, "vi": 4.524, "tr": 3.623, "fi": 4.729, "ru": 5.56, "el": 4.519, "hi": 5.04, "ta": 5.633, "th": 4.656, "zh": 5.39, "my": 6.81}, "bpb_mean": 1.76627, "t": 296.2} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "global", "lm_loss": 1.3713, "bpb": 1.97836, "chunk_loss": 0.032414, "F": 0.20779, "G": 0.44017, "chunks": 212.78, "bytes_per_chunk": 4.812, "boundary_entropy": 0.6702, "router_noise": 0.0, "grad_norm": 0.7703, "lr": 5e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 296.2} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "global", "lm_loss": 1.38325, "bpb": 1.9956, "chunk_loss": 0.03244, "F": 0.20868, "G": 0.43619, "chunks": 213.69, "bytes_per_chunk": 4.792, "boundary_entropy": 0.66725, "router_noise": 0.0, "grad_norm": 0.8894, "lr": 4.81e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 298.3} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "global", "lm_loss": 1.48494, "bpb": 2.14232, "chunk_loss": 0.03296, "F": 0.21674, "G": 0.43998, "chunks": 221.94, "bytes_per_chunk": 4.614, "boundary_entropy": 0.67035, "router_noise": 0.0, "grad_norm": 0.7762, "lr": 4.63e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 300.4} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "global", "lm_loss": 1.33849, "bpb": 1.93104, "chunk_loss": 0.033181, "F": 0.22125, "G": 0.43648, "chunks": 226.56, "bytes_per_chunk": 4.52, "boundary_entropy": 0.66673, "router_noise": 0.0, "grad_norm": 0.7567, "lr": 4.45e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 302.6} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4157.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.238867, "cps": 32.476562, "mean_p": 0.433697, "boundary_entropy": 0.655717, "bpb": 2.03431}, "de": {"chunks": 4757.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.231676, "cps": 37.164062, "mean_p": 0.441796, "boundary_entropy": 0.666654, "bpb": 2.245325}, "vi": {"chunks": 4575.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.248169, "cps": 35.742188, "mean_p": 0.437784, "boundary_entropy": 0.668, "bpb": 1.910046}, "tr": {"chunks": 5072.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.280175, "cps": 39.625, "mean_p": 0.45026, "boundary_entropy": 0.666765, "bpb": 2.191333}, "fi": {"chunks": 4048.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.215193, "cps": 31.625, "mean_p": 0.444554, "boundary_entropy": 0.669238, "bpb": 2.389798}, "ru": {"chunks": 5952.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.31409, "cps": 46.5, "mean_p": 0.428907, "boundary_entropy": 0.663027, "bpb": 1.417502}, "el": {"chunks": 7856.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.38279, "cps": 61.375, "mean_p": 0.444798, "boundary_entropy": 0.673525, "bpb": 1.453866}, "hi": {"chunks": 9154.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.544071, "cps": 71.515625, "mean_p": 0.442276, "boundary_entropy": 0.668315, "bpb": 1.16404}, "ta": {"chunks": 9411.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.503397, "cps": 73.523438, "mean_p": 0.431783, "boundary_entropy": 0.662494, "bpb": 0.956875}, "th": {"chunks": 9491.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.590053, "cps": 74.148438, "mean_p": 0.430704, "boundary_entropy": 0.667845, "bpb": 1.089941}, "zh": {"chunks": 3065.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.518964, "cps": 23.945312, "mean_p": 0.459665, "boundary_entropy": 0.676987, "bpb": 3.141883}, "my": {"chunks": 7652.