| {"event": "start", "run": "spec_E_s1", "condition": "E", "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": "parity_b", "p3_objective": "none", "freeze_in_p3": true, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 2.1} |
| {"event": "probe_schedule", "n": 21, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 11000], "t": 2.1} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 14124.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.811584, "cps": 110.34375, "mean_p": 0.52085, "boundary_entropy": 0.686717, "bpb": 8.06011}, "de": {"chunks": 18700.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.910729, "cps": 146.09375, "mean_p": 0.525595, "boundary_entropy": 0.687217, "bpb": 8.1963}, "vi": {"chunks": 21351.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.158177, "cps": 166.804688, "mean_p": 0.524018, "boundary_entropy": 0.687874, "bpb": 8.061647}, "tr": {"chunks": 17846.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.985803, "cps": 139.421875, "mean_p": 0.529939, "boundary_entropy": 0.686373, "bpb": 8.199}, "fi": {"chunks": 17574.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.934241, "cps": 137.296875, "mean_p": 0.526553, "boundary_entropy": 0.686826, "bpb": 8.038283}, "ru": {"chunks": 34128.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.80095, "cps": 266.625, "mean_p": 0.540146, "boundary_entropy": 0.687211, "bpb": 8.262097}, "el": {"chunks": 36913.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.798616, "cps": 288.382812, "mean_p": 0.535776, "boundary_entropy": 0.688142, "bpb": 8.178552}, "hi": {"chunks": 34684.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.061456, "cps": 270.96875, "mean_p": 0.509835, "boundary_entropy": 0.69075, "bpb": 8.397747}, "ta": {"chunks": 50678.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.710778, "cps": 395.921875, "mean_p": 0.522012, "boundary_entropy": 0.690366, "bpb": 8.106024}, "th": {"chunks": 45409.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.823065, "cps": 354.757812, "mean_p": 0.531604, "boundary_entropy": 0.689071, "bpb": 8.030889}, "zh": {"chunks": 4820.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.816119, "cps": 37.65625, "mean_p": 0.492076, "boundary_entropy": 0.68644, "bpb": 8.099598}, "my": {"chunks": 28215.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.489075, "cps": 220.429688, "mean_p": 0.506229, "boundary_entropy": 0.690851, "bpb": 8.204861}}, "bpb": {"en": 8.06011, "de": 8.1963, "vi": 8.06165, "tr": 8.199, "fi": 8.03828, "ru": 8.2621, "el": 8.17855, "hi": 8.39775, "ta": 8.10602, "th": 8.03089, "zh": 8.0996, "my": 8.20486}, "gini_cpc": 0.247425, "gini_cps": 0.271932, "bpic": 1.182896, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1221621360827407, "vi": 1.4270575576222595, "tr": 1.2146656456674263, "fi": 1.1511320610100555, "ru": 2.2190548395702145, "el": 2.2161793753498933, "hi": 2.540039766651812, "ta": 3.3401072258858213, "th": 3.478462165603346, "zh": 1.0055878258102513, "my": 1.8347761655608366}, "t": 10.5} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.63341, "bpb": 8.1273, "chunk_loss": 0.084693, "F": 0.86952, "G": 0.52249, "chunks": 890.39, "bytes_per_chunk": 1.15, "boundary_entropy": 0.69066, "router_noise": 0.0, "grad_norm": 9.4225, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 11.1} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 14061.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.807964, "cps": 109.851562, "mean_p": 0.520569, "boundary_entropy": 0.686738, "bpb": 8.024397}, "de": {"chunks": 18669.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.909219, "cps": 145.851562, "mean_p": 0.525329, "boundary_entropy": 0.687242, "bpb": 8.163782}, "vi": {"chunks": 21327.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.156876, "cps": 166.617188, "mean_p": 0.523901, "boundary_entropy": 0.687884, "bpb": 8.046571}, "tr": {"chunks": 17817.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.984202, "cps": 139.195312, "mean_p": 0.529733, "boundary_entropy": 0.686396, "bpb": 8.174369}, "fi": {"chunks": 17531.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.931955, "cps": 136.960938, "mean_p": 0.526264, "boundary_entropy": 0.686855, "bpb": 8.007852}, "ru": {"chunks": 34126.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.800844, "cps": 266.609375, "mean_p": 0.540055, "boundary_entropy": 0.687225, "bpb": 8.228455}, "el": {"chunks": 36911.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.798519, "cps": 288.367188, "mean_p": 0.535687, "boundary_entropy": 0.688154, "bpb": 8.166303}, "hi": {"chunks": 34497.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.050342, "cps": 269.507812, "mean_p": 0.509703, "boundary_entropy": 0.690754, "bpb": 8.350597}, "ta": {"chunks": 50662.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.709922, "cps": 395.796875, "mean_p": 0.521837, "boundary_entropy": 0.69038, "bpb": 8.072043}, "th": {"chunks": 45400.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.822505, "cps": 354.6875, "mean_p": 0.531487, "boundary_entropy": 0.689085, "bpb": 7.985638}, "zh": {"chunks": 4806.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.813749, "cps": 37.546875, "mean_p": 0.492034, "boundary_entropy": 0.686438, "bpb": 8.094779}, "my": {"chunks": 28181.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.487281, "cps": 220.164062, "mean_p": 0.506213, "boundary_entropy": 0.690851, "bpb": 8.190905}}, "bpb": {"en": 8.0244, "de": 8.16378, "vi": 8.04657, "tr": 8.17437, "fi": 8.00785, "ru": 8.22845, "el": 8.1663, "hi": 8.3506, "ta": 8.07204, "th": 7.98564, "zh": 8.09478, "my": 8.19091}, "gini_cpc": 0.24777, "gini_cps": 0.272323, "bpic": 1.184553, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1253213550785375, "vi": 1.4318401589633083, "tr": 1.2181252357945451, "fi": 1.1534604773521278, "ru": 2.228866640058141, "el": 2.2259883042854365, "hi": 2.537664286567982, "ta": 3.3540132429138705, "th": 3.4933548232503626, "zh": 1.0071594587879191, "my": 1.840775968041897}, "t": 17.9} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 13916.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.799632, "cps": 108.71875, "mean_p": 0.519955, "boundary_entropy": 0.686783, "bpb": 7.952964}, "de": {"chunks": 18576.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.90469, "cps": 145.125, "mean_p": 0.524733, "boundary_entropy": 0.687298, "bpb": 8.096476}, "vi": {"chunks": 21266.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.153567, "cps": 166.140625, "mean_p": 0.523643, "boundary_entropy": 0.687907, "bpb": 8.014633}, "tr": {"chunks": 17760.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.981053, "cps": 138.75, "mean_p": 0.529283, "boundary_entropy": 0.686444, "bpb": 8.123396}, "fi": {"chunks": 17439.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.927064, "cps": 136.242188, "mean_p": 0.525686, "boundary_entropy": 0.686911, "bpb": 7.947436}, "ru": {"chunks": 34114.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.800211, "cps": 266.515625, "mean_p": 0.539855, "boundary_entropy": 0.687256, "bpb": 8.161366}, "el": {"chunks": 36884.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.797203, "cps": 288.15625, "mean_p": 0.535477, "boundary_entropy": 0.688183, "bpb": 8.135093}, "hi": {"chunks": 34174.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.031144, "cps": 266.984375, "mean_p": 0.509479, "boundary_entropy": 0.690762, "bpb": 8.266855}, "ta": {"chunks": 50639.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.708692, "cps": 395.617188, "mean_p": 0.521529, "boundary_entropy": 0.690406, "bpb": 8.009393}, "th": {"chunks": 45381.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.821324, "cps": 354.539062, "mean_p": 0.531274, "boundary_entropy": 0.689111, "bpb": 7.90974}, "zh": {"chunks": 4778.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.809008, "cps": 37.328125, "mean_p": 0.491951, "boundary_entropy": 0.686434, "bpb": 8.086238}, "my": {"chunks": 28128.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.484484, "cps": 219.75, "mean_p": 0.506178, "boundary_entropy": 0.690852, "bpb": 8.160152}}, "bpb": {"en": 7.95296, "de": 8.09648, "vi": 8.01463, "tr": 8.1234, "fi": 7.94744, "ru": 8.16137, "el": 8.13509, "hi": 8.26686, "ta": 8.00939, "th": 7.90974, "zh": 8.08624, "my": 8.16015}, "gini_cpc": 0.248682, "gini_cps": 0.273255, "bpic": 1.187974, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1313826002399658, "vi": 1.442621391489703, "tr": 1.2268800657576666, "fi": 1.1593628884495717, "ru": 2.2512987536982165, "el": 2.247537094683708, "hi": 2.5400978231947406, "ta": 3.3874223716962684, "th": 3.5282771856504276, "zh": 1.011724816510694, "my": 1.8564582100401195}, "t": 25.3} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 13350.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.767109, "cps": 104.296875, "mean_p": 0.517768, "boundary_entropy": 0.686929, "bpb": 7.704435}, "de": {"chunks": 18153.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.884089, "cps": 141.820312, "mean_p": 0.522622, "boundary_entropy": 0.687485, "bpb": 7.858551}, "vi": {"chunks": 21042.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.141416, "cps": 164.390625, "mean_p": 0.52272, "boundary_entropy": 0.687987, "bpb": 7.903217}, "tr": {"chunks": 17547.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.969287, "cps": 137.085938, "mean_p": 0.52774, "boundary_entropy": 0.686605, "bpb": 7.943379}, "fi": {"chunks": 17152.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.911807, "cps": 134.0, "mean_p": 0.523682, "boundary_entropy": 0.687098, "bpb": 7.743087}, "ru": {"chunks": 34086.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.798734, "cps": 266.296875, "mean_p": 0.539064, "boundary_entropy": 0.687376, "bpb": 7.918348}, "el": {"chunks": 36844.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.795254, "cps": 287.84375, "mean_p": 0.534795, "boundary_entropy": 0.688273, "bpb": 8.04204}, "hi": {"chunks": 33226.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.974799, "cps": 259.578125, "mean_p": 0.508874, "boundary_entropy": 0.690781, "bpb": 8.045427}, "ta": {"chunks": 50545.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.703664, "cps": 394.882812, "mean_p": 0.520586, "boundary_entropy": 0.690482, "bpb": 7.829557}, "th": {"chunks": 45307.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.816724, "cps": 353.960938, "mean_p": 0.530584, "boundary_entropy": 0.689193, "bpb": 7.678164}, "zh": {"chunks": 4689.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.793938, "cps": 36.632812, "mean_p": 0.491642, "boundary_entropy": 0.686419, "bpb": 8.055078}, "my": {"chunks": 28069.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.48137, "cps": 219.289062, "mean_p": 0.506147, "boundary_entropy": 0.690853, "bpb": 8.073691}}, "bpb": {"en": 7.70444, "de": 7.85855, "vi": 7.90322, "tr": 7.94338, "fi": 7.74309, "ru": 7.91835, "el": 8.04204, "hi": 8.04543, "ta": 7.82956, "th": 7.67816, "zh": 8.05508, "my": 8.07369}, "gini_cpc": 0.252338, "gini_cps": 0.277287, "bpic": 1.199278, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1524944827099375, "vi": 1.4879444876166792, "tr": 1.263557992436538, "fi": 1.1886274048354446, "ru": 2.3448209184429754, "el": 2.34028518291618, "hi": 2.5743396296976475, "ta": 3.5244843450901673, "th": 3.6718682979815798, "zh": 1.0349744879516927, "my": 1.9311073594675434}, "t": 33.2} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 11225.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.645004, "cps": 87.695312, "mean_p": 0.510744, "boundary_entropy": 0.68727, "bpb": 7.083483}, "de": {"chunks": 16341.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.795841, "cps": 127.664062, "mean_p": 0.515591, "boundary_entropy": 0.687979, "bpb": 7.216128}, "vi": {"chunks": 20124.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.091619, "cps": 157.21875, "mean_p": 0.519541, "boundary_entropy": 0.688235, "bpb": 7.564622}, "tr": {"chunks": 16631.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.918688, "cps": 129.929688, "mean_p": 0.522667, "boundary_entropy": 0.68707, "bpb": 7.423774}, "fi": {"chunks": 15674.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.833236, "cps": 122.453125, "mean_p": 0.517384, "boundary_entropy": 0.687575, "bpb": 7.214082}, "ru": {"chunks": 33932.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.790607, "cps": 265.09375, "mean_p": 0.535857, "boundary_entropy": 0.687842, "bpb": 7.187904}, "el": {"chunks": 36660.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.786289, "cps": 286.40625, "mean_p": 0.531837, "boundary_entropy": 0.688646, "bpb": 7.596304}, "hi": {"chunks": 30037.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.78526, "cps": 234.664062, "mean_p": 0.506803, "boundary_entropy": 0.690834, "bpb": 7.248865}, "ta": {"chunks": 49986.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.673763, "cps": 390.515625, "mean_p": 0.516478, "boundary_entropy": 0.69077, "bpb": 7.101353}, "th": {"chunks": 44833.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.787255, "cps": 350.257812, "mean_p": 0.527615, "boundary_entropy": 0.689527, "bpb": 6.808814}, "zh": {"chunks": 4343.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.735354, "cps": 33.929688, "mean_p": 0.490315, "boundary_entropy": 0.686349, "bpb": 7.921311}, "my": {"chunks": 27311.