| {"event": "start", "run": "pilot_E_s1", "condition": "E", "params": 22516992, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "global", "p2_objective": "parity_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": 1.9} |
| {"event": "probe_schedule", "n": 40, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 200, 256, 400, 600, 800, 1000, 1200, 1400, 1600, 1800, 2000, 2200, 2400, 2600, 2800, 3000, 3200, 3400, 3600, 3800, 4000, 4200, 4400, 4600, 4800, 5000, 5200, 5400, 5600, 5800, 6000], "t": 1.9} |
| {"event": "probe", "step": 0, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2674.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.153652, "cps": 20.890625, "mean_p": 0.478126, "boundary_entropy": 0.685438, "bpb": 8.041379}, "de": {"chunks": 2056.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.100131, "cps": 16.0625, "mean_p": 0.469682, "boundary_entropy": 0.685273, "bpb": 8.067143}, "vi": {"chunks": 14185.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.76946, "cps": 110.820312, "mean_p": 0.507219, "boundary_entropy": 0.688283, "bpb": 7.894113}, "tr": {"chunks": 5645.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.311827, "cps": 44.101562, "mean_p": 0.486063, "boundary_entropy": 0.686375, "bpb": 8.01701}, "fi": {"chunks": 4732.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.251555, "cps": 36.96875, "mean_p": 0.48451, "boundary_entropy": 0.686595, "bpb": 8.075294}, "ru": {"chunks": 17548.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.926016, "cps": 137.09375, "mean_p": 0.502342, "boundary_entropy": 0.690066, "bpb": 8.095633}, "el": {"chunks": 31473.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.533548, "cps": 245.882812, "mean_p": 0.517965, "boundary_entropy": 0.689704, "bpb": 7.804942}, "hi": {"chunks": 1659.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.098603, "cps": 12.960938, "mean_p": 0.476241, "boundary_entropy": 0.689492, "bpb": 8.047677}, "ta": {"chunks": 39693.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.123188, "cps": 310.101562, "mean_p": 0.514597, "boundary_entropy": 0.690542, "bpb": 8.096414}, "th": {"chunks": 44424.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.761828, "cps": 347.0625, "mean_p": 0.52078, "boundary_entropy": 0.690224, "bpb": 8.201461}, "zh": {"chunks": 12933.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.189807, "cps": 101.039062, "mean_p": 0.528309, "boundary_entropy": 0.684752, "bpb": 8.121995}, "my": {"chunks": 7048.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.371965, "cps": 55.0625, "mean_p": 0.491145, "boundary_entropy": 0.691021, "bpb": 8.112229}}, "bpb": {"en": 8.04138, "de": 8.06714, "vi": 7.89411, "tr": 8.01701, "fi": 8.07529, "ru": 8.09563, "el": 7.80494, "hi": 8.04768, "ta": 8.09641, "th": 8.20146, "zh": 8.12199, "my": 8.11223}, "gini_cpc": 0.507116, "gini_cps": 0.500863, "bpic": 2.084973, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6516785410034256, "vi": 5.0078223656304, "tr": 2.029439514461233, "fi": 1.6371767549398106, "ru": 6.026721581613623, "el": 9.980677296913099, "hi": 0.6417324941848254, "ta": 13.818190381639045, "th": 17.9746032262692, "zh": 14.25176170627287, "my": 2.4208352616119475}, "t": 3.8} |
| {"event": "train", "step": 0, "phase": 1, "objective": "global", "lm_loss": 5.57556, "bpb": 8.04383, "chunk_loss": 0.043969, "F": 0.34537, "G": 0.49096, "chunks": 353.66, "bytes_per_chunk": 2.895, "boundary_entropy": 0.6904, "router_noise": 0.0, "grad_norm": 7.3225, "lr": 1e-06, "router_trainable": true, "sec_per_step": 0.005, "alpha": 0.03, "t": 4.1} |
| {"event": "probe", "step": 1, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2651.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.15233, "cps": 20.710938, "mean_p": 0.478021, "boundary_entropy": 0.685429, "bpb": 8.011976}, "de": {"chunks": 2039.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.099304, "cps": 15.929688, "mean_p": 0.469566, "boundary_entropy": 0.685259, "bpb": 8.041936}, "vi": {"chunks": 14106.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.765175, "cps": 110.203125, "mean_p": 0.507011, "boundary_entropy": 0.688289, "bpb": 7.873419}, "tr": {"chunks": 5610.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.309893, "cps": 43.828125, "mean_p": 0.48594, "boundary_entropy": 0.686368, "bpb": 7.996264}, "fi": {"chunks": 4695.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.249588, "cps": 36.679688, "mean_p": 0.484352, "boundary_entropy": 0.686583, "bpb": 8.050004}, "ru": {"chunks": 17481.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.92248, "cps": 136.570312, "mean_p": 0.502276, "boundary_entropy": 0.690066, "bpb": 8.083968}, "el": {"chunks": 31421.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.531014, "cps": 245.476562, "mean_p": 0.517879, "boundary_entropy": 0.68971, "bpb": 7.792285}, "hi": {"chunks": 1659.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.098603, "cps": 12.960938, "mean_p": 0.476277, "boundary_entropy": 0.689497, "bpb": 8.025306}, "ta": {"chunks": 39580.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.117144, "cps": 309.21875, "mean_p": 0.514531, "boundary_entropy": 0.690545, "bpb": 8.080295}, "th": {"chunks": 44407.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.760771, "cps": 346.929688, "mean_p": 0.52071, "boundary_entropy": 0.69023, "bpb": 8.182804}, "zh": {"chunks": 12935.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.190146, "cps": 101.054688, "mean_p": 0.528283, "boundary_entropy": 0.684755, "bpb": 8.116872}, "my": {"chunks": 7056.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.372388, "cps": 55.125, "mean_p": 0.491149, "boundary_entropy": 0.691021, "bpb": 8.097955}}, "bpb": {"en": 8.01198, "de": 8.04194, "vi": 7.87342, "tr": 7.99626, "fi": 8.05, "ru": 8.08397, "el": 7.79229, "hi": 8.02531, "ta": 8.0803, "th": 8.1828, "zh": 8.11687, "my": 8.09796}, "gini_cpc": 0.507778, "gini_cps": 0.501336, "bpic": 2.089855, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6518973431972304, "vi": 5.023138235751804, "tr": 2.0343548202620187, "fi": 1.6384685349376928, "ru": 6.055798986592405, "el": 10.050636125594837, "hi": 0.6473001469069117, "ta": 13.898396997847781, "th": 18.12361224150229, "zh": 14.377632641038025, "my": 2.4446100256358863}, "t": 5.5} |
| {"event": "probe", "step": 2, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2630.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.151123, "cps": 20.546875, "mean_p": 0.477868, "boundary_entropy": 0.685416, "bpb": 7.961856}, "de": {"chunks": 2015.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.098135, "cps": 15.742188, "mean_p": 0.469405, "boundary_entropy": 0.685239, "bpb": 7.999923}, "vi": {"chunks": 13977.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.758177, "cps": 109.195312, "mean_p": 0.506726, "boundary_entropy": 0.688298, "bpb": 7.840287}, "tr": {"chunks": 5554.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.3068, "cps": 43.390625, "mean_p": 0.485759, "boundary_entropy": 0.686358, "bpb": 7.962016}, "fi": {"chunks": 4630.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.246133, "cps": 36.171875, "mean_p": 0.484107, "boundary_entropy": 0.686566, "bpb": 8.009832}, "ru": {"chunks": 17257.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.91066, "cps": 134.820312, "mean_p": 0.502049, "boundary_entropy": 0.690066, "bpb": 8.040047}, "el": {"chunks": 31369.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.52848, "cps": 245.070312, "mean_p": 0.517785, "boundary_entropy": 0.689716, "bpb": 7.769433}, "hi": {"chunks": 1670.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.099257, "cps": 13.046875, "mean_p": 0.476369, "boundary_entropy": 0.689506, "bpb": 7.98382}, "ta": {"chunks": 39395.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.107248, "cps": 307.773438, "mean_p": 0.514417, "boundary_entropy": 0.69055, "bpb": 8.042752}, "th": {"chunks": 44379.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.75903, "cps": 346.710938, "mean_p": 0.520513, "boundary_entropy": 0.690245, "bpb": 8.123062}, "zh": {"chunks": 12928.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.18896, "cps": 101.0, "mean_p": 0.52826, "boundary_entropy": 0.684757, "bpb": 8.106808}, "my": {"chunks": 7075.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.37339, "cps": 55.273438, "mean_p": 0.49116, "boundary_entropy": 0.691021, "bpb": 8.083872}}, "bpb": {"en": 7.96186, "de": 7.99992, "vi": 7.84029, "tr": 7.96202, "fi": 8.00983, "ru": 8.04005, "el": 7.76943, "hi": 7.98382, "ta": 8.04275, "th": 8.12306, "zh": 8.10681, "my": 8.08387}, "gini_cpc": 0.508775, "gini_cps": 0.501994, "bpic": 2.098552, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6493681968690297, "vi": 5.016943324660378, "tr": 2.0301292834475473, "fi": 1.6286864351878898, "ru": 6.025935190665851, "el": 10.11412229884342, "hi": 0.6567948971463116, "ta": 13.943891892771395, "th": 18.256806745533176, "zh": 14.484592197919497, "my": 2.4707649954127007}, "t": 7.1} |