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.403842, "cps": 59.78125, "mean_p": 0.463459, "boundary_entropy": 0.684674, "bpb": 1.095179}}, "bpb": {"en": 2.03431, "de": 2.24532, "vi": 1.91005, "tr": 2.19133, "fi": 2.3898, "ru": 1.4175, "el": 1.45387, "hi": 1.16404, "ta": 0.95687, "th": 1.08994, "zh": 3.14188, "my": 1.09518}, "gini_cpc": 0.197772, "gini_cps": 0.199386, "bpic": 5.104149, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9698952668422856, "vi": 1.0389437914187114, "tr": 1.172931876427155, "fi": 0.9008919703866605, "ru": 1.3149153762322257, "el": 1.6025246749626214, "hi": 2.277717879328482, "ta": 2.107436024510805, "th": 2.4702164177871215, "zh": 2.172606787061934, "my": 1.690657677299619}, "t": 305.6} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.04646, "de": 2.25338, "vi": 1.91057, "tr": 2.19942, "fi": 2.39955, "ru": 1.39666, "el": 1.44623, "hi": 1.15999, "ta": 0.96496, "th": 1.10652, "zh": 3.13998, "my": 1.10371}, "bytes_per_chunk": {"en": 4.205, "de": 4.254, "vi": 5.047, "tr": 3.744, "fi": 4.885, "ru": 5.675, "el": 4.639, "hi": 4.828, "ta": 5.172, "th": 4.713, "zh": 5.348, "my": 7.272}, "bpb_mean": 1.76062, "t": 306.9} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "global", "lm_loss": 1.25546, "bpb": 1.81124, "chunk_loss": 0.032442, "F": 0.20819, "G": 0.43755, "chunks": 213.19, "bytes_per_chunk": 4.803, "boundary_entropy": 0.66843, "router_noise": 0.0, "grad_norm": 0.818, "lr": 4.29e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 307.0} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "global", "lm_loss": 1.30268, "bpb": 1.87937, "chunk_loss": 0.032291, "F": 0.20554, "G": 0.43714, "chunks": 210.47, "bytes_per_chunk": 4.865, "boundary_entropy": 0.66853, "router_noise": 0.0, "grad_norm": 0.7356, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 309.1} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "global", "lm_loss": 1.23384, "bpb": 1.78005, "chunk_loss": 0.032981, "F": 0.21725, "G": 0.43806, "chunks": 222.47, "bytes_per_chunk": 4.603, "boundary_entropy": 0.6681, "router_noise": 0.0, "grad_norm": 0.747, "lr": 3.99e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 311.1} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "global", "lm_loss": 1.45954, "bpb": 2.10568, "chunk_loss": 0.031832, "F": 0.19843, "G": 0.44089, "chunks": 203.19, "bytes_per_chunk": 5.04, "boundary_entropy": 0.67206, "router_noise": 0.0, "grad_norm": 0.8066, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 313.2} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4152.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.23858, "cps": 32.4375, "mean_p": 0.434331, "boundary_entropy": 0.656217, "bpb": 2.028863}, "de": {"chunks": 4617.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.224858, "cps": 36.070312, "mean_p": 0.441105, "boundary_entropy": 0.666504, "bpb": 2.238575}, "vi": {"chunks": 4466.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.242257, "cps": 34.890625, "mean_p": 0.438517, "boundary_entropy": 0.668535, "bpb": 1.905668}, "tr": {"chunks": 5175.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.285864, "cps": 40.429688, "mean_p": 0.450907, "boundary_entropy": 0.666854, "bpb": 2.18058}, "fi": {"chunks": 4090.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.217426, "cps": 31.953125, "mean_p": 0.445013, "boundary_entropy": 0.669451, "bpb": 2.378401}, "ru": {"chunks": 6197.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327018, "cps": 48.414062, "mean_p": 0.431091, "boundary_entropy": 0.663957, "bpb": 1.40729}, "el": {"chunks": 7685.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.374458, "cps": 60.039062, "mean_p": 0.444868, "boundary_entropy": 0.673878, "bpb": 1.445372}, "hi": {"chunks": 9032.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.53682, "cps": 70.5625, "mean_p": 0.442299, "boundary_entropy": 0.668165, "bpb": 1.156623}, "ta": {"chunks": 9413.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.503504, "cps": 73.539062, "mean_p": 0.431812, "boundary_entropy": 0.661753, "bpb": 0.952024}, "th": {"chunks": 8086.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.502704, "cps": 63.171875, "mean_p": 0.429974, "boundary_entropy": 0.667938, "bpb": 1.087449}, "zh": {"chunks": 3076.