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.441366, "cps": 213.367188, "mean_p": 0.505756, "boundary_entropy": 0.690858, "bpb": 7.664709}}, "bpb": {"en": 7.08348, "de": 7.21613, "vi": 7.56462, "tr": 7.42377, "fi": 7.21408, "ru": 7.1879, "el": 7.5963, "hi": 7.24886, "ta": 7.10135, "th": 6.80881, "zh": 7.92131, "my": 7.66471}, "gini_cpc": 0.268493, "gini_cps": 0.294914, "bpic": 1.249706, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2338546250226563, "vi": 1.6924230719744264, "tr": 1.4243134737416323, "fi": 1.2918310816338685, "ru": 2.7761185912993405, "el": 2.7694235822332858, "hi": 2.7678289797039484, "ta": 4.145345047179224, "th": 4.321300878614654, "zh": 1.1400769290525585, "my": 2.2346627859249533}, "t": 42.5} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 6991.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.401712, "cps": 54.617188, "mean_p": 0.499855, "boundary_entropy": 0.687492, "bpb": 6.269802}, "de": {"chunks": 11035.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.537428, "cps": 86.210938, "mean_p": 0.503732, "boundary_entropy": 0.688415, "bpb": 6.398054}, "vi": {"chunks": 17543.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.951614, "cps": 137.054688, "mean_p": 0.51213, "boundary_entropy": 0.688682, "bpb": 7.012579}, "tr": {"chunks": 13786.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.761531, "cps": 107.703125, "mean_p": 0.512339, "boundary_entropy": 0.687788, "bpb": 6.706716}, "fi": {"chunks": 11309.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.601191, "cps": 88.351562, "mean_p": 0.505807, "boundary_entropy": 0.688064, "bpb": 6.474048}, "ru": {"chunks": 29192.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.540475, "cps": 228.0625, "mean_p": 0.513676, "boundary_entropy": 0.689981, "bpb": 5.767856}, "el": {"chunks": 34953.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.703114, "cps": 273.070312, "mean_p": 0.520553, "boundary_entropy": 0.689776, "bpb": 6.463279}, "hi": {"chunks": 19798.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.176701, "cps": 154.671875, "mean_p": 0.500105, "boundary_entropy": 0.690895, "bpb": 5.80526}, "ta": {"chunks": 25607.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.369725, "cps": 200.054688, "mean_p": 0.503135, "boundary_entropy": 0.691248, "bpb": 5.97689}, "th": {"chunks": 39836.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.476593, "cps": 311.21875, "mean_p": 0.514496, "boundary_entropy": 0.69061, "bpb": 5.542781}, "zh": {"chunks": 3297.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.558246, "cps": 25.757812, "mean_p": 0.485913, "boundary_entropy": 0.686069, "bpb": 7.54449}, "my": {"chunks": 23463.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.238284, "cps": 183.304688, "mean_p": 0.503382, "boundary_entropy": 0.690879, "bpb": 6.393653}}, "bpb": {"en": 6.2698, "de": 6.39805, "vi": 7.01258, "tr": 6.70672, "fi": 6.47405, "ru": 5.76786, "el": 6.46328, "hi": 5.80526, "ta": 5.97689, "th": 5.54278, "zh": 7.54449, "my": 6.39365}, "gini_cpc": 0.285912, "gini_cps": 0.31089, "bpic": 1.620628, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.3378417609555302, "vi": 2.368893517516826, "tr": 1.895712792656064, "fi": 1.4965703567428343, "ru": 3.834771173944336, "el": 4.239634620334106, "hi": 2.9292137566810057, "ta": 3.4097147637456464, "th": 6.16509078882761, "zh": 1.3896656496526931, "my": 3.082513466830465}, "t": 52.3} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 1864.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.107108, "cps": 14.5625, "mean_p": 0.488487, "boundary_entropy": 0.6873, "bpb": 5.796618}, "de": {"chunks": 2664.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.129742, "cps": 20.8125, "mean_p": 0.489788, "boundary_entropy": 0.688226, "bpb": 5.937998}, "vi": {"chunks": 4222.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229021, "cps": 32.984375, "mean_p": 0.48805, "boundary_entropy": 0.688564, "bpb": 6.649947}, "tr": {"chunks": 5491.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.30332, "cps": 42.898438, "mean_p": 0.494979, "boundary_entropy": 0.688102, "bpb": 6.308285}, "fi": {"chunks": 3674.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.195311, "cps": 28.703125, "mean_p": 0.492337, "boundary_entropy": 0.68804, "bpb": 6.0013}, "ru": {"chunks": 395.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.020844, "cps": 3.085938, "mean_p": 0.477277, "boundary_entropy": 0.689172, "bpb": 5.395143}, "el": {"chunks": 8442.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.411343, "cps": 65.953125, "mean_p": 0.495186, "boundary_entropy": 0.69051, "bpb": 5.639564}, "hi": {"chunks": 239.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.014205, "cps": 1.867188, "mean_p": 0.471932, "boundary_entropy": 0.689193, "bpb": 5.445783}, "ta": {"chunks": 715.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.038246, "cps": 5.585938, "mean_p": 0.479757, "boundary_entropy": 0.690353, "bpb": 5.369684}, "th": {"chunks": 817.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.050793, "cps": 6.382812, "mean_p": 0.481405, "boundary_entropy": 0.690266, "bpb": 5.238279}, "zh": {"chunks": 3828.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.648154, "cps": 29.90625, "mean_p": 0.488869, "boundary_entropy": 0.686334, "bpb": 7.002663}, "my": {"chunks": 4211.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.22224, "cps": 32.898438, "mean_p": 0.479752, "boundary_entropy": 0.690115, "bpb": 5.050235}}, "bpb": {"en": 5.79662, "de": 5.938, "vi": 6.64995, "tr": 6.30829, "fi": 6.0013, "ru": 5.39514, "el": 5.63956, "hi": 5.44578, "ta": 5.36968, "th": 5.23828, "zh": 7.00266, "my": 5.05023}, "gini_cpc": 0.479735, "gini_cps": 0.42056, "bpic": 10.496718, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2113231730602618, "vi": 2.1382244890119675, "tr": 2.8319077659809784, "fi": 1.8234987952214063, "ru": 0.19461042159738187, "el": 3.840455302007023, "hi": 0.13262366955978294, "ta": 0.35707445703390894, "th": 0.47421927634975675, "zh": 6.0514116727151706, "my": 2.07491388788471}, "t": 63.1} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 1785.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.102569, "cps": 13.945312, "mean_p": 0.490047, "boundary_entropy": 0.687449, "bpb": 5.405949}, "de": {"chunks": 1890.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.092047, "cps": 14.765625, "mean_p": 0.48895, "boundary_entropy": 0.688238, "bpb": 5.582199}, "vi": {"chunks": 4176.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226526, "cps": 32.625, "mean_p": 0.489708, "boundary_entropy": 0.688769, "bpb": 6.052058}, "tr": {"chunks": 3089.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170635, "cps": 24.132812, "mean_p": 0.490803, "boundary_entropy": 0.688001, "bpb": 5.941606}, "fi": {"chunks": 2184.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.116102, "cps": 17.0625, "mean_p": 0.489642, "boundary_entropy": 0.687981, "bpb": 5.59307}, "ru": {"chunks": 2452.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.129393, "cps": 19.15625, "mean_p": 0.491813, "boundary_entropy": 0.690268, "bpb": 4.852049}, "el": {"chunks": 7737.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.376992, "cps": 60.445312, "mean_p": 0.49722, "boundary_entropy": 0.690649, "bpb": 5.070115}, "hi": {"chunks": 27768.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.650401, "cps": 216.9375, "mean_p": 0.503883, "boundary_entropy": 0.690992, "bpb": 4.460248}, "ta": {"chunks": 22933.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.226692, "cps": 179.164062, "mean_p": 0.50124, "boundary_entropy": 0.691294, "bpb": 4.462468}, "th": {"chunks": 10211.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.634815, "cps": 79.773438, "mean_p": 0.496113, "boundary_entropy": 0.691041, "bpb": 4.480114}, "zh": {"chunks": 3678.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.622757, "cps": 28.734375, "mean_p": 0.492708, "boundary_entropy": 0.686945, "bpb": 6.675317}, "my": {"chunks": 661.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.034885, "cps": 5.164062, "mean_p": 0.481056, "boundary_entropy": 0.690462, "bpb": 4.552249}}, "bpb": {"en": 5.40595, "de": 5.5822, "vi": 6.05206, "tr": 5.94161, "fi": 5.59307, "ru": 4.85205, "el": 5.07012, "hi": 4.46025, "ta": 4.46247, "th": 4.48011, "zh": 6.67532, "my": 4.55225}, "gini_cpc": 0.542369, "gini_cps": 0.54601, "bpic": 4.333375, "worst_lang": "hi", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8974190757489379, "vi": 2.20852971945491, "tr": 1.6636166436966433, "fi": 1.1319484531891542, "ru": 1.2615287465909848, "el": 3.6755103620164626, "hi": 16.090718144018382, "ta": 11.959727963932297, "th": 6.189179960451697, "zh": 6.071614395842998, "my": 0.3401135881838486}, "t": 78.9} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.35916, "bpb": 4.84624, "chunk_loss": 0.03283, "F": 0.20851, "G": 0.48079, "chunks": 213.52, "bytes_per_chunk": 4.796, "boundary_entropy": 0.69037, "router_noise": 0.0, "grad_norm": 2.8504, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.779, "alpha": 0.03, "t": 89.1} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3120.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.179279, "cps": 24.375, "mean_p": 0.436773, "boundary_entropy": 0.669397, "bpb": 4.38278}, "de": {"chunks": 3859.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.187941, "cps": 30.148438, "mean_p": 0.438984, "boundary_entropy": 0.671529, "bpb": 4.615238}, "vi": {"chunks": 6395.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.346894, "cps": 49.960938, "mean_p": 0.478433, "boundary_entropy": 0.685362, "bpb": 4.836056}, "tr": {"chunks": 3442.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.190134, "cps": 26.890625, "mean_p": 0.452994, "boundary_entropy": 0.677358, "bpb": 4.836068}, "fi": {"chunks": 2375.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.126256, "cps": 18.554688, "mean_p": 0.445023, "boundary_entropy": 0.676399, "bpb": 4.658498}, "ru": {"chunks": 11964.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.631346, "cps": 93.46875, "mean_p": 0.430123, "boundary_entropy": 0.655247, "bpb": 3.237345}, "el": {"chunks": 7487.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.36481, "cps": 58.492188, "mean_p": 0.482496, "boundary_entropy": 0.689032, "bpb": 4.265646}, "hi": {"chunks": 11136.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.661872, "cps": 87.0, "mean_p": 0.497403, "boundary_entropy": 0.690729, "bpb": 2.88349}, "ta": {"chunks": 17337.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.92736, "cps": 135.445312, "mean_p": 0.488311, "boundary_entropy": 0.690306, "bpb": 3.161746}, "th": {"chunks": 5979.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.371713, "cps": 46.710938, "mean_p": 0.491196, "boundary_entropy": 0.690776, "bpb": 2.917169}, "zh": {"chunks": 4130.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.699289, "cps": 32.265625, "mean_p": 0.493471, "boundary_entropy": 0.686941, "bpb": 6.245423}, "my": {"chunks": 17034.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.898987, "cps": 133.078125, "mean_p": 0.473214, "boundary_entropy": 0.688486, "bpb": 3.513366}}, "bpb": {"en": 4.38278, "de": 4.61524, "vi": 4.83606, "tr": 4.83607, "fi": 4.6585, "ru": 3.23735, "el": 4.26565, "hi": 2.88349, "ta": 3.16175, "th": 2.91717, "zh": 6.24542, "my": 3.51337}, "gini_cpc": 0.32952, "gini_cps": 0.350398, "bpic": 4.071601, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 1.048315235512065, "vi": 1.9349374621852244, "tr": 1.060546806946575, "fi": 0.7042409208196513, "ru": 3.5215731682565456, "el": 2.0348692742476544, "hi": 3.6918468396388158, "ta": 5.172708509062481, "th": 2.0733709739126276, "zh": 3.900552567141629, "my": 5.014444085838164}, "t": 98.9} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.38559, "bpb": 3.44168, "chunk_loss": 0.032397, "F": 0.20837, "G": 0.38026, "chunks": 213.38, "bytes_per_chunk": 4.799, "boundary_entropy": 0.61059, "router_noise": 0.0, "grad_norm": 1.7225, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.272, "alpha": 0.03, "t": 116.3} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3073.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.176579, "cps": 24.007812, "mean_p": 0.30948, "boundary_entropy": 0.548012, "bpb": 3.598708}, "de": {"chunks": 3658.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178152, "cps": 28.578125, "mean_p": 0.329327, "boundary_entropy": 0.559355, "bpb": 3.791982}, "vi": {"chunks": 6501.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.352644, "cps": 50.789062, "mean_p": 0.355018, "boundary_entropy": 0.532271, "bpb": 3.05633}, "tr": {"chunks": 4045.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.223444, "cps": 31.601562, "mean_p": 0.316085, "boundary_entropy": 0.529792, "bpb": 3.781016}, "fi": {"chunks": 3410.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.181277, "cps": 26.640625, "mean_p": 0.311775, "boundary_entropy": 0.548563, "bpb": 3.801969}, "ru": {"chunks": 7955.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.419789, "cps": 62.148438, "mean_p": 0.32542, "boundary_entropy": 0.558929, "bpb": 2.63144}, "el": {"chunks": 8768.