| {"event": "probe", "step": 4, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2559.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.147044, "cps": 19.992188, "mean_p": 0.477479, "boundary_entropy": 0.685387, "bpb": 7.791445}, "de": {"chunks": 1968.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.095846, "cps": 15.375, "mean_p": 0.468986, "boundary_entropy": 0.685193, "bpb": 7.852553}, "vi": {"chunks": 13539.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.734418, "cps": 105.773438, "mean_p": 0.50567, "boundary_entropy": 0.688326, "bpb": 7.72993}, "tr": {"chunks": 5388.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.29763, "cps": 42.09375, "mean_p": 0.485237, "boundary_entropy": 0.686332, "bpb": 7.842848}, "fi": {"chunks": 4451.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.236617, "cps": 34.773438, "mean_p": 0.483395, "boundary_entropy": 0.686515, "bpb": 7.870576}, "ru": {"chunks": 16106.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.849921, "cps": 125.828125, "mean_p": 0.500932, "boundary_entropy": 0.690065, "bpb": 7.865171}, "el": {"chunks": 31138.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.517225, "cps": 243.265625, "mean_p": 0.517308, "boundary_entropy": 0.689748, "bpb": 7.686045}, "hi": {"chunks": 1731.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.102883, "cps": 13.523438, "mean_p": 0.476706, "boundary_entropy": 0.689542, "bpb": 7.830273}, "ta": {"chunks": 38741.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.072265, "cps": 302.664062, "mean_p": 0.513988, "boundary_entropy": 0.690571, "bpb": 7.912802}, "th": {"chunks": 44273.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.75244, "cps": 345.882812, "mean_p": 0.519893, "boundary_entropy": 0.690291, "bpb": 7.955049}, "zh": {"chunks": 12922.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.187944, "cps": 100.953125, "mean_p": 0.528241, "boundary_entropy": 0.684757, "bpb": 8.079672}, "my": {"chunks": 7115.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.375501, "cps": 55.585938, "mean_p": 0.491186, "boundary_entropy": 0.691022, "bpb": 8.000732}}, "bpb": {"en": 7.79144, "de": 7.85255, "vi": 7.72993, "tr": 7.84285, "fi": 7.87058, "ru": 7.86517, "el": 7.68605, "hi": 7.83027, "ta": 7.9128, "th": 7.95505, "zh": 8.07967, "my": 8.00073}, "gini_cpc": 0.512204, "gini_cps": 0.503669, "bpic": 2.132934, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6518182579622349, "vi": 4.994560527769219, "tr": 2.0240948968644252, "fi": 1.6091611733749607, "ru": 5.780059575547424, "el": 10.318194341075825, "hi": 0.6996741506682994, "ta": 14.092861749682617, "th": 18.71852916492505, "zh": 14.879561349774843, "my": 2.5536734583772307}, "t": 8.7} |
| {"event": "probe", "step": 8, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2299.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.132104, "cps": 17.960938, "mean_p": 0.476348, "boundary_entropy": 0.685317, "bpb": 7.2973}, "de": {"chunks": 1785.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.086933, "cps": 13.945312, "mean_p": 0.467597, "boundary_entropy": 0.685052, "bpb": 7.408808}, "vi": {"chunks": 12231.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.663466, "cps": 95.554688, "mean_p": 0.502572, "boundary_entropy": 0.688389, "bpb": 7.434494}, "tr": {"chunks": 4855.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.268188, "cps": 37.929688, "mean_p": 0.483427, "boundary_entropy": 0.686243, "bpb": 7.494923}, "fi": {"chunks": 3815.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.202807, "cps": 29.804688, "mean_p": 0.480874, "boundary_entropy": 0.68633, "bpb": 7.453907}, "ru": {"chunks": 12107.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.638892, "cps": 94.585938, "mean_p": 0.496814, "boundary_entropy": 0.69002, "bpb": 7.244501}, "el": {"chunks": 30503.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.486284, "cps": 238.304688, "mean_p": 0.516124, "boundary_entropy": 0.689821, "bpb": 7.511974}, "hi": {"chunks": 1893.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.112511, "cps": 14.789062, "mean_p": 0.477699, "boundary_entropy": 0.689646, "bpb": 7.379884}, "ta": {"chunks": 34878.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.865633, "cps": 272.484375, "mean_p": 0.511348, "boundary_entropy": 0.690681, "bpb": 7.492678}, "th": {"chunks": 43408.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.698663, "cps": 339.125, "mean_p": 0.517047, "boundary_entropy": 0.690483, "bpb": 7.519987}, "zh": {"chunks": 12915.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.186759, "cps": 100.898438, "mean_p": 0.528224, "boundary_entropy": 0.684759, "bpb": 7.972146}, "my": {"chunks": 7553.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.398617, "cps": 59.007812, "mean_p": 0.491562, "boundary_entropy": 0.691038, "bpb": 7.746936}}, "bpb": {"en": 7.2973, "de": 7.40881, "vi": 7.43449, "tr": 7.49492, "fi": 7.45391, "ru": 7.2445, "el": 7.51197, "hi": 7.37988, "ta": 7.49268, "th": 7.51999, "zh": 7.97215, "my": 7.74694}, "gini_cpc": 0.522735, "gini_cps": 0.507368, "bpic": 2.281125, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6580682870233802, "vi": 5.0223152883183015, "tr": 2.0301299194207845, "fi": 1.5352100606363674, "ru": 4.836291584846457, "el": 11.25088947934946, "hi": 0.8516883172281765, "ta": 14.122489248148053, "th": 20.428376814506123, "zh": 16.553358348503824, "my": 3.017458164617653}, "t": 10.5} |
| {"event": "probe", "step": 16, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2297.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.131989, "cps": 17.945312, "mean_p": 0.478687, "boundary_entropy": 0.68575, "bpb": 6.63594}, "de": {"chunks": 1725.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.084011, "cps": 13.476562, "mean_p": 0.470352, "boundary_entropy": 0.685667, "bpb": 6.75646}, "vi": {"chunks": 9746.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.528668, "cps": 76.140625, "mean_p": 0.496647, "boundary_entropy": 0.688418, "bpb": 7.04879}, "tr": {"chunks": 4406.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.243385, "cps": 34.421875, "mean_p": 0.482565, "boundary_entropy": 0.686346, "bpb": 7.014989}, "fi": {"chunks": 2974.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158099, "cps": 23.234375, "mean_p": 0.478155, "boundary_entropy": 0.686223, "bpb": 6.854203}, "ru": {"chunks": 6012.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.317256, "cps": 46.96875, "mean_p": 0.487689, "boundary_entropy": 0.689678, "bpb": 6.435837}, "el": {"chunks": 23091.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.125128, "cps": 180.398438, "mean_p": 0.506382, "boundary_entropy": 0.69025, "bpb": 6.833777}, "hi": {"chunks": 3514.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.208856, "cps": 27.453125, "mean_p": 0.483415, "boundary_entropy": 0.690143, "bpb": 6.499478}, "ta": {"chunks": 21611.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.155978, "cps": 168.835938, "mean_p": 0.498093, "boundary_entropy": 0.690847, "bpb": 6.584864}, "th": {"chunks": 24763.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.539509, "cps": 193.460938, "mean_p": 0.503029, "boundary_entropy": 0.690961, "bpb": 6.461894}, "zh": {"chunks": 13087.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.215882, "cps": 102.242188, "mean_p": 0.528842, "boundary_entropy": 0.684759, "bpb": 7.681208}, "my": {"chunks": 8509.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.449071, "cps": 66.476562, "mean_p": 0.492464, "boundary_entropy": 0.691082, "bpb": 6.94985}}, "bpb": {"en": 6.63594, "de": 6.75646, "vi": 7.04879, "tr": 7.01499, "fi": 6.8542, "ru": 6.43584, "el": 6.83378, "hi": 6.49948, "ta": 6.58486, "th": 6.46189, "zh": 7.68121, "my": 6.94985}, "gini_cpc": 0.497195, "gini_cps": 0.436334, "bpic": 3.152594, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.6365020654074784, "vi": 4.005404580141856, "tr": 1.8439836517161714, "fi": 1.1978217804229858, "ru": 2.4036591493091253, "el": 8.524423567930231, "hi": 1.5823764442411803, "ta": 8.758152819304643, "th": 11.66394108681175, "zh": 16.788418424715406, "my": 3.4023443994328035}, "t": 12.5} |