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.520826, "cps": 24.03125, "mean_p": 0.459759, "boundary_entropy": 0.677151, "bpb": 3.128857}, "my": {"chunks": 9681.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.510925, "cps": 75.632812, "mean_p": 0.465131, "boundary_entropy": 0.684643, "bpb": 1.090761}}, "bpb": {"en": 2.02886, "de": 2.23858, "vi": 1.90567, "tr": 2.18058, "fi": 2.3784, "ru": 1.40729, "el": 1.44537, "hi": 1.15662, "ta": 0.95202, "th": 1.08745, "zh": 3.12886, "my": 1.09076}, "gini_cpc": 0.187423, "gini_cps": 0.201586, "bpic": 5.071772, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9424845567926887, "vi": 1.0154121417070758, "tr": 1.1981924482225823, "fi": 0.9113353151197046, "ru": 1.370689409306511, "el": 1.569530653469574, "hi": 2.250067938400227, "ta": 2.1104222868777316, "th": 2.107072345074307, "zh": 2.1830297982457343, "my": 2.1415273211991197}, "t": 316.3} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.04036, "de": 2.24682, "vi": 1.90545, "tr": 2.18744, "fi": 2.38853, "ru": 1.38657, "el": 1.43811, "hi": 1.15267, "ta": 0.95981, "th": 1.10364, "zh": 3.12721, "my": 1.09948}, "bytes_per_chunk": {"en": 4.198, "de": 4.391, "vi": 5.147, "tr": 3.646, "fi": 4.749, "ru": 5.432, "el": 4.721, "hi": 4.946, "ta": 5.225, "th": 5.527, "zh": 5.457, "my": 5.727}, "bpb_mean": 1.75301, "t": 317.6} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "global", "lm_loss": 1.35602, "bpb": 1.95632, "chunk_loss": 0.032514, "F": 0.20975, "G": 0.43682, "chunks": 214.78, "bytes_per_chunk": 4.768, "boundary_entropy": 0.6686, "router_noise": 0.0, "grad_norm": 0.8313, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 317.7} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "global", "lm_loss": 1.2127, "bpb": 1.74955, "chunk_loss": 0.03288, "F": 0.21555, "G": 0.43722, "chunks": 220.72, "bytes_per_chunk": 4.639, "boundary_entropy": 0.6683, "router_noise": 0.0, "grad_norm": 0.771, "lr": 3.62e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 319.7} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "global", "lm_loss": 1.32068, "bpb": 1.90534, "chunk_loss": 0.032958, "F": 0.21667, "G": 0.43929, "chunks": 221.88, "bytes_per_chunk": 4.615, "boundary_entropy": 0.66822, "router_noise": 0.0, "grad_norm": 0.8034, "lr": 3.51e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 321.8} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "global", "lm_loss": 1.39018, "bpb": 2.00561, "chunk_loss": 0.033037, "F": 0.21793, "G": 0.43924, "chunks": 223.16, "bytes_per_chunk": 4.589, "boundary_entropy": 0.66844, "router_noise": 0.0, "grad_norm": 0.8331, "lr": 3.41e-05, "router_trainable": true, "sec_per_step": 0.041, "alpha": 0.03, "t": 323.9} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4001.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.229903, "cps": 31.257812, "mean_p": 0.433011, "boundary_entropy": 0.655795, "bpb": 2.026054}, "de": {"chunks": 4627.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.225345, "cps": 36.148438, "mean_p": 0.440842, "boundary_entropy": 0.666287, "bpb": 2.234248}, "vi": {"chunks": 4526.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.245511, "cps": 35.359375, "mean_p": 0.438292, "boundary_entropy": 0.668372, "bpb": 1.902506}, "tr": {"chunks": 5117.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.28266, "cps": 39.976562, "mean_p": 0.450248, "boundary_entropy": 0.666641, "bpb": 2.173687}, "fi": {"chunks": 4056.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.215619, "cps": 31.6875, "mean_p": 0.44457, "boundary_entropy": 0.669158, "bpb": 2.376066}, "ru": {"chunks": 6164.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.325277, "cps": 48.15625, "mean_p": 0.430462, "boundary_entropy": 0.663551, "bpb": 1.400037}, "el": {"chunks": 8011.