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.427228, "cps": 68.5, "mean_p": 0.325722, "boundary_entropy": 0.55186, "bpb": 2.665677}, "hi": {"chunks": 7859.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.467103, "cps": 61.398438, "mean_p": 0.374693, "boundary_entropy": 0.612239, "bpb": 2.045037}, "ta": {"chunks": 10069.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.538593, "cps": 78.664062, "mean_p": 0.39567, "boundary_entropy": 0.583722, "bpb": 1.880953}, "th": {"chunks": 6783.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.421697, "cps": 52.992188, "mean_p": 0.421727, "boundary_entropy": 0.6667, "bpb": 2.085451}, "zh": {"chunks": 3548.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.600745, "cps": 27.71875, "mean_p": 0.420665, "boundary_entropy": 0.649514, "bpb": 4.937034}, "my": {"chunks": 18216.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.961368, "cps": 142.3125, "mean_p": 0.466506, "boundary_entropy": 0.680355, "bpb": 2.182683}}, "bpb": {"en": 3.59871, "de": 3.79198, "vi": 3.05633, "tr": 3.78102, "fi": 3.80197, "ru": 2.63144, "el": 2.66568, "hi": 2.04504, "ta": 1.88095, "th": 2.08545, "zh": 4.93703, "my": 2.18268}, "gini_cpc": 0.275854, "gini_cps": 0.294305, "bpic": 4.575085, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0089109925972348, "vi": 1.9970943547001272, "tr": 1.2654049659324587, "fi": 1.0266066082908092, "ru": 2.3773467437473546, "el": 2.4194756515112243, "hi": 2.6452929580311477, "ta": 3.0501586647875745, "th": 2.388153906177532, "zh": 3.402136454083935, "my": 5.4444147446956235}, "t": 130.2} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.12909, "bpb": 3.07163, "chunk_loss": 0.030398, "F": 0.17747, "G": 0.33063, "chunks": 181.73, "bytes_per_chunk": 5.635, "boundary_entropy": 0.55805, "router_noise": 0.0, "grad_norm": 0.982, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.238, "alpha": 0.03, "t": 140.0} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.00633, "bpb": 2.89452, "chunk_loss": 0.030744, "F": 0.18994, "G": 0.30904, "chunks": 194.5, "bytes_per_chunk": 5.265, "boundary_entropy": 0.51472, "router_noise": 0.0, "grad_norm": 1.263, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 159.4} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 1.78861, "bpb": 2.58041, "chunk_loss": 0.030747, "F": 0.1893, "G": 0.30128, "chunks": 193.84, "bytes_per_chunk": 5.283, "boundary_entropy": 0.50026, "router_noise": 0.0, "grad_norm": 1.0338, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.198, "alpha": 0.03, "t": 179.2} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.7946, "bpb": 2.58907, "chunk_loss": 0.030431, "F": 0.17892, "G": 0.28699, "chunks": 183.22, "bytes_per_chunk": 5.589, "boundary_entropy": 0.47298, "router_noise": 0.0, "grad_norm": 0.8853, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 198.7} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.76085, "bpb": 2.54038, "chunk_loss": 0.030549, "F": 0.18315, "G": 0.28161, "chunks": 187.55, "bytes_per_chunk": 5.46, "boundary_entropy": 0.44281, "router_noise": 0.0, "grad_norm": 0.8274, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.2, "alpha": 0.03, "t": 218.6} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.61919, "bpb": 2.33599, "chunk_loss": 0.030364, "F": 0.17943, "G": 0.26692, "chunks": 183.73, "bytes_per_chunk": 5.573, "boundary_entropy": 0.42382, "router_noise": 0.0, "grad_norm": 0.722, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 237.9} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.56376, "bpb": 2.25603, "chunk_loss": 0.030345, "F": 0.17973, "G": 0.26411, "chunks": 184.05, "bytes_per_chunk": 5.564, "boundary_entropy": 0.41323, "router_noise": 0.0, "grad_norm": 0.8332, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.197, "alpha": 0.03, "t": 257.6} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2799.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.160834, "cps": 21.867188, "mean_p": 0.242605, "boundary_entropy": 0.402874, "bpb": 2.339607}, "de": {"chunks": 3457.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.168363, "cps": 27.007812, "mean_p": 0.259605, "boundary_entropy": 0.382147, "bpb": 2.627527}, "vi": {"chunks": 4441.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.2409, "cps": 34.695312, "mean_p": 0.273905, "boundary_entropy": 0.443133, "bpb": 2.117144}, "tr": {"chunks": 3832.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.211678, "cps": 29.9375, "mean_p": 0.233691, "boundary_entropy": 0.366745, "bpb": 2.613394}, "fi": {"chunks": 3158.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.16788, "cps": 24.671875, "mean_p": 0.262119, "boundary_entropy": 0.413591, "bpb": 2.851304}, "ru": {"chunks": 5792.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.305646, "cps": 45.25, "mean_p": 0.297803, "boundary_entropy": 0.477967, "bpb": 1.713159}, "el": {"chunks": 6542.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.318764, "cps": 51.109375, "mean_p": 0.273084, "boundary_entropy": 0.447724, "bpb": 1.811954}, "hi": {"chunks": 7861.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.467221, "cps": 61.414062, "mean_p": 0.261133, "boundary_entropy": 0.421578, "bpb": 1.38997}, "ta": {"chunks": 9358.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.500562, "cps": 73.109375, "mean_p": 0.236109, "boundary_entropy": 0.346527, "bpb": 1.125895}, "th": {"chunks": 7339.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.456264, "cps": 57.335938, "mean_p": 0.301846, "boundary_entropy": 0.518378, "bpb": 1.372467}, "zh": {"chunks": 3582.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.606502, "cps": 27.984375, "mean_p": 0.254176, "boundary_entropy": 0.251984, "bpb": 3.480879}, "my": {"chunks": 13218.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.697593, "cps": 103.265625, "mean_p": 0.341759, "boundary_entropy": 0.579477, "bpb": 1.291921}}, "bpb": {"en": 2.33961, "de": 2.62753, "vi": 2.11714, "tr": 2.61339, "fi": 2.8513, "ru": 1.71316, "el": 1.81195, "hi": 1.38997, "ta": 1.12589, "th": 1.37247, "zh": 3.48088, "my": 1.29192}, "gini_cpc": 0.273392, "gini_cps": 0.267939, "bpic": 5.376665, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0468107987513264, "vi": 1.4978173362507998, "tr": 1.3161220665432452, "fi": 1.0438100258662593, "ru": 1.9003804788932346, "el": 1.9819418873834023, "hi": 2.9049853402324226, "ta": 3.112280940132499, "th": 2.8368543852477175, "zh": 3.7709724592027514, "my": 4.337341256184637}, "t": 280.8} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.35197, "de": 2.62091, "vi": 2.11942, "tr": 2.60637, "fi": 2.86099, "ru": 1.7116, "el": 1.8061, "hi": 1.39765, "ta": 1.12608, "th": 1.37088, "zh": 3.4838, "my": 1.30338}, "bytes_per_chunk": {"en": 5.865, "de": 5.772, "vi": 5.564, "tr": 5.0, "fi": 5.982, "ru": 5.677, "el": 5.654, "hi": 5.467, "ta": 5.342, "th": 6.069, "zh": 4.564, "my": 3.899}, "bpb_mean": 2.06326, "t": 286.7} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.54364, "bpb": 2.227, "chunk_loss": 0.030554, "F": 0.18633, "G": 0.26667, "chunks": 190.8, "bytes_per_chunk": 5.367, "boundary_entropy": 0.39943, "router_noise": 0.0, "grad_norm": 0.6394, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.294, "alpha": 0.03, "t": 287.0} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.45905, "bpb": 2.10496, "chunk_loss": 0.030136, "F": 0.17361, "G": 0.26941, "chunks": 177.78, "bytes_per_chunk": 5.76, "boundary_entropy": 0.41001, "router_noise": 0.0, "grad_norm": 0.6652, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.184, "alpha": 0.03, "t": 305.4} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.52, "bpb": 2.1929, "chunk_loss": 0.030178, "F": 0.17473, "G": 0.262, "chunks": 178.92, "bytes_per_chunk": 5.723, "boundary_entropy": 0.38548, "router_noise": 0.0, "grad_norm": 0.6782, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.187, "alpha": 0.03, "t": 324.1} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.38935, "bpb": 2.00441, "chunk_loss": 0.030221, "F": 0.1769, "G": 0.26676, "chunks": 181.14, "bytes_per_chunk": 5.653, "boundary_entropy": 0.3765, "router_noise": 0.0, "grad_norm": 0.742, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.185, "alpha": 0.03, "t": 342.6} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.34918, "bpb": 1.94646, "chunk_loss": 0.030067, "F": 0.16899, "G": 0.2633, "chunks": 173.05, "bytes_per_chunk": 5.917, "boundary_entropy": 0.39107, "router_noise": 0.0, "grad_norm": 0.566, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.195, "alpha": 0.03, "t": 362.1} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.42788, "bpb": 2.05999, "chunk_loss": 0.030371, "F": 0.17844, "G": 0.26228, "chunks": 182.72, "bytes_per_chunk": 5.604, "boundary_entropy": 0.36389, "router_noise": 0.0, "grad_norm": 0.6026, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 381.4} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.4611, "bpb": 2.10792, "chunk_loss": 0.030369, "F": 0.18376, "G": 0.25331, "chunks": 188.17, "bytes_per_chunk": 5.442, "boundary_entropy": 0.33348, "router_noise": 0.0, "grad_norm": 0.5742, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 400.6} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.49198, "bpb": 2.15248, "chunk_loss": 0.030317, "F": 0.18184, "G": 0.25005, "chunks": 186.2, "bytes_per_chunk": 5.499, "boundary_entropy": 0.31513, "router_noise": 0.0, "grad_norm": 0.5894, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.188, "alpha": 0.03, "t": 419.5} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.29783, "bpb": 1.87237, "chunk_loss": 0.030059, "F": 0.16924, "G": 0.24699, "chunks": 173.3, "bytes_per_chunk": 5.909, "boundary_entropy": 0.33943, "router_noise": 0.0, "grad_norm": 0.5156, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 438.8} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.2795, "bpb": 1.84593, "chunk_loss": 0.030053, "F": 0.17137, "G": 0.24004, "chunks": 175.48, "bytes_per_chunk": 5.835, "boundary_entropy": 0.31429, "router_noise": 0.0, "grad_norm": 0.4922, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.195, "alpha": 0.03, "t": 458.3} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3148.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.180888, "cps": 24.59375, "mean_p": 0.211943, "boundary_entropy": 0.23069, "bpb": 2.078229}, "de": {"chunks": 3524.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.171626, "cps": 27.53125, "mean_p": 0.223786, "boundary_entropy": 0.248549, "bpb": 2.31118}, "vi": {"chunks": 3166.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.171739, "cps": 24.734375, "mean_p": 0.230131, "boundary_entropy": 0.38113, "bpb": 1.968458}, "tr": {"chunks": 3229.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.178368, "cps": 25.226562, "mean_p": 0.214976, "boundary_entropy": 0.259346, "bpb": 2.285982}, "fi": {"chunks": 3326.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.176811, "cps": 25.984375, "mean_p": 0.224646, "boundary_entropy": 0.281431, "bpb": 2.49869}, "ru": {"chunks": 5749.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.303377, "cps": 44.914062, "mean_p": 0.277286, "boundary_entropy": 0.406844, "bpb": 1.416273}, "el": {"chunks": 6208.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.30249, "cps": 48.5, "mean_p": 0.279937, "boundary_entropy": 0.426022, "bpb": 1.490169}, "hi": {"chunks": 6946.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.412838, "cps": 54.265625, "mean_p": 0.242871, "boundary_entropy": 0.370627, "bpb": 1.200841}, "ta": {"chunks": 8497.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.454507, "cps": 66.382812, "mean_p": 0.247848, "boundary_entropy": 0.371126, "bpb": 1.010785}, "th": {"chunks": 7237.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.449922, "cps": 56.539062, "mean_p": 0.324574, "boundary_entropy": 0.520555, "bpb": 1.153387}, "zh": {"chunks": 3567.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.603962, "cps": 27.867188, "mean_p": 0.24968, "boundary_entropy": 0.204709, "bpb": 3.170448}, "my": {"chunks": 9395.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.495831, "cps": 73.398438, "mean_p": 0.367128, "boundary_entropy": 0.59428, "bpb": 1.166333}}, "bpb": {"en": 2.07823, "de": 2.31118, "vi": 1.96846, "tr": 2.28598, "fi": 2.49869, "ru": 1.41627, "el": 1.49017, "hi": 1.20084, "ta": 1.01079, "th": 1.15339, "zh": 3.17045, "my": 1.16633}, "gini_cpc": 0.249227, "gini_cps": 0.225401, "bpic": 5.997328, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9487960961098293, "vi": 0.9494173525650238, "tr": 0.9860680564223036, "fi": 0.9774617192879048, "ru": 1.67715225648986, "el": 1.6722463612121174, "hi": 2.2822809359370604, "ta": 2.5126358113405423, "th": 2.4872927494856385, "zh": 3.338866565175466, "my": 2.7410869050023616}, "t": 480.8} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.08672, "de": 2.30524, "vi": 1.97358, "tr": 2.27826, "fi": 2.51261, "ru": 1.41419, "el": 1.48657, "hi": 1.20926, "ta": 1.00984, "th": 1.14728, "zh": 3.17918, "my": 1.17456}, "bytes_per_chunk": {"en": 5.223, "de": 5.738, "vi": 7.179, "tr": 5.732, "fi": 5.696, "ru": 5.823, "el": 5.885, "hi": 6.045, "ta": 5.911, "th": 5.972, "zh": 4.576, "my": 5.562}, "bpb_mean": 1.81477, "t": 486.9} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.37137, "bpb": 1.97846, "chunk_loss": 0.030629, "F": 0.18884, "G": 0.24861, "chunks": 193.38, "bytes_per_chunk": 5.295, "boundary_entropy": 0.2951, "router_noise": 0.0, "grad_norm": 0.4971, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.288, "alpha": 0.03, "t": 487.1} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.39184, "bpb": 2.00801, "chunk_loss": 0.030284, "F": 0.18338, "G": 0.24673, "chunks": 187.78, "bytes_per_chunk": 5.453, "boundary_entropy": 0.29021, "router_noise": 0.0, "grad_norm": 0.4941, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.192, "alpha": 0.03, "t": 506.3} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.25491, "bpb": 1.81045, "chunk_loss": 0.030136, "F": 0.17383, "G": 0.2446, "chunks": 178.0, "bytes_per_chunk": 5.753, "boundary_entropy": 0.30628, "router_noise": 0.0, "grad_norm": 0.4883, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.19, "alpha": 0.03, "t": 525.3} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.39954, "bpb": 2.0191, "chunk_loss": 0.030402, "F": 0.18622, "G": 0.24728, "chunks": 190.69, "bytes_per_chunk": 5.37, "boundary_entropy": 0.28165, "router_noise": 0.0, "grad_norm": 0.4861, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 544.7} |
