Download log.jsonl from AdaptiveChunking/specfull_A_s0: direct link, hf CLI and curl.
- Browser
- Download file 151 kB
-
https://huggingface.co/AdaptiveChunking/specfull_A_s0/resolve/main/log.jsonl
- Command line
-
hf download hf://AdaptiveChunking/specfull_A_s0/log.jsonl
-
curl -L -o log.jsonl https://huggingface.co/AdaptiveChunking/specfull_A_s0/resolve/main/log.jsonl
151 kB
| {"event": "start", "run": "specfull_A_s0", "condition": "A", "params": 82470528, "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": 2.3} | |
| {"event": "probe_schedule", "n": 30, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 4000, 8000, 12000, 16000, 20000, 24000, 28000, 32000, 36000, 40000, 44000, 48000, 52000, 56000, 60000, 64000, 68000, 72000, 76000, 76300], "t": 2.3} | |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8739.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.502155, "cps": 68.273438, "mean_p": 0.50363, "boundary_entropy": 0.687415, "bpb": 8.248358}, "de": {"chunks": 12222.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.595237, "cps": 95.484375, "mean_p": 0.507356, "boundary_entropy": 0.688163, "bpb": 8.204506}, "vi": {"chunks": 6386.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.346406, "cps": 49.890625, "mean_p": 0.491233, "boundary_entropy": 0.688507, "bpb": 8.13873}, "tr": {"chunks": 11403.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.629896, "cps": 89.085938, "mean_p": 0.50659, "boundary_entropy": 0.687851, "bpb": 8.143105}, "fi": {"chunks": 7137.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.379406, "cps": 55.757812, "mean_p": 0.497164, "boundary_entropy": 0.68792, "bpb": 8.234202}, "ru": {"chunks": 10473.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.552665, "cps": 81.820312, "mean_p": 0.496692, "boundary_entropy": 0.690192, "bpb": 8.071624}, "el": {"chunks": 36555.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.781172, "cps": 285.585938, "mean_p": 0.530217, "boundary_entropy": 0.688772, "bpb": 8.142597}, "hi": {"chunks": 42584.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.530996, "cps": 332.6875, "mean_p": 0.521073, "boundary_entropy": 0.69018, "bpb": 7.880816}, "ta": {"chunks": 49758.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.661567, "cps": 388.734375, "mean_p": 0.521091, "boundary_entropy": 0.690409, "bpb": 7.814466}, "th": {"chunks": 45716.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.842151, "cps": 357.15625, "mean_p": 0.525713, "boundary_entropy": 0.689893, "bpb": 8.063302}, "zh": {"chunks": 6218.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.052828, "cps": 48.578125, "mean_p": 0.496822, "boundary_entropy": 0.686552, "bpb": 8.034403}, "my": {"chunks": 53266.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.811167, "cps": 416.140625, "mean_p": 0.521325, "boundary_entropy": 0.690575, "bpb": 8.24483}}, "bpb": {"en": 8.24836, "de": 8.20451, "vi": 8.13873, "tr": 8.1431, "fi": 8.2342, "ru": 8.07162, "el": 8.1426, "hi": 7.88082, "ta": 7.81447, "th": 8.0633, "zh": 8.0344, "my": 8.24483}, "gini_cpc": 0.391191, "gini_cps": 0.409188, "bpic": 1.321301, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1853654183968407, "vi": 0.6898396505624507, "tr": 1.2543852937189204, "fi": 0.755555191784985, "ru": 1.1005867247811874, "el": 3.54705827773647, "hi": 5.040269530092341, "ta": 5.300292378412309, "th": 5.659910185771798, "zh": 2.096619669942964, "my": 5.598208760550946}, "t": 8.9} | |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.63905, "bpb": 8.13544, "chunk_loss": 0.057711, "F": 0.53552, "G": 0.50248, "chunks": 548.38, "bytes_per_chunk": 1.867, "boundary_entropy": 0.69172, "router_noise": 0.0, "grad_norm": 9.2839, "lr": 1e-07, "router_trainable": true, "sec_per_step": 0.001, "alpha": 0.03, "t": 9.6} | |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8736.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.501982, "cps": 68.25, "mean_p": 0.503619, "boundary_entropy": 0.687415, "bpb": 8.238673}, "de": {"chunks": 12217.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.594993, "cps": 95.445312, "mean_p": 0.50734, "boundary_entropy": 0.688163, "bpb": 8.195138}, "vi": {"chunks": 6383.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.346244, "cps": 49.867188, "mean_p": 0.491224, "boundary_entropy": 0.688506, "bpb": 8.133712}, "tr": {"chunks": 11396.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.629509, "cps": 89.03125, "mean_p": 0.506574, "boundary_entropy": 0.687851, "bpb": 8.136467}, "fi": {"chunks": 7121.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.378555, "cps": 55.632812, "mean_p": 0.497134, "boundary_entropy": 0.687919, "bpb": 8.227025}, "ru": {"chunks": 10480.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.553034, "cps": 81.875, "mean_p": 0.496706, "boundary_entropy": 0.690192, "bpb": 8.060737}, "el": {"chunks": 36553.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.781075, "cps": 285.570312, "mean_p": 0.530186, "boundary_entropy": 0.688776, "bpb": 8.137917}, "hi": {"chunks": 42579.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.530698, "cps": 332.648438, "mean_p": 0.521058, "boundary_entropy": 0.690181, "bpb": 7.874979}, "ta": {"chunks": 49741.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.660658, "cps": 388.601562, "mean_p": 0.521036, "boundary_entropy": 0.690413, "bpb": 7.805915}, "th": {"chunks": 45716.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.842151, "cps": 357.15625, "mean_p": 0.52568, "boundary_entropy": 0.689897, "bpb": 8.054953}, "zh": {"chunks": 6218.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.052828, "cps": 48.578125, "mean_p": 0.496817, "boundary_entropy": 0.686551, "bpb": 8.033335}, "my": {"chunks": 53266.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.811167, "cps": 416.140625, "mean_p": 0.521287, "boundary_entropy": 0.690578, "bpb": 8.23821}}, "bpb": {"en": 8.23867, "de": 8.19514, "vi": 8.13371, "tr": 8.13647, "fi": 8.22702, "ru": 8.06074, "el": 8.13792, "hi": 7.87498, "ta": 7.80592, "th": 8.05495, "zh": 8.03333, "my": 8.23821}, "gini_cpc": 0.391265, "gini_cps": 0.409246, "bpic": 1.321533, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1852873831350552, "vi": 0.6897523635166029, "tr": 1.2540457595213714, "fi": 0.7541202401739455, "ru": 1.1017005422018615, "el": 3.5480822273643136, "hi": 5.0414083854481335, "ta": 5.300301047909032, "th": 5.661853836247682, "zh": 2.097339662961488, "my": 5.600131222350586}, "t": 15.9} | |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8728.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.501523, "cps": 68.1875, "mean_p": 0.503592, "boundary_entropy": 0.687414, "bpb": 8.219769}, "de": {"chunks": 12202.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.594263, "cps": 95.328125, "mean_p": 0.507296, "boundary_entropy": 0.688163, "bpb": 8.177084}, "vi": {"chunks": 6372.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.345647, "cps": 49.78125, "mean_p": 0.491198, "boundary_entropy": 0.688504, "bpb": 8.124028}, "tr": {"chunks": 11376.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.628404, "cps": 88.875, "mean_p": 0.506531, "boundary_entropy": 0.687851, "bpb": 8.123315}, "fi": {"chunks": 7093.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.377067, "cps": 55.414062, "mean_p": 0.497066, "boundary_entropy": 0.687917, "bpb": 8.2125}, "ru": {"chunks": 10503.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.554248, "cps": 82.054688, "mean_p": 0.496721, "boundary_entropy": 0.690193, "bpb": 8.040745}, "el": {"chunks": 36548.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.780831, "cps": 285.53125, "mean_p": 0.5301, "boundary_entropy": 0.688785, "bpb": 8.129253}, "hi": {"chunks": 42566.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.529926, "cps": 332.546875, "mean_p": 0.521007, "boundary_entropy": 0.690185, "bpb": 7.862052}, "ta": {"chunks": 49714.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.659214, "cps": 388.390625, "mean_p": 0.520914, "boundary_entropy": 0.690423, "bpb": 7.789935}, "th": {"chunks": 45716.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.842151, "cps": 357.15625, "mean_p": 0.525612, "boundary_entropy": 0.689903, "bpb": 8.034908}, "zh": {"chunks": 6211.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.051642, "cps": 48.523438, "mean_p": 0.496801, "boundary_entropy": 0.686551, "bpb": 8.030337}, "my": {"chunks": 53264.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.811062, "cps": 416.125, "mean_p": 0.521249, "boundary_entropy": 0.690581, "bpb": 8.226986}}, "bpb": {"en": 8.21977, "de": 8.17708, "vi": 8.12403, "tr": 8.12332, "fi": 8.2125, "ru": 8.04074, "el": 8.12925, "hi": 7.86205, "ta": 7.78993, "th": 8.03491, "zh": 8.03034, "my": 8.22699}, "gini_cpc": 0.391445, "gini_cps": 0.409406, "bpic": 1.322047, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.18491717944104, "vi": 0.6891948250311931, "tr": 1.2529923358915804, "fi": 0.7518435166330365, "ru": 1.1051304206399686, "el": 3.5508485862790677, "hi": 5.044488663834724, "ta": 5.302279549495507, "th": 5.6670434364642235, "zh": 2.0968987906471166, "my": 5.605053785099922}, "t": 22.7} | |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8677.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.498592, "cps": 67.789062, "mean_p": 0.503457, "boundary_entropy": 0.687413, "bpb": 8.15468}, "de": {"chunks": 12101.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.589344, "cps": 94.539062, "mean_p": 0.50711, "boundary_entropy": 0.688165, "bpb": 8.11549}, "vi": {"chunks": 6333.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.343531, "cps": 49.476562, "mean_p": 0.491108, "boundary_entropy": 0.688499, "bpb": 8.090236}, "tr": {"chunks": 11317.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.625145, "cps": 88.414062, "mean_p": 0.506349, "boundary_entropy": 0.687853, "bpb": 8.077874}, "fi": {"chunks": 6996.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.37191, "cps": 54.65625, "mean_p": 0.496807, "boundary_entropy": 0.68791, "bpb": 8.163408}, "ru": {"chunks": 10500.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.55409, "cps": 82.03125, "mean_p": 0.496719, "boundary_entropy": 0.690193, "bpb": 7.980457}, "el": {"chunks": 36519.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.779418, "cps": 285.304688, "mean_p": 0.529727, "boundary_entropy": 0.688827, "bpb": 8.093503}, "hi": {"chunks": 42533.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.527964, "cps": 332.289062, "mean_p": 0.520796, "boundary_entropy": 0.690201, "bpb": 7.815139}, "ta": {"chunks": 49593.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.652741, "cps": 387.445312, "mean_p": 0.520486, "boundary_entropy": 0.690457, "bpb": 7.724832}, "th": {"chunks": 45710.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.841778, "cps": 357.109375, "mean_p": 0.525378, "boundary_entropy": 0.689926, "bpb": 7.96705}, "zh": {"chunks": 6199.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.049611, "cps": 48.429688, "mean_p": 0.496758, "boundary_entropy": 0.686551, "bpb": 8.021791}, "my": {"chunks": 53248.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.810217, "cps": 416.0, "mean_p": 0.52105, "boundary_entropy": 0.690597, "bpb": 8.18544}}, "bpb": {"en": 8.15468, "de": 8.11549, "vi": 8.09024, "tr": 8.07787, "fi": 8.16341, "ru": 7.98046, "el": 8.0935, "hi": 7.81514, "ta": 7.72483, "th": 7.96705, "zh": 8.02179, "my": 8.18544}, "gini_cpc": 0.392368, "gini_cps": 0.410157, "bpic": 1.324634, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.182016057814684, "vi": 0.689002612801367, "tr": 1.2538202716937425, "fi": 0.7459203233719807, "ru": 1.111308426760097, "el": 3.5688850032523094, "hi": 5.070204377989599, "ta": 5.320463079853491, "th": 5.699603950131098, "zh": 2.10514840058305, "my": 5.636304489977106}, "t": 29.9} | |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 8470.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.486698, "cps": 66.171875, "mean_p": 0.502942, "boundary_entropy": 0.687406, "bpb": 7.911953}, "de": {"chunks": 11726.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.571081, "cps": 91.609375, "mean_p": 0.506407, "boundary_entropy": 0.688172, "bpb": 7.885978}, "vi": {"chunks": 6200.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.336317, "cps": 48.4375, "mean_p": 0.490766, "boundary_entropy": 0.688479, "bpb": 7.966102}, "tr": {"chunks": 11064.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.611169, "cps": 86.4375, "mean_p": 0.50573, "boundary_entropy": 0.687855, "bpb": 7.906643}, "fi": {"chunks": 6566.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.349051, "cps": 51.296875, "mean_p": 0.495876, "boundary_entropy": 0.687882, "bpb": 7.97619}, "ru": {"chunks": 10789.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.56934, "cps": 84.289062, "mean_p": 0.496963, "boundary_entropy": 0.690198, "bpb": 7.747567}, "el": {"chunks": 36357.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.771525, "cps": 284.039062, "mean_p": 0.528388, "boundary_entropy": 0.688974, "bpb": 7.963861}, "hi": {"chunks": 42433.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.522021, "cps": 331.507812, "mean_p": 0.520125, "boundary_entropy": 0.690251, "bpb": 7.651222}, "ta": {"chunks": 49168.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.630008, "cps": 384.125, "mean_p": 0.519094, "boundary_entropy": 0.690562, "bpb": 7.526498}, "th": {"chunks": 45689.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.840472, "cps": 356.945312, "mean_p": 0.524564, "boundary_entropy": 0.690004, "bpb": 7.76159}, "zh": {"chunks": 6149.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.041145, "cps": 48.039062, "mean_p": 0.496602, "boundary_entropy": 0.686548, "bpb": 7.99291}, "my": {"chunks": 53200.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.807684, "cps": 415.625, "mean_p": 0.520537, "boundary_entropy": 0.690636, "bpb": 8.077284}}, "bpb": {"en": 7.91195, "de": 7.88598, "vi": 7.9661, "tr": 7.90664, "fi": 7.97619, "ru": 7.74757, "el": 7.96386, "hi": 7.65122, "ta": 7.5265, "th": 7.76159, "zh": 7.99291, "my": 8.07728}, "gini_cpc": 0.395445, "gini_cps": 0.412688, "bpic": 1.333448, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1733786789842895, "vi": 0.6910178363688239, "tr": 1.2557475096100035, "fi": 0.717182534748659, "ru": 1.169802886487463, "el": 3.6398870308838287, "hi": 5.18190413503827, "ta": 5.40378153879479, "th": 5.836215199726659, "zh": 2.139201824250927, "my": 5.76884627261736}, "t": 38.0} | |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 7676.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.441073, "cps": 59.96875, "mean_p": 0.500947, "boundary_entropy": 0.687374, "bpb": 7.191681}, "de": {"chunks": 10478.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.5103, "cps": 81.859375, "mean_p": 0.503874, "boundary_entropy": 0.688183, "bpb": 7.216396}, "vi": {"chunks": 5708.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.309628, "cps": 44.59375, "mean_p": 0.489411, "boundary_entropy": 0.688398, "bpb": 7.598114}, "tr": {"chunks": 10112.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.558581, "cps": 79.0, "mean_p": 0.503385, "boundary_entropy": 0.687851, "bpb": 7.394031}, "fi": {"chunks": 5144.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.273457, "cps": 40.1875, "mean_p": 0.492335, "boundary_entropy": 0.687743, "bpb": 7.393161}, "ru": {"chunks": 12377.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.65314, "cps": 96.695312, "mean_p": 0.498383, "boundary_entropy": 0.690226, "bpb": 7.026717}, "el": {"chunks": 35599.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.73459, "cps": 278.117188, "mean_p": 0.52423, "boundary_entropy": 0.689385, "bpb": 7.570875}, "hi": {"chunks": 41476.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.465141, "cps": 324.03125, "mean_p": 0.516873, "boundary_entropy": 0.690468, "bpb": 7.018477}, "ta": {"chunks": 45440.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.430596, "cps": 355.0, "mean_p": 0.513429, "boundary_entropy": 0.69091, "bpb": 6.877527}, "th": {"chunks": 45452.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.825738, "cps": 355.09375, "mean_p": 0.520806, "boundary_entropy": 0.690327, "bpb": 6.926662}, "zh": {"chunks": 5983.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 1.013038, "cps": 46.742188, "mean_p": 0.496051, "boundary_entropy": 0.686542, "bpb": 7.878399}, "my": {"chunks": 53003.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.797287, "cps": 414.085938, "mean_p": 0.518662, "boundary_entropy": 0.690772, "bpb": 7.677892}}, "bpb": {"en": 7.19168, "de": 7.2164, "vi": 7.59811, "tr": 7.39403, "fi": 7.39316, "ru": 7.02672, "el": 7.57088, "hi": 7.01848, "ta": 6.87753, "th": 6.92666, "zh": 7.8784, "my": 7.67789}, "gini_cpc": 0.406565, "gini_cps": 0.421295, "bpic": 1.378286, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1569514669594647, "vi": 0.7019884661585083, "tr": 1.266413885433584, "fi": 0.6199807863576102, "ru": 1.4807963346674897, "el": 3.9326573432006797, "hi": 5.588959300403333, "ta": 5.510639581856301, "th": 6.406503782423463, "zh": 2.296755231826318, "my": 6.3419998830072}, "t": 47.4} | |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 4613.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.265069, "cps": 36.039062, "mean_p": 0.492821, "boundary_entropy": 0.68719, "bpb": 6.261184}, "de": {"chunks": 6466.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.314908, "cps": 50.515625, "mean_p": 0.495452, "boundary_entropy": 0.688144, "bpb": 6.33878}, "vi": {"chunks": 3135.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.170057, "cps": 24.492188, "mean_p": 0.480249, "boundary_entropy": 0.687716, "bpb": 6.963518}, "tr": {"chunks": 6246.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.345026, "cps": 48.796875, "mean_p": 0.493741, "boundary_entropy": 0.687704, "bpb": 6.645685}, "fi": {"chunks": 2724.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.144809, "cps": 21.28125, "mean_p": 0.484457, "boundary_entropy": 0.687308, "bpb": 6.520236}, "ru": {"chunks": 13227.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.697995, "cps": 103.335938, "mean_p": 0.498914, "boundary_entropy": 0.690282, "bpb": 5.792033}, "el": {"chunks": 24251.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.18165, "cps": 189.460938, "mean_p": 0.506777, "boundary_entropy": 0.690354, "bpb": 6.653742}, "hi": {"chunks": 29968.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.781159, "cps": 234.125, "mean_p": 0.505577, "boundary_entropy": 0.6909, "bpb": 5.758387}, "ta": {"chunks": 13351.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.714148, "cps": 104.304688, "mean_p": 0.49495, "boundary_entropy": 0.691147, "bpb": 5.860254}, "th": {"chunks": 27253.