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.390343, "cps": 62.585938, "mean_p": 0.445467, "boundary_entropy": 0.673869, "bpb": 1.436412}, "hi": {"chunks": 8590.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.51055, "cps": 67.109375, "mean_p": 0.441983, "boundary_entropy": 0.66836, "bpb": 1.151878}, "ta": {"chunks": 8976.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.480128, "cps": 70.125, "mean_p": 0.430303, "boundary_entropy": 0.662042, "bpb": 0.949624}, "th": {"chunks": 8594.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.534287, "cps": 67.140625, "mean_p": 0.431313, "boundary_entropy": 0.668529, "bpb": 1.083277}, "zh": {"chunks": 3042.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.515069, "cps": 23.765625, "mean_p": 0.458977, "boundary_entropy": 0.676849, "bpb": 3.123858}, "my": {"chunks": 8117.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.428383, "cps": 63.414062, "mean_p": 0.464475, "boundary_entropy": 0.684708, "bpb": 1.086518}}, "bpb": {"en": 2.02605, "de": 2.23425, "vi": 1.90251, "tr": 2.17369, "fi": 2.37607, "ru": 1.40004, "el": 1.43641, "hi": 1.15188, "ta": 0.94962, "th": 1.08328, "zh": 3.12386, "my": 1.08652}, "gini_cpc": 0.185424, "gini_cps": 0.189087, "bpic": 5.198805, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9801728328936613, "vi": 1.0678911232322106, "tr": 1.2294770725370914, "fi": 0.9378678136271046, "ru": 1.414845391554486, "el": 1.6978583522515804, "hi": 2.2207192629777177, "ta": 2.0883964352586633, "th": 2.32396644388794, "zh": 2.2403781885547915, "my": 1.863321258032547}, "t": 326.9} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.03726, "de": 2.24155, "vi": 1.90255, "tr": 2.18047, "fi": 2.3852, "ru": 1.38017, "el": 1.4291, "hi": 1.14825, "ta": 0.95791, "th": 1.0996, "zh": 3.12142, "my": 1.09493}, "bytes_per_chunk": {"en": 4.381, "de": 4.401, "vi": 5.073, "tr": 3.689, "fi": 4.796, "ru": 5.444, "el": 4.581, "hi": 5.155, "ta": 5.517, "th": 5.174, "zh": 5.379, "my": 6.744}, "bpb_mean": 1.7482, "t": 328.2} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "global", "lm_loss": 1.33855, "bpb": 1.93112, "chunk_loss": 0.032904, "F": 0.21625, "G": 0.43693, "chunks": 221.44, "bytes_per_chunk": 4.624, "boundary_entropy": 0.66767, "router_noise": 0.0, "grad_norm": 0.7773, "lr": 3.33e-05, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 328.3} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "global", "lm_loss": 1.31634, "bpb": 1.89908, "chunk_loss": 0.032812, "F": 0.21432, "G": 0.438, "chunks": 219.47, "bytes_per_chunk": 4.666, "boundary_entropy": 0.66838, "router_noise": 0.0, "grad_norm": 0.8309, "lr": 3.25e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 330.4} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "global", "lm_loss": 1.2942, "bpb": 1.86714, "chunk_loss": 0.032918, "F": 0.21652, "G": 0.43417, "chunks": 221.72, "bytes_per_chunk": 4.618, "boundary_entropy": 0.66566, "router_noise": 0.0, "grad_norm": 0.7879, "lr": 3.18e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 332.5} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "global", "lm_loss": 1.208, "bpb": 1.74278, "chunk_loss": 0.032614, "F": 0.21115, "G": 0.43653, "chunks": 216.22, "bytes_per_chunk": 4.736, "boundary_entropy": 0.66674, "router_noise": 0.0, "grad_norm": 0.7547, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 334.5} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4126.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.237086, "cps": 32.234375, "mean_p": 0.434512, "boundary_entropy": 0.656174, "bpb": 2.02094}, "de": {"chunks": 4663.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.227098, "cps": 36.429688, "mean_p": 0.441313, "boundary_entropy": 0.666324, "bpb": 2.228244}, "vi": {"chunks": 4721.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.256089, "cps": 36.882812, "mean_p": 0.438691, "boundary_entropy": 0.668228, "bpb": 1.898199}, "tr": {"chunks": 5023.