| {"event": "train", "step": 2400, "phase": 1, "objective": "global", "lm_loss": 1.30009, "bpb": 1.87564, "chunk_loss": 0.030207, "F": 0.1799, "G": 0.23631, "chunks": 184.22, "bytes_per_chunk": 5.559, "boundary_entropy": 0.27203, "router_noise": 0.0, "grad_norm": 0.5016, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.198, "alpha": 0.03, "t": 564.5} |
| {"event": "train", "step": 2500, "phase": 1, "objective": "global", "lm_loss": 1.26792, "bpb": 1.82922, "chunk_loss": 0.030363, "F": 0.18182, "G": 0.23718, "chunks": 186.19, "bytes_per_chunk": 5.5, "boundary_entropy": 0.26876, "router_noise": 0.0, "grad_norm": 0.4634, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.192, "alpha": 0.03, "t": 583.7} |
| {"event": "train", "step": 2600, "phase": 1, "objective": "global", "lm_loss": 1.371, "bpb": 1.97794, "chunk_loss": 0.030382, "F": 0.18556, "G": 0.2376, "chunks": 190.02, "bytes_per_chunk": 5.389, "boundary_entropy": 0.25258, "router_noise": 0.0, "grad_norm": 0.4544, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.192, "alpha": 0.03, "t": 602.9} |
| {"event": "train", "step": 2700, "phase": 1, "objective": "global", "lm_loss": 1.22661, "bpb": 1.76963, "chunk_loss": 0.030301, "F": 0.18192, "G": 0.24752, "chunks": 186.28, "bytes_per_chunk": 5.497, "boundary_entropy": 0.28275, "router_noise": 0.0, "grad_norm": 0.4448, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 622.2} |
| {"event": "train", "step": 2800, "phase": 1, "objective": "global", "lm_loss": 1.29597, "bpb": 1.86968, "chunk_loss": 0.030305, "F": 0.18614, "G": 0.23448, "chunks": 190.61, "bytes_per_chunk": 5.372, "boundary_entropy": 0.24995, "router_noise": 0.0, "grad_norm": 0.4301, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 641.5} |
| {"event": "train", "step": 2900, "phase": 1, "objective": "global", "lm_loss": 1.36104, "bpb": 1.96356, "chunk_loss": 0.030372, "F": 0.18761, "G": 0.23595, "chunks": 192.11, "bytes_per_chunk": 5.33, "boundary_entropy": 0.24684, "router_noise": 0.0, "grad_norm": 0.4718, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 660.8} |
| {"event": "probe", "step": 3000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3280.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188473, "cps": 25.625, "mean_p": 0.205645, "boundary_entropy": 0.146584, "bpb": 1.959216}, "de": {"chunks": 3664.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178444, "cps": 28.625, "mean_p": 0.204033, "boundary_entropy": 0.168103, "bpb": 2.1454}, "vi": {"chunks": 4696.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.254733, "cps": 36.6875, "mean_p": 0.24941, "boundary_entropy": 0.335598, "bpb": 1.856013}, "tr": {"chunks": 3395.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.187538, "cps": 26.523438, "mean_p": 0.207116, "boundary_entropy": 0.194411, "bpb": 2.144947}, "fi": {"chunks": 3273.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.173994, "cps": 25.570312, "mean_p": 0.201351, "boundary_entropy": 0.19542, "bpb": 2.339935}, "ru": {"chunks": 5651.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.298206, "cps": 44.148438, "mean_p": 0.256572, "boundary_entropy": 0.349182, "bpb": 1.299942}, "el": {"chunks": 6420.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.31282, "cps": 50.15625, "mean_p": 0.290378, "boundary_entropy": 0.41145, "bpb": 1.35138}, "hi": {"chunks": 7140.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.424368, "cps": 55.78125, "mean_p": 0.237848, "boundary_entropy": 0.32895, "bpb": 1.106798}, "ta": {"chunks": 9097.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.486601, "cps": 71.070312, "mean_p": 0.241563, "boundary_entropy": 0.322051, "bpb": 0.940255}, "th": {"chunks": 7416.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.461051, "cps": 57.9375, "mean_p": 0.335776, "boundary_entropy": 0.530611, "bpb": 1.069445}, "zh": {"chunks": 3545.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.600237, "cps": 27.695312, "mean_p": 0.239486, "boundary_entropy": 0.17655, "bpb": 2.983665}, "my": {"chunks": 7481.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.394817, "cps": 58.445312, "mean_p": 0.351946, "boundary_entropy": 0.580813, "bpb": 1.073896}}, "bpb": {"en": 1.95922, "de": 2.1454, "vi": 1.85601, "tr": 2.14495, "fi": 2.33994, "ru": 1.29994, "el": 1.35138, "hi": 1.1068, "ta": 0.94025, "th": 1.06944, "zh": 2.98367, "my": 1.0739}, "gini_cpc": 0.231217, "gini_cps": 0.201382, "bpic": 5.899059, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9467892850652073, "vi": 1.3515596658000755, "tr": 0.9950376268318282, "fi": 0.92317577221942, "ru": 1.582218176845357, "el": 1.6597568106217533, "hi": 2.2516112782227378, "ta": 2.5818024090176714, "th": 2.446239262454794, "zh": 3.1847333323697273, "my": 2.094819245354423}, "t": 683.2} |
| {"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 1.96649, "de": 2.14002, "vi": 1.86278, "tr": 2.13802, "fi": 2.35394, "ru": 1.29859, "el": 1.34708, "hi": 1.11497, "ta": 0.94087, "th": 1.06473, "zh": 2.99156, "my": 1.08324}, "bytes_per_chunk": {"en": 4.991, "de": 5.456, "vi": 5.188, "tr": 5.547, "fi": 5.725, "ru": 5.926, "el": 5.746, "hi": 5.92, "ta": 5.543, "th": 5.868, "zh": 4.621, "my": 6.756}, "bpb_mean": 1.69186, "t": 689.5} |
| {"event": "train", "step": 3000, "phase": 1, "objective": "global", "lm_loss": 1.39559, "bpb": 2.01341, "chunk_loss": 0.030446, "F": 0.19412, "G": 0.23971, "chunks": 198.78, "bytes_per_chunk": 5.151, "boundary_entropy": 0.22918, "router_noise": 0.0, "grad_norm": 0.4553, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.288, "alpha": 0.03, "t": 689.6} |
| {"event": "train", "step": 3100, "phase": 1, "objective": "global", "lm_loss": 1.27735, "bpb": 1.84283, "chunk_loss": 0.030239, "F": 0.1824, "G": 0.23803, "chunks": 186.78, "bytes_per_chunk": 5.482, "boundary_entropy": 0.25675, "router_noise": 0.0, "grad_norm": 0.4432, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.184, "alpha": 0.03, "t": 708.1} |
| {"event": "train", "step": 3200, "phase": 1, "objective": "global", "lm_loss": 1.30688, "bpb": 1.88542, "chunk_loss": 0.030354, "F": 0.18799, "G": 0.23461, "chunks": 192.5, "bytes_per_chunk": 5.319, "boundary_entropy": 0.23754, "router_noise": 0.0, "grad_norm": 0.4385, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.192, "alpha": 0.03, "t": 727.3} |
| {"event": "train", "step": 3300, "phase": 1, "objective": "global", "lm_loss": 1.17439, "bpb": 1.69428, "chunk_loss": 0.030474, "F": 0.18867, "G": 0.24554, "chunks": 193.2, "bytes_per_chunk": 5.3, "boundary_entropy": 0.26576, "router_noise": 0.0, "grad_norm": 0.4076, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 746.6} |
| {"event": "train", "step": 3400, "phase": 1, "objective": "global", "lm_loss": 1.22513, "bpb": 1.76749, "chunk_loss": 0.030403, "F": 0.18999, "G": 0.23684, "chunks": 194.55, "bytes_per_chunk": 5.264, "boundary_entropy": 0.23396, "router_noise": 0.0, "grad_norm": 0.4389, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.191, "alpha": 0.03, "t": 765.7} |
| {"event": "train", "step": 3500, "phase": 1, "objective": "global", "lm_loss": 1.20806, "bpb": 1.74286, "chunk_loss": 0.030551, "F": 0.19194, "G": 0.24372, "chunks": 196.55, "bytes_per_chunk": 5.21, "boundary_entropy": 0.24412, "router_noise": 0.0, "grad_norm": 0.4335, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.187, "alpha": 0.03, "t": 784.4} |
| {"event": "train", "step": 3600, "phase": 1, "objective": "global", "lm_loss": 1.26907, "bpb": 1.83088, "chunk_loss": 0.030285, "F": 0.18394, "G": 0.22467, "chunks": 188.36, "bytes_per_chunk": 5.436, "boundary_entropy": 0.20976, "router_noise": 0.0, "grad_norm": 0.4541, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.191, "alpha": 0.03, "t": 803.5} |
| {"event": "train", "step": 3700, "phase": 1, "objective": "global", "lm_loss": 1.19048, "bpb": 1.7175, "chunk_loss": 0.030073, "F": 0.17624, "G": 0.23684, "chunks": 180.47, "bytes_per_chunk": 5.674, "boundary_entropy": 0.25536, "router_noise": 0.0, "grad_norm": 0.4473, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 822.8} |
| {"event": "train", "step": 3800, "phase": 1, "objective": "global", "lm_loss": 1.32719, "bpb": 1.91472, "chunk_loss": 0.030678, "F": 0.19722, "G": 0.24879, "chunks": 201.95, "bytes_per_chunk": 5.07, "boundary_entropy": 0.23896, "router_noise": 0.0, "grad_norm": 0.4096, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 842.1} |
| {"event": "train", "step": 3900, "phase": 1, "objective": "global", "lm_loss": 1.17988, "bpb": 1.70221, "chunk_loss": 0.030307, "F": 0.18588, "G": 0.23357, "chunks": 190.34, "bytes_per_chunk": 5.38, "boundary_entropy": 0.22414, "router_noise": 0.0, "grad_norm": 0.3904, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.188, "alpha": 0.03, "t": 860.9} |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3312.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.190312, "cps": 25.875, "mean_p": 0.201291, "boundary_entropy": 0.115502, "bpb": 1.878084}, "de": {"chunks": 3663.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.178396, "cps": 28.617188, "mean_p": 0.196944, "boundary_entropy": 0.13529, "bpb": 2.058455}, "vi": {"chunks": 4324.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234554, "cps": 33.78125, "mean_p": 0.228529, "boundary_entropy": 0.289935, "bpb": 1.778348}, "tr": {"chunks": 3351.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185107, "cps": 26.179688, "mean_p": 0.194834, "boundary_entropy": 0.155644, "bpb": 2.050788}, "fi": {"chunks": 3303.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.175589, "cps": 25.804688, "mean_p": 0.19511, "boundary_entropy": 0.165127, "bpb": 2.20666}, "ru": {"chunks": 5740.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.302902, "cps": 44.84375, "mean_p": 0.252374, "boundary_entropy": 0.327001, "bpb": 1.241688}, "el": {"chunks": 6436.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.313599, "cps": 50.28125, "mean_p": 0.292446, "boundary_entropy": 0.406841, "bpb": 1.269847}, "hi": {"chunks": 7452.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.442912, "cps": 58.21875, "mean_p": 0.235787, "boundary_entropy": 0.298749, "bpb": 1.045912}, "ta": {"chunks": 8850.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.473389, "cps": 69.140625, "mean_p": 0.235585, "boundary_entropy": 0.296927, "bpb": 0.891702}, "th": {"chunks": 7868.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.489151, "cps": 61.46875, "mean_p": 0.320711, "boundary_entropy": 0.503192, "bpb": 1.011324}, "zh": {"chunks": 3540.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.59939, "cps": 27.65625, "mean_p": 0.237911, "boundary_entropy": 0.166135, "bpb": 2.874445}, "my": {"chunks": 9393.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.495725, "cps": 73.382812, "mean_p": 0.380703, "boundary_entropy": 0.593009, "bpb": 1.020069}}, "bpb": {"en": 1.87808, "de": 2.05846, "vi": 1.77835, "tr": 2.05079, "fi": 2.20666, "ru": 1.24169, "el": 1.26985, "hi": 1.04591, "ta": 0.8917, "th": 1.01132, "zh": 2.87444, "my": 1.02007}, "gini_cpc": 0.238854, "gini_cps": 0.220314, "bpic": 5.708309, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9373856559642061, "vi": 1.2324699394268148, "tr": 0.9726524128837643, "fi": 0.922636183038139, "ru": 1.5916093074834614, "el": 1.6478170126552958, "hi": 2.327295690936107, "ta": 2.487434154679814, "th": 2.570260121726516, "zh": 3.149514495207527, "my": 2.604802130086002}, "t": 883.3} |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.88317, "de": 2.05308, "vi": 1.78755, "tr": 2.04425, "fi": 2.22318, "ru": 1.24072, "el": 1.2674, "hi": 1.05488, "ta": 0.89295, "th": 1.00861, "zh": 2.88322, "my": 1.0275}, "bytes_per_chunk": {"en": 4.991, "de": 5.468, "vi": 5.554, "tr": 5.667, "fi": 5.68, "ru": 5.895, "el": 5.815, "hi": 5.77, "ta": 5.746, "th": 5.532, "zh": 4.633, "my": 5.509}, "bpb_mean": 1.61388, "t": 889.2} |
| {"event": "train", "step": 4000, "phase": 1, "objective": "global", "lm_loss": 1.1896, "bpb": 1.71624, "chunk_loss": 0.030387, "F": 0.18901, "G": 0.24165, "chunks": 193.55, "bytes_per_chunk": 5.291, "boundary_entropy": 0.23177, "router_noise": 0.0, "grad_norm": 0.4095, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.285, "alpha": 0.03, "t": 889.4} |