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.694311, "cps": 212.914062, "mean_p": 0.50315, "boundary_entropy": 0.691081, "bpb": 5.277064}, "zh": {"chunks": 5237.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.886725, "cps": 40.914062, "mean_p": 0.49343, "boundary_entropy": 0.686472, "bpb": 7.508629}, "my": {"chunks": 46799.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 2.469865, "cps": 365.617188, "mean_p": 0.508749, "boundary_entropy": 0.691256, "bpb": 5.96393}}, "bpb": {"en": 6.26118, "de": 6.33878, "vi": 6.96352, "tr": 6.64569, "fi": 6.52024, "ru": 5.79203, "el": 6.65374, "hi": 5.75839, "ta": 5.86025, "th": 5.27706, "zh": 7.50863, "my": 5.96393}, "gini_cpc": 0.43797, "gini_cps": 0.458317, "bpic": 2.094074, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1880205656053515, "vi": 0.6415567354226046, "tr": 1.3016436201072554, "fi": 0.5463058931519134, "ru": 2.6332543168411857, "el": 4.457891274327432, "hi": 6.719598936914831, "ta": 2.6941947901574945, "th": 6.39195805720221, "zh": 3.3452593779164275, "my": 9.317810262239037}, "t": 59.3} | |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6083.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.349537, "cps": 47.523438, "mean_p": 0.498485, "boundary_entropy": 0.687569, "bpb": 5.730191}, "de": {"chunks": 7677.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.373886, "cps": 59.976562, "mean_p": 0.499043, "boundary_entropy": 0.688481, "bpb": 5.858665}, "vi": {"chunks": 4695.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.254679, "cps": 36.679688, "mean_p": 0.488073, "boundary_entropy": 0.688449, "bpb": 6.351775}, "tr": {"chunks": 5648.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.311992, "cps": 44.125, "mean_p": 0.494292, "boundary_entropy": 0.68796, "bpb": 6.228498}, "fi": {"chunks": 4800.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.25517, "cps": 37.5, "mean_p": 0.493728, "boundary_entropy": 0.688013, "bpb": 5.929793}, "ru": {"chunks": 7155.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.377573, "cps": 55.898438, "mean_p": 0.494773, "boundary_entropy": 0.690308, "bpb": 5.075876}, "el": {"chunks": 4000.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.194903, "cps": 31.25, "mean_p": 0.487811, "boundary_entropy": 0.690114, "bpb": 5.444692}, "hi": {"chunks": 6907.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.41052, "cps": 53.960938, "mean_p": 0.493593, "boundary_entropy": 0.690817, "bpb": 4.871888}, "ta": {"chunks": 18104.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.968387, "cps": 141.4375, "mean_p": 0.49791, "boundary_entropy": 0.691218, "bpb": 4.73351}, "th": {"chunks": 8211.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.510476, "cps": 64.148438, "mean_p": 0.493981, "boundary_entropy": 0.690964, "bpb": 4.71539}, "zh": {"chunks": 2776.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.47003, "cps": 21.6875, "mean_p": 0.485071, "boundary_entropy": 0.686178, "bpb": 6.97237}, "my": {"chunks": 9601.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.506703, "cps": 75.007812, "mean_p": 0.494446, "boundary_entropy": 0.691297, "bpb": 4.587654}}, "bpb": {"en": 5.73019, "de": 5.85867, "vi": 6.35177, "tr": 6.2285, "fi": 5.92979, "ru": 5.07588, "el": 5.44469, "hi": 4.87189, "ta": 4.73351, "th": 4.71539, "zh": 6.97237, "my": 4.58765}, "gini_cpc": 0.226532, "gini_cps": 0.256215, "bpic": 4.480439, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.069659215636639, "vi": 0.7286160914509576, "tr": 0.8925867596246727, "fi": 0.7300215115599192, "ru": 1.0802063538812476, "el": 0.5576034471062402, "hi": 1.1744666438194378, "ta": 2.7704822699382268, "th": 1.4604318327584365, "zh": 1.3447214202849964, "my": 1.4496374409855142}, "t": 73.5} | |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5064.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.290984, "cps": 39.5625, "mean_p": 0.491169, "boundary_entropy": 0.686773, "bpb": 5.283032}, "de": {"chunks": 5606.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.273024, "cps": 43.796875, "mean_p": 0.489581, "boundary_entropy": 0.687605, "bpb": 5.443113}, "vi": {"chunks": 6091.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.330404, "cps": 47.585938, "mean_p": 0.491168, "boundary_entropy": 0.688615, "bpb": 5.811832}, "tr": {"chunks": 3544.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.195769, "cps": 27.6875, "mean_p": 0.485616, "boundary_entropy": 0.687278, "bpb": 5.737864}, "fi": {"chunks": 3474.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184679, "cps": 27.140625, "mean_p": 0.486717, "boundary_entropy": 0.687439, "bpb": 5.445679}, "ru": {"chunks": 9737.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.513826, "cps": 76.070312, "mean_p": 0.49597, "boundary_entropy": 0.690304, "bpb": 4.542408}, "el": {"chunks": 7492.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.365054, "cps": 58.53125, "mean_p": 0.494486, "boundary_entropy": 0.690477, "bpb": 4.998417}, "hi": {"chunks": 8913.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.529747, "cps": 69.632812, "mean_p": 0.490542, "boundary_entropy": 0.690562, "bpb": 4.109309}, "ta": {"chunks": 13232.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.707783, "cps": 103.375, "mean_p": 0.478918, "boundary_entropy": 0.689405, "bpb": 4.124594}, "th": {"chunks": 9944.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.618216, "cps": 77.6875, "mean_p": 0.483689, "boundary_entropy": 0.690005, "bpb": 4.044175}, "zh": {"chunks": 3318.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.561802, "cps": 25.921875, "mean_p": 0.49128, "boundary_entropy": 0.686829, "bpb": 6.647426}, "my": {"chunks": 9335.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.492664, "cps": 72.929688, "mean_p": 0.493861, "boundary_entropy": 0.691274, "bpb": 4.138422}}, "bpb": {"en": 5.28303, "de": 5.44311, "vi": 5.81183, "tr": 5.73786, "fi": 5.44568, "ru": 4.54241, "el": 4.99842, "hi": 4.10931, "ta": 4.12459, "th": 4.04418, "zh": 6.64743, "my": 4.13842}, "gini_cpc": 0.221324, "gini_cps": 0.236455, "bpic": 4.47558, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9382770839589397, "vi": 1.1354705658538968, "tr": 0.672780793840826, "fi": 0.6346705605598821, "ru": 1.7658197864172365, "el": 1.254548185794553, "hi": 1.820535939381562, "ta": 2.432374522929186, "th": 2.1245672059129004, "zh": 1.9306936234819436, "my": 1.6930951646661803}, "t": 94.5} | |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3616.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20778, "cps": 28.25, "mean_p": 0.450634, "boundary_entropy": 0.620403, "bpb": 4.182081}, "de": {"chunks": 3840.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187016, "cps": 30.0, "mean_p": 0.447537, "boundary_entropy": 0.633493, "bpb": 4.424519}, "vi": {"chunks": 7138.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.387198, "cps": 55.765625, "mean_p": 0.493326, "boundary_entropy": 0.654091, "bpb": 4.15127}, "tr": {"chunks": 3705.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.204662, "cps": 28.945312, "mean_p": 0.473238, "boundary_entropy": 0.661518, "bpb": 4.510791}, "fi": {"chunks": 3278.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.17426, "cps": 25.609375, "mean_p": 0.46232, "boundary_entropy": 0.660283, "bpb": 4.43768}, "ru": {"chunks": 5793.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305699, "cps": 45.257812, "mean_p": 0.478711, "boundary_entropy": 0.685781, "bpb": 3.107385}, "el": {"chunks": 6031.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.293865, "cps": 47.117188, "mean_p": 0.473112, "boundary_entropy": 0.687889, "bpb": 3.338712}, "hi": {"chunks": 8251.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.490401, "cps": 64.460938, "mean_p": 0.468385, "boundary_entropy": 0.687198, "bpb": 2.578231}, "ta": {"chunks": 11405.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.610056, "cps": 89.101562, "mean_p": 0.431756, "boundary_entropy": 0.675862, "bpb": 2.576375}, "th": {"chunks": 9931.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.617408, "cps": 77.585938, "mean_p": 0.455479, "boundary_entropy": 0.684343, "bpb": 2.545263}, "zh": {"chunks": 2181.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.369285, "cps": 17.039062, "mean_p": 0.482788, "boundary_entropy": 0.685935, "bpb": 5.909729}, "my": {"chunks": 10192.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.537893, "cps": 79.625, "mean_p": 0.45574, "boundary_entropy": 0.684889, "bpb": 2.646345}}, "bpb": {"en": 4.18208, "de": 4.42452, "vi": 4.15127, "tr": 4.51079, "fi": 4.43768, "ru": 3.10738, "el": 3.33871, "hi": 2.57823, "ta": 2.57638, "th": 2.54526, "zh": 5.90973, "my": 2.64634}, "gini_cpc": 0.242679, "gini_cps": 0.266226, "bpic": 5.092568, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9000663296597017, "vi": 1.863498725490902, "tr": 0.9849934879761287, "fi": 0.8386731979452806, "ru": 1.4712619813435448, "el": 1.4143085984039492, "hi": 2.360191340451551, "ta": 2.936064003081606, "th": 2.9714444454378155, "zh": 1.7772884613367537, "my": 2.5887596888363307}, "t": 124.1} | |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.2641, "bpb": 3.26641, "chunk_loss": 0.030662, "F": 0.17868, "G": 0.40871, "chunks": 182.97, "bytes_per_chunk": 5.597, "boundary_entropy": 0.58401, "router_noise": 0.0, "grad_norm": 2.2063, "lr": 7.51e-05, "router_trainable": true, "sec_per_step": 0.319, "alpha": 0.03, "t": 169.2} | |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.75312, "bpb": 2.52922, "chunk_loss": 0.030596, "F": 0.18941, "G": 0.3049, "chunks": 193.95, "bytes_per_chunk": 5.28, "boundary_entropy": 0.39652, "router_noise": 0.0, "grad_norm": 1.7539, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.175, "alpha": 0.03, "t": 256.5} | |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.58658, "bpb": 2.28894, "chunk_loss": 0.030133, "F": 0.17839, "G": 0.26761, "chunks": 182.67, "bytes_per_chunk": 5.606, "boundary_entropy": 0.34283, "router_noise": 0.0, "grad_norm": 0.8669, "lr": 0.0002251, "router_trainable": true, "sec_per_step": 0.172, "alpha": 0.03, "t": 342.5} | |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.47215, "bpb": 2.12387, "chunk_loss": 0.030252, "F": 0.1869, "G": 0.25359, "chunks": 191.39, "bytes_per_chunk": 5.35, "boundary_entropy": 0.26705, "router_noise": 0.0, "grad_norm": 0.676, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.171, "alpha": 0.03, "t": 427.9} | |
| {"event": "train", "step": 2500, "phase": 1, "objective": "global", "lm_loss": 1.34547, "bpb": 1.9411, "chunk_loss": 0.030199, "F": 0.18414, "G": 0.25545, "chunks": 188.56, "bytes_per_chunk": 5.431, "boundary_entropy": 0.2819, "router_noise": 0.0, "grad_norm": 0.4709, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.169, "alpha": 0.03, "t": 512.5} | |
| {"event": "train", "step": 3000, "phase": 1, "objective": "global", "lm_loss": 1.38452, "bpb": 1.99744, "chunk_loss": 0.030305, "F": 0.19083, "G": 0.24013, "chunks": 195.41, "bytes_per_chunk": 5.24, "boundary_entropy": 0.22129, "router_noise": 0.0, "grad_norm": 0.4379, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.169, "alpha": 0.03, "t": 597.1} | |
| {"event": "train", "step": 3500, "phase": 1, "objective": "global", "lm_loss": 1.36335, "bpb": 1.9669, "chunk_loss": 0.030311, "F": 0.18503, "G": 0.23102, "chunks": 189.47, "bytes_per_chunk": 5.405, "boundary_entropy": 0.20998, "router_noise": 0.0, "grad_norm": 0.4005, "lr": 0.0002997, "router_trainable": true, "sec_per_step": 0.169, "alpha": 0.03, "t": 681.4} | |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3533.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203011, "cps": 27.601562, "mean_p": 0.204943, "boundary_entropy": 0.092659, "bpb": 1.916734}, "de": {"chunks": 3846.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187308, "cps": 30.046875, "mean_p": 0.197576, "boundary_entropy": 0.102853, "bpb": 2.099293}, "vi": {"chunks": 4226.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229238, "cps": 33.015625, "mean_p": 0.190474, "boundary_entropy": 0.120147, "bpb": 1.768984}, "tr": {"chunks": 3173.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.175275, "cps": 24.789062, "mean_p": 0.183536, "boundary_entropy": 0.152933, "bpb": 2.071698}, "fi": {"chunks": 3213.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.170804, "cps": 25.101562, "mean_p": 0.186639, "boundary_entropy": 0.137235, "bpb": 2.26145}, "ru": {"chunks": 5547.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.292718, "cps": 43.335938, "mean_p": 0.212896, "boundary_entropy": 0.27392, "bpb": 1.297492}, "el": {"chunks": 5581.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.271939, "cps": 43.601562, "mean_p": 0.280402, "boundary_entropy": 0.411949, "bpb": 1.337307}, "hi": {"chunks": 8706.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.517444, "cps": 68.015625, "mean_p": 0.418129, "boundary_entropy": 0.591514, "bpb": 1.095842}, "ta": {"chunks": 9174.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.490719, "cps": 71.671875, "mean_p": 0.395765, "boundary_entropy": 0.594475, "bpb": 0.92768}, "th": {"chunks": 7490.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.465651, "cps": 58.515625, "mean_p": 0.394672, "boundary_entropy": 0.584295, "bpb": 1.055812}, "zh": {"chunks": 2538.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.429732, "cps": 19.828125, "mean_p": 0.249808, "boundary_entropy": 0.355543, "bpb": 2.991654}, "my": {"chunks": 8894.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.46939, "cps": 69.484375, "mean_p": 0.37033, "boundary_entropy": 0.561716, "bpb": 1.075558}}, "bpb": {"en": 1.91673, "de": 2.09929, "vi": 1.76898, "tr": 2.0717, "fi": 2.26145, "ru": 1.29749, "el": 1.33731, "hi": 1.09584, "ta": 0.92768, "th": 1.05581, "zh": 2.99165, "my": 1.07556}, "gini_cpc": 0.225499, "gini_cps": 0.239235, "bpic": 5.821832, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9226507848321704, "vi": 1.1291895062639665, "tr": 0.863376062431319, "fi": 0.8413550869500313, "ru": 1.4418810506591824, "el": 1.339527580765529, "hi": 2.5488487952058048, "ta": 2.417206475725341, "th": 2.2937244037104394, "zh": 2.1167943023588056, "my": 2.312140557669683}, "t": 768.4} | |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.92101, "de": 2.09712, "vi": 1.77852, "tr": 2.06449, "fi": 2.27753, "ru": 1.29456, "el": 1.33473, "hi": 1.10498, "ta": 0.92958, "th": 1.05306, "zh": 2.99921, "my": 1.08317}, "bytes_per_chunk": {"en": 4.725, "de": 5.263, "vi": 5.583, "tr": 5.952, "fi": 5.685, "ru": 6.055, "el": 6.444, "hi": 4.932, "ta": 5.405, "th": 6.081, "zh": 6.392, "my": 6.039}, "bpb_mean": 1.6615, "t": 773.7} | |
| {"event": "train", "step": 4000, "phase": 1, "objective": "global", "lm_loss": 1.34264, "bpb": 1.93703, "chunk_loss": 0.030242, "F": 0.18079, "G": 0.24791, "chunks": 185.12, "bytes_per_chunk": 5.531, "boundary_entropy": 0.25758, "router_noise": 0.0, "grad_norm": 0.3896, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.185, "alpha": 0.03, "t": 773.9} | |
| {"event": "train", "step": 4500, "phase": 1, "objective": "global", "lm_loss": 1.26307, "bpb": 1.82222, "chunk_loss": 0.029961, "F": 0.17447, "G": 0.23984, "chunks": 178.66, "bytes_per_chunk": 5.732, "boundary_entropy": 0.24158, "router_noise": 0.0, "grad_norm": 0.381, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 858.0} | |
| {"event": "train", "step": 5000, "phase": 1, "objective": "global", "lm_loss": 1.21681, "bpb": 1.75549, "chunk_loss": 0.030145, "F": 0.18097, "G": 0.24407, "chunks": 185.31, "bytes_per_chunk": 5.526, "boundary_entropy": 0.24667, "router_noise": 0.0, "grad_norm": 0.3667, "lr": 0.0002989, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 942.0} | |
| {"event": "train", "step": 5500, "phase": 1, "objective": "global", "lm_loss": 1.18393, "bpb": 1.70805, "chunk_loss": 0.030086, "F": 0.18015, "G": 0.25069, "chunks": 184.47, "bytes_per_chunk": 5.551, "boundary_entropy": 0.2513, "router_noise": 0.0, "grad_norm": 0.365, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 1025.7} | |
| {"event": "train", "step": 6000, "phase": 1, "objective": "global", "lm_loss": 1.14481, "bpb": 1.65161, "chunk_loss": 0.030217, "F": 0.18796, "G": 0.25596, "chunks": 192.47, "bytes_per_chunk": 5.32, "boundary_entropy": 0.23389, "router_noise": 0.0, "grad_norm": 0.3485, "lr": 0.0002981, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 1109.3} | |
| {"event": "train", "step": 6500, "phase": 1, "objective": "global", "lm_loss": 1.08108, "bpb": 1.55967, "chunk_loss": 0.030099, "F": 0.18037, "G": 0.25028, "chunks": 184.7, "bytes_per_chunk": 5.544, "boundary_entropy": 0.24018, "router_noise": 0.0, "grad_norm": 0.3186, "lr": 0.0002976, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 1193.5} | |
| {"event": "train", "step": 7000, "phase": 1, "objective": "global", "lm_loss": 1.10284, "bpb": 1.59106, "chunk_loss": 0.030145, "F": 0.18066, "G": 0.24575, "chunks": 185.0, "bytes_per_chunk": 5.535, "boundary_entropy": 0.23154, "router_noise": 0.0, "grad_norm": 0.326, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 1277.3} | |
| {"event": "train", "step": 7500, "phase": 1, "objective": "global", "lm_loss": 1.0968, "bpb": 1.58235, "chunk_loss": 0.030112, "F": 0.17993, "G": 0.2452, "chunks": 184.25, "bytes_per_chunk": 5.558, "boundary_entropy": 0.23271, "router_noise": 0.0, "grad_norm": 0.3031, "lr": 0.0002964, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 1360.9} | |