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.277468, "cps": 39.242188, "mean_p": 0.449994, "boundary_entropy": 0.6668, "bpb": 2.17099}, "fi": {"chunks": 4246.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.225719, "cps": 33.171875, "mean_p": 0.445504, "boundary_entropy": 0.669284, "bpb": 2.365018}, "ru": {"chunks": 6308.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.332876, "cps": 49.28125, "mean_p": 0.430919, "boundary_entropy": 0.663607, "bpb": 1.398426}, "el": {"chunks": 8145.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.396872, "cps": 63.632812, "mean_p": 0.445463, "boundary_entropy": 0.673489, "bpb": 1.431217}, "hi": {"chunks": 8174.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.485825, "cps": 63.859375, "mean_p": 0.44096, "boundary_entropy": 0.668554, "bpb": 1.148761}, "ta": {"chunks": 8658.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.463118, "cps": 67.640625, "mean_p": 0.42922, "boundary_entropy": 0.661649, "bpb": 0.948452}, "th": {"chunks": 7519.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.467454, "cps": 58.742188, "mean_p": 0.430456, "boundary_entropy": 0.668438, "bpb": 1.081217}, "zh": {"chunks": 3122.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.528615, "cps": 24.390625, "mean_p": 0.46063, "boundary_entropy": 0.67724, "bpb": 3.120444}, "my": {"chunks": 7546.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.398248, "cps": 58.953125, "mean_p": 0.463345, "boundary_entropy": 0.684554, "bpb": 1.087187}}, "bpb": {"en": 2.02094, "de": 2.22824, "vi": 1.8982, "tr": 2.17099, "fi": 2.36502, "ru": 1.39843, "el": 1.43122, "hi": 1.14876, "ta": 0.94845, "th": 1.08122, "zh": 3.12044, "my": 1.08719}, "gini_cpc": 0.169412, "gini_cps": 0.171526, "bpic": 5.311774, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9578729406163361, "vi": 1.0801541909903145, "tr": 1.1703278051117467, "fi": 0.9520571149196875, "ru": 1.4040331663420205, "el": 1.6739602463915673, "hi": 2.0491533341640262, "ta": 1.953381222178043, "th": 1.9716685820345772, "zh": 2.2296379750687745, "my": 1.6797642009444733}, "t": 337.6} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.03165, "de": 2.23751, "vi": 1.89832, "tr": 2.17805, "fi": 2.37516, "ru": 1.37759, "el": 1.42397, "hi": 1.14519, "ta": 0.95641, "th": 1.09691, "zh": 3.11802, "my": 1.09499}, "bytes_per_chunk": {"en": 4.239, "de": 4.338, "vi": 4.88, "tr": 3.783, "fi": 4.586, "ru": 5.324, "el": 4.483, "hi": 5.445, "ta": 5.712, "th": 5.925, "zh": 5.413, "my": 7.377}, "bpb_mean": 1.74448, "t": 338.9} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "global", "lm_loss": 1.29138, "bpb": 1.86306, "chunk_loss": 0.032046, "F": 0.20163, "G": 0.43584, "chunks": 206.47, "bytes_per_chunk": 4.96, "boundary_entropy": 0.66778, "router_noise": 0.0, "grad_norm": 0.8303, "lr": 3.08e-05, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 338.9} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "global", "lm_loss": 1.28525, "bpb": 1.85422, "chunk_loss": 0.033256, "F": 0.22095, "G": 0.44105, "chunks": 226.25, "bytes_per_chunk": 4.526, "boundary_entropy": 0.66931, "router_noise": 0.0, "grad_norm": 0.7657, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 341.0} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "global", "lm_loss": 1.35534, "bpb": 1.95534, "chunk_loss": 0.03284, "F": 0.21405, "G": 0.44063, "chunks": 219.19, "bytes_per_chunk": 4.672, "boundary_entropy": 0.67028, "router_noise": 0.0, "grad_norm": 0.839, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 343.1} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "global", "lm_loss": 1.39842, "bpb": 2.0175, "chunk_loss": 0.033113, "F": 0.21835, "G": 0.44236, "chunks": 223.59, "bytes_per_chunk": 4.58, "boundary_entropy": 0.67035, "router_noise": 0.0, "grad_norm": 0.798, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 345.2} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4143.