| {"event": "probe", "step": 32, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3538.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203298, "cps": 27.640625, "mean_p": 0.488211, "boundary_entropy": 0.686847, "bpb": 6.226881}, "de": {"chunks": 3105.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15122, "cps": 24.257812, "mean_p": 0.483319, "boundary_entropy": 0.687363, "bpb": 6.333103}, "vi": {"chunks": 6114.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.331652, "cps": 47.765625, "mean_p": 0.489058, "boundary_entropy": 0.688304, "bpb": 6.743838}, "tr": {"chunks": 4962.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.274098, "cps": 38.765625, "mean_p": 0.487113, "boundary_entropy": 0.68706, "bpb": 6.642124}, "fi": {"chunks": 3334.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177237, "cps": 26.046875, "mean_p": 0.483075, "boundary_entropy": 0.686941, "bpb": 6.410096}, "ru": {"chunks": 2372.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.125172, "cps": 18.53125, "mean_p": 0.481987, "boundary_entropy": 0.689459, "bpb": 5.795755}, "el": {"chunks": 1340.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.065293, "cps": 10.46875, "mean_p": 0.473723, "boundary_entropy": 0.688751, "bpb": 6.222343}, "hi": {"chunks": 4964.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.295037, "cps": 38.78125, "mean_p": 0.488146, "boundary_entropy": 0.690516, "bpb": 5.769716}, "ta": {"chunks": 4418.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.23632, "cps": 34.515625, "mean_p": 0.482218, "boundary_entropy": 0.6903, "bpb": 6.128576}, "th": {"chunks": 4788.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.297669, "cps": 37.40625, "mean_p": 0.489189, "boundary_entropy": 0.690753, "bpb": 5.553158}, "zh": {"chunks": 4838.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.819167, "cps": 37.796875, "mean_p": 0.491141, "boundary_entropy": 0.686036, "bpb": 7.308051}, "my": {"chunks": 4127.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.217807, "cps": 32.242188, "mean_p": 0.489036, "boundary_entropy": 0.690968, "bpb": 5.725026}}, "bpb": {"en": 6.22688, "de": 6.3331, "vi": 6.74384, "tr": 6.64212, "fi": 6.4101, "ru": 5.79575, "el": 6.22234, "hi": 5.76972, "ta": 6.12858, "th": 5.55316, "zh": 7.30805, "my": 5.72503}, "gini_cpc": 0.312986, "gini_cps": 0.174673, "bpic": 8.012129, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.7438330807331149, "vi": 1.6313553970123744, "tr": 1.348256429893799, "fi": 0.8718063012275361, "ru": 0.6157036979585384, "el": 0.3211665043895794, "hi": 1.4512525356204804, "ta": 1.1624292611256444, "th": 1.464196512556158, "zh": 4.0293845141626194, "my": 1.071364827221041}, "t": 14.9} |
| {"event": "train", "step": 50, "phase": 1, "objective": "global", "lm_loss": 4.18199, "bpb": 6.03334, "chunk_loss": 0.028757, "F": 0.14752, "G": 0.48501, "chunks": 151.06, "bytes_per_chunk": 6.779, "boundary_entropy": 0.69151, "router_noise": 0.0, "grad_norm": 1.6427, "lr": 5.1e-05, "router_trainable": true, "sec_per_step": 0.238, "alpha": 0.03, "t": 16.0} |
| {"event": "probe", "step": 64, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 5828.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.334885, "cps": 45.53125, "mean_p": 0.496736, "boundary_entropy": 0.68737, "bpb": 5.70089}, "de": {"chunks": 5785.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.281742, "cps": 45.195312, "mean_p": 0.493328, "boundary_entropy": 0.688122, "bpb": 5.815395}, "vi": {"chunks": 4974.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.269813, "cps": 38.859375, "mean_p": 0.485702, "boundary_entropy": 0.688067, "bpb": 6.194321}, "tr": {"chunks": 4536.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.250566, "cps": 35.4375, "mean_p": 0.487411, "boundary_entropy": 0.687251, "bpb": 6.16141}, "fi": {"chunks": 4537.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.241189, "cps": 35.445312, "mean_p": 0.489664, "boundary_entropy": 0.687558, "bpb": 5.889794}, "ru": {"chunks": 10036.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.529604, "cps": 78.40625, "mean_p": 0.494853, "boundary_entropy": 0.690227, "bpb": 5.366093}, "el": {"chunks": 2396.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.116747, "cps": 18.71875, "mean_p": 0.484177, "boundary_entropy": 0.689909, "bpb": 5.639584}, "hi": {"chunks": 8896.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.528737, "cps": 69.5, "mean_p": 0.491185, "boundary_entropy": 0.690635, "bpb": 5.211551}, "ta": {"chunks": 9807.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.524579, "cps": 76.617188, "mean_p": 0.486993, "boundary_entropy": 0.690597, "bpb": 5.235738}, "th": {"chunks": 2736.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170096, "cps": 21.375, "mean_p": 0.486813, "boundary_entropy": 0.690644, "bpb": 5.048564}, "zh": {"chunks": 2609.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.441754, "cps": 20.382812, "mean_p": 0.48503, "boundary_entropy": 0.68618, "bpb": 6.981163}, "my": {"chunks": 21406.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.129723, "cps": 167.234375, "mean_p": 0.498039, "boundary_entropy": 0.69133, "bpb": 4.952445}}, "bpb": {"en": 5.70089, "de": 5.81539, "vi": 6.19432, "tr": 6.16141, "fi": 5.88979, "ru": 5.36609, "el": 5.63958, "hi": 5.21155, "ta": 5.23574, "th": 5.04856, "zh": 6.98116, "my": 4.95245}, "gini_cpc": 0.316949, "gini_cps": 0.351038, "bpic": 4.593649, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8413089968402472, "vi": 0.8056885951661211, "tr": 0.7482161387873123, "fi": 0.7202138568946521, "ru": 1.581452002252795, "el": 0.348618592910022, "hi": 1.5788623859198971, "ta": 1.5664454763944495, "th": 0.5079250182786241, "zh": 1.3191227596564898, "my": 3.3734689883478377}, "t": 17.9} |
| {"event": "train", "step": 100, "phase": 1, "objective": "global", "lm_loss": 3.33768, "bpb": 4.81526, "chunk_loss": 0.03274, "F": 0.20618, "G": 0.48409, "chunks": 211.12, "bytes_per_chunk": 4.85, "boundary_entropy": 0.69112, "router_noise": 0.0, "grad_norm": 1.5693, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.078, "alpha": 0.03, "t": 19.9} |