| {"event": "probe", "step": 8000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3506.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20146, "cps": 27.390625, "mean_p": 0.196644, "boundary_entropy": 0.071185, "bpb": 1.708392}, "de": {"chunks": 3760.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.18312, "cps": 29.375, "mean_p": 0.185974, "boundary_entropy": 0.062877, "bpb": 1.846673}, "vi": {"chunks": 4224.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229129, "cps": 33.0, "mean_p": 0.178394, "boundary_entropy": 0.078319, "bpb": 1.477329}, "tr": {"chunks": 3064.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.169254, "cps": 23.9375, "mean_p": 0.173715, "boundary_entropy": 0.124455, "bpb": 1.773208}, "fi": {"chunks": 3091.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.164319, "cps": 24.148438, "mean_p": 0.173646, "boundary_entropy": 0.105471, "bpb": 1.972202}, "ru": {"chunks": 5666.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.298997, "cps": 44.265625, "mean_p": 0.217537, "boundary_entropy": 0.239407, "bpb": 1.126293}, "el": {"chunks": 5953.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.290065, "cps": 46.507812, "mean_p": 0.310474, "boundary_entropy": 0.428136, "bpb": 1.157945}, "hi": {"chunks": 7829.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.465319, "cps": 61.164062, "mean_p": 0.401756, "boundary_entropy": 0.579468, "bpb": 0.955891}, "ta": {"chunks": 8182.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.437657, "cps": 63.921875, "mean_p": 0.404166, "boundary_entropy": 0.60093, "bpb": 0.813034}, "th": {"chunks": 7486.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.465403, "cps": 58.484375, "mean_p": 0.41899, "boundary_entropy": 0.592049, "bpb": 0.921967}, "zh": {"chunks": 2737.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.463427, "cps": 21.382812, "mean_p": 0.245584, "boundary_entropy": 0.318817, "bpb": 2.602752}, "my": {"chunks": 8938.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.471712, "cps": 69.828125, "mean_p": 0.385216, "boundary_entropy": 0.574756, "bpb": 0.92475}}, "bpb": {"en": 1.70839, "de": 1.84667, "vi": 1.47733, "tr": 1.77321, "fi": 1.9722, "ru": 1.12629, "el": 1.15795, "hi": 0.95589, "ta": 0.81303, "th": 0.92197, "zh": 2.60275, "my": 0.92475}, "gini_cpc": 0.216293, "gini_cps": 0.226788, "bpic": 5.956003, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9089660157487499, "vi": 1.1373469727822165, "tr": 0.8401375925847907, "fi": 0.8156415289411898, "ru": 1.4841560415841972, "el": 1.4398168303873669, "hi": 2.3097417714823836, "ta": 2.172432113854567, "th": 2.3101541812685427, "zh": 2.3003481139873614, "my": 2.3414731535950133}, "t": 1447.1} | |
| {"event": "eval", "step": 8000, "phase": 1, "bpb": {"en": 1.70741, "de": 1.85315, "vi": 1.49427, "tr": 1.76684, "fi": 1.985, "ru": 1.12474, "el": 1.15674, "hi": 0.96154, "ta": 0.81755, "th": 0.92325, "zh": 2.6045, "my": 0.93182}, "bytes_per_chunk": {"en": 4.785, "de": 5.381, "vi": 5.597, "tr": 6.161, "fi": 5.866, "ru": 6.005, "el": 6.016, "hi": 5.435, "ta": 5.884, "th": 6.183, "zh": 5.928, "my": 5.634}, "bpb_mean": 1.4439, "t": 1452.5} | |
| {"event": "train", "step": 8000, "phase": 1, "objective": "global", "lm_loss": 1.11671, "bpb": 1.61107, "chunk_loss": 0.030159, "F": 0.18407, "G": 0.23286, "chunks": 188.48, "bytes_per_chunk": 5.433, "boundary_entropy": 0.19286, "router_noise": 0.0, "grad_norm": 0.3057, "lr": 0.0002957, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 1452.6} | |
| {"event": "train", "step": 8500, "phase": 1, "objective": "global", "lm_loss": 1.09721, "bpb": 1.58294, "chunk_loss": 0.030067, "F": 0.17979, "G": 0.23466, "chunks": 184.11, "bytes_per_chunk": 5.562, "boundary_entropy": 0.21223, "router_noise": 0.0, "grad_norm": 0.305, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 1536.2} | |
| {"event": "train", "step": 9000, "phase": 1, "objective": "global", "lm_loss": 1.08262, "bpb": 1.56189, "chunk_loss": 0.030167, "F": 0.17723, "G": 0.23892, "chunks": 181.48, "bytes_per_chunk": 5.642, "boundary_entropy": 0.22288, "router_noise": 0.0, "grad_norm": 0.3069, "lr": 0.0002941, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 1619.9} | |
| {"event": "train", "step": 9500, "phase": 1, "objective": "global", "lm_loss": 1.13401, "bpb": 1.63604, "chunk_loss": 0.030108, "F": 0.18071, "G": 0.23136, "chunks": 185.05, "bytes_per_chunk": 5.534, "boundary_entropy": 0.19832, "router_noise": 0.0, "grad_norm": 0.3281, "lr": 0.0002933, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 1703.3} | |
| {"event": "train", "step": 10000, "phase": 1, "objective": "global", "lm_loss": 1.06362, "bpb": 1.53448, "chunk_loss": 0.03023, "F": 0.18533, "G": 0.25267, "chunks": 189.78, "bytes_per_chunk": 5.396, "boundary_entropy": 0.2423, "router_noise": 0.0, "grad_norm": 0.2919, "lr": 0.0002924, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 1787.1} | |
| {"event": "train", "step": 10500, "phase": 1, "objective": "global", "lm_loss": 1.06235, "bpb": 1.53265, "chunk_loss": 0.030138, "F": 0.18434, "G": 0.24092, "chunks": 188.77, "bytes_per_chunk": 5.425, "boundary_entropy": 0.20251, "router_noise": 0.0, "grad_norm": 0.2872, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 1871.1} | |
| {"event": "train", "step": 11000, "phase": 1, "objective": "global", "lm_loss": 1.09824, "bpb": 1.58442, "chunk_loss": 0.030105, "F": 0.18259, "G": 0.23175, "chunks": 186.97, "bytes_per_chunk": 5.477, "boundary_entropy": 0.18662, "router_noise": 0.0, "grad_norm": 0.2962, "lr": 0.0002903, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 1954.4} | |
| {"event": "train", "step": 11500, "phase": 1, "objective": "global", "lm_loss": 1.00297, "bpb": 1.44698, "chunk_loss": 0.030133, "F": 0.18309, "G": 0.24866, "chunks": 187.48, "bytes_per_chunk": 5.462, "boundary_entropy": 0.21546, "router_noise": 0.0, "grad_norm": 0.2773, "lr": 0.0002893, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2037.8} | |
| {"event": "probe", "step": 12000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3462.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198931, "cps": 27.046875, "mean_p": 0.193423, "boundary_entropy": 0.06877, "bpb": 1.640953}, "de": {"chunks": 3750.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182633, "cps": 29.296875, "mean_p": 0.184163, "boundary_entropy": 0.058781, "bpb": 1.775543}, "vi": {"chunks": 4227.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229292, "cps": 33.023438, "mean_p": 0.176428, "boundary_entropy": 0.074945, "bpb": 1.402683}, "tr": {"chunks": 2975.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164337, "cps": 23.242188, "mean_p": 0.17053, "boundary_entropy": 0.124259, "bpb": 1.701888}, "fi": {"chunks": 3032.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161182, "cps": 23.6875, "mean_p": 0.171188, "boundary_entropy": 0.105577, "bpb": 1.880074}, "ru": {"chunks": 5647.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.297995, "cps": 44.117188, "mean_p": 0.213723, "boundary_entropy": 0.223628, "bpb": 1.070723}, "el": {"chunks": 5626.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.274131, "cps": 43.953125, "mean_p": 0.299224, "boundary_entropy": 0.410199, "bpb": 1.089587}, "hi": {"chunks": 6271.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.372719, "cps": 48.992188, "mean_p": 0.384916, "boundary_entropy": 0.566208, "bpb": 0.895133}, "ta": {"chunks": 8181.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.437604, "cps": 63.914062, "mean_p": 0.412012, "boundary_entropy": 0.599427, "bpb": 0.777444}, "th": {"chunks": 6810.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.423376, "cps": 53.203125, "mean_p": 0.413339, "boundary_entropy": 0.585165, "bpb": 0.871391}, "zh": {"chunks": 2389.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.404504, "cps": 18.664062, "mean_p": 0.235783, "boundary_entropy": 0.315808, "bpb": 2.52804}, "my": {"chunks": 8020.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.423264, "cps": 62.65625, "mean_p": 0.380966, "boundary_entropy": 0.58057, "bpb": 0.869555}}, "bpb": {"en": 1.64095, "de": 1.77554, "vi": 1.40268, "tr": 1.70189, "fi": 1.88007, "ru": 1.07072, "el": 1.08959, "hi": 0.89513, "ta": 0.77744, "th": 0.87139, "zh": 2.52804, "my": 0.86956}, "gini_cpc": 0.199311, "gini_cps": 0.216231, "bpic": 6.355042, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9180702560978687, "vi": 1.1526200303751202, "tr": 0.8261016228052246, "fi": 0.8102412882613118, "ru": 1.4979786723247808, "el": 1.3780213280597977, "hi": 1.8736082261840075, "ta": 2.1997735703528734, "th": 2.1282522603866267, "zh": 2.0333856941887376, "my": 2.1276885128285055}, "t": 2124.0} | |
| {"event": "eval", "step": 12000, "phase": 1, "bpb": {"en": 1.6414, "de": 1.78347, "vi": 1.41944, "tr": 1.69566, "fi": 1.8942, "ru": 1.06908, "el": 1.08674, "hi": 0.90114, "ta": 0.78141, "th": 0.87076, "zh": 2.52838, "my": 0.87415}, "bytes_per_chunk": {"en": 4.819, "de": 5.402, "vi": 5.591, "tr": 6.288, "fi": 5.943, "ru": 6.056, "el": 6.282, "hi": 6.449, "ta": 5.995, "th": 6.595, "zh": 6.622, "my": 6.399}, "bpb_mean": 1.37882, "t": 2129.2} | |
| {"event": "train", "step": 12000, "phase": 1, "objective": "global", "lm_loss": 0.97369, "bpb": 1.40473, "chunk_loss": 0.030025, "F": 0.17694, "G": 0.24294, "chunks": 181.19, "bytes_per_chunk": 5.652, "boundary_entropy": 0.22613, "router_noise": 0.0, "grad_norm": 0.2731, "lr": 0.0002881, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 2129.4} | |
| {"event": "train", "step": 12500, "phase": 1, "objective": "global", "lm_loss": 0.9914, "bpb": 1.43029, "chunk_loss": 0.030064, "F": 0.18294, "G": 0.24231, "chunks": 187.33, "bytes_per_chunk": 5.466, "boundary_entropy": 0.20966, "router_noise": 0.0, "grad_norm": 0.2723, "lr": 0.0002869, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2212.9} | |
| {"event": "train", "step": 13000, "phase": 1, "objective": "global", "lm_loss": 1.08214, "bpb": 1.5612, "chunk_loss": 0.030166, "F": 0.18481, "G": 0.23829, "chunks": 189.25, "bytes_per_chunk": 5.411, "boundary_entropy": 0.20535, "router_noise": 0.0, "grad_norm": 0.2877, "lr": 0.0002857, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2296.4} | |
| {"event": "train", "step": 13500, "phase": 1, "objective": "global", "lm_loss": 0.97954, "bpb": 1.41317, "chunk_loss": 0.030049, "F": 0.1772, "G": 0.24703, "chunks": 181.45, "bytes_per_chunk": 5.643, "boundary_entropy": 0.22974, "router_noise": 0.0, "grad_norm": 0.2949, "lr": 0.0002844, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2380.1} | |
| {"event": "train", "step": 14000, "phase": 1, "objective": "global", "lm_loss": 1.03481, "bpb": 1.49291, "chunk_loss": 0.030038, "F": 0.1785, "G": 0.24819, "chunks": 182.78, "bytes_per_chunk": 5.602, "boundary_entropy": 0.23723, "router_noise": 0.0, "grad_norm": 0.2737, "lr": 0.000283, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2463.5} | |
| {"event": "train", "step": 14500, "phase": 1, "objective": "global", "lm_loss": 0.99549, "bpb": 1.4362, "chunk_loss": 0.03018, "F": 0.18253, "G": 0.2384, "chunks": 186.91, "bytes_per_chunk": 5.479, "boundary_entropy": 0.20687, "router_noise": 0.0, "grad_norm": 0.2959, "lr": 0.0002816, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2547.0} | |
| {"event": "train", "step": 15000, "phase": 1, "objective": "global", "lm_loss": 1.02491, "bpb": 1.47863, "chunk_loss": 0.030188, "F": 0.18027, "G": 0.24634, "chunks": 184.59, "bytes_per_chunk": 5.547, "boundary_entropy": 0.23035, "router_noise": 0.0, "grad_norm": 0.2806, "lr": 0.0002801, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2630.6} | |
| {"event": "train", "step": 15500, "phase": 1, "objective": "global", "lm_loss": 0.96949, "bpb": 1.39868, "chunk_loss": 0.030024, "F": 0.1795, "G": 0.26127, "chunks": 183.81, "bytes_per_chunk": 5.571, "boundary_entropy": 0.24916, "router_noise": 0.0, "grad_norm": 0.2829, "lr": 0.0002786, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2714.0} | |
| {"event": "probe", "step": 16000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3458.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.198701, "cps": 27.015625, "mean_p": 0.193567, "boundary_entropy": 0.062448, "bpb": 1.602042}, "de": {"chunks": 3741.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182195, "cps": 29.226562, "mean_p": 0.183663, "boundary_entropy": 0.053311, "bpb": 1.733319}, "vi": {"chunks": 4222.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229021, "cps": 32.984375, "mean_p": 0.176927, "boundary_entropy": 0.072309, "bpb": 1.359444}, "tr": {"chunks": 3071.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16964, "cps": 23.992188, "mean_p": 0.177367, "boundary_entropy": 0.127789, "bpb": 1.654078}, "fi": {"chunks": 3006.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.1598, "cps": 23.484375, "mean_p": 0.169575, "boundary_entropy": 0.096653, "bpb": 1.825086}, "ru": {"chunks": 5672.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.299314, "cps": 44.3125, "mean_p": 0.223137, "boundary_entropy": 0.238853, "bpb": 1.038747}, "el": {"chunks": 5586.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.272182, "cps": 43.640625, "mean_p": 0.310805, "boundary_entropy": 0.422261, "bpb": 1.05415}, "hi": {"chunks": 7118.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.423061, "cps": 55.609375, "mean_p": 0.401021, "boundary_entropy": 0.572287, "bpb": 0.868777}, "ta": {"chunks": 8149.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.435892, "cps": 63.664062, "mean_p": 0.410307, "boundary_entropy": 0.596397, "bpb": 0.756659}, "th": {"chunks": 7802.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.485048, "cps": 60.953125, "mean_p": 0.423622, "boundary_entropy": 0.588978, "bpb": 0.847408}, "zh": {"chunks": 2791.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.47257, "cps": 21.804688, "mean_p": 0.290035, "boundary_entropy": 0.385054, "bpb": 2.471992}, "my": {"chunks": 7853.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.41445, "cps": 61.351562, "mean_p": 0.37786, "boundary_entropy": 0.56699, "bpb": 0.83948}}, "bpb": {"en": 1.60204, "de": 1.73332, "vi": 1.35944, "tr": 1.65408, "fi": 1.82509, "ru": 1.03875, "el": 1.05415, "hi": 0.86878, "ta": 0.75666, "th": 0.84741, "zh": 2.47199, "my": 0.83948}, "gini_cpc": 0.217921, "gini_cps": 0.214953, "bpic": 6.143543, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.916926305513868, "vi": 1.1525883306877696, "tr": 0.8537454384711455, "fi": 0.8042225087710791, "ru": 1.506350857863148, "el": 1.369806483532547, "hi": 2.1291293269177056, "ta": 2.19370375967693, "th": 2.441091238653531, "zh": 2.3782939172150854, "my": 2.085793706143861}, "t": 2800.0} | |
| {"event": "eval", "step": 16000, "phase": 1, "bpb": {"en": 1.60131, "de": 1.73955, "vi": 1.37733, "tr": 1.64554, "fi": 1.84092, "ru": 1.03688, "el": 1.05185, "hi": 0.87327, "ta": 0.76101, "th": 0.84592, "zh": 2.47391, "my": 0.8448}, "bytes_per_chunk": {"en": 4.82, "de": 5.413, "vi": 5.601, "tr": 6.091, "fi": 5.964, "ru": 6.0, "el": 6.3, "hi": 5.866, "ta": 6.038, "th": 5.631, "zh": 5.87, "my": 6.586}, "bpb_mean": 1.34102, "t": 2805.4} | |
| {"event": "train", "step": 16000, "phase": 1, "objective": "global", "lm_loss": 0.93862, "bpb": 1.35414, "chunk_loss": 0.030181, "F": 0.18277, "G": 0.25136, "chunks": 187.16, "bytes_per_chunk": 5.471, "boundary_entropy": 0.22514, "router_noise": 0.0, "grad_norm": 0.2499, "lr": 0.000277, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 2805.5} | |
| {"event": "train", "step": 16500, "phase": 1, "objective": "global", "lm_loss": 0.98496, "bpb": 1.421, "chunk_loss": 0.030175, "F": 0.17914, "G": 0.24132, "chunks": 183.44, "bytes_per_chunk": 5.582, "boundary_entropy": 0.21078, "router_noise": 0.0, "grad_norm": 0.2647, "lr": 0.0002754, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2889.2} | |
| {"event": "train", "step": 17000, "phase": 1, "objective": "global", "lm_loss": 1.03103, "bpb": 1.48746, "chunk_loss": 0.030175, "F": 0.18211, "G": 0.24665, "chunks": 186.48, "bytes_per_chunk": 5.491, "boundary_entropy": 0.22527, "router_noise": 0.0, "grad_norm": 0.3095, "lr": 0.0002737, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 2972.8} | |
| {"event": "train", "step": 17500, "phase": 1, "objective": "global", "lm_loss": 0.9064, "bpb": 1.30766, "chunk_loss": 0.030132, "F": 0.17963, "G": 0.2572, "chunks": 183.94, "bytes_per_chunk": 5.567, "boundary_entropy": 0.25256, "router_noise": 0.0, "grad_norm": 0.258, "lr": 0.000272, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 3056.7} | |
| {"event": "train", "step": 18000, "phase": 1, "objective": "global", "lm_loss": 0.98751, "bpb": 1.42467, "chunk_loss": 0.030143, "F": 0.1785, "G": 0.2623, "chunks": 182.78, "bytes_per_chunk": 5.602, "boundary_entropy": 0.25769, "router_noise": 0.0, "grad_norm": 0.2672, "lr": 0.0002703, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 3139.8} | |
| {"event": "train", "step": 18500, "phase": 1, "objective": "global", "lm_loss": 1.00112, "bpb": 1.44431, "chunk_loss": 0.030196, "F": 0.1808, "G": 0.25643, "chunks": 185.14, "bytes_per_chunk": 5.531, "boundary_entropy": 0.24289, "router_noise": 0.0, "grad_norm": 0.2659, "lr": 0.0002685, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 3223.3} | |
| {"event": "train", "step": 19000, "phase": 1, "objective": "global", "lm_loss": 1.03164, "bpb": 1.48834, "chunk_loss": 0.030152, "F": 0.18169, "G": 0.25358, "chunks": 186.05, "bytes_per_chunk": 5.504, "boundary_entropy": 0.23532, "router_noise": 0.0, "grad_norm": 0.2816, "lr": 0.0002666, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 3306.5} | |
| {"event": "train", "step": 19500, "phase": 1, "objective": "global", "lm_loss": 1.01923, "bpb": 1.47044, "chunk_loss": 0.030163, "F": 0.18472, "G": 0.24182, "chunks": 189.16, "bytes_per_chunk": 5.414, "boundary_entropy": 0.20986, "router_noise": 0.0, "grad_norm": 0.2708, "lr": 0.0002647, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 3390.0} | |
| {"event": "probe", "step": 20000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3437.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.197495, "cps": 26.851562, "mean_p": 0.191181, "boundary_entropy": 0.068035, "bpb": 1.575861}, "de": {"chunks": 3726.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.181464, "cps": 29.109375, "mean_p": 0.182386, "boundary_entropy": 0.0558, "bpb": 1.702839}, "vi": {"chunks": 4226.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229238, "cps": 33.015625, "mean_p": 0.176959, "boundary_entropy": 0.069178, "bpb": 1.326282}, "tr": {"chunks": 3055.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.168757, "cps": 23.867188, "mean_p": 0.183661, "boundary_entropy": 0.146495, "bpb": 1.625893}, "fi": {"chunks": 3037.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161448, "cps": 23.726562, "mean_p": 0.173392, "boundary_entropy": 0.106554, "bpb": 1.782259}, "ru": {"chunks": 5632.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.297203, "cps": 44.0, "mean_p": 0.22959, "boundary_entropy": 0.255927, "bpb": 1.015768}, "el": {"chunks": 5814.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.283292, "cps": 45.421875, "mean_p": 0.317921, "boundary_entropy": 0.426845, "bpb": 1.03363}, "hi": {"chunks": 6628.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.393938, "cps": 51.78125, "mean_p": 0.376764, "boundary_entropy": 0.552687, "bpb": 0.84906}, "ta": {"chunks": 8297.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.443809, "cps": 64.820312, "mean_p": 0.411218, "boundary_entropy": 0.593858, "bpb": 0.73526}, "th": {"chunks": 7796.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.484675, "cps": 60.90625, "mean_p": 0.418748, "boundary_entropy": 0.582283, "bpb": 0.828385}, "zh": {"chunks": 2613.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.442431, "cps": 20.414062, "mean_p": 0.304653, "boundary_entropy": 0.410579, "bpb": 2.433599}, "my": {"chunks": 8873.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.468282, "cps": 69.320312, "mean_p": 0.390061, "boundary_entropy": 0.56391, "bpb": 0.814216}}, "bpb": {"en": 1.57586, "de": 1.70284, "vi": 1.32628, "tr": 1.62589, "fi": 1.78226, "ru": 1.01577, "el": 1.03363, "hi": 0.84906, "ta": 0.73526, "th": 0.82838, "zh": 2.4336, "my": 0.81422}, "gini_cpc": 0.218111, "gini_cps": 0.228366, "bpic": 6.078832, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9188297141578446, "vi": 1.1607292771692155, "tr": 0.8544865895314611, "fi": 0.8174806781644159, "ru": 1.5048666575056635, "el": 1.434428053789392, "hi": 1.9946744046195992, "ta": 2.2471921476890837, "th": 2.45411750511833, "zh": 2.2402194067251453, "my": 2.3711099529152535}, "t": 3476.1} | |