| {"event": "train", "step": 4100, "phase": 1, "objective": "global", "lm_loss": 1.25916, "bpb": 1.81658, "chunk_loss": 0.030229, "F": 0.18439, "G": 0.22922, "chunks": 188.81, "bytes_per_chunk": 5.423, "boundary_entropy": 0.21359, "router_noise": 0.0, "grad_norm": 0.4074, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.193, "alpha": 0.03, "t": 908.7} |
| {"event": "train", "step": 4200, "phase": 1, "objective": "global", "lm_loss": 1.07897, "bpb": 1.55662, "chunk_loss": 0.030204, "F": 0.18015, "G": 0.24792, "chunks": 184.47, "bytes_per_chunk": 5.551, "boundary_entropy": 0.27035, "router_noise": 0.0, "grad_norm": 0.4008, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.194, "alpha": 0.03, "t": 928.1} |
| {"event": "train", "step": 4300, "phase": 1, "objective": "global", "lm_loss": 1.23018, "bpb": 1.77477, "chunk_loss": 0.030319, "F": 0.18982, "G": 0.22983, "chunks": 194.38, "bytes_per_chunk": 5.268, "boundary_entropy": 0.19511, "router_noise": 0.0, "grad_norm": 0.3805, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.186, "alpha": 0.03, "t": 946.7} |
| {"event": "phase_transition", "step": 4400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 965.2} |
| {"event": "train", "step": 4400, "phase": 2, "objective": "parity_b", "lm_loss": 1.25282, "bpb": 1.80744, "chunk_loss": 0.059196, "F": 0.18523, "G": 0.22997, "chunks": 189.67, "bytes_per_chunk": 5.399, "boundary_entropy": 0.20827, "router_noise": 0.0, "grad_norm": 0.5355, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.188, "worst_lang": "my", "n_target_mean": 21.145, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 965.5} |
| {"event": "train", "step": 4500, "phase": 2, "objective": "parity_b", "lm_loss": 1.25277, "bpb": 1.80737, "chunk_loss": 0.030534, "F": 0.15994, "G": 0.17551, "chunks": 163.78, "bytes_per_chunk": 6.252, "boundary_entropy": 0.14315, "router_noise": 0.0, "grad_norm": 0.4353, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.193, "worst_lang": "th", "n_target_mean": 16.5724, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 984.8} |
| {"event": "train", "step": 4600, "phase": 2, "objective": "parity_b", "lm_loss": 1.18797, "bpb": 1.71388, "chunk_loss": 0.030228, "F": 0.15082, "G": 0.16875, "chunks": 154.44, "bytes_per_chunk": 6.631, "boundary_entropy": 0.14676, "router_noise": 0.0, "grad_norm": 0.4175, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.189, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1003.7} |
| {"event": "train", "step": 4700, "phase": 2, "objective": "parity_b", "lm_loss": 1.20827, "bpb": 1.74316, "chunk_loss": 0.030828, "F": 0.14377, "G": 0.16395, "chunks": 147.22, "bytes_per_chunk": 6.956, "boundary_entropy": 0.14661, "router_noise": 0.0, "grad_norm": 0.4195, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.189, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1022.6} |
| {"event": "train", "step": 4800, "phase": 2, "objective": "parity_b", "lm_loss": 1.21062, "bpb": 1.74655, "chunk_loss": 0.030326, "F": 0.1483, "G": 0.16331, "chunks": 151.86, "bytes_per_chunk": 6.743, "boundary_entropy": 0.13169, "router_noise": 0.0, "grad_norm": 0.402, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.186, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1041.1} |
| {"event": "train", "step": 4900, "phase": 2, "objective": "parity_b", "lm_loss": 1.17022, "bpb": 1.68826, "chunk_loss": 0.031103, "F": 0.16357, "G": 0.17284, "chunks": 167.5, "bytes_per_chunk": 6.113, "boundary_entropy": 0.12462, "router_noise": 0.0, "grad_norm": 0.4214, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.188, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1059.9} |
| {"event": "probe", "step": 5000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3290.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189048, "cps": 25.703125, "mean_p": 0.192711, "boundary_entropy": 0.074889, "bpb": 1.787017}, "de": {"chunks": 3491.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.170019, "cps": 27.273438, "mean_p": 0.157593, "boundary_entropy": 0.099899, "bpb": 1.962171}, "vi": {"chunks": 3212.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.174234, "cps": 25.09375, "mean_p": 0.157609, "boundary_entropy": 0.171591, "bpb": 1.706847}, "tr": {"chunks": 3333.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.184113, "cps": 26.039062, "mean_p": 0.179921, "boundary_entropy": 0.103863, "bpb": 1.922162}, "fi": {"chunks": 3220.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171176, "cps": 25.15625, "mean_p": 0.173038, "boundary_entropy": 0.105015, "bpb": 2.107262}, "ru": {"chunks": 3303.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.174301, "cps": 25.804688, "mean_p": 0.109523, "boundary_entropy": 0.134228, "bpb": 1.248982}, "el": {"chunks": 3433.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.167276, "cps": 26.820312, "mean_p": 0.104961, "boundary_entropy": 0.138829, "bpb": 1.234271}, "hi": {"chunks": 3319.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.197266, "cps": 25.929688, "mean_p": 0.105579, "boundary_entropy": 0.176151, "bpb": 1.046932}, "ta": {"chunks": 2596.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.138861, "cps": 20.28125, "mean_p": 0.096646, "boundary_entropy": 0.178514, "bpb": 0.902135}, "th": {"chunks": 1644.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.102207, "cps": 12.84375, "mean_p": 0.082928, "boundary_entropy": 0.185696, "bpb": 1.016278}, "zh": {"chunks": 3542.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.599729, "cps": 27.671875, "mean_p": 0.249278, "boundary_entropy": 0.184413, "bpb": 2.792816}, "my": {"chunks": 1493.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.078795, "cps": 11.664062, "mean_p": 0.06931, "boundary_entropy": 0.171385, "bpb": 1.010143}}, "bpb": {"en": 1.78702, "de": 1.96217, "vi": 1.70685, "tr": 1.92216, "fi": 2.10726, "ru": 1.24898, "el": 1.23427, "hi": 1.04693, "ta": 0.90214, "th": 1.01628, "zh": 2.79282, "my": 1.01014}, "gini_cpc": 0.249863, "gini_cps": 0.106701, "bpic": 10.69743, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8993436320241847, "vi": 0.9216385179835832, "tr": 0.9738969023421701, "fi": 0.9054661317450067, "ru": 0.9219929104746935, "el": 0.8848327334429591, "hi": 1.04347104332549, "ta": 0.7345264528737184, "th": 0.5406409906339139, "zh": 3.172366363813216, "my": 0.41679706674417843}, "t": 1082.0} |
| {"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.79014, "de": 1.96103, "vi": 1.71904, "tr": 1.91323, "fi": 2.12187, "ru": 1.2478, "el": 1.23316, "hi": 1.05482, "ta": 0.90445, "th": 1.01425, "zh": 2.80166, "my": 1.01814}, "bytes_per_chunk": {"en": 5.027, "de": 5.784, "vi": 7.614, "tr": 5.718, "fi": 5.834, "ru": 9.561, "el": 10.082, "hi": 12.128, "ta": 19.064, "th": 25.555, "zh": 4.679, "my": 30.555}, "bpb_mean": 1.56497, "t": 1087.3} |
| {"event": "train", "step": 5000, "phase": 2, "objective": "parity_b", "lm_loss": 1.13663, "bpb": 1.63981, "chunk_loss": 0.031203, "F": 0.14645, "G": 0.16223, "chunks": 149.97, "bytes_per_chunk": 6.828, "boundary_entropy": 0.12808, "router_noise": 0.0, "grad_norm": 0.4372, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.275, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1087.4} |
| {"event": "train", "step": 5100, "phase": 2, "objective": "parity_b", "lm_loss": 1.12872, "bpb": 1.62841, "chunk_loss": 0.030398, "F": 0.14072, "G": 0.15732, "chunks": 144.09, "bytes_per_chunk": 7.106, "boundary_entropy": 0.13033, "router_noise": 0.0, "grad_norm": 0.4079, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.19, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 1106.4} |
| {"event": "train", "step": 5200, "phase": 2, "objective": "parity_b", "lm_loss": 1.24348, "bpb": 1.79396, "chunk_loss": 0.030541, "F": 0.15752, "G": 0.17091, "chunks": 161.3, "bytes_per_chunk": 6.349, "boundary_entropy": 0.11835, "router_noise": 0.0, "grad_norm": 0.4211, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.188, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1125.2} |
| {"event": "train", "step": 5300, "phase": 2, "objective": "parity_b", "lm_loss": 1.14667, "bpb": 1.65429, "chunk_loss": 0.030193, "F": 0.15132, "G": 0.16019, "chunks": 154.95, "bytes_per_chunk": 6.608, "boundary_entropy": 0.11281, "router_noise": 0.0, "grad_norm": 0.4121, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.193, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1144.5} |
| {"event": "train", "step": 5400, "phase": 2, "objective": "parity_b", "lm_loss": 1.08384, "bpb": 1.56365, "chunk_loss": 0.03006, "F": 0.14482, "G": 0.15782, "chunks": 148.3, "bytes_per_chunk": 6.905, "boundary_entropy": 0.10879, "router_noise": 0.0, "grad_norm": 0.4227, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 1163.2} |
| {"event": "train", "step": 5500, "phase": 2, "objective": "parity_b", "lm_loss": 1.10446, "bpb": 1.59339, "chunk_loss": 0.030402, "F": 0.13994, "G": 0.1517, "chunks": 143.3, "bytes_per_chunk": 7.146, "boundary_entropy": 0.10713, "router_noise": 0.0, "grad_norm": 0.3919, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.191, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1182.3} |
| {"event": "train", "step": 5600, "phase": 2, "objective": "parity_b", "lm_loss": 1.11845, "bpb": 1.61358, "chunk_loss": 0.030521, "F": 0.15166, "G": 0.16184, "chunks": 155.3, "bytes_per_chunk": 6.594, "boundary_entropy": 0.10925, "router_noise": 0.0, "grad_norm": 0.4154, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "th", "n_target_mean": 16.5724, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1201.0} |
| {"event": "train", "step": 5700, "phase": 2, "objective": "parity_b", "lm_loss": 1.21618, "bpb": 1.75457, "chunk_loss": 0.030015, "F": 0.16203, "G": 0.17269, "chunks": 165.92, "bytes_per_chunk": 6.172, "boundary_entropy": 0.11108, "router_noise": 0.0, "grad_norm": 0.4355, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.189, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1219.9} |
| {"event": "train", "step": 5800, "phase": 2, "objective": "parity_b", "lm_loss": 1.21601, "bpb": 1.75434, "chunk_loss": 0.03059, "F": 0.16637, "G": 0.17589, "chunks": 170.36, "bytes_per_chunk": 6.011, "boundary_entropy": 0.09738, "router_noise": 0.0, "grad_norm": 0.4334, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.194, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1239.2} |
| {"event": "train", "step": 5900, "phase": 2, "objective": "parity_b", "lm_loss": 1.15054, "bpb": 1.65988, "chunk_loss": 0.030174, "F": 0.15977, "G": 0.17004, "chunks": 163.61, "bytes_per_chunk": 6.259, "boundary_entropy": 0.09854, "router_noise": 0.0, "grad_norm": 0.4289, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.188, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 1258.0} |
| {"event": "probe", "step": 6000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3279.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188416, "cps": 25.617188, "mean_p": 0.191226, "boundary_entropy": 0.058516, "bpb": 1.722709}, "de": {"chunks": 3437.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.167389, "cps": 26.851562, "mean_p": 0.154854, "boundary_entropy": 0.082944, "bpb": 1.882238}, "vi": {"chunks": 3251.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.176349, "cps": 25.398438, "mean_p": 0.152787, "boundary_entropy": 0.139923, "bpb": 1.633951}, "tr": {"chunks": 3218.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177761, "cps": 25.140625, "mean_p": 0.166495, "boundary_entropy": 0.078184, "bpb": 1.827387}, "fi": {"chunks": 3224.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171389, "cps": 25.1875, "mean_p": 0.170801, "boundary_entropy": 0.076766, "bpb": 2.020054}, "ru": {"chunks": 2932.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.154723, "cps": 22.90625, "mean_p": 0.097689, "boundary_entropy": 0.106834, "bpb": 1.210636}, "el": {"chunks": 2700.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.13156, "cps": 21.09375, "mean_p": 0.088828, "boundary_entropy": 0.123371, "bpb": 1.209086}, "hi": {"chunks": 2912.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.173076, "cps": 22.75, "mean_p": 0.099632, "boundary_entropy": 0.169897, "bpb": 1.01874}, "ta": {"chunks": 2693.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.144049, "cps": 21.039062, "mean_p": 0.087067, "boundary_entropy": 0.155326, "bpb": 0.872785}, "th": {"chunks": 3131.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.194653, "cps": 24.460938, "mean_p": 0.101281, "boundary_entropy": 0.176616, "bpb": 0.966107}, "zh": {"chunks": 3463.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.586353, "cps": 27.054688, "mean_p": 0.234274, "boundary_entropy": 0.147043, "bpb": 2.741678}, "my": {"chunks": 1506.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.079481, "cps": 11.765625, "mean_p": 0.078032, "boundary_entropy": 0.186864, "bpb": 0.973355}}, "bpb": {"en": 1.72271, "de": 1.88224, "vi": 1.63395, "tr": 1.82739, "fi": 2.02005, "ru": 1.21064, "el": 1.20909, "hi": 1.01874, "ta": 0.87279, "th": 0.96611, "zh": 2.74168, "my": 0.97336}, "gini_cpc": 0.234442, "gini_cps": 0.08144, "bpic": 10.736334, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8884026155829129, "vi": 0.9359583671121625, "tr": 0.9434484859661691, "fi": 0.9096322580522783, "ru": 0.8211782825222633, "el": 0.6982414437036901, "hi": 0.918584263766859, "ta": 0.7645283377137744, "th": 1.0331055767606958, "zh": 3.112015507358497, "my": 0.4218366402120234}, "t": 1279.7} |