| {"event": "probe", "step": 128, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2593.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.148997, "cps": 20.257812, "mean_p": 0.474744, "boundary_entropy": 0.685099, "bpb": 4.484147}, "de": {"chunks": 3097.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15083, "cps": 24.195312, "mean_p": 0.473208, "boundary_entropy": 0.685661, "bpb": 4.652212}, "vi": {"chunks": 4902.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.265907, "cps": 38.296875, "mean_p": 0.475084, "boundary_entropy": 0.684798, "bpb": 4.906009}, "tr": {"chunks": 3308.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.182732, "cps": 25.84375, "mean_p": 0.47162, "boundary_entropy": 0.684639, "bpb": 4.971756}, "fi": {"chunks": 2422.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.128754, "cps": 18.921875, "mean_p": 0.473801, "boundary_entropy": 0.685606, "bpb": 4.735386}, "ru": {"chunks": 5265.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.277836, "cps": 41.132812, "mean_p": 0.478288, "boundary_entropy": 0.687726, "bpb": 3.664492}, "el": {"chunks": 5591.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.272426, "cps": 43.679688, "mean_p": 0.484079, "boundary_entropy": 0.689565, "bpb": 4.433533}, "hi": {"chunks": 3548.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.210877, "cps": 27.71875, "mean_p": 0.480272, "boundary_entropy": 0.689831, "bpb": 3.452695}, "ta": {"chunks": 6401.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.342391, "cps": 50.007812, "mean_p": 0.483344, "boundary_entropy": 0.690068, "bpb": 3.586765}, "th": {"chunks": 7940.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.493628, "cps": 62.03125, "mean_p": 0.479011, "boundary_entropy": 0.68905, "bpb": 3.320824}, "zh": {"chunks": 2275.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.385201, "cps": 17.773438, "mean_p": 0.482624, "boundary_entropy": 0.685918, "bpb": 6.305572}, "my": {"chunks": 14965.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.789793, "cps": 116.914062, "mean_p": 0.491769, "boundary_entropy": 0.690832, "bpb": 3.929612}}, "bpb": {"en": 4.48415, "de": 4.65221, "vi": 4.90601, "tr": 4.97176, "fi": 4.73539, "ru": 3.66449, "el": 4.43353, "hi": 3.4527, "ta": 3.58676, "th": 3.32082, "zh": 6.30557, "my": 3.92961}, "gini_cpc": 0.297297, "gini_cps": 0.31511, "bpic": 6.159517, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0123027149827706, "vi": 1.7846447075229457, "tr": 1.2264124563476695, "fi": 0.8641395030186715, "ru": 1.8647077019823008, "el": 1.8283959495821374, "hi": 1.4153053282669252, "ta": 2.2979679176155443, "th": 3.312996984660526, "zh": 2.5852917588302358, "my": 5.300721030728466}, "t": 22.3} |
| {"event": "train", "step": 150, "phase": 1, "objective": "global", "lm_loss": 2.91824, "bpb": 4.21013, "chunk_loss": 0.030229, "F": 0.17062, "G": 0.46313, "chunks": 174.72, "bytes_per_chunk": 5.861, "boundary_entropy": 0.68609, "router_noise": 0.0, "grad_norm": 1.5833, "lr": 0.000151, "router_trainable": true, "sec_per_step": 0.072, "alpha": 0.03, "t": 23.5} |
| {"event": "probe", "step": 200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3535.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203126, "cps": 27.617188, "mean_p": 0.370816, "boundary_entropy": 0.546965, "bpb": 3.881286}, "de": {"chunks": 3959.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.192812, "cps": 30.929688, "mean_p": 0.349554, "boundary_entropy": 0.544277, "bpb": 4.09761}, "vi": {"chunks": 7371.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.399837, "cps": 57.585938, "mean_p": 0.44078, "boundary_entropy": 0.575002, "bpb": 3.586596}, "tr": {"chunks": 4867.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.26885, "cps": 38.023438, "mean_p": 0.378113, "boundary_entropy": 0.572162, "bpb": 4.151264}, "fi": {"chunks": 3736.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.198607, "cps": 29.1875, "mean_p": 0.351272, "boundary_entropy": 0.561957, "bpb": 4.14796}, "ru": {"chunks": 7942.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.419103, "cps": 62.046875, "mean_p": 0.431487, "boundary_entropy": 0.661571, "bpb": 2.804064}, "el": {"chunks": 8190.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.399064, "cps": 63.984375, "mean_p": 0.446529, "boundary_entropy": 0.675686, "bpb": 2.976351}, "hi": {"chunks": 7103.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.422169, "cps": 55.492188, "mean_p": 0.478162, "boundary_entropy": 0.68817, "bpb": 2.428864}, "ta": {"chunks": 16073.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.859749, "cps": 125.570312, "mean_p": 0.448152, "boundary_entropy": 0.676888, "bpb": 2.374042}, "th": {"chunks": 5665.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.352191, "cps": 44.257812, "mean_p": 0.450006, "boundary_entropy": 0.680478, "bpb": 2.322478}, "zh": {"chunks": 1950.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.330173, "cps": 15.234375, "mean_p": 0.44153, "boundary_entropy": 0.674133, "bpb": 5.259916}, "my": {"chunks": 10732.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.566392, "cps": 83.84375, "mean_p": 0.379609, "boundary_entropy": 0.640761, "bpb": 2.462235}}, "bpb": {"en": 3.88129, "de": 4.09761, "vi": 3.5866, "tr": 4.15126, "fi": 4.14796, "ru": 2.80406, "el": 2.97635, "hi": 2.42886, "ta": 2.37404, "th": 2.32248, "zh": 5.25992, "my": 2.46223}, "gini_cpc": 0.238266, "gini_cps": 0.287803, "bpic": 4.730853, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9492220030666473, "vi": 1.968420916949303, "tr": 1.323565678432412, "fi": 0.977754191042182, "ru": 2.0632667072210427, "el": 1.9646163710070759, "hi": 2.0783631988331503, "ta": 4.232589763542128, "th": 1.7338580973866295, "zh": 1.625458443478715, "my": 2.788380198736421}, "t": 26.8} |
| {"event": "eval", "step": 200, "phase": 1, "bpb": {"en": 3.8845, "de": 4.0977, "vi": 3.59039, "tr": 4.16176, "fi": 4.1488, "ru": 2.78968, "el": 2.96955, "hi": 2.42447, "ta": 2.39063, "th": 2.35422, "zh": 5.26776, "my": 2.47911}, "bytes_per_chunk": {"en": 4.963, "de": 5.299, "vi": 3.219, "tr": 4.068, "fi": 5.179, "ru": 4.217, "el": 4.747, "hi": 6.295, "ta": 3.421, "th": 8.545, "zh": 8.538, "my": 4.924}, "bpb_mean": 3.37988, "t": 28.2} |
| {"event": "train", "step": 200, "phase": 1, "objective": "global", "lm_loss": 2.53313, "bpb": 3.65453, "chunk_loss": 0.032256, "F": 0.21005, "G": 0.39482, "chunks": 215.09, "bytes_per_chunk": 4.761, "boundary_entropy": 0.59378, "router_noise": 0.0, "grad_norm": 1.1927, "lr": 0.000201, "router_trainable": true, "sec_per_step": 0.095, "alpha": 0.03, "t": 28.3} |
| {"event": "train", "step": 250, "phase": 1, "objective": "global", "lm_loss": 2.18153, "bpb": 3.14728, "chunk_loss": 0.03431, "F": 0.24274, "G": 0.35632, "chunks": 248.56, "bytes_per_chunk": 4.12, "boundary_entropy": 0.51744, "router_noise": 0.0, "grad_norm": 1.0691, "lr": 0.000251, "router_trainable": true, "sec_per_step": 0.047, "alpha": 0.03, "t": 30.6} |
| {"event": "probe", "step": 256, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3026.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.173878, "cps": 23.640625, "mean_p": 0.287707, "boundary_entropy": 0.393076, "bpb": 3.577964}, "de": {"chunks": 3528.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.171821, "cps": 27.5625, "mean_p": 0.272928, "boundary_entropy": 0.401311, "bpb": 3.828468}, "vi": {"chunks": 7112.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.385788, "cps": 55.5625, "mean_p": 0.359231, "boundary_entropy": 0.464641, "bpb": 3.176966}, "tr": {"chunks": 4256.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.235099, "cps": 33.25, "mean_p": 0.308561, "boundary_entropy": 0.453207, "bpb": 3.872989}, "fi": {"chunks": 3706.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.197012, "cps": 28.953125, "mean_p": 0.289049, "boundary_entropy": 0.445824, "bpb": 3.92625}, "ru": {"chunks": 8982.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.473984, "cps": 70.171875, "mean_p": 0.308092, "boundary_entropy": 0.500317, "bpb": 2.645008}, "el": {"chunks": 7873.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.383618, "cps": 61.507812, "mean_p": 0.381059, "boundary_entropy": 0.61821, "bpb": 2.713261}, "hi": {"chunks": 11405.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.67786, "cps": 89.101562, "mean_p": 0.454382, "boundary_entropy": 0.672651, "bpb": 2.134484}, "ta": {"chunks": 13636.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.729393, "cps": 106.53125, "mean_p": 0.466179, "boundary_entropy": 0.672749, "bpb": 1.929934}, "th": {"chunks": 11182.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.695182, "cps": 87.359375, "mean_p": 0.437349, "boundary_entropy": 0.671981, "bpb": 2.082228}, "zh": {"chunks": 3078.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.521165, "cps": 24.046875, "mean_p": 0.417987, "boundary_entropy": 0.651534, "bpb": 4.974382}, "my": {"chunks": 6368.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.336078, "cps": 49.75, "mean_p": 0.340666, "boundary_entropy": 0.61183, "bpb": 2.218416}}, "bpb": {"en": 3.57796, "de": 3.82847, "vi": 3.17697, "tr": 3.87299, "fi": 3.92625, "ru": 2.64501, "el": 2.71326, "hi": 2.13448, "ta": 1.92993, "th": 2.08223, "zh": 4.97438, "my": 2.21842}, "gini_cpc": 0.268312, "gini_cps": 0.280451, "bpic": 4.560569, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9881693193639989, "vi": 2.2187266634650493, "tr": 1.3520920659665068, "fi": 1.1330490945113993, "ru": 2.7259572011781796, "el": 2.2062493907259726, "hi": 3.8984809216606156, "ta": 4.194852740101964, "th": 3.9980996938562297, "zh": 2.9973010744000126, "my": 1.9328354179337446}, "t": 31.9} |