| {"event": "eval", "step": 20000, "phase": 1, "bpb": {"en": 1.57401, "de": 1.70956, "vi": 1.34571, "tr": 1.6162, "fi": 1.79839, "ru": 1.01432, "el": 1.03264, "hi": 0.85471, "ta": 0.74087, "th": 0.82904, "zh": 2.43611, "my": 0.81835}, "bytes_per_chunk": {"en": 4.847, "de": 5.447, "vi": 5.597, "tr": 6.128, "fi": 5.898, "ru": 6.028, "el": 6.134, "hi": 6.152, "ta": 5.963, "th": 5.743, "zh": 6.14, "my": 5.94}, "bpb_mean": 1.31416, "t": 3481.3} | |
| {"event": "train", "step": 20000, "phase": 1, "objective": "global", "lm_loss": 1.01003, "bpb": 1.45716, "chunk_loss": 0.030185, "F": 0.18204, "G": 0.24656, "chunks": 186.41, "bytes_per_chunk": 5.493, "boundary_entropy": 0.22121, "router_noise": 0.0, "grad_norm": 0.2586, "lr": 0.0002628, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 3481.5} | |
| {"event": "train", "step": 20500, "phase": 1, "objective": "global", "lm_loss": 0.9573, "bpb": 1.38109, "chunk_loss": 0.030085, "F": 0.179, "G": 0.26008, "chunks": 183.3, "bytes_per_chunk": 5.587, "boundary_entropy": 0.25228, "router_noise": 0.0, "grad_norm": 0.2645, "lr": 0.0002608, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 3565.0} | |
| {"event": "train", "step": 21000, "phase": 1, "objective": "global", "lm_loss": 1.0024, "bpb": 1.44615, "chunk_loss": 0.030269, "F": 0.1826, "G": 0.24814, "chunks": 186.98, "bytes_per_chunk": 5.476, "boundary_entropy": 0.22822, "router_noise": 0.0, "grad_norm": 0.2744, "lr": 0.0002587, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 3648.9} | |
| {"event": "train", "step": 21500, "phase": 1, "objective": "global", "lm_loss": 0.96919, "bpb": 1.39825, "chunk_loss": 0.030142, "F": 0.17856, "G": 0.2584, "chunks": 182.84, "bytes_per_chunk": 5.6, "boundary_entropy": 0.26046, "router_noise": 0.0, "grad_norm": 0.269, "lr": 0.0002567, "router_trainable": true, "sec_per_step": 0.168, "alpha": 0.03, "t": 3732.7} | |
| {"event": "train", "step": 22000, "phase": 1, "objective": "global", "lm_loss": 1.00257, "bpb": 1.4464, "chunk_loss": 0.029942, "F": 0.17529, "G": 0.26026, "chunks": 179.5, "bytes_per_chunk": 5.705, "boundary_entropy": 0.2611, "router_noise": 0.0, "grad_norm": 0.2665, "lr": 0.0002545, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 3815.9} | |
| {"event": "train", "step": 22500, "phase": 1, "objective": "global", "lm_loss": 0.97305, "bpb": 1.40381, "chunk_loss": 0.030099, "F": 0.17831, "G": 0.24946, "chunks": 182.59, "bytes_per_chunk": 5.608, "boundary_entropy": 0.23248, "router_noise": 0.0, "grad_norm": 0.2654, "lr": 0.0002524, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 3899.2} | |
| {"event": "train", "step": 23000, "phase": 1, "objective": "global", "lm_loss": 1.00798, "bpb": 1.45421, "chunk_loss": 0.03006, "F": 0.17889, "G": 0.2571, "chunks": 183.19, "bytes_per_chunk": 5.59, "boundary_entropy": 0.2487, "router_noise": 0.0, "grad_norm": 0.2729, "lr": 0.0002502, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 3982.7} | |
| {"event": "train", "step": 23500, "phase": 1, "objective": "global", "lm_loss": 0.94878, "bpb": 1.3688, "chunk_loss": 0.030124, "F": 0.17931, "G": 0.26056, "chunks": 183.61, "bytes_per_chunk": 5.577, "boundary_entropy": 0.25413, "router_noise": 0.0, "grad_norm": 0.2795, "lr": 0.000248, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 4066.1} | |
| {"event": "probe", "step": 24000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3439.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19761, "cps": 26.867188, "mean_p": 0.191612, "boundary_entropy": 0.064628, "bpb": 1.554773}, "de": {"chunks": 3728.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.181561, "cps": 29.125, "mean_p": 0.182618, "boundary_entropy": 0.054003, "bpb": 1.680209}, "vi": {"chunks": 4230.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229455, "cps": 33.046875, "mean_p": 0.176718, "boundary_entropy": 0.070637, "bpb": 1.306617}, "tr": {"chunks": 3120.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172347, "cps": 24.375, "mean_p": 0.189284, "boundary_entropy": 0.150241, "bpb": 1.599652}, "fi": {"chunks": 3058.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162564, "cps": 23.890625, "mean_p": 0.175635, "boundary_entropy": 0.106193, "bpb": 1.745043}, "ru": {"chunks": 5671.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.299261, "cps": 44.304688, "mean_p": 0.233566, "boundary_entropy": 0.257301, "bpb": 1.002198}, "el": {"chunks": 6055.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.295035, "cps": 47.304688, "mean_p": 0.3264, "boundary_entropy": 0.434302, "bpb": 1.012224}, "hi": {"chunks": 7240.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.430312, "cps": 56.5625, "mean_p": 0.385955, "boundary_entropy": 0.552406, "bpb": 0.841765}, "ta": {"chunks": 8466.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.452848, "cps": 66.140625, "mean_p": 0.412256, "boundary_entropy": 0.591013, "bpb": 0.725424}, "th": {"chunks": 7608.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.472987, "cps": 59.4375, "mean_p": 0.411019, "boundary_entropy": 0.580722, "bpb": 0.816652}, "zh": {"chunks": 2608.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.441585, "cps": 20.375, "mean_p": 0.318714, "boundary_entropy": 0.431638, "bpb": 2.415028}, "my": {"chunks": 8450.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.445957, "cps": 66.015625, "mean_p": 0.376296, "boundary_entropy": 0.555562, "bpb": 0.792756}}, "bpb": {"en": 1.55477, "de": 1.68021, "vi": 1.30662, "tr": 1.59965, "fi": 1.74504, "ru": 1.0022, "el": 1.01222, "hi": 0.84177, "ta": 0.72542, "th": 0.81665, "zh": 2.41503, "my": 0.79276}, "gini_cpc": 0.211921, "gini_cps": 0.224859, "bpic": 6.027374, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9187882676596021, "vi": 1.1611522546504451, "tr": 0.8721596431084144, "fi": 0.8226546218641846, "ru": 1.5144061943514597, "el": 1.4930186980055724, "hi": 2.1775866114022056, "ta": 2.2916312663895515, "th": 2.3935438215651206, "zh": 2.234632386993006, "my": 2.2567594892324956}, "t": 4152.3} | |
| {"event": "eval", "step": 24000, "phase": 1, "bpb": {"en": 1.55305, "de": 1.68682, "vi": 1.32594, "tr": 1.59244, "fi": 1.76277, "ru": 0.99988, "el": 1.01119, "hi": 0.8469, "ta": 0.73109, "th": 0.81651, "zh": 2.4171, "my": 0.79694}, "bytes_per_chunk": {"en": 4.85, "de": 5.444, "vi": 5.599, "tr": 6.013, "fi": 5.891, "ru": 6.066, "el": 5.941, "hi": 5.734, "ta": 5.766, "th": 5.988, "zh": 6.182, "my": 6.218}, "bpb_mean": 1.29505, "t": 4157.7} | |
| {"event": "train", "step": 24000, "phase": 1, "objective": "global", "lm_loss": 0.98472, "bpb": 1.42065, "chunk_loss": 0.030204, "F": 0.18433, "G": 0.24807, "chunks": 188.75, "bytes_per_chunk": 5.425, "boundary_entropy": 0.21499, "router_noise": 0.0, "grad_norm": 0.2799, "lr": 0.0002457, "router_trainable": true, "sec_per_step": 0.184, "alpha": 0.03, "t": 4157.8} | |
| {"event": "train", "step": 24500, "phase": 1, "objective": "global", "lm_loss": 0.97832, "bpb": 1.41142, "chunk_loss": 0.030059, "F": 0.18108, "G": 0.24588, "chunks": 185.42, "bytes_per_chunk": 5.523, "boundary_entropy": 0.21818, "router_noise": 0.0, "grad_norm": 0.2697, "lr": 0.0002434, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 4241.4} | |
| {"event": "train", "step": 25000, "phase": 1, "objective": "global", "lm_loss": 0.98863, "bpb": 1.42629, "chunk_loss": 0.030097, "F": 0.17757, "G": 0.24997, "chunks": 181.83, "bytes_per_chunk": 5.632, "boundary_entropy": 0.23558, "router_noise": 0.0, "grad_norm": 0.2682, "lr": 0.000241, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 4324.6} | |
| {"event": "train", "step": 25500, "phase": 1, "objective": "global", "lm_loss": 0.97992, "bpb": 1.41372, "chunk_loss": 0.03005, "F": 0.17992, "G": 0.24941, "chunks": 184.23, "bytes_per_chunk": 5.558, "boundary_entropy": 0.22401, "router_noise": 0.0, "grad_norm": 0.2864, "lr": 0.0002387, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 4408.1} | |
| {"event": "train", "step": 26000, "phase": 1, "objective": "global", "lm_loss": 0.94767, "bpb": 1.3672, "chunk_loss": 0.030032, "F": 0.1786, "G": 0.24298, "chunks": 182.89, "bytes_per_chunk": 5.599, "boundary_entropy": 0.21642, "router_noise": 0.0, "grad_norm": 0.2666, "lr": 0.0002363, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 4491.7} | |
| {"event": "train", "step": 26500, "phase": 1, "objective": "global", "lm_loss": 0.94182, "bpb": 1.35876, "chunk_loss": 0.030012, "F": 0.17506, "G": 0.24291, "chunks": 179.27, "bytes_per_chunk": 5.712, "boundary_entropy": 0.22541, "router_noise": 0.0, "grad_norm": 0.2653, "lr": 0.0002338, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 4575.1} | |
| {"event": "train", "step": 27000, "phase": 1, "objective": "global", "lm_loss": 0.88785, "bpb": 1.2809, "chunk_loss": 0.03006, "F": 0.17918, "G": 0.24037, "chunks": 183.48, "bytes_per_chunk": 5.581, "boundary_entropy": 0.21392, "router_noise": 0.0, "grad_norm": 0.2614, "lr": 0.0002313, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 4658.0} | |
| {"event": "train", "step": 27500, "phase": 1, "objective": "global", "lm_loss": 0.96392, "bpb": 1.39064, "chunk_loss": 0.030201, "F": 0.18324, "G": 0.24739, "chunks": 187.64, "bytes_per_chunk": 5.457, "boundary_entropy": 0.22043, "router_noise": 0.0, "grad_norm": 0.2662, "lr": 0.0002288, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 4741.1} | |
| {"event": "probe", "step": 28000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3427.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.19692, "cps": 26.773438, "mean_p": 0.191471, "boundary_entropy": 0.064139, "bpb": 1.537815}, "de": {"chunks": 3749.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182584, "cps": 29.289062, "mean_p": 0.183457, "boundary_entropy": 0.052689, "bpb": 1.658444}, "vi": {"chunks": 4231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229509, "cps": 33.054688, "mean_p": 0.176992, "boundary_entropy": 0.070815, "bpb": 1.285694}, "tr": {"chunks": 3080.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170138, "cps": 24.0625, "mean_p": 0.185885, "boundary_entropy": 0.144824, "bpb": 1.577336}, "fi": {"chunks": 3038.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161501, "cps": 23.734375, "mean_p": 0.173403, "boundary_entropy": 0.10101, "bpb": 1.728207}, "ru": {"chunks": 5809.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306544, "cps": 45.382812, "mean_p": 0.234636, "boundary_entropy": 0.252331, "bpb": 0.982946}, "el": {"chunks": 6232.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.303659, "cps": 48.6875, "mean_p": 0.322098, "boundary_entropy": 0.421223, "bpb": 0.995021}, "hi": {"chunks": 7342.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.436374, "cps": 57.359375, "mean_p": 0.386239, "boundary_entropy": 0.547446, "bpb": 0.821458}, "ta": {"chunks": 9365.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.500936, "cps": 73.164062, "mean_p": 0.41553, "boundary_entropy": 0.583469, "bpb": 0.712147}, "th": {"chunks": 7872.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.4894, "cps": 61.5, "mean_p": 0.416915, "boundary_entropy": 0.582326, "bpb": 0.805018}, "zh": {"chunks": 2486.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.420928, "cps": 19.421875, "mean_p": 0.321956, "boundary_entropy": 0.442436, "bpb": 2.387925}, "my": {"chunks": 9553.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.504169, "cps": 74.632812, "mean_p": 0.394695, "boundary_entropy": 0.562669, "bpb": 0.778923}}, "bpb": {"en": 1.53782, "de": 1.65844, "vi": 1.28569, "tr": 1.57734, "fi": 1.72821, "ru": 0.98295, "el": 0.99502, "hi": 0.82146, "ta": 0.71215, "th": 0.80502, "zh": 2.38793, "my": 0.77892}, "gini_cpc": 0.227807, "gini_cps": 0.247827, "bpic": 5.798698, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9271992016779668, "vi": 1.1654936163615268, "tr": 0.8639929143903896, "fi": 0.8201360493227853, "ru": 1.5566901518517007, "el": 1.5420434653365132, "hi": 2.2159977904682293, "ta": 2.5438548543093393, "th": 2.4852726238516767, "zh": 2.1375569655563855, "my": 2.5602738322301035}, "t": 4827.1} | |
| {"event": "eval", "step": 28000, "phase": 1, "bpb": {"en": 1.53454, "de": 1.66595, "vi": 1.3038, "tr": 1.56922, "fi": 1.74672, "ru": 0.98223, "el": 0.99269, "hi": 0.82777, "ta": 0.71712, "th": 0.80416, "zh": 2.39164, "my": 0.78374}, "bytes_per_chunk": {"en": 4.857, "de": 5.42, "vi": 5.595, "tr": 6.08, "fi": 5.904, "ru": 5.922, "el": 5.738, "hi": 5.67, "ta": 5.294, "th": 5.788, "zh": 6.374, "my": 5.629}, "bpb_mean": 1.27663, "t": 4832.5} | |
| {"event": "train", "step": 28000, "phase": 1, "objective": "global", "lm_loss": 1.00993, "bpb": 1.45702, "chunk_loss": 0.030291, "F": 0.18167, "G": 0.2676, "chunks": 186.03, "bytes_per_chunk": 5.504, "boundary_entropy": 0.26961, "router_noise": 0.0, "grad_norm": 0.2826, "lr": 0.0002263, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 4832.7} | |
| {"event": "train", "step": 28500, "phase": 1, "objective": "global", "lm_loss": 1.00582, "bpb": 1.45109, "chunk_loss": 0.030146, "F": 0.18373, "G": 0.23172, "chunks": 188.14, "bytes_per_chunk": 5.443, "boundary_entropy": 0.18456, "router_noise": 0.0, "grad_norm": 0.2708, "lr": 0.0002238, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 4915.8} | |
| {"event": "train", "step": 29000, "phase": 1, "objective": "global", "lm_loss": 0.92499, "bpb": 1.33448, "chunk_loss": 0.030052, "F": 0.17842, "G": 0.24296, "chunks": 182.7, "bytes_per_chunk": 5.605, "boundary_entropy": 0.21759, "router_noise": 0.0, "grad_norm": 0.2695, "lr": 0.0002212, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 4998.6} | |
| {"event": "train", "step": 29500, "phase": 1, "objective": "global", "lm_loss": 1.00447, "bpb": 1.44914, "chunk_loss": 0.030113, "F": 0.18347, "G": 0.24173, "chunks": 187.88, "bytes_per_chunk": 5.45, "boundary_entropy": 0.2029, "router_noise": 0.0, "grad_norm": 0.2794, "lr": 0.0002186, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5081.4} | |
| {"event": "train", "step": 30000, "phase": 1, "objective": "global", "lm_loss": 0.95387, "bpb": 1.37614, "chunk_loss": 0.030205, "F": 0.18745, "G": 0.24526, "chunks": 191.95, "bytes_per_chunk": 5.335, "boundary_entropy": 0.20543, "router_noise": 0.0, "grad_norm": 0.2615, "lr": 0.0002159, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5164.3} | |
| {"event": "train", "step": 30500, "phase": 1, "objective": "global", "lm_loss": 1.02996, "bpb": 1.48592, "chunk_loss": 0.030145, "F": 0.18341, "G": 0.24534, "chunks": 187.81, "bytes_per_chunk": 5.452, "boundary_entropy": 0.2165, "router_noise": 0.0, "grad_norm": 0.2867, "lr": 0.0002133, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5247.5} | |
| {"event": "phase_transition", "step": 30520, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 5250.6} | |
| {"event": "train", "step": 31000, "phase": 2, "objective": "global", "lm_loss": 0.96072, "bpb": 1.38602, "chunk_loss": 0.030096, "F": 0.17996, "G": 0.25627, "chunks": 184.28, "bytes_per_chunk": 5.557, "boundary_entropy": 0.24106, "router_noise": 0.0, "grad_norm": 0.2833, "lr": 0.0002106, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 5330.0} | |
| {"event": "train", "step": 31500, "phase": 2, "objective": "global", "lm_loss": 0.88372, "bpb": 1.27494, "chunk_loss": 0.030055, "F": 0.17992, "G": 0.25061, "chunks": 184.23, "bytes_per_chunk": 5.558, "boundary_entropy": 0.22839, "router_noise": 0.0, "grad_norm": 0.2727, "lr": 0.0002079, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5412.8} | |
| {"event": "probe", "step": 32000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3403.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195541, "cps": 26.585938, "mean_p": 0.189445, "boundary_entropy": 0.063081, "bpb": 1.519979}, "de": {"chunks": 3731.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.181707, "cps": 29.148438, "mean_p": 0.181571, "boundary_entropy": 0.049651, "bpb": 1.642866}, "vi": {"chunks": 4231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229509, "cps": 33.054688, "mean_p": 0.175412, "boundary_entropy": 0.066898, "bpb": 1.268398}, "tr": {"chunks": 3124.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172568, "cps": 24.40625, "mean_p": 0.186225, "boundary_entropy": 0.141436, "bpb": 1.556022}, "fi": {"chunks": 3079.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.163681, "cps": 24.054688, "mean_p": 0.172947, "boundary_entropy": 0.09428, "bpb": 1.706259}, "ru": {"chunks": 5627.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.296939, "cps": 43.960938, "mean_p": 0.220191, "boundary_entropy": 0.22804, "bpb": 0.973133}, "el": {"chunks": 5650.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.275301, "cps": 44.140625, "mean_p": 0.295938, "boundary_entropy": 0.392278, "bpb": 0.981769}, "hi": {"chunks": 6709.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.398752, "cps": 52.414062, "mean_p": 0.367029, "boundary_entropy": 0.533171, "bpb": 0.812458}, "ta": {"chunks": 7525.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.402514, "cps": 58.789062, "mean_p": 0.388434, "boundary_entropy": 0.572193, "bpb": 0.703553}, "th": {"chunks": 7548.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.469257, "cps": 58.96875, "mean_p": 0.394358, "boundary_entropy": 0.560268, "bpb": 0.792603}, "zh": {"chunks": 2517.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.426177, "cps": 19.664062, "mean_p": 0.315048, "boundary_entropy": 0.428145, "bpb": 2.356045}, "my": {"chunks": 8173.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.431338, "cps": 63.851562, "mean_p": 0.383774, "boundary_entropy": 0.556124, "bpb": 0.765796}}, "bpb": {"en": 1.51998, "de": 1.64287, "vi": 1.2684, "tr": 1.55602, "fi": 1.70626, "ru": 0.97313, "el": 0.98177, "hi": 0.81246, "ta": 0.70355, "th": 0.7926, "zh": 2.35605, "my": 0.7658}, "gini_cpc": 0.205076, "gini_cps": 0.213273, "bpic": 6.258966, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9292552276998707, "vi": 1.1737133773937563, "tr": 0.882516115825223, "fi": 0.8370665290422462, "ru": 1.5185527126848342, "el": 1.4078934023965628, "hi": 2.0392237947549994, "ta": 2.058463666995525, "th": 2.3997886631326324, "zh": 2.17947526231063, "my": 2.205871928375235}, "t": 5498.5} | |