| {"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.72476, "de": 1.88464, "vi": 1.64987, "tr": 1.81936, "fi": 2.03377, "ru": 1.20927, "el": 1.2084, "hi": 1.02464, "ta": 0.87538, "th": 0.96364, "zh": 2.75032, "my": 0.98149}, "bytes_per_chunk": {"en": 5.041, "de": 5.92, "vi": 7.627, "tr": 5.977, "fi": 5.847, "ru": 10.681, "el": 12.214, "hi": 13.663, "ta": 18.404, "th": 14.604, "zh": 4.703, "my": 30.942}, "bpb_mean": 1.51046, "t": 1284.9} |
| {"event": "train", "step": 6000, "phase": 2, "objective": "parity_b", "lm_loss": 1.18139, "bpb": 1.70439, "chunk_loss": 0.030707, "F": 0.14975, "G": 0.16258, "chunks": 153.34, "bytes_per_chunk": 6.678, "boundary_entropy": 0.10972, "router_noise": 0.0, "grad_norm": 0.4067, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.271, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1406, "fallback": 0.0, "alpha": 0.03, "t": 1285.1} |
| {"event": "train", "step": 6100, "phase": 2, "objective": "parity_b", "lm_loss": 1.12538, "bpb": 1.62358, "chunk_loss": 0.031603, "F": 0.15263, "G": 0.16112, "chunks": 156.3, "bytes_per_chunk": 6.552, "boundary_entropy": 0.10003, "router_noise": 0.0, "grad_norm": 0.4067, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.191, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1719, "fallback": 0.0, "alpha": 0.03, "t": 1304.2} |
| {"event": "train", "step": 6200, "phase": 2, "objective": "parity_b", "lm_loss": 1.06985, "bpb": 1.54347, "chunk_loss": 0.030156, "F": 0.13795, "G": 0.14865, "chunks": 141.27, "bytes_per_chunk": 7.249, "boundary_entropy": 0.09948, "router_noise": 0.0, "grad_norm": 0.4143, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1323.0} |
| {"event": "train", "step": 6300, "phase": 2, "objective": "parity_b", "lm_loss": 1.13334, "bpb": 1.63506, "chunk_loss": 0.030014, "F": 0.1655, "G": 0.17718, "chunks": 169.47, "bytes_per_chunk": 6.042, "boundary_entropy": 0.10796, "router_noise": 0.0, "grad_norm": 0.4195, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.194, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1342.4} |
| {"event": "train", "step": 6400, "phase": 2, "objective": "parity_b", "lm_loss": 1.10003, "bpb": 1.58701, "chunk_loss": 0.030642, "F": 0.14252, "G": 0.15461, "chunks": 145.94, "bytes_per_chunk": 7.017, "boundary_entropy": 0.10293, "router_noise": 0.0, "grad_norm": 0.4127, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.191, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1361.4} |
| {"event": "train", "step": 6500, "phase": 2, "objective": "parity_b", "lm_loss": 1.0772, "bpb": 1.55407, "chunk_loss": 0.030094, "F": 0.14754, "G": 0.15975, "chunks": 151.08, "bytes_per_chunk": 6.778, "boundary_entropy": 0.09829, "router_noise": 0.0, "grad_norm": 0.3973, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.185, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0469, "fallback": 0.0, "alpha": 0.03, "t": 1379.9} |
| {"event": "train", "step": 6600, "phase": 2, "objective": "parity_b", "lm_loss": 1.13132, "bpb": 1.63214, "chunk_loss": 0.030118, "F": 0.14799, "G": 0.15679, "chunks": 151.55, "bytes_per_chunk": 6.757, "boundary_entropy": 0.09277, "router_noise": 0.0, "grad_norm": 0.4029, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.191, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1399.0} |
| {"event": "train", "step": 6700, "phase": 2, "objective": "parity_b", "lm_loss": 1.18272, "bpb": 1.7063, "chunk_loss": 0.030331, "F": 0.17593, "G": 0.18319, "chunks": 180.16, "bytes_per_chunk": 5.684, "boundary_entropy": 0.09163, "router_noise": 0.0, "grad_norm": 0.4319, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1417.7} |
| {"event": "train", "step": 6800, "phase": 2, "objective": "parity_b", "lm_loss": 1.06344, "bpb": 1.53423, "chunk_loss": 0.030334, "F": 0.15492, "G": 0.16478, "chunks": 158.64, "bytes_per_chunk": 6.455, "boundary_entropy": 0.0917, "router_noise": 0.0, "grad_norm": 0.4088, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.191, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 1436.8} |
| {"event": "train", "step": 6900, "phase": 2, "objective": "parity_b", "lm_loss": 1.12688, "bpb": 1.62574, "chunk_loss": 0.030226, "F": 0.14636, "G": 0.15704, "chunks": 149.88, "bytes_per_chunk": 6.832, "boundary_entropy": 0.09705, "router_noise": 0.0, "grad_norm": 0.4049, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0781, "fallback": 0.0, "alpha": 0.03, "t": 1455.5} |
| {"event": "probe", "step": 7000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 3273.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188071, "cps": 25.570312, "mean_p": 0.190594, "boundary_entropy": 0.050887, "bpb": 1.681173}, "de": {"chunks": 3430.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.167048, "cps": 26.796875, "mean_p": 0.154221, "boundary_entropy": 0.076487, "bpb": 1.831381}, "vi": {"chunks": 3410.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.184974, "cps": 26.640625, "mean_p": 0.156435, "boundary_entropy": 0.128992, "bpb": 1.567533}, "tr": {"chunks": 3285.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181462, "cps": 25.664062, "mean_p": 0.172631, "boundary_entropy": 0.071373, "bpb": 1.773366}, "fi": {"chunks": 3221.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.17123, "cps": 25.164062, "mean_p": 0.170515, "boundary_entropy": 0.069154, "bpb": 1.959725}, "ru": {"chunks": 3116.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.164433, "cps": 24.34375, "mean_p": 0.102029, "boundary_entropy": 0.098064, "bpb": 1.176244}, "el": {"chunks": 3105.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.151294, "cps": 24.257812, "mean_p": 0.096591, "boundary_entropy": 0.107308, "bpb": 1.166267}, "hi": {"chunks": 3061.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.181932, "cps": 23.914062, "mean_p": 0.097462, "boundary_entropy": 0.147101, "bpb": 0.978169}, "ta": {"chunks": 2805.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.15004, "cps": 21.914062, "mean_p": 0.092406, "boundary_entropy": 0.153861, "bpb": 0.850073}, "th": {"chunks": 2965.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.184333, "cps": 23.164062, "mean_p": 0.097233, "boundary_entropy": 0.173548, "bpb": 0.944909}, "zh": {"chunks": 3392.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.574331, "cps": 26.5, "mean_p": 0.220059, "boundary_entropy": 0.134245, "bpb": 2.691756}, "my": {"chunks": 1476.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.077897, "cps": 11.53125, "mean_p": 0.066903, "boundary_entropy": 0.157362, "bpb": 0.938702}}, "bpb": {"en": 1.68117, "de": 1.83138, "vi": 1.56753, "tr": 1.77337, "fi": 1.95972, "ru": 1.17624, "el": 1.16627, "hi": 0.97817, "ta": 0.85007, "th": 0.94491, "zh": 2.69176, "my": 0.9387}, "gini_cpc": 0.217118, "gini_cps": 0.072199, "bpic": 10.503325, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8882185271229347, "vi": 0.9835339415594764, "tr": 0.9648569675114491, "fi": 0.910451794567418, "ru": 0.8743118196614663, "el": 0.8044496633027987, "hi": 0.9673560932824692, "ta": 0.7977843470697671, "th": 0.9801256329872211, "zh": 3.0537994731390627, "my": 0.4141914181840135}, "t": 1477.2} |
| {"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.68276, "de": 1.83556, "vi": 1.58324, "tr": 1.76431, "fi": 1.97374, "ru": 1.17427, "el": 1.16607, "hi": 0.98556, "ta": 0.85464, "th": 0.9422, "zh": 2.70047, "my": 0.94778}, "bytes_per_chunk": {"en": 5.034, "de": 5.948, "vi": 7.237, "tr": 5.847, "fi": 5.858, "ru": 10.101, "el": 10.922, "hi": 13.124, "ta": 17.71, "th": 15.433, "zh": 4.794, "my": 31.338}, "bpb_mean": 1.46755, "t": 1482.5} |
| {"event": "train", "step": 7000, "phase": 2, "objective": "parity_b", "lm_loss": 1.14089, "bpb": 1.64596, "chunk_loss": 0.030299, "F": 0.14845, "G": 0.15918, "chunks": 152.02, "bytes_per_chunk": 6.736, "boundary_entropy": 0.10091, "router_noise": 0.0, "grad_norm": 0.4435, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.272, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1406, "fallback": 0.0, "alpha": 0.03, "t": 1482.7} |
| {"event": "train", "step": 7100, "phase": 2, "objective": "parity_b", "lm_loss": 1.03382, "bpb": 1.49149, "chunk_loss": 0.030504, "F": 0.13922, "G": 0.15006, "chunks": 142.56, "bytes_per_chunk": 7.183, "boundary_entropy": 0.09598, "router_noise": 0.0, "grad_norm": 0.4046, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.189, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 1501.6} |
| {"event": "train", "step": 7200, "phase": 2, "objective": "parity_b", "lm_loss": 1.07901, "bpb": 1.55669, "chunk_loss": 0.03038, "F": 0.14922, "G": 0.15758, "chunks": 152.8, "bytes_per_chunk": 6.702, "boundary_entropy": 0.08788, "router_noise": 0.0, "grad_norm": 0.4173, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.186, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1520.2} |
| {"event": "train", "step": 7300, "phase": 2, "objective": "parity_b", "lm_loss": 1.12106, "bpb": 1.61734, "chunk_loss": 0.030067, "F": 0.15634, "G": 0.16499, "chunks": 160.09, "bytes_per_chunk": 6.396, "boundary_entropy": 0.0867, "router_noise": 0.0, "grad_norm": 0.4367, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.187, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 1538.9} |
| {"event": "train", "step": 7400, "phase": 2, "objective": "parity_b", "lm_loss": 1.07593, "bpb": 1.55224, "chunk_loss": 0.030461, "F": 0.14114, "G": 0.14995, "chunks": 144.53, "bytes_per_chunk": 7.085, "boundary_entropy": 0.09272, "router_noise": 0.0, "grad_norm": 0.4152, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.192, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1558.1} |
| {"event": "train", "step": 7500, "phase": 2, "objective": "parity_b", "lm_loss": 1.12695, "bpb": 1.62585, "chunk_loss": 0.030061, "F": 0.15323, "G": 0.16253, "chunks": 156.91, "bytes_per_chunk": 6.526, "boundary_entropy": 0.08938, "router_noise": 0.0, "grad_norm": 0.4097, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.189, "worst_lang": "de", "n_target_mean": 7.102, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 1577.0} |
| {"event": "train", "step": 7600, "phase": 2, "objective": "parity_b", "lm_loss": 1.07753, "bpb": 1.55455, "chunk_loss": 0.030245, "F": 0.14044, "G": 0.15043, "chunks": 143.81, "bytes_per_chunk": 7.12, "boundary_entropy": 0.09405, "router_noise": 0.0, "grad_norm": 0.4159, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.192, "worst_lang": "zh", "n_target_mean": 5.4871, "frac_in_batch": 0.1094, "fallback": 0.0, "alpha": 0.03, "t": 1596.1} |
| {"event": "phase_transition", "step": 7700, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 295680, "t": 1614.7} |
| {"event": "train", "step": 7700, "phase": 3, "objective": "none", "lm_loss": 1.08346, "bpb": 1.5631, "chunk_loss": 0.0, "F": 0.14569, "G": 0.15404, "chunks": 149.19, "bytes_per_chunk": 6.864, "boundary_entropy": 0.09183, "router_noise": 0.0, "grad_norm": 0.4121, "lr": 9.06e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 1614.9} |
| {"event": "train", "step": 7800, "phase": 3, "objective": "none", "lm_loss": 1.10198, "bpb": 1.58982, "chunk_loss": 0.0, "F": 0.161, "G": 0.1694, "chunks": 164.86, "bytes_per_chunk": 6.211, "boundary_entropy": 0.08205, "router_noise": 0.0, "grad_norm": 0.4081, "lr": 8.73e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 1633.8} |
| {"event": "train", "step": 7900, "phase": 3, "objective": "none", "lm_loss": 1.01927, "bpb": 1.47049, "chunk_loss": 0.0, "F": 0.13913, "G": 0.15293, "chunks": 142.47, "bytes_per_chunk": 7.188, "boundary_entropy": 0.10359, "router_noise": 0.0, "grad_norm": 0.4028, "lr": 8.4e-05, "router_trainable": false, "sec_per_step": 0.185, "alpha": 0.0, "t": 1652.3} |