| {"event": "train", "step": 300, "phase": 1, "objective": "global", "lm_loss": 2.18455, "bpb": 3.15164, "chunk_loss": 0.033434, "F": 0.24164, "G": 0.33171, "chunks": 247.44, "bytes_per_chunk": 4.138, "boundary_entropy": 0.47146, "router_noise": 0.0, "grad_norm": 0.8607, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.077, "alpha": 0.03, "t": 34.5} |
| {"event": "train", "step": 350, "phase": 1, "objective": "global", "lm_loss": 2.43533, "bpb": 3.51343, "chunk_loss": 0.033402, "F": 0.22913, "G": 0.344, "chunks": 234.62, "bytes_per_chunk": 4.364, "boundary_entropy": 0.42635, "router_noise": 0.0, "grad_norm": 1.0827, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.053, "alpha": 0.03, "t": 37.1} |
| {"event": "probe", "step": 400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 3018.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.173418, "cps": 23.578125, "mean_p": 0.286122, "boundary_entropy": 0.370308, "bpb": 3.198769}, "de": {"chunks": 3549.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.172844, "cps": 27.726562, "mean_p": 0.27444, "boundary_entropy": 0.376046, "bpb": 3.452982}, "vi": {"chunks": 4623.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.250773, "cps": 36.117188, "mean_p": 0.299258, "boundary_entropy": 0.382645, "bpb": 2.758102}, "tr": {"chunks": 4117.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.227421, "cps": 32.164062, "mean_p": 0.287267, "boundary_entropy": 0.388148, "bpb": 3.436574}, "fi": {"chunks": 3553.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.188879, "cps": 27.757812, "mean_p": 0.278758, "boundary_entropy": 0.393611, "bpb": 3.623892}, "ru": {"chunks": 8172.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.43124, "cps": 63.84375, "mean_p": 0.279615, "boundary_entropy": 0.288105, "bpb": 2.544146}, "el": {"chunks": 8037.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.391609, "cps": 62.789062, "mean_p": 0.273372, "boundary_entropy": 0.330201, "bpb": 2.563421}, "hi": {"chunks": 5875.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.349183, "cps": 45.898438, "mean_p": 0.357617, "boundary_entropy": 0.556004, "bpb": 1.948676}, "ta": {"chunks": 9159.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.489917, "cps": 71.554688, "mean_p": 0.329732, "boundary_entropy": 0.510928, "bpb": 1.731031}, "th": {"chunks": 5814.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.361455, "cps": 45.421875, "mean_p": 0.42177, "boundary_entropy": 0.652376, "bpb": 1.942214}, "zh": {"chunks": 3670.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.621402, "cps": 28.671875, "mean_p": 0.31093, "boundary_entropy": 0.486138, "bpb": 4.16264}, "my": {"chunks": 3369.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.177802, "cps": 26.320312, "mean_p": 0.31998, "boundary_entropy": 0.573105, "bpb": 1.975451}}, "bpb": {"en": 3.19877, "de": 3.45298, "vi": 2.7581, "tr": 3.43657, "fi": 3.62389, "ru": 2.54415, "el": 2.56342, "hi": 1.94868, "ta": 1.73103, "th": 1.94221, "zh": 4.16264, "my": 1.97545}, "gini_cpc": 0.240507, "gini_cps": 0.21277, "bpic": 6.096019, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9966862730498613, "vi": 1.4460577462806563, "tr": 1.3114000634100715, "fi": 1.089151286528873, "ru": 2.486703630459984, "el": 2.2581771467380487, "hi": 2.0135280442161294, "ta": 2.825058687723884, "th": 2.084293369407429, "zh": 3.583253274124303, "my": 1.0252800801272164}, "t": 40.4} |
| {"event": "eval", "step": 400, "phase": 1, "bpb": {"en": 3.20825, "de": 3.45008, "vi": 2.76436, "tr": 3.4475, "fi": 3.61789, "ru": 2.53066, "el": 2.55851, "hi": 1.94645, "ta": 1.74583, "th": 1.9721, "zh": 4.17271, "my": 1.99198}, "bytes_per_chunk": {"en": 5.678, "de": 5.93, "vi": 5.367, "tr": 4.761, "fi": 5.522, "ru": 4.172, "el": 4.359, "hi": 6.716, "ta": 5.374, "th": 7.026, "zh": 3.759, "my": 15.64}, "bpb_mean": 2.78386, "t": 41.8} |
| {"event": "train", "step": 400, "phase": 1, "objective": "global", "lm_loss": 2.01275, "bpb": 2.90378, "chunk_loss": 0.030586, "F": 0.18677, "G": 0.30133, "chunks": 191.25, "bytes_per_chunk": 5.354, "boundary_entropy": 0.40064, "router_noise": 0.0, "grad_norm": 1.47, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.094, "alpha": 0.03, "t": 41.8} |
| {"event": "train", "step": 450, "phase": 1, "objective": "global", "lm_loss": 2.06821, "bpb": 2.9838, "chunk_loss": 0.029994, "F": 0.16599, "G": 0.27866, "chunks": 169.97, "bytes_per_chunk": 6.025, "boundary_entropy": 0.38699, "router_noise": 0.0, "grad_norm": 1.1043, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.044, "alpha": 0.03, "t": 44.0} |
| {"event": "train", "step": 500, "phase": 1, "objective": "global", "lm_loss": 2.05837, "bpb": 2.9696, "chunk_loss": 0.030029, "F": 0.16873, "G": 0.27979, "chunks": 172.78, "bytes_per_chunk": 5.927, "boundary_entropy": 0.38545, "router_noise": 0.0, "grad_norm": 0.9509, "lr": 0.0002992, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 46.1} |
| {"event": "train", "step": 550, "phase": 1, "objective": "global", "lm_loss": 1.95397, "bpb": 2.81898, "chunk_loss": 0.029557, "F": 0.15497, "G": 0.29503, "chunks": 158.69, "bytes_per_chunk": 6.453, "boundary_entropy": 0.42532, "router_noise": 0.0, "grad_norm": 0.8251, "lr": 0.0002987, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 48.3} |
| {"event": "probe", "step": 600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.28234, "boundary_entropy": 0.351455, "bpb": 2.91654}, "de": {"chunks": 3329.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162129, "cps": 26.007812, "mean_p": 0.293582, "boundary_entropy": 0.398535, "bpb": 3.184404}, "vi": {"chunks": 4387.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237971, "cps": 34.273438, "mean_p": 0.291473, "boundary_entropy": 0.364961, "bpb": 2.51033}, "tr": {"chunks": 3875.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.214053, "cps": 30.273438, "mean_p": 0.306839, "boundary_entropy": 0.386346, "bpb": 3.143484}, "fi": {"chunks": 3692.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.196268, "cps": 28.84375, "mean_p": 0.343034, "boundary_entropy": 0.460752, "bpb": 3.35313}, "ru": {"chunks": 6564.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.346385, "cps": 51.28125, "mean_p": 0.244735, "boundary_entropy": 0.267941, "bpb": 2.163121}, "el": {"chunks": 7367.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.358963, "cps": 57.554688, "mean_p": 0.229123, "boundary_entropy": 0.225859, "bpb": 2.394969}, "hi": {"chunks": 6870.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.408321, "cps": 53.671875, "mean_p": 0.315837, "boundary_entropy": 0.45704, "bpb": 1.7393}, "ta": {"chunks": 8838.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.472747, "cps": 69.046875, "mean_p": 0.24062, "boundary_entropy": 0.348699, "bpb": 1.452758}, "th": {"chunks": 6090.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.378614, "cps": 47.578125, "mean_p": 0.318529, "boundary_entropy": 0.522765, "bpb": 1.777363}, "zh": {"chunks": 3186.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.539451, "cps": 24.890625, "mean_p": 0.333651, "boundary_entropy": 0.548891, "bpb": 3.868711}, "my": {"chunks": 9540.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.503483, "cps": 74.53125, "mean_p": 0.411596, "boundary_entropy": 0.608958, "bpb": 1.808705}}, "bpb": {"en": 2.91654, "de": 3.1844, "vi": 2.51033, "tr": 3.14348, "fi": 3.35313, "ru": 2.16312, "el": 2.39497, "hi": 1.7393, "ta": 1.45276, "th": 1.77736, "zh": 3.86871, "my": 1.80871}, "gini_cpc": 0.220186, "gini_cps": 0.222542, "bpic": 5.759969, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9759721310169421, "vi": 1.4325194291422074, "tr": 1.2885378610916858, "fi": 1.1814785394321246, "ru": 2.0851408018151134, "el": 2.160856133721987, "hi": 2.4579763091173454, "ta": 2.8458011874610816, "th": 2.2791465737783763, "zh": 3.2473444508179163, "my": 3.030826160285606}, "t": 51.3} |