| {"event": "eval", "step": 32000, "phase": 2, "bpb": {"en": 1.51791, "de": 1.65132, "vi": 1.28645, "tr": 1.5488, "fi": 1.72286, "ru": 0.97247, "el": 0.98039, "hi": 0.81846, "ta": 0.70908, "th": 0.79099, "zh": 2.359, "my": 0.7712}, "bytes_per_chunk": {"en": 4.883, "de": 5.435, "vi": 5.595, "tr": 6.061, "fi": 5.863, "ru": 6.095, "el": 6.271, "hi": 6.111, "ta": 6.5, "th": 6.043, "zh": 6.282, "my": 6.432}, "bpb_mean": 1.26074, "t": 5503.8} | |
| {"event": "train", "step": 32000, "phase": 2, "objective": "global", "lm_loss": 0.95841, "bpb": 1.3827, "chunk_loss": 0.030023, "F": 0.17795, "G": 0.25244, "chunks": 182.22, "bytes_per_chunk": 5.62, "boundary_entropy": 0.23866, "router_noise": 0.0, "grad_norm": 0.2761, "lr": 0.0002052, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 5504.0} | |
| {"event": "train", "step": 32500, "phase": 2, "objective": "global", "lm_loss": 0.87297, "bpb": 1.25943, "chunk_loss": 0.030133, "F": 0.1819, "G": 0.25069, "chunks": 186.27, "bytes_per_chunk": 5.498, "boundary_entropy": 0.22821, "router_noise": 0.0, "grad_norm": 0.2797, "lr": 0.0002025, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5587.1} | |
| {"event": "train", "step": 33000, "phase": 2, "objective": "global", "lm_loss": 1.01331, "bpb": 1.4619, "chunk_loss": 0.030105, "F": 0.18156, "G": 0.24524, "chunks": 185.92, "bytes_per_chunk": 5.508, "boundary_entropy": 0.21471, "router_noise": 0.0, "grad_norm": 0.317, "lr": 0.0001997, "router_trainable": true, "sec_per_step": 0.167, "alpha": 0.03, "t": 5670.8} | |
| {"event": "train", "step": 33500, "phase": 2, "objective": "global", "lm_loss": 0.96096, "bpb": 1.38637, "chunk_loss": 0.03007, "F": 0.17873, "G": 0.24285, "chunks": 183.02, "bytes_per_chunk": 5.595, "boundary_entropy": 0.21778, "router_noise": 0.0, "grad_norm": 0.2887, "lr": 0.0001969, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5753.8} | |
| {"event": "train", "step": 34000, "phase": 2, "objective": "global", "lm_loss": 0.90588, "bpb": 1.30691, "chunk_loss": 0.030105, "F": 0.181, "G": 0.2466, "chunks": 185.34, "bytes_per_chunk": 5.525, "boundary_entropy": 0.21501, "router_noise": 0.0, "grad_norm": 0.262, "lr": 0.0001942, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5836.8} | |
| {"event": "train", "step": 34500, "phase": 2, "objective": "global", "lm_loss": 0.94417, "bpb": 1.36215, "chunk_loss": 0.030177, "F": 0.18497, "G": 0.22593, "chunks": 189.41, "bytes_per_chunk": 5.406, "boundary_entropy": 0.16476, "router_noise": 0.0, "grad_norm": 0.2683, "lr": 0.0001914, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 5919.9} | |
| {"event": "train", "step": 35000, "phase": 2, "objective": "global", "lm_loss": 0.95141, "bpb": 1.3726, "chunk_loss": 0.030046, "F": 0.17957, "G": 0.25538, "chunks": 183.88, "bytes_per_chunk": 5.569, "boundary_entropy": 0.23957, "router_noise": 0.0, "grad_norm": 0.2753, "lr": 0.0001886, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6003.0} | |
| {"event": "train", "step": 35500, "phase": 2, "objective": "global", "lm_loss": 0.98498, "bpb": 1.42103, "chunk_loss": 0.030196, "F": 0.18448, "G": 0.23797, "chunks": 188.91, "bytes_per_chunk": 5.421, "boundary_entropy": 0.18729, "router_noise": 0.0, "grad_norm": 0.2731, "lr": 0.0001858, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6086.0} | |
| {"event": "probe", "step": 36000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3416.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196288, "cps": 26.6875, "mean_p": 0.191381, "boundary_entropy": 0.059234, "bpb": 1.504669}, "de": {"chunks": 3748.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182535, "cps": 29.28125, "mean_p": 0.182149, "boundary_entropy": 0.04891, "bpb": 1.624498}, "vi": {"chunks": 4233.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229618, "cps": 33.070312, "mean_p": 0.175905, "boundary_entropy": 0.061155, "bpb": 1.251127}, "tr": {"chunks": 3155.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174281, "cps": 24.648438, "mean_p": 0.183981, "boundary_entropy": 0.128923, "bpb": 1.542599}, "fi": {"chunks": 3086.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.164053, "cps": 24.109375, "mean_p": 0.172746, "boundary_entropy": 0.091447, "bpb": 1.682568}, "ru": {"chunks": 5754.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.303641, "cps": 44.953125, "mean_p": 0.222696, "boundary_entropy": 0.222786, "bpb": 0.95973}, "el": {"chunks": 5806.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.282902, "cps": 45.359375, "mean_p": 0.292752, "boundary_entropy": 0.38324, "bpb": 0.967592}, "hi": {"chunks": 6979.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.414799, "cps": 54.523438, "mean_p": 0.369412, "boundary_entropy": 0.528756, "bpb": 0.803455}, "ta": {"chunks": 8025.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.429259, "cps": 62.695312, "mean_p": 0.395557, "boundary_entropy": 0.574154, "bpb": 0.693357}, "th": {"chunks": 7208.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.448119, "cps": 56.3125, "mean_p": 0.384787, "boundary_entropy": 0.553915, "bpb": 0.781812}, "zh": {"chunks": 2541.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.43024, "cps": 19.851562, "mean_p": 0.325998, "boundary_entropy": 0.441356, "bpb": 2.336749}, "my": {"chunks": 8764.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.462529, "cps": 68.46875, "mean_p": 0.384299, "boundary_entropy": 0.548621, "bpb": 0.753813}}, "bpb": {"en": 1.50467, "de": 1.6245, "vi": 1.25113, "tr": 1.5426, "fi": 1.68257, "ru": 0.95973, "el": 0.96759, "hi": 0.80345, "ta": 0.69336, "th": 0.78181, "zh": 2.33675, "my": 0.75381}, "gini_cpc": 0.206702, "gini_cps": 0.221618, "bpic": 6.119445, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9299367979186073, "vi": 1.1697993742175337, "tr": 0.8878816293338885, "fi": 0.8357767709535642, "ru": 1.5469166053895067, "el": 1.441260368072789, "hi": 2.1132184006040973, "ta": 2.1868844159886955, "th": 2.2829687717257343, "zh": 2.191883566586911, "my": 2.35637958242922}, "t": 6171.7} | |
| {"event": "eval", "step": 36000, "phase": 2, "bpb": {"en": 1.5021, "de": 1.63317, "vi": 1.27057, "tr": 1.53462, "fi": 1.70153, "ru": 0.95916, "el": 0.96602, "hi": 0.80818, "ta": 0.69919, "th": 0.78132, "zh": 2.33918, "my": 0.75926}, "bytes_per_chunk": {"en": 4.864, "de": 5.43, "vi": 5.579, "tr": 5.965, "fi": 5.867, "ru": 5.969, "el": 6.144, "hi": 6.024, "ta": 6.101, "th": 6.172, "zh": 6.191, "my": 6.153}, "bpb_mean": 1.24619, "t": 6177.0} | |
| {"event": "train", "step": 36000, "phase": 2, "objective": "global", "lm_loss": 0.94352, "bpb": 1.36121, "chunk_loss": 0.030106, "F": 0.1772, "G": 0.23912, "chunks": 181.45, "bytes_per_chunk": 5.643, "boundary_entropy": 0.21117, "router_noise": 0.0, "grad_norm": 0.285, "lr": 0.0001829, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 6177.2} | |
| {"event": "train", "step": 36500, "phase": 2, "objective": "global", "lm_loss": 0.91836, "bpb": 1.32491, "chunk_loss": 0.030195, "F": 0.1821, "G": 0.2585, "chunks": 186.47, "bytes_per_chunk": 5.492, "boundary_entropy": 0.24037, "router_noise": 0.0, "grad_norm": 0.2729, "lr": 0.0001801, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6260.0} | |
| {"event": "train", "step": 37000, "phase": 2, "objective": "global", "lm_loss": 0.93748, "bpb": 1.35249, "chunk_loss": 0.030103, "F": 0.17996, "G": 0.24706, "chunks": 184.28, "bytes_per_chunk": 5.557, "boundary_entropy": 0.21899, "router_noise": 0.0, "grad_norm": 0.2877, "lr": 0.0001773, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6343.0} | |
| {"event": "train", "step": 37500, "phase": 2, "objective": "global", "lm_loss": 0.91383, "bpb": 1.31837, "chunk_loss": 0.03014, "F": 0.18246, "G": 0.23801, "chunks": 186.84, "bytes_per_chunk": 5.481, "boundary_entropy": 0.1927, "router_noise": 0.0, "grad_norm": 0.2976, "lr": 0.0001744, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6425.8} | |
| {"event": "train", "step": 38000, "phase": 2, "objective": "global", "lm_loss": 0.92153, "bpb": 1.32948, "chunk_loss": 0.030219, "F": 0.18707, "G": 0.23783, "chunks": 191.56, "bytes_per_chunk": 5.346, "boundary_entropy": 0.19021, "router_noise": 0.0, "grad_norm": 0.267, "lr": 0.0001716, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6508.8} | |
| {"event": "train", "step": 38500, "phase": 2, "objective": "global", "lm_loss": 0.91654, "bpb": 1.32229, "chunk_loss": 0.030299, "F": 0.18503, "G": 0.25222, "chunks": 189.47, "bytes_per_chunk": 5.405, "boundary_entropy": 0.22328, "router_noise": 0.0, "grad_norm": 0.2788, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6591.7} | |
| {"event": "train", "step": 39000, "phase": 2, "objective": "global", "lm_loss": 0.94157, "bpb": 1.3584, "chunk_loss": 0.030289, "F": 0.18726, "G": 0.26369, "chunks": 191.75, "bytes_per_chunk": 5.34, "boundary_entropy": 0.24765, "router_noise": 0.0, "grad_norm": 0.2911, "lr": 0.0001659, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6674.6} | |
| {"event": "train", "step": 39500, "phase": 2, "objective": "global", "lm_loss": 0.87996, "bpb": 1.26951, "chunk_loss": 0.030234, "F": 0.18413, "G": 0.2435, "chunks": 188.55, "bytes_per_chunk": 5.431, "boundary_entropy": 0.20498, "router_noise": 0.0, "grad_norm": 0.2908, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6757.6} | |
| {"event": "probe", "step": 40000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3416.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196288, "cps": 26.6875, "mean_p": 0.190142, "boundary_entropy": 0.061571, "bpb": 1.490608}, "de": {"chunks": 3738.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182048, "cps": 29.203125, "mean_p": 0.182083, "boundary_entropy": 0.050943, "bpb": 1.609424}, "vi": {"chunks": 4228.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229346, "cps": 33.03125, "mean_p": 0.174867, "boundary_entropy": 0.059276, "bpb": 1.23693}, "tr": {"chunks": 3116.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172126, "cps": 24.34375, "mean_p": 0.18395, "boundary_entropy": 0.131154, "bpb": 1.524386}, "fi": {"chunks": 3057.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162511, "cps": 23.882812, "mean_p": 0.171782, "boundary_entropy": 0.092955, "bpb": 1.659532}, "ru": {"chunks": 5688.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.300158, "cps": 44.4375, "mean_p": 0.225377, "boundary_entropy": 0.229562, "bpb": 0.947928}, "el": {"chunks": 5903.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.287629, "cps": 46.117188, "mean_p": 0.292587, "boundary_entropy": 0.380031, "bpb": 0.956281}, "hi": {"chunks": 6992.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.415572, "cps": 54.625, "mean_p": 0.3758, "boundary_entropy": 0.531403, "bpb": 0.791623}, "ta": {"chunks": 7983.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.427013, "cps": 62.367188, "mean_p": 0.394129, "boundary_entropy": 0.565553, "bpb": 0.685158}, "th": {"chunks": 7273.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.45216, "cps": 56.820312, "mean_p": 0.37469, "boundary_entropy": 0.543218, "bpb": 0.771871}, "zh": {"chunks": 2543.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.430579, "cps": 19.867188, "mean_p": 0.33314, "boundary_entropy": 0.44918, "bpb": 2.329916}, "my": {"chunks": 9510.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.5019, "cps": 74.296875, "mean_p": 0.391263, "boundary_entropy": 0.554136, "bpb": 0.741845}}, "bpb": {"en": 1.49061, "de": 1.60942, "vi": 1.23693, "tr": 1.52439, "fi": 1.65953, "ru": 0.94793, "el": 0.95628, "hi": 0.79162, "ta": 0.68516, "th": 0.77187, "zh": 2.32992, "my": 0.74184}, "gini_cpc": 0.215537, "gini_cps": 0.231039, "bpic": 6.048844, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9274556431749612, "vi": 1.1684176126132135, "tr": 0.8769062304292857, "fi": 0.8279227442660549, "ru": 1.5291730364017226, "el": 1.465339296027157, "hi": 2.1171547581349546, "ta": 2.1754390396059513, "th": 2.303556031737134, "zh": 2.1936087799411705, "my": 2.5569568494867503}, "t": 6843.4} | |
| {"event": "eval", "step": 40000, "phase": 2, "bpb": {"en": 1.48762, "de": 1.61704, "vi": 1.25571, "tr": 1.51734, "fi": 1.67814, "ru": 0.94573, "el": 0.9547, "hi": 0.79683, "ta": 0.69069, "th": 0.77127, "zh": 2.33133, "my": 0.74786}, "bytes_per_chunk": {"en": 4.861, "de": 5.446, "vi": 5.59, "tr": 6.019, "fi": 5.897, "ru": 6.021, "el": 6.123, "hi": 5.961, "ta": 6.224, "th": 6.148, "zh": 6.253, "my": 5.685}, "bpb_mean": 1.23286, "t": 6848.7} | |
| {"event": "train", "step": 40000, "phase": 2, "objective": "global", "lm_loss": 0.95264, "bpb": 1.37436, "chunk_loss": 0.029958, "F": 0.17462, "G": 0.26408, "chunks": 178.81, "bytes_per_chunk": 5.727, "boundary_entropy": 0.26765, "router_noise": 0.0, "grad_norm": 0.2889, "lr": 0.0001601, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 6848.9} | |
| {"event": "train", "step": 40500, "phase": 2, "objective": "global", "lm_loss": 0.90444, "bpb": 1.30483, "chunk_loss": 0.030076, "F": 0.18167, "G": 0.23137, "chunks": 186.03, "bytes_per_chunk": 5.504, "boundary_entropy": 0.17938, "router_noise": 0.0, "grad_norm": 0.291, "lr": 0.0001573, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 6931.7} | |
| {"event": "train", "step": 41000, "phase": 2, "objective": "global", "lm_loss": 0.91379, "bpb": 1.31832, "chunk_loss": 0.030173, "F": 0.1817, "G": 0.24579, "chunks": 186.06, "bytes_per_chunk": 5.504, "boundary_entropy": 0.21536, "router_noise": 0.0, "grad_norm": 0.2768, "lr": 0.0001545, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 7014.4} | |
| {"event": "train", "step": 41500, "phase": 2, "objective": "global", "lm_loss": 0.92435, "bpb": 1.33355, "chunk_loss": 0.030166, "F": 0.17982, "G": 0.24956, "chunks": 184.14, "bytes_per_chunk": 5.561, "boundary_entropy": 0.22609, "router_noise": 0.0, "grad_norm": 0.3023, "lr": 0.0001516, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 7096.8} | |
| {"event": "train", "step": 42000, "phase": 2, "objective": "global", "lm_loss": 0.86493, "bpb": 1.24783, "chunk_loss": 0.030096, "F": 0.18025, "G": 0.24517, "chunks": 184.58, "bytes_per_chunk": 5.548, "boundary_entropy": 0.21457, "router_noise": 0.0, "grad_norm": 0.2842, "lr": 0.0001488, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 7179.7} | |
| {"event": "train", "step": 42500, "phase": 2, "objective": "global", "lm_loss": 0.98573, "bpb": 1.4221, "chunk_loss": 0.030135, "F": 0.18121, "G": 0.24228, "chunks": 185.56, "bytes_per_chunk": 5.518, "boundary_entropy": 0.21037, "router_noise": 0.0, "grad_norm": 0.2913, "lr": 0.0001459, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 7262.3} | |
| {"event": "train", "step": 43000, "phase": 2, "objective": "global", "lm_loss": 0.95435, "bpb": 1.37683, "chunk_loss": 0.030096, "F": 0.18176, "G": 0.23728, "chunks": 186.12, "bytes_per_chunk": 5.502, "boundary_entropy": 0.19453, "router_noise": 0.0, "grad_norm": 0.2979, "lr": 0.0001431, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 7345.1} | |
| {"event": "train", "step": 43500, "phase": 2, "objective": "global", "lm_loss": 0.96999, "bpb": 1.39939, "chunk_loss": 0.030179, "F": 0.18529, "G": 0.24557, "chunks": 189.73, "bytes_per_chunk": 5.397, "boundary_entropy": 0.20738, "router_noise": 0.0, "grad_norm": 0.2953, "lr": 0.0001403, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 7428.3} | |
| {"event": "probe", "step": 44000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3411.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196001, "cps": 26.648438, "mean_p": 0.190719, "boundary_entropy": 0.057466, "bpb": 1.480875}, "de": {"chunks": 3740.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182146, "cps": 29.21875, "mean_p": 0.181607, "boundary_entropy": 0.045875, "bpb": 1.593959}, "vi": {"chunks": 4235.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229726, "cps": 33.085938, "mean_p": 0.17537, "boundary_entropy": 0.059977, "bpb": 1.221263}, "tr": {"chunks": 3203.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176932, "cps": 25.023438, "mean_p": 0.183863, "boundary_entropy": 0.121972, "bpb": 1.507941}, "fi": {"chunks": 3082.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.16384, "cps": 24.078125, "mean_p": 0.171279, "boundary_entropy": 0.085553, "bpb": 1.637404}, "ru": {"chunks": 5838.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.308074, "cps": 45.609375, "mean_p": 0.226575, "boundary_entropy": 0.221735, "bpb": 0.938294}, "el": {"chunks": 5885.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.286751, "cps": 45.976562, "mean_p": 0.296714, "boundary_entropy": 0.381752, "bpb": 0.942948}, "hi": {"chunks": 6943.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.41266, "cps": 54.242188, "mean_p": 0.371236, "boundary_entropy": 0.523866, "bpb": 0.785835}, "ta": {"chunks": 8378.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.448141, "cps": 65.453125, "mean_p": 0.404016, "boundary_entropy": 0.573744, "bpb": 0.680327}, "th": {"chunks": 7344.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.456574, "cps": 57.375, "mean_p": 0.38146, "boundary_entropy": 0.549761, "bpb": 0.764092}, "zh": {"chunks": 2566.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.434473, "cps": 20.046875, "mean_p": 0.330495, "boundary_entropy": 0.443285, "bpb": 2.305012}, "my": {"chunks": 8747.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.461632, "cps": 68.335938, "mean_p": 0.379717, "boundary_entropy": 0.543958, "bpb": 0.732692}}, "bpb": {"en": 1.48087, "de": 1.59396, "vi": 1.22126, "tr": 1.50794, "fi": 1.6374, "ru": 0.93829, "el": 0.94295, "hi": 0.78583, "ta": 0.68033, "th": 0.76409, "zh": 2.30501, "my": 0.73269}, "gini_cpc": 0.208878, "gini_cps": 0.223995, "bpic": 6.056003, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.929312108474502, "vi": 1.1720676345304128, "tr": 0.9027111110657664, "fi": 0.8359169888200018, "ru": 1.5717999735451575, "el": 1.4630124530253312, "hi": 2.10539939179379, "ta": 2.2864267227255652, "th": 2.329453275516037, "zh": 2.216693308029251, "my": 2.355256175815345}, "t": 7514.1} | |
| {"event": "eval", "step": 44000, "phase": 2, "bpb": {"en": 1.47806, "de": 1.60371, "vi": 1.24048, "tr": 1.50065, "fi": 1.65722, "ru": 0.93643, "el": 0.94234, "hi": 0.79085, "ta": 0.68533, "th": 0.76359, "zh": 2.30861, "my": 0.73848}, "bytes_per_chunk": {"en": 4.87, "de": 5.441, "vi": 5.589, "tr": 5.874, "fi": 5.879, "ru": 5.929, "el": 6.038, "hi": 5.871, "ta": 5.827, "th": 5.992, "zh": 6.138, "my": 6.146}, "bpb_mean": 1.22048, "t": 7519.4} | |
| {"event": "train", "step": 44000, "phase": 2, "objective": "global", "lm_loss": 0.87192, "bpb": 1.25792, "chunk_loss": 0.03018, "F": 0.18471, "G": 0.2475, "chunks": 189.14, "bytes_per_chunk": 5.414, "boundary_entropy": 0.21217, "router_noise": 0.0, "grad_norm": 0.2909, "lr": 0.0001375, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 7519.6} | |