| {"event": "probe", "step": 8000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3277.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188301, "cps": 25.601562, "mean_p": 0.190807, "boundary_entropy": 0.048406, "bpb": 1.650347}, "de": {"chunks": 3480.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.169483, "cps": 27.1875, "mean_p": 0.159957, "boundary_entropy": 0.071403, "bpb": 1.794901}, "vi": {"chunks": 3699.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.200651, "cps": 28.898438, "mean_p": 0.16769, "boundary_entropy": 0.129948, "bpb": 1.530274}, "tr": {"chunks": 3276.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.180964, "cps": 25.59375, "mean_p": 0.171026, "boundary_entropy": 0.067179, "bpb": 1.732951}, "fi": {"chunks": 3224.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171389, "cps": 25.1875, "mean_p": 0.17098, "boundary_entropy": 0.066101, "bpb": 1.91467}, "ru": {"chunks": 3477.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.183483, "cps": 27.164062, "mean_p": 0.110175, "boundary_entropy": 0.095536, "bpb": 1.148182}, "el": {"chunks": 3249.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.15831, "cps": 25.382812, "mean_p": 0.099013, "boundary_entropy": 0.106867, "bpb": 1.139415}, "hi": {"chunks": 3548.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.210877, "cps": 27.71875, "mean_p": 0.106167, "boundary_entropy": 0.141304, "bpb": 0.958237}, "ta": {"chunks": 2762.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.14774, "cps": 21.578125, "mean_p": 0.090419, "boundary_entropy": 0.144623, "bpb": 0.826811}, "th": {"chunks": 3054.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.189866, "cps": 23.859375, "mean_p": 0.099324, "boundary_entropy": 0.170676, "bpb": 0.924771}, "zh": {"chunks": 3429.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.580596, "cps": 26.789062, "mean_p": 0.224429, "boundary_entropy": 0.120719, "bpb": 2.652557}, "my": {"chunks": 1527.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.080589, "cps": 11.929688, "mean_p": 0.073488, "boundary_entropy": 0.1692, "bpb": 0.923047}}, "bpb": {"en": 1.65035, "de": 1.7949, "vi": 1.53027, "tr": 1.73295, "fi": 1.91467, "ru": 1.14818, "el": 1.13942, "hi": 0.95824, "ta": 0.82681, "th": 0.92477, "zh": 2.65256, "my": 0.92305}, "gini_cpc": 0.220661, "gini_cps": 0.078404, "bpic": 10.098968, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9000663296597017, "vi": 1.0655868887204332, "tr": 0.961039018395584, "fi": 0.9101874196379068, "ru": 0.9744131924511045, "el": 0.8407300005080749, "hi": 1.1198921929191752, "ta": 0.7845956275933563, "th": 1.0083136495742928, "zh": 3.0833421601220463, "my": 0.42797986735809856}, "t": 1675.0} |
| {"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.65153, "de": 1.80008, "vi": 1.54684, "tr": 1.72487, "fi": 1.92779, "ru": 1.14658, "el": 1.13796, "hi": 0.96401, "ta": 0.83239, "th": 0.92257, "zh": 2.664, "my": 0.9324}, "bytes_per_chunk": {"en": 5.014, "de": 5.807, "vi": 6.638, "tr": 5.877, "fi": 5.845, "ru": 9.15, "el": 10.49, "hi": 11.744, "ta": 18.136, "th": 14.943, "zh": 4.738, "my": 30.225}, "bpb_mean": 1.43758, "t": 1680.4} |
| {"event": "train", "step": 8000, "phase": 3, "objective": "none", "lm_loss": 1.03214, "bpb": 1.48906, "chunk_loss": 0.0, "F": 0.15196, "G": 0.1628, "chunks": 155.61, "bytes_per_chunk": 6.581, "boundary_entropy": 0.09465, "router_noise": 0.0, "grad_norm": 0.4006, "lr": 8.08e-05, "router_trainable": false, "sec_per_step": 0.284, "alpha": 0.0, "t": 1680.6} |
| {"event": "train", "step": 8100, "phase": 3, "objective": "none", "lm_loss": 1.10739, "bpb": 1.59763, "chunk_loss": 0.0, "F": 0.15001, "G": 0.16187, "chunks": 153.61, "bytes_per_chunk": 6.666, "boundary_entropy": 0.09327, "router_noise": 0.0, "grad_norm": 0.4212, "lr": 7.77e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 1699.6} |
| {"event": "train", "step": 8200, "phase": 3, "objective": "none", "lm_loss": 1.09669, "bpb": 1.58218, "chunk_loss": 0.0, "F": 0.16496, "G": 0.17274, "chunks": 168.92, "bytes_per_chunk": 6.062, "boundary_entropy": 0.08406, "router_noise": 0.0, "grad_norm": 0.4095, "lr": 7.47e-05, "router_trainable": false, "sec_per_step": 0.187, "alpha": 0.0, "t": 1718.3} |
| {"event": "train", "step": 8300, "phase": 3, "objective": "none", "lm_loss": 1.05329, "bpb": 1.51957, "chunk_loss": 0.0, "F": 0.15073, "G": 0.16213, "chunks": 154.34, "bytes_per_chunk": 6.635, "boundary_entropy": 0.09441, "router_noise": 0.0, "grad_norm": 0.4187, "lr": 7.17e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 1737.2} |
| {"event": "train", "step": 8400, "phase": 3, "objective": "none", "lm_loss": 1.19287, "bpb": 1.72094, "chunk_loss": 0.0, "F": 0.16878, "G": 0.17907, "chunks": 172.83, "bytes_per_chunk": 5.925, "boundary_entropy": 0.09244, "router_noise": 0.0, "grad_norm": 0.4376, "lr": 6.88e-05, "router_trainable": false, "sec_per_step": 0.186, "alpha": 0.0, "t": 1755.8} |
| {"event": "train", "step": 8500, "phase": 3, "objective": "none", "lm_loss": 1.07797, "bpb": 1.55518, "chunk_loss": 0.0, "F": 0.15909, "G": 0.16912, "chunks": 162.91, "bytes_per_chunk": 6.286, "boundary_entropy": 0.09328, "router_noise": 0.0, "grad_norm": 0.4031, "lr": 6.6e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 1774.7} |
| {"event": "train", "step": 8600, "phase": 3, "objective": "none", "lm_loss": 1.0961, "bpb": 1.58133, "chunk_loss": 0.0, "F": 0.16267, "G": 0.17106, "chunks": 166.58, "bytes_per_chunk": 6.147, "boundary_entropy": 0.08452, "router_noise": 0.0, "grad_norm": 0.4305, "lr": 6.33e-05, "router_trainable": false, "sec_per_step": 0.185, "alpha": 0.0, "t": 1793.3} |
| {"event": "train", "step": 8700, "phase": 3, "objective": "none", "lm_loss": 1.05026, "bpb": 1.5152, "chunk_loss": 0.0, "F": 0.15425, "G": 0.16434, "chunks": 157.95, "bytes_per_chunk": 6.483, "boundary_entropy": 0.08818, "router_noise": 0.0, "grad_norm": 0.446, "lr": 6.07e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 1812.1} |
| {"event": "train", "step": 8800, "phase": 3, "objective": "none", "lm_loss": 1.12405, "bpb": 1.62166, "chunk_loss": 0.0, "F": 0.16513, "G": 0.17584, "chunks": 169.09, "bytes_per_chunk": 6.056, "boundary_entropy": 0.09233, "router_noise": 0.0, "grad_norm": 0.432, "lr": 5.82e-05, "router_trainable": false, "sec_per_step": 0.187, "alpha": 0.0, "t": 1830.9} |
| {"event": "train", "step": 8900, "phase": 3, "objective": "none", "lm_loss": 1.0754, "bpb": 1.55147, "chunk_loss": 0.0, "F": 0.14841, "G": 0.16319, "chunks": 151.97, "bytes_per_chunk": 6.738, "boundary_entropy": 0.10202, "router_noise": 0.0, "grad_norm": 0.4362, "lr": 5.58e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 1849.7} |
| {"event": "probe", "step": 9000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3282.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188588, "cps": 25.640625, "mean_p": 0.191966, "boundary_entropy": 0.04794, "bpb": 1.629481}, "de": {"chunks": 3600.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.175328, "cps": 28.125, "mean_p": 0.168639, "boundary_entropy": 0.0621, "bpb": 1.769964}, "vi": {"chunks": 3589.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.194684, "cps": 28.039062, "mean_p": 0.164993, "boundary_entropy": 0.128554, "bpb": 1.496723}, "tr": {"chunks": 3286.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.181517, "cps": 25.671875, "mean_p": 0.172645, "boundary_entropy": 0.066448, "bpb": 1.705446}, "fi": {"chunks": 3225.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171442, "cps": 25.195312, "mean_p": 0.172383, "boundary_entropy": 0.065023, "bpb": 1.880124}, "ru": {"chunks": 3744.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.197573, "cps": 29.25, "mean_p": 0.117681, "boundary_entropy": 0.096877, "bpb": 1.126607}, "el": {"chunks": 3598.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.175315, "cps": 28.109375, "mean_p": 0.10921, "boundary_entropy": 0.108001, "bpb": 1.110832}, "hi": {"chunks": 3727.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.221516, "cps": 29.117188, "mean_p": 0.113897, "boundary_entropy": 0.144802, "bpb": 0.936223}, "ta": {"chunks": 2721.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.145547, "cps": 21.257812, "mean_p": 0.090712, "boundary_entropy": 0.145496, "bpb": 0.809871}, "th": {"chunks": 3228.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.200684, "cps": 25.21875, "mean_p": 0.106304, "boundary_entropy": 0.182722, "bpb": 0.910758}, "zh": {"chunks": 3441.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.582628, "cps": 26.882812, "mean_p": 0.227289, "boundary_entropy": 0.122477, "bpb": 2.618505}, "my": {"chunks": 1508.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.079586, "cps": 11.78125, "mean_p": 0.07678, "boundary_entropy": 0.178801, "bpb": 0.904561}}, "bpb": {"en": 1.62948, "de": 1.76996, "vi": 1.49672, "tr": 1.70545, "fi": 1.88012, "ru": 1.12661, "el": 1.11083, "hi": 0.93622, "ta": 0.80987, "th": 0.91076, "zh": 2.6185, "my": 0.90456}, "gini_cpc": 0.218683, "gini_cps": 0.083872, "bpic": 9.853424, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9296846001055419, "vi": 1.0323236094603847, "tr": 0.9625040179892815, "fi": 0.9090826704239816, "ru": 1.047640234814025, "el": 0.9296208534394148, "hi": 1.1745996398021357, "ta": 0.771771284738707, "th": 1.0641381281685107, "zh": 3.0894187177418804, "my": 0.4220107450816031}, "t": 1871.4} |
| {"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.63089, "de": 1.77519, "vi": 1.51325, "tr": 1.69685, "fi": 1.89401, "ru": 1.12383, "el": 1.11048, "hi": 0.94319, "ta": 0.81606, "th": 0.90785, "zh": 2.62892, "my": 0.91165}, "bytes_per_chunk": {"en": 4.997, "de": 5.588, "vi": 6.834, "tr": 5.861, "fi": 5.827, "ru": 8.566, "el": 9.798, "hi": 11.313, "ta": 18.376, "th": 14.203, "zh": 4.72, "my": 30.653}, "bpb_mean": 1.41268, "t": 1877.0} |
| {"event": "train", "step": 9000, "phase": 3, "objective": "none", "lm_loss": 1.09122, "bpb": 1.5743, "chunk_loss": 0.0, "F": 0.1669, "G": 0.17678, "chunks": 170.91, "bytes_per_chunk": 5.992, "boundary_entropy": 0.08601, "router_noise": 0.0, "grad_norm": 0.4666, "lr": 5.35e-05, "router_trainable": false, "sec_per_step": 0.275, "alpha": 0.0, "t": 1877.2} |
| {"event": "train", "step": 9100, "phase": 3, "objective": "none", "lm_loss": 1.07698, "bpb": 1.55376, "chunk_loss": 0.0, "F": 0.16304, "G": 0.17279, "chunks": 166.95, "bytes_per_chunk": 6.133, "boundary_entropy": 0.08967, "router_noise": 0.0, "grad_norm": 0.434, "lr": 5.12e-05, "router_trainable": false, "sec_per_step": 0.19, "alpha": 0.0, "t": 1896.2} |
| {"event": "train", "step": 9200, "phase": 3, "objective": "none", "lm_loss": 1.01167, "bpb": 1.45953, "chunk_loss": 0.0, "F": 0.15134, "G": 0.16311, "chunks": 154.97, "bytes_per_chunk": 6.608, "boundary_entropy": 0.09262, "router_noise": 0.0, "grad_norm": 0.4197, "lr": 4.91e-05, "router_trainable": false, "sec_per_step": 0.19, "alpha": 0.0, "t": 1915.2} |
| {"event": "train", "step": 9300, "phase": 3, "objective": "none", "lm_loss": 1.04814, "bpb": 1.51215, "chunk_loss": 0.0, "F": 0.15385, "G": 0.16751, "chunks": 157.55, "bytes_per_chunk": 6.5, "boundary_entropy": 0.10619, "router_noise": 0.0, "grad_norm": 0.4265, "lr": 4.71e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 1934.1} |
| {"event": "train", "step": 9400, "phase": 3, "objective": "none", "lm_loss": 1.01352, "bpb": 1.46221, "chunk_loss": 0.0, "F": 0.14423, "G": 0.15985, "chunks": 147.69, "bytes_per_chunk": 6.934, "boundary_entropy": 0.10802, "router_noise": 0.0, "grad_norm": 0.4121, "lr": 4.52e-05, "router_trainable": false, "sec_per_step": 0.186, "alpha": 0.0, "t": 1952.7} |
| {"event": "train", "step": 9500, "phase": 3, "objective": "none", "lm_loss": 1.08275, "bpb": 1.56208, "chunk_loss": 0.0, "F": 0.16138, "G": 0.17189, "chunks": 165.25, "bytes_per_chunk": 6.197, "boundary_entropy": 0.09043, "router_noise": 0.0, "grad_norm": 0.4521, "lr": 4.34e-05, "router_trainable": false, "sec_per_step": 0.187, "alpha": 0.0, "t": 1971.5} |
| {"event": "train", "step": 9600, "phase": 3, "objective": "none", "lm_loss": 1.04153, "bpb": 1.50262, "chunk_loss": 0.0, "F": 0.16487, "G": 0.17516, "chunks": 168.83, "bytes_per_chunk": 6.065, "boundary_entropy": 0.09367, "router_noise": 0.0, "grad_norm": 0.4243, "lr": 4.17e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 1990.3} |
| {"event": "train", "step": 9700, "phase": 3, "objective": "none", "lm_loss": 1.06914, "bpb": 1.54244, "chunk_loss": 0.0, "F": 0.15041, "G": 0.16228, "chunks": 154.02, "bytes_per_chunk": 6.649, "boundary_entropy": 0.09219, "router_noise": 0.0, "grad_norm": 0.4411, "lr": 4.01e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 2009.1} |
| {"event": "train", "step": 9800, "phase": 3, "objective": "none", "lm_loss": 0.99643, "bpb": 1.43755, "chunk_loss": 0.0, "F": 0.14915, "G": 0.16117, "chunks": 152.73, "bytes_per_chunk": 6.704, "boundary_entropy": 0.09619, "router_noise": 0.0, "grad_norm": 0.4211, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 2028.0} |
| {"event": "train", "step": 9900, "phase": 3, "objective": "none", "lm_loss": 1.04929, "bpb": 1.5138, "chunk_loss": 0.0, "F": 0.16661, "G": 0.17476, "chunks": 170.61, "bytes_per_chunk": 6.002, "boundary_entropy": 0.08062, "router_noise": 0.0, "grad_norm": 0.4177, "lr": 3.72e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 2046.8} |