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.238062, "cps": 32.367188, "mean_p": 0.434981, "boundary_entropy": 0.656825, "bpb": 2.015748}, "de": {"chunks": 4691.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.228462, "cps": 36.648438, "mean_p": 0.441814, "boundary_entropy": 0.666642, "bpb": 2.225238}, "vi": {"chunks": 4946.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.268294, "cps": 38.640625, "mean_p": 0.439252, "boundary_entropy": 0.668068, "bpb": 1.893113}, "tr": {"chunks": 5078.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.280506, "cps": 39.671875, "mean_p": 0.450418, "boundary_entropy": 0.666665, "bpb": 2.164685}, "fi": {"chunks": 4151.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.220669, "cps": 32.429688, "mean_p": 0.44506, "boundary_entropy": 0.669403, "bpb": 2.363643}, "ru": {"chunks": 6242.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.329393, "cps": 48.765625, "mean_p": 0.430773, "boundary_entropy": 0.663705, "bpb": 1.394709}, "el": {"chunks": 8080.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.393705, "cps": 63.125, "mean_p": 0.444247, "boundary_entropy": 0.673188, "bpb": 1.4264}, "hi": {"chunks": 7926.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.471085, "cps": 61.921875, "mean_p": 0.439737, "boundary_entropy": 0.668426, "bpb": 1.148743}, "ta": {"chunks": 8827.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.472158, "cps": 68.960938, "mean_p": 0.430048, "boundary_entropy": 0.661364, "bpb": 0.943812}, "th": {"chunks": 8725.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.542431, "cps": 68.164062, "mean_p": 0.431588, "boundary_entropy": 0.668429, "bpb": 1.077694}, "zh": {"chunks": 3193.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.540637, "cps": 24.945312, "mean_p": 0.460971, "boundary_entropy": 0.677229, "bpb": 3.111338}, "my": {"chunks": 9775.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.515886, "cps": 76.367188, "mean_p": 0.465821, "boundary_entropy": 0.68483, "bpb": 1.083693}}, "bpb": {"en": 2.01575, "de": 2.22524, "vi": 1.89311, "tr": 2.16469, "fi": 2.36364, "ru": 1.39471, "el": 1.4264, "hi": 1.14874, "ta": 0.94381, "th": 1.07769, "zh": 3.11134, "my": 1.08369}, "gini_cpc": 0.18419, "gini_cps": 0.190939, "bpic": 5.064611, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9596706497405335, "vi": 1.1269902451948222, "tr": 1.1782876670438764, "fi": 0.9269366133816891, "ru": 1.3836420016087154, "el": 1.6537874904343381, "hi": 1.978828615480954, "ta": 1.9833385074132006, "th": 2.2785237369270823, "zh": 2.270987076557597, "my": 2.1670182902343194}, "t": 348.2} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.02805, "de": 2.23469, "vi": 1.89366, "tr": 2.17162, "fi": 2.37391, "ru": 1.37392, "el": 1.41977, "hi": 1.14405, "ta": 0.95208, "th": 1.09389, "zh": 3.10911, "my": 1.09176}, "bytes_per_chunk": {"en": 4.215, "de": 4.343, "vi": 4.633, "tr": 3.721, "fi": 4.709, "ru": 5.351, "el": 4.512, "hi": 5.576, "ta": 5.525, "th": 5.215, "zh": 5.2, "my": 5.565}, "bpb_mean": 1.74054, "t": 349.5} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4143.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.238062, "cps": 32.367188, "mean_p": 0.434981, "boundary_entropy": 0.656825, "bpb": 2.015748}, "de": {"chunks": 4691.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.228462, "cps": 36.648438, "mean_p": 0.441814, "boundary_entropy": 0.666642, "bpb": 2.225238}, "vi": {"chunks": 4946.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.268294, "cps": 38.640625, "mean_p": 0.439252, "boundary_entropy": 0.668068, "bpb": 1.893113}, "tr": {"chunks": 5078.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.280506, "cps": 39.671875, "mean_p": 0.450418, "boundary_entropy": 0.666665, "bpb": 2.164685}, "fi": {"chunks": 4151.