| {"event": "eval", "step": 600, "phase": 1, "bpb": {"en": 2.92492, "de": 3.18489, "vi": 2.51658, "tr": 3.15343, "fi": 3.34732, "ru": 2.14884, "el": 2.38864, "hi": 1.73501, "ta": 1.46398, "th": 1.80605, "zh": 3.87933, "my": 1.82282}, "bytes_per_chunk": {"en": 5.728, "de": 6.276, "vi": 5.556, "tr": 5.082, "fi": 5.52, "ru": 5.443, "el": 5.153, "hi": 5.544, "ta": 5.435, "th": 7.33, "zh": 5.148, "my": 6.028}, "bpb_mean": 2.53098, "t": 52.6} |
| {"event": "train", "step": 600, "phase": 1, "objective": "global", "lm_loss": 1.86543, "bpb": 2.69124, "chunk_loss": 0.030356, "F": 0.17999, "G": 0.29505, "chunks": 184.31, "bytes_per_chunk": 5.556, "boundary_entropy": 0.39975, "router_noise": 0.0, "grad_norm": 0.6314, "lr": 0.0002982, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 52.7} |
| {"event": "train", "step": 650, "phase": 1, "objective": "global", "lm_loss": 1.88767, "bpb": 2.72333, "chunk_loss": 0.030172, "F": 0.17435, "G": 0.28232, "chunks": 178.53, "bytes_per_chunk": 5.736, "boundary_entropy": 0.36279, "router_noise": 0.0, "grad_norm": 0.7842, "lr": 0.0002975, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 54.8} |
| {"event": "train", "step": 700, "phase": 1, "objective": "global", "lm_loss": 1.841, "bpb": 2.65601, "chunk_loss": 0.030026, "F": 0.16702, "G": 0.27566, "chunks": 171.03, "bytes_per_chunk": 5.987, "boundary_entropy": 0.36023, "router_noise": 0.0, "grad_norm": 0.7663, "lr": 0.0002967, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 56.9} |
| {"event": "train", "step": 750, "phase": 1, "objective": "global", "lm_loss": 1.91363, "bpb": 2.76079, "chunk_loss": 0.030313, "F": 0.17761, "G": 0.28738, "chunks": 181.88, "bytes_per_chunk": 5.63, "boundary_entropy": 0.37604, "router_noise": 0.0, "grad_norm": 0.7933, "lr": 0.0002959, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 59.1} |
| {"event": "probe", "step": 800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2883.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165661, "cps": 22.523438, "mean_p": 0.272947, "boundary_entropy": 0.330031, "bpb": 2.733365}, "de": {"chunks": 3304.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160912, "cps": 25.8125, "mean_p": 0.28569, "boundary_entropy": 0.379457, "bpb": 3.037963}, "vi": {"chunks": 4461.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.241985, "cps": 34.851562, "mean_p": 0.279332, "boundary_entropy": 0.342089, "bpb": 2.39705}, "tr": {"chunks": 3877.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.214163, "cps": 30.289062, "mean_p": 0.292634, "boundary_entropy": 0.361915, "bpb": 2.988565}, "fi": {"chunks": 3236.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.172027, "cps": 25.28125, "mean_p": 0.3132, "boundary_entropy": 0.420025, "bpb": 3.219245}, "ru": {"chunks": 6446.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.340158, "cps": 50.359375, "mean_p": 0.249922, "boundary_entropy": 0.28428, "bpb": 2.004709}, "el": {"chunks": 7278.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.354627, "cps": 56.859375, "mean_p": 0.238418, "boundary_entropy": 0.302581, "bpb": 2.131077}, "hi": {"chunks": 8044.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.478098, "cps": 62.84375, "mean_p": 0.330481, "boundary_entropy": 0.47184, "bpb": 1.619798}, "ta": {"chunks": 10998.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.588286, "cps": 85.921875, "mean_p": 0.267573, "boundary_entropy": 0.345917, "bpb": 1.350356}, "th": {"chunks": 9183.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.570905, "cps": 71.742188, "mean_p": 0.394787, "boundary_entropy": 0.552216, "bpb": 1.589597}, "zh": {"chunks": 3465.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.586692, "cps": 27.070312, "mean_p": 0.338493, "boundary_entropy": 0.546449, "bpb": 3.708105}, "my": {"chunks": 20657.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.090194, "cps": 161.382812, "mean_p": 0.423804, "boundary_entropy": 0.596904, "bpb": 1.732656}}, "bpb": {"en": 2.73337, "de": 3.03796, "vi": 2.39705, "tr": 2.98857, "fi": 3.21924, "ru": 2.00471, "el": 2.13108, "hi": 1.6198, "ta": 1.35036, "th": 1.5896, "zh": 3.70811, "my": 1.73266}, "gini_cpc": 0.326283, "gini_cps": 0.339576, "bpic": 4.577977, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9713306865276591, "vi": 1.460725325954707, "tr": 1.2927803056426508, "fi": 1.0384273244157018, "ru": 2.053338568278975, "el": 2.1406747375080255, "hi": 2.886000930803568, "ta": 3.551139422778548, "th": 3.4462199870161667, "zh": 3.541516537442478, "my": 6.580870598913902}, "t": 62.1} |
| {"event": "eval", "step": 800, "phase": 1, "bpb": {"en": 2.74289, "de": 3.03511, "vi": 2.40206, "tr": 2.99838, "fi": 3.21447, "ru": 1.98834, "el": 2.11837, "hi": 1.61472, "ta": 1.36134, "th": 1.62145, "zh": 3.72094, "my": 1.74059}, "bytes_per_chunk": {"en": 5.741, "de": 6.277, "vi": 5.51, "tr": 5.134, "fi": 6.034, "ru": 5.552, "el": 5.39, "hi": 5.378, "ta": 4.591, "th": 4.807, "zh": 4.881, "my": 2.888}, "bpb_mean": 2.37989, "t": 63.5} |
| {"event": "train", "step": 800, "phase": 1, "objective": "global", "lm_loss": 1.64873, "bpb": 2.37861, "chunk_loss": 0.030857, "F": 0.18869, "G": 0.2803, "chunks": 193.22, "bytes_per_chunk": 5.3, "boundary_entropy": 0.34331, "router_noise": 0.0, "grad_norm": 0.8387, "lr": 0.0002949, "router_trainable": true, "sec_per_step": 0.089, "alpha": 0.03, "t": 63.5} |
| {"event": "train", "step": 850, "phase": 1, "objective": "global", "lm_loss": 1.71082, "bpb": 2.46819, "chunk_loss": 0.030042, "F": 0.17056, "G": 0.28398, "chunks": 174.66, "bytes_per_chunk": 5.863, "boundary_entropy": 0.37829, "router_noise": 0.0, "grad_norm": 0.6169, "lr": 0.0002938, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 65.6} |
| {"event": "train", "step": 900, "phase": 1, "objective": "global", "lm_loss": 1.62554, "bpb": 2.34515, "chunk_loss": 0.030295, "F": 0.18027, "G": 0.2621, "chunks": 184.59, "bytes_per_chunk": 5.547, "boundary_entropy": 0.31443, "router_noise": 0.0, "grad_norm": 0.6774, "lr": 0.0002927, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 67.8} |
| {"event": "train", "step": 950, "phase": 1, "objective": "global", "lm_loss": 1.7236, "bpb": 2.48663, "chunk_loss": 0.030118, "F": 0.17505, "G": 0.27485, "chunks": 179.25, "bytes_per_chunk": 5.713, "boundary_entropy": 0.352, "router_noise": 0.0, "grad_norm": 0.6881, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 69.9} |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2876.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165259, "cps": 22.46875, "mean_p": 0.253189, "boundary_entropy": 0.290891, "bpb": 2.599278}, "de": {"chunks": 3319.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161642, "cps": 25.929688, "mean_p": 0.277215, "boundary_entropy": 0.359997, "bpb": 2.914663}, "vi": {"chunks": 4277.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.232004, "cps": 33.414062, "mean_p": 0.259542, "boundary_entropy": 0.298869, "bpb": 2.283504}, "tr": {"chunks": 3873.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213942, "cps": 30.257812, "mean_p": 0.271347, "boundary_entropy": 0.325631, "bpb": 2.884085}, "fi": {"chunks": 3125.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.166126, "cps": 24.414062, "mean_p": 0.279708, "boundary_entropy": 0.372081, "bpb": 3.167717}, "ru": {"chunks": 6200.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327177, "cps": 48.4375, "mean_p": 0.249231, "boundary_entropy": 0.302961, "bpb": 1.904682}, "el": {"chunks": 7063.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.34415, "cps": 55.179688, "mean_p": 0.240226, "boundary_entropy": 0.309929, "bpb": 2.008806}, "hi": {"chunks": 7114.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.422823, "cps": 55.578125, "mean_p": 0.305425, "boundary_entropy": 0.454829, "bpb": 1.551038}, "ta": {"chunks": 8931.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.477721, "cps": 69.773438, "mean_p": 0.241291, "boundary_entropy": 0.341298, "bpb": 1.292252}, "th": {"chunks": 8734.