| {"event": "train", "step": 44500, "phase": 2, "objective": "global", "lm_loss": 0.87028, "bpb": 1.25555, "chunk_loss": 0.030159, "F": 0.18291, "G": 0.25218, "chunks": 187.3, "bytes_per_chunk": 5.467, "boundary_entropy": 0.22912, "router_noise": 0.0, "grad_norm": 0.2775, "lr": 0.0001347, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 7602.3} | |
| {"event": "train", "step": 45000, "phase": 2, "objective": "global", "lm_loss": 0.91042, "bpb": 1.31346, "chunk_loss": 0.030138, "F": 0.18105, "G": 0.24181, "chunks": 185.39, "bytes_per_chunk": 5.523, "boundary_entropy": 0.20697, "router_noise": 0.0, "grad_norm": 0.2989, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 7684.7} | |
| {"event": "train", "step": 45500, "phase": 2, "objective": "global", "lm_loss": 0.89219, "bpb": 1.28716, "chunk_loss": 0.030116, "F": 0.18109, "G": 0.24064, "chunks": 185.44, "bytes_per_chunk": 5.522, "boundary_entropy": 0.20495, "router_noise": 0.0, "grad_norm": 0.2921, "lr": 0.0001292, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 7767.6} | |
| {"event": "train", "step": 46000, "phase": 2, "objective": "global", "lm_loss": 0.9321, "bpb": 1.34473, "chunk_loss": 0.030065, "F": 0.18106, "G": 0.23834, "chunks": 185.41, "bytes_per_chunk": 5.523, "boundary_entropy": 0.19952, "router_noise": 0.0, "grad_norm": 0.3027, "lr": 0.0001264, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 7850.2} | |
| {"event": "train", "step": 46500, "phase": 2, "objective": "global", "lm_loss": 0.90135, "bpb": 1.30037, "chunk_loss": 0.030169, "F": 0.18393, "G": 0.23067, "chunks": 188.34, "bytes_per_chunk": 5.437, "boundary_entropy": 0.17484, "router_noise": 0.0, "grad_norm": 0.2867, "lr": 0.0001237, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 7933.2} | |
| {"event": "train", "step": 47000, "phase": 2, "objective": "global", "lm_loss": 0.91472, "bpb": 1.31966, "chunk_loss": 0.030111, "F": 0.18294, "G": 0.24459, "chunks": 187.33, "bytes_per_chunk": 5.466, "boundary_entropy": 0.20744, "router_noise": 0.0, "grad_norm": 0.3009, "lr": 0.000121, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 8015.8} | |
| {"event": "train", "step": 47500, "phase": 2, "objective": "global", "lm_loss": 0.92331, "bpb": 1.33205, "chunk_loss": 0.030077, "F": 0.17744, "G": 0.25367, "chunks": 181.7, "bytes_per_chunk": 5.636, "boundary_entropy": 0.2358, "router_noise": 0.0, "grad_norm": 0.3122, "lr": 0.0001183, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 8098.7} | |
| {"event": "probe", "step": 48000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3411.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196001, "cps": 26.648438, "mean_p": 0.189103, "boundary_entropy": 0.059026, "bpb": 1.465021}, "de": {"chunks": 3749.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182584, "cps": 29.289062, "mean_p": 0.181487, "boundary_entropy": 0.045057, "bpb": 1.576791}, "vi": {"chunks": 4235.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229726, "cps": 33.085938, "mean_p": 0.174886, "boundary_entropy": 0.057817, "bpb": 1.205137}, "tr": {"chunks": 3235.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.1787, "cps": 25.273438, "mean_p": 0.177399, "boundary_entropy": 0.096959, "bpb": 1.493437}, "fi": {"chunks": 3055.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162405, "cps": 23.867188, "mean_p": 0.169416, "boundary_entropy": 0.082164, "bpb": 1.623322}, "ru": {"chunks": 5788.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305435, "cps": 45.21875, "mean_p": 0.22432, "boundary_entropy": 0.216314, "bpb": 0.926288}, "el": {"chunks": 5913.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.288116, "cps": 46.195312, "mean_p": 0.286739, "boundary_entropy": 0.365435, "bpb": 0.931207}, "hi": {"chunks": 7200.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.427935, "cps": 56.25, "mean_p": 0.371009, "boundary_entropy": 0.518599, "bpb": 0.772371}, "ta": {"chunks": 8295.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.443702, "cps": 64.804688, "mean_p": 0.392412, "boundary_entropy": 0.558243, "bpb": 0.669889}, "th": {"chunks": 7351.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.45701, "cps": 57.429688, "mean_p": 0.373827, "boundary_entropy": 0.540546, "bpb": 0.752925}, "zh": {"chunks": 2541.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.43024, "cps": 19.851562, "mean_p": 0.333827, "boundary_entropy": 0.446934, "bpb": 2.291198}, "my": {"chunks": 8507.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.448966, "cps": 66.460938, "mean_p": 0.368651, "boundary_entropy": 0.531107, "bpb": 0.722069}}, "bpb": {"en": 1.46502, "de": 1.57679, "vi": 1.20514, "tr": 1.49344, "fi": 1.62332, "ru": 0.92629, "el": 0.93121, "hi": 0.77237, "ta": 0.66989, "th": 0.75292, "zh": 2.2912, "my": 0.72207}, "gini_cpc": 0.206266, "gini_cps": 0.222013, "bpic": 6.064807, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.931548421035002, "vi": 1.1720676345304128, "tr": 0.9117297671863112, "fi": 0.8285939003390996, "ru": 1.5583381717847502, "el": 1.4699732599386208, "hi": 2.1833322225140845, "ta": 2.2637753240640444, "th": 2.331673614967101, "zh": 2.1950965298917873, "my": 2.2906327069465124}, "t": 8184.0} | |
| {"event": "eval", "step": 48000, "phase": 2, "bpb": {"en": 1.46216, "de": 1.58675, "vi": 1.22406, "tr": 1.48658, "fi": 1.64293, "ru": 0.92497, "el": 0.93025, "hi": 0.77871, "ta": 0.67476, "th": 0.7523, "zh": 2.29562, "my": 0.7282}, "bytes_per_chunk": {"en": 4.874, "de": 5.429, "vi": 5.587, "tr": 5.813, "fi": 5.88, "ru": 5.971, "el": 6.075, "hi": 5.836, "ta": 6.014, "th": 6.002, "zh": 6.246, "my": 6.358}, "bpb_mean": 1.20728, "t": 8189.2} | |
| {"event": "train", "step": 48000, "phase": 2, "objective": "global", "lm_loss": 0.90784, "bpb": 1.30973, "chunk_loss": 0.030159, "F": 0.18343, "G": 0.24071, "chunks": 187.83, "bytes_per_chunk": 5.452, "boundary_entropy": 0.20071, "router_noise": 0.0, "grad_norm": 0.2901, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.181, "alpha": 0.03, "t": 8189.4} | |
| {"event": "train", "step": 48500, "phase": 2, "objective": "global", "lm_loss": 0.93703, "bpb": 1.35185, "chunk_loss": 0.030015, "F": 0.1761, "G": 0.23927, "chunks": 180.33, "bytes_per_chunk": 5.679, "boundary_entropy": 0.20831, "router_noise": 0.0, "grad_norm": 0.3018, "lr": 0.000113, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 8272.0} | |
| {"event": "train", "step": 49000, "phase": 2, "objective": "global", "lm_loss": 0.82823, "bpb": 1.19488, "chunk_loss": 0.030177, "F": 0.18427, "G": 0.24764, "chunks": 188.69, "bytes_per_chunk": 5.427, "boundary_entropy": 0.21427, "router_noise": 0.0, "grad_norm": 0.2878, "lr": 0.0001104, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 8355.0} | |
| {"event": "train", "step": 49500, "phase": 2, "objective": "global", "lm_loss": 0.97764, "bpb": 1.41044, "chunk_loss": 0.030047, "F": 0.17728, "G": 0.24843, "chunks": 181.53, "bytes_per_chunk": 5.641, "boundary_entropy": 0.22588, "router_noise": 0.0, "grad_norm": 0.3086, "lr": 0.0001078, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 8437.5} | |
| {"event": "train", "step": 50000, "phase": 2, "objective": "global", "lm_loss": 0.90035, "bpb": 1.29893, "chunk_loss": 0.030126, "F": 0.18201, "G": 0.24103, "chunks": 186.38, "bytes_per_chunk": 5.494, "boundary_entropy": 0.20054, "router_noise": 0.0, "grad_norm": 0.2849, "lr": 0.0001052, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 8520.3} | |
| {"event": "train", "step": 50500, "phase": 2, "objective": "global", "lm_loss": 0.93405, "bpb": 1.34754, "chunk_loss": 0.030059, "F": 0.17502, "G": 0.26551, "chunks": 179.22, "bytes_per_chunk": 5.714, "boundary_entropy": 0.2703, "router_noise": 0.0, "grad_norm": 0.317, "lr": 0.0001027, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 8602.8} | |
| {"event": "train", "step": 51000, "phase": 2, "objective": "global", "lm_loss": 0.84725, "bpb": 1.22232, "chunk_loss": 0.030049, "F": 0.1756, "G": 0.24638, "chunks": 179.81, "bytes_per_chunk": 5.695, "boundary_entropy": 0.22874, "router_noise": 0.0, "grad_norm": 0.3004, "lr": 0.0001002, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 8685.3} | |
| {"event": "train", "step": 51500, "phase": 2, "objective": "global", "lm_loss": 0.87613, "bpb": 1.26399, "chunk_loss": 0.030166, "F": 0.18355, "G": 0.24332, "chunks": 187.95, "bytes_per_chunk": 5.448, "boundary_entropy": 0.21001, "router_noise": 0.0, "grad_norm": 0.2978, "lr": 9.77e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 8768.1} | |
| {"event": "probe", "step": 52000, "phase": 2, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3407.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195771, "cps": 26.617188, "mean_p": 0.189339, "boundary_entropy": 0.057314, "bpb": 1.453239}, "de": {"chunks": 3737.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182, "cps": 29.195312, "mean_p": 0.180998, "boundary_entropy": 0.04368, "bpb": 1.565166}, "vi": {"chunks": 4234.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229672, "cps": 33.078125, "mean_p": 0.174605, "boundary_entropy": 0.052583, "bpb": 1.195059}, "tr": {"chunks": 3250.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179528, "cps": 25.390625, "mean_p": 0.175359, "boundary_entropy": 0.087398, "bpb": 1.478565}, "fi": {"chunks": 3048.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162033, "cps": 23.8125, "mean_p": 0.168837, "boundary_entropy": 0.080421, "bpb": 1.604337}, "ru": {"chunks": 5802.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306174, "cps": 45.328125, "mean_p": 0.223421, "boundary_entropy": 0.213128, "bpb": 0.916291}, "el": {"chunks": 6103.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.297374, "cps": 47.679688, "mean_p": 0.2847, "boundary_entropy": 0.354917, "bpb": 0.920231}, "hi": {"chunks": 7374.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.438276, "cps": 57.609375, "mean_p": 0.376544, "boundary_entropy": 0.518648, "bpb": 0.764159}, "ta": {"chunks": 8863.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.474084, "cps": 69.242188, "mean_p": 0.406932, "boundary_entropy": 0.565269, "bpb": 0.661331}, "th": {"chunks": 7660.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.47622, "cps": 59.84375, "mean_p": 0.374122, "boundary_entropy": 0.535436, "bpb": 0.742777}, "zh": {"chunks": 2594.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.439214, "cps": 20.265625, "mean_p": 0.334712, "boundary_entropy": 0.442405, "bpb": 2.271364}, "my": {"chunks": 9508.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.501794, "cps": 74.28125, "mean_p": 0.375907, "boundary_entropy": 0.5308, "bpb": 0.712429}}, "bpb": {"en": 1.45324, "de": 1.56517, "vi": 1.19506, "tr": 1.47856, "fi": 1.60434, "ru": 0.91629, "el": 0.92023, "hi": 0.76416, "ta": 0.66133, "th": 0.74278, "zh": 2.27136, "my": 0.71243}, "gini_cpc": 0.221184, "gini_cps": 0.238589, "bpic": 5.852104, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9296568578295384, "vi": 1.173166622205802, "tr": 0.9170326449986049, "fi": 0.8276659063134892, "ru": 1.5639414739017063, "el": 1.5189885892768216, "hi": 2.2387213808124256, "ta": 2.4216270907724002, "th": 2.432538323017788, "zh": 2.2435126143650983, "my": 2.5631721971658767}, "t": 8854.0} | |
| {"event": "eval", "step": 52000, "phase": 2, "bpb": {"en": 1.45011, "de": 1.57382, "vi": 1.21383, "tr": 1.47153, "fi": 1.62409, "ru": 0.91492, "el": 0.91972, "hi": 0.77014, "ta": 0.66658, "th": 0.74279, "zh": 2.27415, "my": 0.71919}, "bytes_per_chunk": {"en": 4.88, "de": 5.441, "vi": 5.589, "tr": 5.792, "fi": 5.894, "ru": 5.952, "el": 5.876, "hi": 5.645, "ta": 5.592, "th": 5.805, "zh": 6.061, "my": 5.707}, "bpb_mean": 1.19507, "t": 8859.2} | |
| {"event": "train", "step": 52000, "phase": 2, "objective": "global", "lm_loss": 0.86868, "bpb": 1.25324, "chunk_loss": 0.030141, "F": 0.18307, "G": 0.22974, "chunks": 187.47, "bytes_per_chunk": 5.462, "boundary_entropy": 0.1753, "router_noise": 0.0, "grad_norm": 0.2911, "lr": 9.52e-05, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 8859.4} | |
| {"event": "train", "step": 52500, "phase": 2, "objective": "global", "lm_loss": 0.88277, "bpb": 1.27357, "chunk_loss": 0.030116, "F": 0.1843, "G": 0.23784, "chunks": 188.72, "bytes_per_chunk": 5.426, "boundary_entropy": 0.1885, "router_noise": 0.0, "grad_norm": 0.2958, "lr": 9.28e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 8942.4} | |
| {"event": "train", "step": 53000, "phase": 2, "objective": "global", "lm_loss": 0.89785, "bpb": 1.29533, "chunk_loss": 0.030045, "F": 0.17697, "G": 0.25032, "chunks": 181.22, "bytes_per_chunk": 5.651, "boundary_entropy": 0.23204, "router_noise": 0.0, "grad_norm": 0.3195, "lr": 9.04e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9025.4} | |
| {"event": "phase_transition", "step": 53410, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 9093.4} | |
| {"event": "train", "step": 53500, "phase": 3, "objective": "global", "lm_loss": 0.92561, "bpb": 1.33537, "chunk_loss": 0.030071, "F": 0.18065, "G": 0.24141, "chunks": 184.98, "bytes_per_chunk": 5.536, "boundary_entropy": 0.20039, "router_noise": 0.0, "grad_norm": 0.3142, "lr": 8.8e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9108.5} | |
| {"event": "train", "step": 54000, "phase": 3, "objective": "global", "lm_loss": 0.87092, "bpb": 1.25648, "chunk_loss": 0.030148, "F": 0.18198, "G": 0.22966, "chunks": 186.34, "bytes_per_chunk": 5.495, "boundary_entropy": 0.1802, "router_noise": 0.0, "grad_norm": 0.3327, "lr": 8.57e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9191.4} | |
| {"event": "train", "step": 54500, "phase": 3, "objective": "global", "lm_loss": 0.92283, "bpb": 1.33136, "chunk_loss": 0.030055, "F": 0.18105, "G": 0.23769, "chunks": 185.39, "bytes_per_chunk": 5.523, "boundary_entropy": 0.19677, "router_noise": 0.0, "grad_norm": 0.3078, "lr": 8.34e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 9274.0} | |
| {"event": "train", "step": 55000, "phase": 3, "objective": "global", "lm_loss": 0.85254, "bpb": 1.22996, "chunk_loss": 0.030079, "F": 0.1823, "G": 0.21928, "chunks": 186.67, "bytes_per_chunk": 5.486, "boundary_entropy": 0.14816, "router_noise": 0.0, "grad_norm": 0.3071, "lr": 8.11e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9357.0} | |
| {"event": "train", "step": 55500, "phase": 3, "objective": "global", "lm_loss": 0.92917, "bpb": 1.3405, "chunk_loss": 0.03023, "F": 0.18425, "G": 0.24505, "chunks": 188.67, "bytes_per_chunk": 5.427, "boundary_entropy": 0.20703, "router_noise": 0.0, "grad_norm": 0.3145, "lr": 7.89e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9439.8} | |
| {"event": "probe", "step": 56000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3403.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195541, "cps": 26.585938, "mean_p": 0.187815, "boundary_entropy": 0.058176, "bpb": 1.44289}, "de": {"chunks": 3730.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.181659, "cps": 29.140625, "mean_p": 0.179932, "boundary_entropy": 0.043017, "bpb": 1.551686}, "vi": {"chunks": 4238.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229889, "cps": 33.109375, "mean_p": 0.173924, "boundary_entropy": 0.053122, "bpb": 1.18107}, "tr": {"chunks": 3242.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.179086, "cps": 25.328125, "mean_p": 0.173531, "boundary_entropy": 0.081123, "bpb": 1.468134}, "fi": {"chunks": 3049.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162086, "cps": 23.820312, "mean_p": 0.166609, "boundary_entropy": 0.075986, "bpb": 1.589028}, "ru": {"chunks": 5783.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305172, "cps": 45.179688, "mean_p": 0.217704, "boundary_entropy": 0.200458, "bpb": 0.905496}, "el": {"chunks": 5902.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.28758, "cps": 46.109375, "mean_p": 0.27269, "boundary_entropy": 0.340653, "bpb": 0.908274}, "hi": {"chunks": 6970.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.414264, "cps": 54.453125, "mean_p": 0.361334, "boundary_entropy": 0.506914, "bpb": 0.755366}, "ta": {"chunks": 8317.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.444878, "cps": 64.976562, "mean_p": 0.395721, "boundary_entropy": 0.558194, "bpb": 0.654889}, "th": {"chunks": 7544.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.469008, "cps": 58.9375, "mean_p": 0.367224, "boundary_entropy": 0.52869, "bpb": 0.73886}, "zh": {"chunks": 2516.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.426007, "cps": 19.65625, "mean_p": 0.325977, "boundary_entropy": 0.432874, "bpb": 2.258621}, "my": {"chunks": 8666.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.457357, "cps": 67.703125, "mean_p": 0.374967, "boundary_entropy": 0.533648, "bpb": 0.702944}}, "bpb": {"en": 1.44289, "de": 1.55169, "vi": 1.18107, "tr": 1.46813, "fi": 1.58903, "ru": 0.9055, "el": 0.90827, "hi": 0.75537, "ta": 0.65489, "th": 0.73886, "zh": 2.25862, "my": 0.70294}, "gini_cpc": 0.209773, "gini_cps": 0.225, "bpic": 6.05715, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9290061643850223, "vi": 1.1756552336078323, "tr": 0.9158505913909646, "fi": 0.8289106356121497, "ru": 1.5606522725175753, "el": 1.4706879399901795, "hi": 2.1185556490449167, "ta": 2.275115258259373, "th": 2.398516915033463, "zh": 2.1786093603391126, "my": 2.3389313754190364}, "t": 9525.5} | |
| {"event": "eval", "step": 56000, "phase": 3, "bpb": {"en": 1.43934, "de": 1.56123, "vi": 1.20142, "tr": 1.46144, "fi": 1.60897, "ru": 0.90429, "el": 0.90801, "hi": 0.76112, "ta": 0.66009, "th": 0.73883, "zh": 2.26235, "my": 0.70922}, "bytes_per_chunk": {"en": 4.885, "de": 5.451, "vi": 5.59, "tr": 5.794, "fi": 5.902, "ru": 6.012, "el": 6.077, "hi": 5.958, "ta": 5.967, "th": 5.89, "zh": 6.272, "my": 6.134}, "bpb_mean": 1.18469, "t": 9530.8} | |
| {"event": "train", "step": 56000, "phase": 3, "objective": "global", "lm_loss": 0.91173, "bpb": 1.31535, "chunk_loss": 0.030111, "F": 0.18219, "G": 0.24211, "chunks": 186.56, "bytes_per_chunk": 5.489, "boundary_entropy": 0.2076, "router_noise": 0.0, "grad_norm": 0.324, "lr": 7.68e-05, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 9531.0} | |
| {"event": "train", "step": 56500, "phase": 3, "objective": "global", "lm_loss": 0.89447, "bpb": 1.29045, "chunk_loss": 0.030061, "F": 0.18001, "G": 0.24179, "chunks": 184.33, "bytes_per_chunk": 5.555, "boundary_entropy": 0.20746, "router_noise": 0.0, "grad_norm": 0.3183, "lr": 7.46e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9614.0} | |
| {"event": "train", "step": 57000, "phase": 3, "objective": "global", "lm_loss": 0.91441, "bpb": 1.31921, "chunk_loss": 0.030223, "F": 0.18701, "G": 0.2393, "chunks": 191.5, "bytes_per_chunk": 5.347, "boundary_entropy": 0.1921, "router_noise": 0.0, "grad_norm": 0.3177, "lr": 7.25e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 9696.7} | |