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.220669, "cps": 32.429688, "mean_p": 0.44506, "boundary_entropy": 0.669403, "bpb": 2.363643}, "ru": {"chunks": 6242.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.329393, "cps": 48.765625, "mean_p": 0.430773, "boundary_entropy": 0.663705, "bpb": 1.394709}, "el": {"chunks": 8080.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.393705, "cps": 63.125, "mean_p": 0.444247, "boundary_entropy": 0.673188, "bpb": 1.4264}, "hi": {"chunks": 7926.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.471085, "cps": 61.921875, "mean_p": 0.439737, "boundary_entropy": 0.668426, "bpb": 1.148743}, "ta": {"chunks": 8827.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.472158, "cps": 68.960938, "mean_p": 0.430048, "boundary_entropy": 0.661364, "bpb": 0.943812}, "th": {"chunks": 8725.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.542431, "cps": 68.164062, "mean_p": 0.431588, "boundary_entropy": 0.668429, "bpb": 1.077694}, "zh": {"chunks": 3193.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.540637, "cps": 24.945312, "mean_p": 0.460971, "boundary_entropy": 0.677229, "bpb": 3.111338}, "my": {"chunks": 9775.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.515886, "cps": 76.367188, "mean_p": 0.465821, "boundary_entropy": 0.68483, "bpb": 1.083693}}, "bpb": {"en": 2.01575, "de": 2.22524, "vi": 1.89311, "tr": 2.16469, "fi": 2.36364, "ru": 1.39471, "el": 1.4264, "hi": 1.14874, "ta": 0.94381, "th": 1.07769, "zh": 3.11134, "my": 1.08369}, "gini_cpc": 0.18419, "gini_cps": 0.190939, "bpic": 5.064611, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9596706497405335, "vi": 1.1269902451948222, "tr": 1.1782876670438764, "fi": 0.9269366133816891, "ru": 1.3836420016087154, "el": 1.6537874904343381, "hi": 1.978828615480954, "ta": 1.9833385074132006, "th": 2.2785237369270823, "zh": 2.270987076557597, "my": 2.1670182902343194}, "t": 350.5} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/emactl_genreg_s2/th0.70"], "t": 355.5} |
| {"event": "final", "run": "emactl_genreg_s2", "condition": "A", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.23806240303395967, "cps": 32.3671875, "bpb": 2.0372584237188307, "bytes_per_chunk": 4.226799635598481}, "de": {"cpc": 0.22846150099839282, "cps": 36.6484375, "bpb": 2.226351327546341, "bytes_per_chunk": 4.347402112804524}, "vi": {"cpc": 0.2682940059669108, "cps": 38.640625, "bpb": 1.8961533875071113, "bytes_per_chunk": 4.608639088623618}, "tr": {"cpc": 0.2805059934817434, "cps": 39.671875, "bpb": 2.1642174461915413, "bytes_per_chunk": 3.7224202320264688}, "fi": {"cpc": 0.22066875764180532, "cps": 32.4296875, "bpb": 2.3910613802988614, "bytes_per_chunk": 4.665106553365662}, "ru": {"cpc": 0.32939313984168866, "cps": 48.765625, "bpb": 1.3742821696914036, "bytes_per_chunk": 5.337351114730734}, "el": {"cpc": 0.39370462408030016, "cps": 63.125, "bpb": 1.4151377179209739, "bytes_per_chunk": 4.5019878581794135}, "hi": {"cpc": 0.4710846953937593, "cps": 61.921875, "bpb": 1.1513827221751163, "bytes_per_chunk": 5.574092581173319}, "ta": {"cpc": 0.4721583311045734, "cps": 68.9609375, "bpb": 0.9525330490836764, "bytes_per_chunk": 5.550194255952066}, "th": {"cpc": 0.542430836182779, "cps": 68.1640625, "bpb": 1.090443856895379, "bytes_per_chunk": 5.2221480721535505}, "zh": {"cpc": 0.5406366407043685, "cps": 24.9453125, "bpb": 3.1167576970487225, "bytes_per_chunk": 5.181940381118052}, "my": {"cpc": 0.5158855815917247, "cps": 76.3671875, "bpb": 1.1004410095350938, "bytes_per_chunk": 5.546905913096838}}, "gini_cpc": 0.18418976005934495, "gini_cps": 0.19093854335748328, "bpic": 5.064610633833485, "seed": 2, "mean_boundary_entropy": 0.6683645248413086, "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": 357.9} |
|
|