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.54299, "cps": 68.234375, "mean_p": 0.403056, "boundary_entropy": 0.578163, "bpb": 1.504507}, "zh": {"chunks": 3336.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.564849, "cps": 26.0625, "mean_p": 0.344937, "boundary_entropy": 0.554086, "bpb": 3.639392}, "my": {"chunks": 8628.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.455351, "cps": 67.40625, "mean_p": 0.373537, "boundary_entropy": 0.561898, "bpb": 1.417532}}, "bpb": {"en": 2.59928, "de": 2.91466, "vi": 2.2835, "tr": 2.88409, "fi": 3.16772, "ru": 1.90468, "el": 2.00881, "hi": 1.55104, "ta": 1.29225, "th": 1.50451, "zh": 3.63939, "my": 1.41753}, "gini_cpc": 0.241622, "gini_cps": 0.228022, "bpic": 5.687667, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9781153707112952, "vi": 1.403884395536247, "tr": 1.294589809340878, "fi": 1.0052483680538176, "ru": 1.9797835604273013, "el": 2.082493236497035, "hi": 2.558550694686983, "ta": 2.8907455006637925, "th": 3.285695862081787, "zh": 3.417966783267011, "my": 2.7553831538948703}, "t": 73.0} |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.61026, "de": 2.91765, "vi": 2.29015, "tr": 2.89237, "fi": 3.16855, "ru": 1.88782, "el": 1.99939, "hi": 1.54578, "ta": 1.30095, "th": 1.52852, "zh": 3.64868, "my": 1.42667}, "bytes_per_chunk": {"en": 5.747, "de": 6.246, "vi": 5.675, "tr": 5.073, "fi": 6.151, "ru": 5.743, "el": 5.358, "hi": 6.182, "ta": 5.827, "th": 5.214, "zh": 5.61, "my": 6.404}, "bpb_mean": 2.26807, "t": 74.3} |
| {"event": "train", "step": 1000, "phase": 1, "objective": "global", "lm_loss": 1.74318, "bpb": 2.51488, "chunk_loss": 0.030366, "F": 0.17917, "G": 0.28501, "chunks": 183.47, "bytes_per_chunk": 5.581, "boundary_entropy": 0.36911, "router_noise": 0.0, "grad_norm": 0.7348, "lr": 0.0002901, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 74.3} |
| {"event": "train", "step": 1050, "phase": 1, "objective": "global", "lm_loss": 1.51811, "bpb": 2.19016, "chunk_loss": 0.030076, "F": 0.17346, "G": 0.27314, "chunks": 177.62, "bytes_per_chunk": 5.765, "boundary_entropy": 0.33774, "router_noise": 0.0, "grad_norm": 0.7461, "lr": 0.0002886, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 76.4} |
| {"event": "train", "step": 1100, "phase": 1, "objective": "global", "lm_loss": 1.77166, "bpb": 2.55597, "chunk_loss": 0.030167, "F": 0.17371, "G": 0.27894, "chunks": 177.88, "bytes_per_chunk": 5.757, "boundary_entropy": 0.35705, "router_noise": 0.0, "grad_norm": 0.8629, "lr": 0.0002871, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 78.6} |
| {"event": "train", "step": 1150, "phase": 1, "objective": "global", "lm_loss": 1.66863, "bpb": 2.40732, "chunk_loss": 0.03003, "F": 0.17084, "G": 0.26717, "chunks": 174.94, "bytes_per_chunk": 5.854, "boundary_entropy": 0.34045, "router_noise": 0.0, "grad_norm": 0.7352, "lr": 0.0002855, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 80.7} |
| {"event": "probe", "step": 1200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2896.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166408, "cps": 22.625, "mean_p": 0.252982, "boundary_entropy": 0.284686, "bpb": 2.502214}, "de": {"chunks": 3346.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162957, "cps": 26.140625, "mean_p": 0.273014, "boundary_entropy": 0.346652, "bpb": 2.803836}, "vi": {"chunks": 4267.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.231462, "cps": 33.335938, "mean_p": 0.254933, "boundary_entropy": 0.288944, "bpb": 2.224999}, "tr": {"chunks": 3853.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212838, "cps": 30.101562, "mean_p": 0.265264, "boundary_entropy": 0.313095, "bpb": 2.79794}, "fi": {"chunks": 3149.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.167402, "cps": 24.601562, "mean_p": 0.28592, "boundary_entropy": 0.37371, "bpb": 3.040506}, "ru": {"chunks": 6209.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.327652, "cps": 48.507812, "mean_p": 0.250898, "boundary_entropy": 0.296112, "bpb": 1.815665}, "el": {"chunks": 6857.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.334113, "cps": 53.570312, "mean_p": 0.236194, "boundary_entropy": 0.292996, "bpb": 1.885155}, "hi": {"chunks": 7486.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.444933, "cps": 58.484375, "mean_p": 0.314813, "boundary_entropy": 0.462725, "bpb": 1.558628}, "ta": {"chunks": 10451.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.559026, "cps": 81.648438, "mean_p": 0.253316, "boundary_entropy": 0.335655, "bpb": 1.173948}, "th": {"chunks": 8850.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.550202, "cps": 69.140625, "mean_p": 0.378652, "boundary_entropy": 0.55462, "bpb": 1.417018}, "zh": {"chunks": 3290.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.557061, "cps": 25.703125, "mean_p": 0.34521, "boundary_entropy": 0.548534, "bpb": 3.577468}, "my": {"chunks": 7763.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.4097, "cps": 60.648438, "mean_p": 0.377715, "boundary_entropy": 0.575309, "bpb": 1.347641}}, "bpb": {"en": 2.50221, "de": 2.80384, "vi": 2.225, "tr": 2.79794, "fi": 3.04051, "ru": 1.81566, "el": 1.88515, "hi": 1.55863, "ta": 1.17395, "th": 1.41702, "zh": 3.57747, "my": 1.34764}, "gini_cpc": 0.247628, "gini_cps": 0.240412, "bpic": 5.609439, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9792624283830395, "vi": 1.3909293253734565, "tr": 1.2790102362387359, "fi": 1.0059730354961582, "ru": 1.9689650541553085, "el": 2.0077926820010323, "hi": 2.6737470138655155, "ta": 3.359370784134382, "th": 3.3063419558045397, "zh": 3.3475572879345474, "my": 2.4620211123027027}, "t": 83.7} |
| {"event": "eval", "step": 1200, "phase": 1, "bpb": {"en": 2.51364, "de": 2.80485, "vi": 2.23068, "tr": 2.80562, "fi": 3.04752, "ru": 1.80054, "el": 1.87302, "hi": 1.55369, "ta": 1.1829, "th": 1.44004, "zh": 3.58718, "my": 1.3564}, "bytes_per_chunk": {"en": 5.728, "de": 6.216, "vi": 5.717, "tr": 5.081, "fi": 6.125, "ru": 5.738, "el": 5.607, "hi": 5.942, "ta": 4.861, "th": 5.255, "zh": 5.085, "my": 6.725}, "bpb_mean": 2.183, "t": 85.0} |
| {"event": "train", "step": 1200, "phase": 1, "objective": "global", "lm_loss": 1.61741, "bpb": 2.33344, "chunk_loss": 0.03034, "F": 0.18008, "G": 0.26696, "chunks": 184.41, "bytes_per_chunk": 5.553, "boundary_entropy": 0.32518, "router_noise": 0.0, "grad_norm": 0.7476, "lr": 0.0002837, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 85.1} |
| {"event": "train", "step": 1250, "phase": 1, "objective": "global", "lm_loss": 1.64268, "bpb": 2.36988, "chunk_loss": 0.029998, "F": 0.16882, "G": 0.26867, "chunks": 172.88, "bytes_per_chunk": 5.923, "boundary_entropy": 0.33177, "router_noise": 0.0, "grad_norm": 0.8081, "lr": 0.0002819, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 87.2} |
| {"event": "train", "step": 1300, "phase": 1, "objective": "global", "lm_loss": 1.56596, "bpb": 2.2592, "chunk_loss": 0.03027, "F": 0.17645, "G": 0.27529, "chunks": 180.69, "bytes_per_chunk": 5.667, "boundary_entropy": 0.34324, "router_noise": 0.0, "grad_norm": 0.7377, "lr": 0.00028, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 89.3} |
| {"event": "train", "step": 1350, "phase": 1, "objective": "global", "lm_loss": 1.71633, "bpb": 2.47614, "chunk_loss": 0.02997, "F": 0.16693, "G": 0.26747, "chunks": 170.94, "bytes_per_chunk": 5.99, "boundary_entropy": 0.33472, "router_noise": 0.0, "grad_norm": 0.7792, "lr": 0.000278, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 91.4} |