| {"event": "train", "step": 57500, "phase": 3, "objective": "global", "lm_loss": 0.87055, "bpb": 1.25594, "chunk_loss": 0.030137, "F": 0.18547, "G": 0.2164, "chunks": 189.92, "bytes_per_chunk": 5.392, "boundary_entropy": 0.13912, "router_noise": 0.0, "grad_norm": 0.3069, "lr": 7.05e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9779.5} | |
| {"event": "train", "step": 58000, "phase": 3, "objective": "global", "lm_loss": 0.92571, "bpb": 1.33551, "chunk_loss": 0.030083, "F": 0.18317, "G": 0.23498, "chunks": 187.56, "bytes_per_chunk": 5.46, "boundary_entropy": 0.18539, "router_noise": 0.0, "grad_norm": 0.3271, "lr": 6.84e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 9862.0} | |
| {"event": "train", "step": 58500, "phase": 3, "objective": "global", "lm_loss": 0.91215, "bpb": 1.31596, "chunk_loss": 0.030116, "F": 0.18155, "G": 0.2333, "chunks": 185.91, "bytes_per_chunk": 5.508, "boundary_entropy": 0.1852, "router_noise": 0.0, "grad_norm": 0.3307, "lr": 6.65e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 9944.9} | |
| {"event": "train", "step": 59000, "phase": 3, "objective": "global", "lm_loss": 0.93861, "bpb": 1.35413, "chunk_loss": 0.030214, "F": 0.18401, "G": 0.23508, "chunks": 188.42, "bytes_per_chunk": 5.435, "boundary_entropy": 0.1888, "router_noise": 0.0, "grad_norm": 0.3427, "lr": 6.45e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 10027.6} | |
| {"event": "train", "step": 59500, "phase": 3, "objective": "global", "lm_loss": 0.89621, "bpb": 1.29296, "chunk_loss": 0.030101, "F": 0.18036, "G": 0.24764, "chunks": 184.69, "bytes_per_chunk": 5.545, "boundary_entropy": 0.22022, "router_noise": 0.0, "grad_norm": 0.3214, "lr": 6.27e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 10110.3} | |
| {"event": "probe", "step": 60000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3408.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195828, "cps": 26.625, "mean_p": 0.188678, "boundary_entropy": 0.055428, "bpb": 1.433126}, "de": {"chunks": 3741.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182195, "cps": 29.226562, "mean_p": 0.180401, "boundary_entropy": 0.040549, "bpb": 1.54118}, "vi": {"chunks": 4235.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229726, "cps": 33.085938, "mean_p": 0.173995, "boundary_entropy": 0.05157, "bpb": 1.171096}, "tr": {"chunks": 3196.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.176545, "cps": 24.96875, "mean_p": 0.170764, "boundary_entropy": 0.075138, "bpb": 1.454099}, "fi": {"chunks": 3062.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162777, "cps": 23.921875, "mean_p": 0.168301, "boundary_entropy": 0.074813, "bpb": 1.572538}, "ru": {"chunks": 5750.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.30343, "cps": 44.921875, "mean_p": 0.213836, "boundary_entropy": 0.190417, "bpb": 0.896377}, "el": {"chunks": 6201.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.302149, "cps": 48.445312, "mean_p": 0.279928, "boundary_entropy": 0.34133, "bpb": 0.900815}, "hi": {"chunks": 7285.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.432987, "cps": 56.914062, "mean_p": 0.369012, "boundary_entropy": 0.507854, "bpb": 0.747862}, "ta": {"chunks": 8506.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.454988, "cps": 66.453125, "mean_p": 0.391325, "boundary_entropy": 0.55208, "bpb": 0.648294}, "th": {"chunks": 7849.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.48797, "cps": 61.320312, "mean_p": 0.366206, "boundary_entropy": 0.521742, "bpb": 0.730463}, "zh": {"chunks": 2575.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.435997, "cps": 20.117188, "mean_p": 0.333109, "boundary_entropy": 0.441226, "bpb": 2.244872}, "my": {"chunks": 9369.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.494459, "cps": 73.195312, "mean_p": 0.38105, "boundary_entropy": 0.534371, "bpb": 0.696205}}, "bpb": {"en": 1.43313, "de": 1.54118, "vi": 1.1711, "tr": 1.4541, "fi": 1.57254, "ru": 0.89638, "el": 0.90082, "hi": 0.74786, "ta": 0.64829, "th": 0.73046, "zh": 2.24487, "my": 0.69621}, "gini_cpc": 0.219674, "gini_cps": 0.236, "bpic": 5.888289, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9303788628130738, "vi": 1.1730993842086965, "tr": 0.9015311834549249, "fi": 0.8312235461570656, "ru": 1.54946997287153, "el": 1.5429271470071708, "hi": 2.21105230939874, "ta": 2.3234024500035786, "th": 2.491826487366849, "zh": 2.2264263067398633, "my": 2.524959386623732}, "t": 10195.8} | |
| {"event": "eval", "step": 60000, "phase": 3, "bpb": {"en": 1.43003, "de": 1.55102, "vi": 1.19062, "tr": 1.44775, "fi": 1.59391, "ru": 0.89546, "el": 0.90017, "hi": 0.75333, "ta": 0.65375, "th": 0.72982, "zh": 2.24881, "my": 0.70247}, "bytes_per_chunk": {"en": 4.881, "de": 5.449, "vi": 5.59, "tr": 5.869, "fi": 5.891, "ru": 6.012, "el": 5.846, "hi": 5.731, "ta": 5.896, "th": 5.722, "zh": 6.153, "my": 5.756}, "bpb_mean": 1.17476, "t": 10201.1} | |
| {"event": "train", "step": 60000, "phase": 3, "objective": "global", "lm_loss": 0.90891, "bpb": 1.31128, "chunk_loss": 0.030115, "F": 0.18138, "G": 0.23797, "chunks": 185.73, "bytes_per_chunk": 5.513, "boundary_entropy": 0.19222, "router_noise": 0.0, "grad_norm": 0.3351, "lr": 6.08e-05, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 10201.3} | |
| {"event": "train", "step": 60500, "phase": 3, "objective": "global", "lm_loss": 0.82444, "bpb": 1.18942, "chunk_loss": 0.030089, "F": 0.18161, "G": 0.23091, "chunks": 185.97, "bytes_per_chunk": 5.506, "boundary_entropy": 0.18051, "router_noise": 0.0, "grad_norm": 0.3132, "lr": 5.9e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 10283.7} | |
| {"event": "train", "step": 61000, "phase": 3, "objective": "global", "lm_loss": 0.83462, "bpb": 1.2041, "chunk_loss": 0.030086, "F": 0.179, "G": 0.23401, "chunks": 183.3, "bytes_per_chunk": 5.587, "boundary_entropy": 0.19319, "router_noise": 0.0, "grad_norm": 0.3459, "lr": 5.73e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 10366.6} | |
| {"event": "train", "step": 61500, "phase": 3, "objective": "global", "lm_loss": 0.89809, "bpb": 1.29566, "chunk_loss": 0.030089, "F": 0.18381, "G": 0.22131, "chunks": 188.22, "bytes_per_chunk": 5.44, "boundary_entropy": 0.15192, "router_noise": 0.0, "grad_norm": 0.3185, "lr": 5.56e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 10449.3} | |
| {"event": "train", "step": 62000, "phase": 3, "objective": "global", "lm_loss": 0.8924, "bpb": 1.28746, "chunk_loss": 0.030041, "F": 0.1803, "G": 0.23268, "chunks": 184.62, "bytes_per_chunk": 5.546, "boundary_entropy": 0.17972, "router_noise": 0.0, "grad_norm": 0.3529, "lr": 5.39e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 10531.9} | |
| {"event": "train", "step": 62500, "phase": 3, "objective": "global", "lm_loss": 0.91931, "bpb": 1.32628, "chunk_loss": 0.030128, "F": 0.17957, "G": 0.24122, "chunks": 183.88, "bytes_per_chunk": 5.569, "boundary_entropy": 0.20593, "router_noise": 0.0, "grad_norm": 0.3458, "lr": 5.23e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 10614.5} | |
| {"event": "train", "step": 63000, "phase": 3, "objective": "global", "lm_loss": 0.88896, "bpb": 1.2825, "chunk_loss": 0.030175, "F": 0.18484, "G": 0.22226, "chunks": 189.28, "bytes_per_chunk": 5.41, "boundary_entropy": 0.15513, "router_noise": 0.0, "grad_norm": 0.3366, "lr": 5.08e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 10697.0} | |
| {"event": "train", "step": 63500, "phase": 3, "objective": "global", "lm_loss": 0.80436, "bpb": 1.16045, "chunk_loss": 0.030104, "F": 0.1824, "G": 0.23823, "chunks": 186.78, "bytes_per_chunk": 5.482, "boundary_entropy": 0.19196, "router_noise": 0.0, "grad_norm": 0.3292, "lr": 4.93e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 10779.9} | |
| {"event": "probe", "step": 64000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3405.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195656, "cps": 26.601562, "mean_p": 0.187997, "boundary_entropy": 0.055546, "bpb": 1.42488}, "de": {"chunks": 3728.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.181561, "cps": 29.125, "mean_p": 0.179149, "boundary_entropy": 0.039561, "bpb": 1.532273}, "vi": {"chunks": 4231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229509, "cps": 33.054688, "mean_p": 0.17317, "boundary_entropy": 0.048798, "bpb": 1.162431}, "tr": {"chunks": 3219.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177816, "cps": 25.148438, "mean_p": 0.170662, "boundary_entropy": 0.07212, "bpb": 1.443033}, "fi": {"chunks": 3048.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162033, "cps": 23.8125, "mean_p": 0.165962, "boundary_entropy": 0.070191, "bpb": 1.561595}, "ru": {"chunks": 5799.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306016, "cps": 45.304688, "mean_p": 0.212861, "boundary_entropy": 0.182919, "bpb": 0.891455}, "el": {"chunks": 6060.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.295278, "cps": 47.34375, "mean_p": 0.268329, "boundary_entropy": 0.324318, "bpb": 0.892274}, "hi": {"chunks": 7029.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.417771, "cps": 54.914062, "mean_p": 0.35813, "boundary_entropy": 0.497806, "bpb": 0.740147}, "ta": {"chunks": 8262.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.441936, "cps": 64.546875, "mean_p": 0.385225, "boundary_entropy": 0.544288, "bpb": 0.641313}, "th": {"chunks": 7528.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.468014, "cps": 58.8125, "mean_p": 0.356123, "boundary_entropy": 0.515388, "bpb": 0.723406}, "zh": {"chunks": 2599.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.440061, "cps": 20.304688, "mean_p": 0.33114, "boundary_entropy": 0.434845, "bpb": 2.23399}, "my": {"chunks": 9411.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.496675, "cps": 73.523438, "mean_p": 0.376491, "boundary_entropy": 0.529138, "bpb": 0.688762}}, "bpb": {"en": 1.42488, "de": 1.53227, "vi": 1.16243, "tr": 1.44303, "fi": 1.5616, "ru": 0.89146, "el": 0.89227, "hi": 0.74015, "ta": 0.64131, "th": 0.72341, "zh": 2.23399, "my": 0.68876}, "gini_cpc": 0.216688, "gini_cps": 0.231535, "bpic": 5.966837, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9279626585848374, "vi": 1.1730239715920563, "tr": 0.90881906545049, "fi": 0.82815205368871, "ru": 1.5640509571908454, "el": 1.5091721526144193, "hi": 2.1352339907443167, "ta": 2.258742508126628, "th": 2.3920240900831073, "zh": 2.2491573565895058, "my": 2.538513055438534}, "t": 10865.5} | |
| {"event": "eval", "step": 64000, "phase": 3, "bpb": {"en": 1.4215, "de": 1.54145, "vi": 1.18169, "tr": 1.43696, "fi": 1.58226, "ru": 0.88944, "el": 0.8915, "hi": 0.74617, "ta": 0.64711, "th": 0.72284, "zh": 2.23768, "my": 0.69486}, "bytes_per_chunk": {"en": 4.877, "de": 5.46, "vi": 5.59, "tr": 5.844, "fi": 5.893, "ru": 5.975, "el": 5.961, "hi": 5.906, "ta": 5.959, "th": 5.875, "zh": 6.099, "my": 5.725}, "bpb_mean": 1.16612, "t": 10870.8} | |
| {"event": "train", "step": 64000, "phase": 3, "objective": "global", "lm_loss": 0.88161, "bpb": 1.2719, "chunk_loss": 0.03011, "F": 0.18031, "G": 0.22159, "chunks": 184.64, "bytes_per_chunk": 5.546, "boundary_entropy": 0.16186, "router_noise": 0.0, "grad_norm": 0.3335, "lr": 4.78e-05, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 10871.0} | |
| {"event": "train", "step": 64500, "phase": 3, "objective": "global", "lm_loss": 0.83557, "bpb": 1.20548, "chunk_loss": 0.030217, "F": 0.18343, "G": 0.24857, "chunks": 187.83, "bytes_per_chunk": 5.452, "boundary_entropy": 0.21661, "router_noise": 0.0, "grad_norm": 0.3394, "lr": 4.65e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 10954.0} | |
| {"event": "train", "step": 65000, "phase": 3, "objective": "global", "lm_loss": 0.88024, "bpb": 1.26992, "chunk_loss": 0.030148, "F": 0.18431, "G": 0.23242, "chunks": 188.73, "bytes_per_chunk": 5.426, "boundary_entropy": 0.17957, "router_noise": 0.0, "grad_norm": 0.3285, "lr": 4.51e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 11036.5} | |
| {"event": "train", "step": 65500, "phase": 3, "objective": "global", "lm_loss": 0.84213, "bpb": 1.21494, "chunk_loss": 0.030132, "F": 0.18277, "G": 0.24681, "chunks": 187.16, "bytes_per_chunk": 5.471, "boundary_entropy": 0.21682, "router_noise": 0.0, "grad_norm": 0.3369, "lr": 4.38e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 11119.2} | |
| {"event": "train", "step": 66000, "phase": 3, "objective": "global", "lm_loss": 0.83676, "bpb": 1.20719, "chunk_loss": 0.030046, "F": 0.17484, "G": 0.2385, "chunks": 179.03, "bytes_per_chunk": 5.72, "boundary_entropy": 0.20693, "router_noise": 0.0, "grad_norm": 0.3343, "lr": 4.26e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 11201.9} | |
| {"event": "train", "step": 66500, "phase": 3, "objective": "global", "lm_loss": 0.93103, "bpb": 1.3432, "chunk_loss": 0.030163, "F": 0.18407, "G": 0.22941, "chunks": 188.48, "bytes_per_chunk": 5.433, "boundary_entropy": 0.16729, "router_noise": 0.0, "grad_norm": 0.3576, "lr": 4.14e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 11284.7} | |
| {"event": "train", "step": 67000, "phase": 3, "objective": "global", "lm_loss": 0.83964, "bpb": 1.21134, "chunk_loss": 0.030136, "F": 0.1825, "G": 0.23567, "chunks": 186.88, "bytes_per_chunk": 5.48, "boundary_entropy": 0.18948, "router_noise": 0.0, "grad_norm": 0.3672, "lr": 4.03e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 11367.3} | |
| {"event": "train", "step": 67500, "phase": 3, "objective": "global", "lm_loss": 0.84912, "bpb": 1.22502, "chunk_loss": 0.030046, "F": 0.17805, "G": 0.24834, "chunks": 182.33, "bytes_per_chunk": 5.616, "boundary_entropy": 0.22727, "router_noise": 0.0, "grad_norm": 0.3343, "lr": 3.92e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 11449.8} | |
| {"event": "probe", "step": 68000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3409.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195886, "cps": 26.632812, "mean_p": 0.188308, "boundary_entropy": 0.054589, "bpb": 1.4183}, "de": {"chunks": 3742.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182243, "cps": 29.234375, "mean_p": 0.179916, "boundary_entropy": 0.038929, "bpb": 1.523703}, "vi": {"chunks": 4239.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229943, "cps": 33.117188, "mean_p": 0.173894, "boundary_entropy": 0.046778, "bpb": 1.155108}, "tr": {"chunks": 3230.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178423, "cps": 25.234375, "mean_p": 0.169898, "boundary_entropy": 0.06826, "bpb": 1.435578}, "fi": {"chunks": 3061.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162724, "cps": 23.914062, "mean_p": 0.1665, "boundary_entropy": 0.066881, "bpb": 1.551426}, "ru": {"chunks": 5812.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306702, "cps": 45.40625, "mean_p": 0.209277, "boundary_entropy": 0.173255, "bpb": 0.886078}, "el": {"chunks": 6078.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.296156, "cps": 47.484375, "mean_p": 0.259807, "boundary_entropy": 0.306526, "bpb": 0.887103}, "hi": {"chunks": 6989.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.415394, "cps": 54.601562, "mean_p": 0.354829, "boundary_entropy": 0.491087, "bpb": 0.736022}, "ta": {"chunks": 8094.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.43295, "cps": 63.234375, "mean_p": 0.381862, "boundary_entropy": 0.543261, "bpb": 0.636339}, "th": {"chunks": 7354.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.457196, "cps": 57.453125, "mean_p": 0.343379, "boundary_entropy": 0.497987, "bpb": 0.718818}, "zh": {"chunks": 2539.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.429902, "cps": 19.835938, "mean_p": 0.3254, "boundary_entropy": 0.428192, "bpb": 2.225941}, "my": {"chunks": 8734.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.460946, "cps": 68.234375, "mean_p": 0.365123, "boundary_entropy": 0.516541, "bpb": 0.683486}}, "bpb": {"en": 1.4183, "de": 1.5237, "vi": 1.15511, "tr": 1.43558, "fi": 1.55143, "ru": 0.88608, "el": 0.8871, "hi": 0.73602, "ta": 0.63634, "th": 0.71882, "zh": 2.22594, "my": 0.68349}, "gini_cpc": 0.206336, "gini_cps": 0.222277, "bpic": 6.064711, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9303545693538825, "vi": 1.1738629449972207, "tr": 0.9108546711625874, "fi": 0.8307083294588421, "ru": 1.5657178770149789, "el": 1.5118787731333119, "hi": 2.120591843023689, "ta": 2.2102166668458043, "th": 2.3339936989553793, "zh": 2.1946555983111575, "my": 2.3531354714107553}, "t": 11535.3} | |
| {"event": "eval", "step": 68000, "phase": 3, "bpb": {"en": 1.41451, "de": 1.53305, "vi": 1.17416, "tr": 1.42961, "fi": 1.57287, "ru": 0.88353, "el": 0.88551, "hi": 0.74127, "ta": 0.64213, "th": 0.71803, "zh": 2.22963, "my": 0.68981}, "bytes_per_chunk": {"en": 4.877, "de": 5.45, "vi": 5.58, "tr": 5.83, "fi": 5.872, "ru": 5.963, "el": 5.944, "hi": 5.948, "ta": 6.158, "th": 5.988, "zh": 6.216, "my": 6.137}, "bpb_mean": 1.15951, "t": 11540.6} | |
| {"event": "train", "step": 68000, "phase": 3, "objective": "global", "lm_loss": 0.89486, "bpb": 1.29101, "chunk_loss": 0.030128, "F": 0.18225, "G": 0.23505, "chunks": 186.62, "bytes_per_chunk": 5.487, "boundary_entropy": 0.19142, "router_noise": 0.0, "grad_norm": 0.3417, "lr": 3.82e-05, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 11540.8} | |
| {"event": "train", "step": 68500, "phase": 3, "objective": "global", "lm_loss": 0.88489, "bpb": 1.27662, "chunk_loss": 0.030103, "F": 0.18045, "G": 0.22282, "chunks": 184.78, "bytes_per_chunk": 5.542, "boundary_entropy": 0.16398, "router_noise": 0.0, "grad_norm": 0.3413, "lr": 3.73e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 11623.5} | |
| {"event": "train", "step": 69000, "phase": 3, "objective": "global", "lm_loss": 0.90273, "bpb": 1.30236, "chunk_loss": 0.030167, "F": 0.1815, "G": 0.24493, "chunks": 185.86, "bytes_per_chunk": 5.51, "boundary_entropy": 0.21567, "router_noise": 0.0, "grad_norm": 0.3612, "lr": 3.64e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 11706.2} | |
| {"event": "train", "step": 69500, "phase": 3, "objective": "global", "lm_loss": 0.85005, "bpb": 1.22636, "chunk_loss": 0.030147, "F": 0.18379, "G": 0.22937, "chunks": 188.2, "bytes_per_chunk": 5.441, "boundary_entropy": 0.17045, "router_noise": 0.0, "grad_norm": 0.3446, "lr": 3.55e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 11789.0} | |
| {"event": "train", "step": 70000, "phase": 3, "objective": "global", "lm_loss": 0.91838, "bpb": 1.32494, "chunk_loss": 0.030119, "F": 0.18256, "G": 0.23443, "chunks": 186.94, "bytes_per_chunk": 5.478, "boundary_entropy": 0.18552, "router_noise": 0.0, "grad_norm": 0.3535, "lr": 3.48e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 11871.8} | |