| {"event": "probe", "step": 10000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3287.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188875, "cps": 25.679688, "mean_p": 0.192265, "boundary_entropy": 0.047722, "bpb": 1.613928}, "de": {"chunks": 3604.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.175522, "cps": 28.15625, "mean_p": 0.169385, "boundary_entropy": 0.061787, "bpb": 1.750809}, "vi": {"chunks": 3810.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.206672, "cps": 29.765625, "mean_p": 0.173419, "boundary_entropy": 0.133105, "bpb": 1.477999}, "tr": {"chunks": 3305.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.182566, "cps": 25.820312, "mean_p": 0.173272, "boundary_entropy": 0.065814, "bpb": 1.685044}, "fi": {"chunks": 3232.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171814, "cps": 25.25, "mean_p": 0.172582, "boundary_entropy": 0.065218, "bpb": 1.857276}, "ru": {"chunks": 3883.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.204908, "cps": 30.335938, "mean_p": 0.121817, "boundary_entropy": 0.10238, "bpb": 1.10838}, "el": {"chunks": 3680.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.179311, "cps": 28.75, "mean_p": 0.110267, "boundary_entropy": 0.107497, "bpb": 1.096569}, "hi": {"chunks": 3789.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.225201, "cps": 29.601562, "mean_p": 0.115403, "boundary_entropy": 0.148541, "bpb": 0.921166}, "ta": {"chunks": 2879.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.153998, "cps": 22.492188, "mean_p": 0.094986, "boundary_entropy": 0.151766, "bpb": 0.799062}, "th": {"chunks": 3496.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.217345, "cps": 27.3125, "mean_p": 0.112698, "boundary_entropy": 0.189157, "bpb": 0.896069}, "zh": {"chunks": 3448.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583813, "cps": 26.9375, "mean_p": 0.228949, "boundary_entropy": 0.12129, "bpb": 2.586847}, "my": {"chunks": 1518.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.080114, "cps": 11.859375, "mean_p": 0.078504, "boundary_entropy": 0.182713, "bpb": 0.891592}}, "bpb": {"en": 1.61393, "de": 1.75081, "vi": 1.478, "tr": 1.68504, "fi": 1.85728, "ru": 1.10838, "el": 1.09657, "hi": 0.92117, "ta": 0.79906, "th": 0.89607, "zh": 2.58685, "my": 0.89159}, "gini_cpc": 0.218505, "gini_cps": 0.085988, "bpic": 9.611104, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9293018276203551, "vi": 1.0942240346677237, "tr": 0.9665967454158683, "fi": 0.9096700260353268, "ru": 1.0848822217209717, "el": 0.9493610061598657, "hi": 1.1923230737865544, "ta": 0.8153434965870479, "th": 1.1507335488310835, "zh": 3.09099448813276, "my": 0.42416302886242907}, "t": 2068.8} |
| {"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.61484, "de": 1.75624, "vi": 1.49442, "tr": 1.67646, "fi": 1.87125, "ru": 1.10641, "el": 1.09598, "hi": 0.92796, "ta": 0.80572, "th": 0.89425, "zh": 2.59829, "my": 0.89957}, "bytes_per_chunk": {"en": 4.99, "de": 5.57, "vi": 6.49, "tr": 5.837, "fi": 5.824, "ru": 8.324, "el": 9.686, "hi": 11.164, "ta": 17.377, "th": 13.183, "zh": 4.705, "my": 30.549}, "bpb_mean": 1.39512, "t": 2074.6} |
| {"event": "train", "step": 10000, "phase": 3, "objective": "none", "lm_loss": 1.07218, "bpb": 1.54683, "chunk_loss": 0.0, "F": 0.16415, "G": 0.17371, "chunks": 168.09, "bytes_per_chunk": 6.092, "boundary_entropy": 0.0841, "router_noise": 0.0, "grad_norm": 0.4273, "lr": 3.6e-05, "router_trainable": false, "sec_per_step": 0.28, "alpha": 0.0, "t": 2074.8} |
| {"event": "train", "step": 10100, "phase": 3, "objective": "none", "lm_loss": 1.12041, "bpb": 1.61641, "chunk_loss": 0.0, "F": 0.16977, "G": 0.17926, "chunks": 173.84, "bytes_per_chunk": 5.89, "boundary_entropy": 0.08787, "router_noise": 0.0, "grad_norm": 0.439, "lr": 3.49e-05, "router_trainable": false, "sec_per_step": 0.191, "alpha": 0.0, "t": 2093.9} |
| {"event": "train", "step": 10200, "phase": 3, "objective": "none", "lm_loss": 1.03309, "bpb": 1.49044, "chunk_loss": 0.0, "F": 0.1638, "G": 0.17458, "chunks": 167.73, "bytes_per_chunk": 6.105, "boundary_entropy": 0.08965, "router_noise": 0.0, "grad_norm": 0.4353, "lr": 3.38e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 2112.7} |
| {"event": "train", "step": 10300, "phase": 3, "objective": "none", "lm_loss": 1.02838, "bpb": 1.48363, "chunk_loss": 0.0, "F": 0.15854, "G": 0.16863, "chunks": 162.34, "bytes_per_chunk": 6.308, "boundary_entropy": 0.08777, "router_noise": 0.0, "grad_norm": 0.4368, "lr": 3.3e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 2131.5} |
| {"event": "train", "step": 10400, "phase": 3, "objective": "none", "lm_loss": 1.03411, "bpb": 1.4919, "chunk_loss": 0.0, "F": 0.15802, "G": 0.16986, "chunks": 161.81, "bytes_per_chunk": 6.328, "boundary_entropy": 0.09582, "router_noise": 0.0, "grad_norm": 0.4476, "lr": 3.22e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 2150.4} |
| {"event": "train", "step": 10500, "phase": 3, "objective": "none", "lm_loss": 0.9972, "bpb": 1.43866, "chunk_loss": 0.0, "F": 0.14656, "G": 0.16012, "chunks": 150.08, "bytes_per_chunk": 6.823, "boundary_entropy": 0.09582, "router_noise": 0.0, "grad_norm": 0.4467, "lr": 3.15e-05, "router_trainable": false, "sec_per_step": 0.186, "alpha": 0.0, "t": 2169.0} |
| {"event": "train", "step": 10600, "phase": 3, "objective": "none", "lm_loss": 1.0793, "bpb": 1.5571, "chunk_loss": 0.0, "F": 0.16435, "G": 0.17623, "chunks": 168.3, "bytes_per_chunk": 6.084, "boundary_entropy": 0.08952, "router_noise": 0.0, "grad_norm": 0.4308, "lr": 3.1e-05, "router_trainable": false, "sec_per_step": 0.187, "alpha": 0.0, "t": 2187.7} |
| {"event": "train", "step": 10700, "phase": 3, "objective": "none", "lm_loss": 1.05356, "bpb": 1.51997, "chunk_loss": 0.0, "F": 0.16685, "G": 0.17831, "chunks": 170.86, "bytes_per_chunk": 5.993, "boundary_entropy": 0.0918, "router_noise": 0.0, "grad_norm": 0.464, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.188, "alpha": 0.0, "t": 2206.5} |
| {"event": "train", "step": 10800, "phase": 3, "objective": "none", "lm_loss": 1.03266, "bpb": 1.48981, "chunk_loss": 0.0, "F": 0.15439, "G": 0.1668, "chunks": 158.09, "bytes_per_chunk": 6.477, "boundary_entropy": 0.09244, "router_noise": 0.0, "grad_norm": 0.4534, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.187, "alpha": 0.0, "t": 2225.2} |
| {"event": "train", "step": 10900, "phase": 3, "objective": "none", "lm_loss": 1.10705, "bpb": 1.59713, "chunk_loss": 0.0, "F": 0.16576, "G": 0.17797, "chunks": 169.73, "bytes_per_chunk": 6.033, "boundary_entropy": 0.09235, "router_noise": 0.0, "grad_norm": 0.4632, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.189, "alpha": 0.0, "t": 2244.1} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3290.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189048, "cps": 25.703125, "mean_p": 0.192932, "boundary_entropy": 0.048392, "bpb": 1.605337}, "de": {"chunks": 3606.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.17562, "cps": 28.171875, "mean_p": 0.170191, "boundary_entropy": 0.060371, "bpb": 1.741951}, "vi": {"chunks": 3734.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.202549, "cps": 29.171875, "mean_p": 0.172003, "boundary_entropy": 0.134018, "bpb": 1.465491}, "tr": {"chunks": 3360.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185605, "cps": 26.25, "mean_p": 0.175404, "boundary_entropy": 0.066737, "bpb": 1.674711}, "fi": {"chunks": 3235.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171974, "cps": 25.273438, "mean_p": 0.173306, "boundary_entropy": 0.067636, "bpb": 1.844871}, "ru": {"chunks": 3936.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.207704, "cps": 30.75, "mean_p": 0.123097, "boundary_entropy": 0.104567, "bpb": 1.096364}, "el": {"chunks": 3733.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.181893, "cps": 29.164062, "mean_p": 0.112573, "boundary_entropy": 0.109971, "bpb": 1.087206}, "hi": {"chunks": 3824.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.227281, "cps": 29.875, "mean_p": 0.117869, "boundary_entropy": 0.151467, "bpb": 0.909708}, "ta": {"chunks": 2881.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.154105, "cps": 22.507812, "mean_p": 0.094826, "boundary_entropy": 0.151176, "bpb": 0.791208}, "th": {"chunks": 3666.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.227914, "cps": 28.640625, "mean_p": 0.11509, "boundary_entropy": 0.188877, "bpb": 0.886451}, "zh": {"chunks": 3445.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583305, "cps": 26.914062, "mean_p": 0.228299, "boundary_entropy": 0.123167, "bpb": 2.566384}, "my": {"chunks": 1518.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.080114, "cps": 11.859375, "mean_p": 0.077564, "boundary_entropy": 0.180143, "bpb": 0.882595}}, "bpb": {"en": 1.60534, "de": 1.74195, "vi": 1.46549, "tr": 1.67471, "fi": 1.84487, "ru": 1.09636, "el": 1.08721, "hi": 0.90971, "ta": 0.79121, "th": 0.88645, "zh": 2.56638, "my": 0.8826}, "gini_cpc": 0.219433, "gini_cps": 0.086739, "bpic": 9.540146, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9289696754738499, "vi": 1.071419123957254, "tr": 0.981786256186526, "fi": 0.9096841416754958, "ru": 1.0986872829634857, "el": 0.9621557220922128, "hi": 1.2022396112312967, "ta": 0.8151659132238761, "th": 1.2055899462675963, "zh": 3.085489024092752, "my": 0.42377625406407426}, "t": 2265.9} |
| {"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.6059, "de": 1.74765, "vi": 1.4826, "tr": 1.66577, "fi": 1.85873, "ru": 1.09477, "el": 1.08723, "hi": 0.9166, "ta": 0.79739, "th": 0.88467, "zh": 2.57667, "my": 0.89134}, "bytes_per_chunk": {"en": 4.986, "de": 5.566, "vi": 6.611, "tr": 5.747, "fi": 5.83, "ru": 8.274, "el": 9.562, "hi": 11.056, "ta": 17.264, "th": 12.518, "zh": 4.719, "my": 30.334}, "bpb_mean": 1.38411, "t": 2271.4} |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 3290.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.189048, "cps": 25.703125, "mean_p": 0.192932, "boundary_entropy": 0.048392, "bpb": 1.605337}, "de": {"chunks": 3606.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.17562, "cps": 28.171875, "mean_p": 0.170191, "boundary_entropy": 0.060371, "bpb": 1.741951}, "vi": {"chunks": 3734.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.202549, "cps": 29.171875, "mean_p": 0.172003, "boundary_entropy": 0.134018, "bpb": 1.465491}, "tr": {"chunks": 3360.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.185605, "cps": 26.25, "mean_p": 0.175404, "boundary_entropy": 0.066737, "bpb": 1.674711}, "fi": {"chunks": 3235.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.171974, "cps": 25.273438, "mean_p": 0.173306, "boundary_entropy": 0.067636, "bpb": 1.844871}, "ru": {"chunks": 3936.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.207704, "cps": 30.75, "mean_p": 0.123097, "boundary_entropy": 0.104567, "bpb": 1.096364}, "el": {"chunks": 3733.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.181893, "cps": 29.164062, "mean_p": 0.112573, "boundary_entropy": 0.109971, "bpb": 1.087206}, "hi": {"chunks": 3824.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.227281, "cps": 29.875, "mean_p": 0.117869, "boundary_entropy": 0.151467, "bpb": 0.909708}, "ta": {"chunks": 2881.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.154105, "cps": 22.507812, "mean_p": 0.094826, "boundary_entropy": 0.151176, "bpb": 0.791208}, "th": {"chunks": 3666.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.227914, "cps": 28.640625, "mean_p": 0.11509, "boundary_entropy": 0.188877, "bpb": 0.886451}, "zh": {"chunks": 3445.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583305, "cps": 26.914062, "mean_p": 0.228299, "boundary_entropy": 0.123167, "bpb": 2.566384}, "my": {"chunks": 1518.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.080114, "cps": 11.859375, "mean_p": 0.077564, "boundary_entropy": 0.180143, "bpb": 0.882595}}, "bpb": {"en": 1.60534, "de": 1.74195, "vi": 1.46549, "tr": 1.67471, "fi": 1.84487, "ru": 1.09636, "el": 1.08721, "hi": 0.90971, "ta": 0.79121, "th": 0.88645, "zh": 2.56638, "my": 0.8826}, "gini_cpc": 0.219433, "gini_cps": 0.086739, "bpic": 9.540146, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9289696754738499, "vi": 1.071419123957254, "tr": 0.981786256186526, "fi": 0.9096841416754958, "ru": 1.0986872829634857, "el": 0.9621557220922128, "hi": 1.2022396112312967, "ta": 0.8151659132238761, "th": 1.2055899462675963, "zh": 3.085489024092752, "my": 0.42377625406407426}, "t": 2274.7} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s1/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s1/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s1/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s1/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_E_s1/th0.70"], "t": 2277.4} |
| {"event": "final", "run": "spec_E_s1", "condition": "E", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 0.18904786531057863, "cps": 25.703125, "bpb": 1.6011961329843718, "bytes_per_chunk": 4.9802703447227685}, "de": {"cpc": 0.17561973408659232, "cps": 28.171875, "bpb": 1.7453144449751115, "bytes_per_chunk": 5.526499978918076}, "vi": {"cpc": 0.2025494982370491, "cps": 29.171875, "bpb": 1.4825155276576893, "bytes_per_chunk": 6.608991610939184}, "tr": {"cpc": 0.18560459592332762, "cps": 26.25, "bpb": 1.664305510683781, "bytes_per_chunk": 5.763243231359445}, "fi": {"cpc": 0.17197384509063846, "cps": 25.2734375, "bpb": 1.8642697173441374, "bytes_per_chunk": 5.846502110386895}, "ru": {"cpc": 0.20770448548812664, "cps": 30.75, "bpb": 1.0954605693232156, "bytes_per_chunk": 8.322361998492005}, "el": {"cpc": 0.18189348535789115, "cps": 29.1640625, "bpb": 1.0836721152600655, "bytes_per_chunk": 9.515381404381205}, "hi": {"cpc": 0.22728083209509659, "cps": 29.875, "bpb": 0.9184566162455753, "bytes_per_chunk": 11.050320894500006}, "ta": {"cpc": 0.15410537576892216, "cps": 22.5078125, "bpb": 0.7952043143490795, "bytes_per_chunk": 17.154898239644}, "th": {"cpc": 0.22791420578178428, "cps": 28.640625, "bpb": 0.8867598159816795, "bytes_per_chunk": 12.423444664289184}, "zh": {"cpc": 0.5833051134439553, "cps": 26.9140625, "bpb": 2.5821120701302998, "bytes_per_chunk": 4.733614124423859}, "my": {"cpc": 0.08011399620012666, "cps": 11.859375, "bpb": 0.8891614462118219, "bytes_per_chunk": 30.233140155119223}}, "gini_cpc": 0.21943267932036825, "gini_cps": 0.08673892148089224, "bpic": 9.540146166848961, "seed": 1, "mean_boundary_entropy": 0.10158485919237137, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 7700, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 295680}], "t": 2288.1} |
|
|