| {"event": "train", "step": 70500, "phase": 3, "objective": "global", "lm_loss": 0.89526, "bpb": 1.29158, "chunk_loss": 0.03024, "F": 0.18671, "G": 0.23074, "chunks": 191.19, "bytes_per_chunk": 5.356, "boundary_entropy": 0.16908, "router_noise": 0.0, "grad_norm": 0.3499, "lr": 3.4e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 11954.7} | |
| {"event": "train", "step": 71000, "phase": 3, "objective": "global", "lm_loss": 0.92014, "bpb": 1.32748, "chunk_loss": 0.030214, "F": 0.19189, "G": 0.22332, "chunks": 196.5, "bytes_per_chunk": 5.211, "boundary_entropy": 0.14462, "router_noise": 0.0, "grad_norm": 0.3544, "lr": 3.34e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 12037.7} | |
| {"event": "train", "step": 71500, "phase": 3, "objective": "global", "lm_loss": 0.8625, "bpb": 1.24433, "chunk_loss": 0.030087, "F": 0.18044, "G": 0.249, "chunks": 184.77, "bytes_per_chunk": 5.542, "boundary_entropy": 0.22437, "router_noise": 0.0, "grad_norm": 0.3594, "lr": 3.28e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 12120.5} | |
| {"event": "probe", "step": 72000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3411.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.196001, "cps": 26.648438, "mean_p": 0.188031, "boundary_entropy": 0.055038, "bpb": 1.412404}, "de": {"chunks": 3740.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182146, "cps": 29.21875, "mean_p": 0.179984, "boundary_entropy": 0.038514, "bpb": 1.51822}, "vi": {"chunks": 4241.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230052, "cps": 33.132812, "mean_p": 0.173947, "boundary_entropy": 0.046958, "bpb": 1.14888}, "tr": {"chunks": 3227.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178258, "cps": 25.210938, "mean_p": 0.170218, "boundary_entropy": 0.067303, "bpb": 1.430183}, "fi": {"chunks": 3057.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162511, "cps": 23.882812, "mean_p": 0.166164, "boundary_entropy": 0.067608, "bpb": 1.543899}, "ru": {"chunks": 5811.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306649, "cps": 45.398438, "mean_p": 0.209107, "boundary_entropy": 0.172401, "bpb": 0.880741}, "el": {"chunks": 6013.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.292988, "cps": 46.976562, "mean_p": 0.25831, "boundary_entropy": 0.303807, "bpb": 0.880945}, "hi": {"chunks": 7117.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.423001, "cps": 55.601562, "mean_p": 0.360002, "boundary_entropy": 0.496073, "bpb": 0.731475}, "ta": {"chunks": 8467.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.452902, "cps": 66.148438, "mean_p": 0.386548, "boundary_entropy": 0.543018, "bpb": 0.632667}, "th": {"chunks": 7525.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.467827, "cps": 58.789062, "mean_p": 0.346624, "boundary_entropy": 0.502411, "bpb": 0.715595}, "zh": {"chunks": 2690.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.455469, "cps": 21.015625, "mean_p": 0.335987, "boundary_entropy": 0.436772, "bpb": 2.216362}, "my": {"chunks": 9111.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.480842, "cps": 71.179688, "mean_p": 0.370734, "boundary_entropy": 0.520816, "bpb": 0.67705}}, "bpb": {"en": 1.4124, "de": 1.51822, "vi": 1.14888, "tr": 1.43018, "fi": 1.5439, "ru": 0.88074, "el": 0.88094, "hi": 0.73147, "ta": 0.63267, "th": 0.71559, "zh": 2.21636, "my": 0.67705}, "gini_cpc": 0.215438, "gini_cps": 0.22813, "bpic": 5.958407, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.929312108474502, "vi": 1.1737281789949188, "tr": 0.909475103156175, "fi": 0.829136351337685, "ru": 1.5645306005945376, "el": 1.494833284628941, "hi": 2.1581632538378805, "ta": 2.310715571892738, "th": 2.3868649098935433, "zh": 2.3238133275910697, "my": 2.4532684369330755}, "t": 12206.4} | |
| {"event": "eval", "step": 72000, "phase": 3, "bpb": {"en": 1.40914, "de": 1.52727, "vi": 1.16841, "tr": 1.42417, "fi": 1.56522, "ru": 0.87833, "el": 0.88012, "hi": 0.73695, "ta": 0.63828, "th": 0.71517, "zh": 2.22031, "my": 0.68322}, "bytes_per_chunk": {"en": 4.879, "de": 5.446, "vi": 5.583, "tr": 5.822, "fi": 5.886, "ru": 5.967, "el": 5.947, "hi": 5.846, "ta": 5.911, "th": 5.903, "zh": 5.949, "my": 5.866}, "bpb_mean": 1.15388, "t": 12211.7} | |
| {"event": "train", "step": 72000, "phase": 3, "objective": "global", "lm_loss": 0.85493, "bpb": 1.2334, "chunk_loss": 0.030247, "F": 0.18425, "G": 0.24375, "chunks": 188.67, "bytes_per_chunk": 5.427, "boundary_entropy": 0.20742, "router_noise": 0.0, "grad_norm": 0.3532, "lr": 3.22e-05, "router_trainable": true, "sec_per_step": 0.183, "alpha": 0.03, "t": 12211.9} | |
| {"event": "train", "step": 72500, "phase": 3, "objective": "global", "lm_loss": 0.83026, "bpb": 1.19781, "chunk_loss": 0.030042, "F": 0.18028, "G": 0.24355, "chunks": 184.61, "bytes_per_chunk": 5.547, "boundary_entropy": 0.21533, "router_noise": 0.0, "grad_norm": 0.3526, "lr": 3.17e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 12294.6} | |
| {"event": "train", "step": 73000, "phase": 3, "objective": "global", "lm_loss": 0.92975, "bpb": 1.34135, "chunk_loss": 0.030102, "F": 0.18271, "G": 0.22693, "chunks": 187.09, "bytes_per_chunk": 5.473, "boundary_entropy": 0.16797, "router_noise": 0.0, "grad_norm": 0.3747, "lr": 3.13e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 12377.3} | |
| {"event": "train", "step": 73500, "phase": 3, "objective": "global", "lm_loss": 0.93337, "bpb": 1.34658, "chunk_loss": 0.030113, "F": 0.18169, "G": 0.22328, "chunks": 186.05, "bytes_per_chunk": 5.504, "boundary_entropy": 0.15737, "router_noise": 0.0, "grad_norm": 0.376, "lr": 3.09e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 12459.9} | |
| {"event": "train", "step": 74000, "phase": 3, "objective": "global", "lm_loss": 0.89895, "bpb": 1.29691, "chunk_loss": 0.030134, "F": 0.1795, "G": 0.23088, "chunks": 183.81, "bytes_per_chunk": 5.571, "boundary_entropy": 0.18104, "router_noise": 0.0, "grad_norm": 0.3677, "lr": 3.06e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 12542.6} | |
| {"event": "train", "step": 74500, "phase": 3, "objective": "global", "lm_loss": 0.85598, "bpb": 1.23492, "chunk_loss": 0.030116, "F": 0.18291, "G": 0.23524, "chunks": 187.3, "bytes_per_chunk": 5.467, "boundary_entropy": 0.1872, "router_noise": 0.0, "grad_norm": 0.3677, "lr": 3.04e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 12625.2} | |
| {"event": "train", "step": 75000, "phase": 3, "objective": "global", "lm_loss": 0.84528, "bpb": 1.21948, "chunk_loss": 0.030148, "F": 0.1826, "G": 0.2374, "chunks": 186.98, "bytes_per_chunk": 5.476, "boundary_entropy": 0.18974, "router_noise": 0.0, "grad_norm": 0.3713, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.166, "alpha": 0.03, "t": 12708.1} | |
| {"event": "train", "step": 75500, "phase": 3, "objective": "global", "lm_loss": 0.88456, "bpb": 1.27616, "chunk_loss": 0.03009, "F": 0.17651, "G": 0.24272, "chunks": 180.75, "bytes_per_chunk": 5.665, "boundary_entropy": 0.21585, "router_noise": 0.0, "grad_norm": 0.3904, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.165, "alpha": 0.03, "t": 12790.8} | |
| {"event": "probe", "step": 76000, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3406.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195713, "cps": 26.609375, "mean_p": 0.187588, "boundary_entropy": 0.054561, "bpb": 1.409036}, "de": {"chunks": 3744.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182341, "cps": 29.25, "mean_p": 0.179836, "boundary_entropy": 0.037528, "bpb": 1.51525}, "vi": {"chunks": 4244.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230214, "cps": 33.15625, "mean_p": 0.173505, "boundary_entropy": 0.046301, "bpb": 1.144724}, "tr": {"chunks": 3240.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178976, "cps": 25.3125, "mean_p": 0.169847, "boundary_entropy": 0.064414, "bpb": 1.424194}, "fi": {"chunks": 3062.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162777, "cps": 23.921875, "mean_p": 0.165222, "boundary_entropy": 0.063237, "bpb": 1.537508}, "ru": {"chunks": 5821.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.307177, "cps": 45.476562, "mean_p": 0.207375, "boundary_entropy": 0.16446, "bpb": 0.877174}, "el": {"chunks": 6018.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.293232, "cps": 47.015625, "mean_p": 0.251657, "boundary_entropy": 0.290691, "bpb": 0.878946}, "hi": {"chunks": 6949.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.413016, "cps": 54.289062, "mean_p": 0.352332, "boundary_entropy": 0.486761, "bpb": 0.728722}, "ta": {"chunks": 8218.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.439583, "cps": 64.203125, "mean_p": 0.382292, "boundary_entropy": 0.539326, "bpb": 0.629767}, "th": {"chunks": 7485.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.46534, "cps": 58.476562, "mean_p": 0.338815, "boundary_entropy": 0.492244, "bpb": 0.713277}, "zh": {"chunks": 2529.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.428209, "cps": 19.757812, "mean_p": 0.320886, "boundary_entropy": 0.420763, "bpb": 2.211958}, "my": {"chunks": 9303.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.490975, "cps": 72.679688, "mean_p": 0.371573, "boundary_entropy": 0.518792, "bpb": 0.675662}}, "bpb": {"en": 1.40904, "de": 1.51525, "vi": 1.14472, "tr": 1.42419, "fi": 1.53751, "ru": 0.87717, "el": 0.87895, "hi": 0.72872, "ta": 0.62977, "th": 0.71328, "zh": 2.21196, "my": 0.67566}, "gini_cpc": 0.213704, "gini_cps": 0.230071, "bpic": 5.994798, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9316717122279049, "vi": 1.1762827002747749, "tr": 0.9144794180128462, "fi": 0.8317116398424192, "ru": 1.5695236405722015, "el": 1.498272522337126, "hi": 2.1103122276133632, "ta": 2.246053744340219, "th": 2.377662542522649, "zh": 2.1879372550064535, "my": 2.5086444980117415}, "t": 12876.2} | |
| {"event": "eval", "step": 76000, "phase": 3, "bpb": {"en": 1.40591, "de": 1.52398, "vi": 1.16424, "tr": 1.41841, "fi": 1.55919, "ru": 0.87572, "el": 0.87776, "hi": 0.73428, "ta": 0.6354, "th": 0.71252, "zh": 2.21588, "my": 0.68185}, "bytes_per_chunk": {"en": 4.882, "de": 5.445, "vi": 5.584, "tr": 5.802, "fi": 5.878, "ru": 5.939, "el": 5.964, "hi": 5.911, "ta": 6.039, "th": 5.92, "zh": 6.214, "my": 5.758}, "bpb_mean": 1.15043, "t": 12881.5} | |
| {"event": "train", "step": 76000, "phase": 3, "objective": "global", "lm_loss": 0.88044, "bpb": 1.2702, "chunk_loss": 0.030108, "F": 0.18231, "G": 0.24498, "chunks": 186.69, "bytes_per_chunk": 5.485, "boundary_entropy": 0.21102, "router_noise": 0.0, "grad_norm": 0.3737, "lr": 3e-05, "router_trainable": true, "sec_per_step": 0.182, "alpha": 0.03, "t": 12881.7} | |
| {"event": "probe", "step": 76300, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3409.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195886, "cps": 26.632812, "mean_p": 0.18797, "boundary_entropy": 0.054534, "bpb": 1.40927}, "de": {"chunks": 3742.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182243, "cps": 29.234375, "mean_p": 0.179631, "boundary_entropy": 0.038099, "bpb": 1.514618}, "vi": {"chunks": 4241.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230052, "cps": 33.132812, "mean_p": 0.173612, "boundary_entropy": 0.045688, "bpb": 1.14501}, "tr": {"chunks": 3223.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178037, "cps": 25.179688, "mean_p": 0.168786, "boundary_entropy": 0.063022, "bpb": 1.426312}, "fi": {"chunks": 3055.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162405, "cps": 23.867188, "mean_p": 0.1647, "boundary_entropy": 0.063286, "bpb": 1.538461}, "ru": {"chunks": 5803.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306227, "cps": 45.335938, "mean_p": 0.205434, "boundary_entropy": 0.161938, "bpb": 0.877759}, "el": {"chunks": 5950.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.289919, "cps": 46.484375, "mean_p": 0.24949, "boundary_entropy": 0.288414, "bpb": 0.880392}, "hi": {"chunks": 6948.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.412957, "cps": 54.28125, "mean_p": 0.348312, "boundary_entropy": 0.481575, "bpb": 0.728447}, "ta": {"chunks": 8173.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.437176, "cps": 63.851562, "mean_p": 0.377275, "boundary_entropy": 0.533876, "bpb": 0.62942}, "th": {"chunks": 7463.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.463973, "cps": 58.304688, "mean_p": 0.334699, "boundary_entropy": 0.486689, "bpb": 0.713209}, "zh": {"chunks": 2561.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.433627, "cps": 20.007812, "mean_p": 0.323133, "boundary_entropy": 0.422965, "bpb": 2.211814}, "my": {"chunks": 8458.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.44638, "cps": 66.078125, "mean_p": 0.35723, "boundary_entropy": 0.509776, "bpb": 0.6736}}, "bpb": {"en": 1.40927, "de": 1.51462, "vi": 1.14501, "tr": 1.42631, "fi": 1.53846, "ru": 0.87776, "el": 0.88039, "hi": 0.72845, "ta": 0.62942, "th": 0.71321, "zh": 2.21181, "my": 0.6736}, "gini_cpc": 0.206626, "gini_cps": 0.220134, "bpic": 6.089249, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9303545693538825, "vi": 1.1744167845560776, "tr": 0.9088806827111515, "fi": 0.8290800217238689, "ru": 1.5632933310939305, "el": 1.4800392728106624, "hi": 2.1081516848374005, "ta": 2.2317890805696514, "th": 2.3685878399923843, "zh": 2.2136719130661184, "my": 2.278774881748588}, "t": 12934.1} | |
| {"event": "eval", "step": 76300, "phase": 3, "bpb": {"en": 1.40663, "de": 1.52365, "vi": 1.16463, "tr": 1.42023, "fi": 1.56017, "ru": 0.8757, "el": 0.87947, "hi": 0.73418, "ta": 0.63531, "th": 0.71231, "zh": 2.2154, "my": 0.68021}, "bytes_per_chunk": {"en": 4.878, "de": 5.448, "vi": 5.582, "tr": 5.828, "fi": 5.887, "ru": 5.953, "el": 6.027, "hi": 5.961, "ta": 6.087, "th": 5.921, "zh": 6.15, "my": 6.295}, "bpb_mean": 1.15066, "t": 12939.4} | |
| {"event": "probe", "step": 76300, "phase": 3, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3409.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.195886, "cps": 26.632812, "mean_p": 0.18797, "boundary_entropy": 0.054534, "bpb": 1.40927}, "de": {"chunks": 3742.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.182243, "cps": 29.234375, "mean_p": 0.179631, "boundary_entropy": 0.038099, "bpb": 1.514618}, "vi": {"chunks": 4241.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.230052, "cps": 33.132812, "mean_p": 0.173612, "boundary_entropy": 0.045688, "bpb": 1.14501}, "tr": {"chunks": 3223.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178037, "cps": 25.179688, "mean_p": 0.168786, "boundary_entropy": 0.063022, "bpb": 1.426312}, "fi": {"chunks": 3055.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162405, "cps": 23.867188, "mean_p": 0.1647, "boundary_entropy": 0.063286, "bpb": 1.538461}, "ru": {"chunks": 5803.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.306227, "cps": 45.335938, "mean_p": 0.205434, "boundary_entropy": 0.161938, "bpb": 0.877759}, "el": {"chunks": 5950.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.289919, "cps": 46.484375, "mean_p": 0.24949, "boundary_entropy": 0.288414, "bpb": 0.880392}, "hi": {"chunks": 6948.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.412957, "cps": 54.28125, "mean_p": 0.348312, "boundary_entropy": 0.481575, "bpb": 0.728447}, "ta": {"chunks": 8173.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.437176, "cps": 63.851562, "mean_p": 0.377275, "boundary_entropy": 0.533876, "bpb": 0.62942}, "th": {"chunks": 7463.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.463973, "cps": 58.304688, "mean_p": 0.334699, "boundary_entropy": 0.486689, "bpb": 0.713209}, "zh": {"chunks": 2561.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.433627, "cps": 20.007812, "mean_p": 0.323133, "boundary_entropy": 0.422965, "bpb": 2.211814}, "my": {"chunks": 8458.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.44638, "cps": 66.078125, "mean_p": 0.35723, "boundary_entropy": 0.509776, "bpb": 0.6736}}, "bpb": {"en": 1.40927, "de": 1.51462, "vi": 1.14501, "tr": 1.42631, "fi": 1.53846, "ru": 0.87776, "el": 0.88039, "hi": 0.72845, "ta": 0.62942, "th": 0.71321, "zh": 2.21181, "my": 0.6736}, "gini_cpc": 0.206626, "gini_cps": 0.220134, "bpic": 6.089249, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9303545693538825, "vi": 1.1744167845560776, "tr": 0.9088806827111515, "fi": 0.8290800217238689, "ru": 1.5632933310939305, "el": 1.4800392728106624, "hi": 2.1081516848374005, "ta": 2.2317890805696514, "th": 2.3685878399923843, "zh": 2.2136719130661184, "my": 2.278774881748588}, "t": 12942.5} | |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/specfull_A_s0/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/specfull_A_s0/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/specfull_A_s0/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/specfull_A_s0/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/specfull_A_s0/th0.70"], "t": 12946.4} | |
| {"event": "final", "run": "specfull_A_s0", "condition": "A", "step": 76300, "params": 82470528, "per_lang": {"en": {"cpc": 0.19588576682181233, "cps": 26.6328125, "bpb": 1.4036495348173146, "bytes_per_chunk": 4.875442291687008}, "de": {"cpc": 0.18224321823406225, "cps": 29.234375, "bpb": 1.520054643064889, "bytes_per_chunk": 5.427355824473866}, "vi": {"cpc": 0.2300515324111744, "cps": 33.1328125, "bpb": 1.162168791481086, "bytes_per_chunk": 5.583858307550576}, "tr": {"cpc": 0.17803678948240623, "cps": 25.1796875, "bpb": 1.4175154640604777, "bytes_per_chunk": 5.832745558312103}, "fi": {"cpc": 0.16240497581202487, "cps": 23.8671875, "bpb": 1.5665407474244302, "bytes_per_chunk": 5.9119668480252585}, "ru": {"cpc": 0.3062269129287599, "cps": 45.3359375, "bpb": 0.8785567863557654, "bytes_per_chunk": 5.940536620739667}, "el": {"cpc": 0.2899186278809141, "cps": 46.484375, "bpb": 0.8778654053503812, "bytes_per_chunk": 6.037193556187603}, "hi": {"cpc": 0.4129569093610698, "cps": 54.28125, "bpb": 0.7355408939059719, "bytes_per_chunk": 5.970471510644718}, "ta": {"cpc": 0.4371757154319337, "cps": 63.8515625, "bpb": 0.6337944581817971, "bytes_per_chunk": 6.081804050762128}, "th": {"cpc": 0.4639726453217283, "cps": 58.3046875, "bpb": 0.7144447199675241, "bytes_per_chunk": 5.935627030759999}, "zh": {"cpc": 0.43362682018286486, "cps": 20.0078125, "bpb": 2.215191857033059, "bytes_per_chunk": 6.144276012398994}, "my": {"cpc": 0.44637956512560695, "cps": 66.078125, "bpb": 0.6806133274729266, "bytes_per_chunk": 6.294464754152485}}, "gini_cpc": 0.2066257765320496, "gini_cps": 0.2201340187647427, "bpic": 6.0892488814140195, "seed": 0, "mean_boundary_entropy": 0.1498613953590393, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 30520, "phase": 2, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 53410, "phase": 3, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 12956.4} | |