| {"event": "probe", "step": 1400, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.239261, "boundary_entropy": 0.256659, "bpb": 2.444201}, "de": {"chunks": 3314.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161399, "cps": 25.890625, "mean_p": 0.25951, "boundary_entropy": 0.321599, "bpb": 2.734889}, "vi": {"chunks": 4384.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.237809, "cps": 34.25, "mean_p": 0.249616, "boundary_entropy": 0.275469, "bpb": 2.180355}, "tr": {"chunks": 3928.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216981, "cps": 30.6875, "mean_p": 0.254116, "boundary_entropy": 0.284341, "bpb": 2.727427}, "fi": {"chunks": 3035.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161342, "cps": 23.710938, "mean_p": 0.256247, "boundary_entropy": 0.334422, "bpb": 2.984487}, "ru": {"chunks": 6095.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.321636, "cps": 47.617188, "mean_p": 0.253074, "boundary_entropy": 0.29769, "bpb": 1.751177}, "el": {"chunks": 6834.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332992, "cps": 53.390625, "mean_p": 0.248855, "boundary_entropy": 0.320735, "bpb": 1.834214}, "hi": {"chunks": 6469.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.384487, "cps": 50.539062, "mean_p": 0.312559, "boundary_entropy": 0.468804, "bpb": 1.438772}, "ta": {"chunks": 8577.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.458786, "cps": 67.007812, "mean_p": 0.236508, "boundary_entropy": 0.337242, "bpb": 1.158394}, "th": {"chunks": 7275.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.452285, "cps": 56.835938, "mean_p": 0.360916, "boundary_entropy": 0.557165, "bpb": 1.345157}, "zh": {"chunks": 3279.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.555198, "cps": 25.617188, "mean_p": 0.324203, "boundary_entropy": 0.52263, "bpb": 3.561937}, "my": {"chunks": 5847.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.308581, "cps": 45.679688, "mean_p": 0.325287, "boundary_entropy": 0.540183, "bpb": 1.402929}}, "bpb": {"en": 2.4442, "de": 2.73489, "vi": 2.18035, "tr": 2.72743, "fi": 2.98449, "ru": 1.75118, "el": 1.83421, "hi": 1.43877, "ta": 1.15839, "th": 1.34516, "zh": 3.56194, "my": 1.40293}, "gini_cpc": 0.225058, "gini_cps": 0.200946, "bpic": 6.197213, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9715745395584698, "vi": 1.43153981704113, "tr": 1.3061617337724236, "fi": 0.9712316812504057, "ru": 1.936156792666532, "el": 2.0045189110704573, "hi": 2.3145049117438297, "ta": 2.7617602155299497, "th": 2.7226258332081583, "zh": 3.3421351080451807, "my": 1.8575723856593227}, "t": 94.4} |
| {"event": "eval", "step": 1400, "phase": 1, "bpb": {"en": 2.45567, "de": 2.73629, "vi": 2.18625, "tr": 2.73229, "fi": 2.98973, "ru": 1.7295, "el": 1.82293, "hi": 1.43632, "ta": 1.16565, "th": 1.3677, "zh": 3.56707, "my": 1.41268}, "bytes_per_chunk": {"en": 5.725, "de": 6.246, "vi": 5.608, "tr": 5.019, "fi": 6.27, "ru": 5.824, "el": 5.615, "hi": 6.556, "ta": 5.844, "th": 6.345, "zh": 5.64, "my": 8.857}, "bpb_mean": 2.1335, "t": 95.7} |
| {"event": "train", "step": 1400, "phase": 1, "objective": "global", "lm_loss": 1.57058, "bpb": 2.26587, "chunk_loss": 0.030121, "F": 0.17303, "G": 0.2611, "chunks": 177.19, "bytes_per_chunk": 5.779, "boundary_entropy": 0.30545, "router_noise": 0.0, "grad_norm": 0.8368, "lr": 0.0002759, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 95.8} |
| {"event": "train", "step": 1450, "phase": 1, "objective": "global", "lm_loss": 1.72284, "bpb": 2.48554, "chunk_loss": 0.03006, "F": 0.16931, "G": 0.25374, "chunks": 173.38, "bytes_per_chunk": 5.906, "boundary_entropy": 0.30483, "router_noise": 0.0, "grad_norm": 0.7782, "lr": 0.0002738, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 97.9} |
| {"event": "train", "step": 1500, "phase": 1, "objective": "global", "lm_loss": 1.53902, "bpb": 2.22034, "chunk_loss": 0.030133, "F": 0.17236, "G": 0.27585, "chunks": 176.5, "bytes_per_chunk": 5.802, "boundary_entropy": 0.35873, "router_noise": 0.0, "grad_norm": 0.6922, "lr": 0.0002715, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 100.0} |
| {"event": "train", "step": 1550, "phase": 1, "objective": "global", "lm_loss": 1.60922, "bpb": 2.32162, "chunk_loss": 0.030057, "F": 0.172, "G": 0.26798, "chunks": 176.12, "bytes_per_chunk": 5.814, "boundary_entropy": 0.33627, "router_noise": 0.0, "grad_norm": 0.7292, "lr": 0.0002692, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 102.1} |
| {"event": "probe", "step": 1600, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2889.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166006, "cps": 22.570312, "mean_p": 0.230753, "boundary_entropy": 0.238552, "bpb": 2.377944}, "de": {"chunks": 3322.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161788, "cps": 25.953125, "mean_p": 0.25541, "boundary_entropy": 0.314938, "bpb": 2.671424}, "vi": {"chunks": 4250.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.23054, "cps": 33.203125, "mean_p": 0.237306, "boundary_entropy": 0.248664, "bpb": 2.133907}, "tr": {"chunks": 3897.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215268, "cps": 30.445312, "mean_p": 0.251629, "boundary_entropy": 0.284433, "bpb": 2.646459}, "fi": {"chunks": 3029.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.161023, "cps": 23.664062, "mean_p": 0.254751, "boundary_entropy": 0.329745, "bpb": 2.904484}, "ru": {"chunks": 6013.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.317309, "cps": 46.976562, "mean_p": 0.255868, "boundary_entropy": 0.30748, "bpb": 1.663355}, "el": {"chunks": 6800.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.331336, "cps": 53.125, "mean_p": 0.241352, "boundary_entropy": 0.296826, "bpb": 1.786847}, "hi": {"chunks": 7746.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.460386, "cps": 60.515625, "mean_p": 0.335919, "boundary_entropy": 0.483741, "bpb": 1.40532}, "ta": {"chunks": 11350.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.607114, "cps": 88.671875, "mean_p": 0.272447, "boundary_entropy": 0.358094, "bpb": 1.141503}, "th": {"chunks": 7395.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.459745, "cps": 57.773438, "mean_p": 0.357706, "boundary_entropy": 0.5507, "bpb": 1.304107}, "zh": {"chunks": 2902.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.491365, "cps": 22.671875, "mean_p": 0.311509, "boundary_entropy": 0.50949, "bpb": 3.512789}, "my": {"chunks": 8074.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.426114, "cps": 63.078125, "mean_p": 0.351449, "boundary_entropy": 0.545509, "bpb": 1.270671}}, "bpb": {"en": 2.37794, "de": 2.67142, "vi": 2.13391, "tr": 2.64646, "fi": 2.90448, "ru": 1.66336, "el": 1.78685, "hi": 1.40532, "ta": 1.1415, "th": 1.30411, "zh": 3.51279, "my": 1.27067}, "gini_cpc": 0.243235, "gini_cps": 0.250584, "bpic": 5.671612, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9745941479913371, "vi": 1.388744546315096, "tr": 1.2967505256783483, "fi": 0.9699826546993421, "ru": 1.9114307476909502, "el": 1.9959269667256285, "hi": 2.7733137067022167, "ta": 3.657185341452963, "th": 2.769451037469363, "zh": 2.9599238889363617, "my": 2.5668586113473153}, "t": 105.2} |
| {"event": "eval", "step": 1600, "phase": 1, "bpb": {"en": 2.38933, "de": 2.67487, "vi": 2.13567, "tr": 2.65042, "fi": 2.91096, "ru": 1.6447, "el": 1.7755, "hi": 1.4047, "ta": 1.14898, "th": 1.32225, "zh": 3.5167, "my": 1.27798}, "bytes_per_chunk": {"en": 5.73, "de": 6.223, "vi": 5.723, "tr": 5.058, "fi": 6.294, "ru": 5.911, "el": 5.555, "hi": 5.705, "ta": 4.464, "th": 6.187, "zh": 6.419, "my": 6.841}, "bpb_mean": 2.07101, "t": 106.5} |
| {"event": "train", "step": 1600, "phase": 1, "objective": "global", "lm_loss": 1.57792, "bpb": 2.27645, "chunk_loss": 0.030176, "F": 0.17206, "G": 0.26188, "chunks": 176.19, "bytes_per_chunk": 5.812, "boundary_entropy": 0.3434, "router_noise": 0.0, "grad_norm": 0.7713, "lr": 0.0002668, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 106.5} |
| {"event": "train", "step": 1650, "phase": 1, "objective": "global", "lm_loss": 1.50338, "bpb": 2.16893, "chunk_loss": 0.030568, "F": 0.18405, "G": 0.27885, "chunks": 188.47, "bytes_per_chunk": 5.433, "boundary_entropy": 0.34598, "router_noise": 0.0, "grad_norm": 0.6802, "lr": 0.0002643, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 108.6} |
| {"event": "train", "step": 1700, "phase": 1, "objective": "global", "lm_loss": 1.42941, "bpb": 2.0622, "chunk_loss": 0.029886, "F": 0.16364, "G": 0.26322, "chunks": 167.56, "bytes_per_chunk": 6.111, "boundary_entropy": 0.33639, "router_noise": 0.0, "grad_norm": 0.7035, "lr": 0.0002618, "router_trainable": true, "sec_per_step": 0.042, "alpha": 0.03, "t": 110.7} |
| {"event": "train", "step": 1750, "phase": 1, "objective": "global", "lm_loss": 1.62134, "bpb": 2.3391, "chunk_loss": 0.030159, "F": 0.1738, "G": 0.2609, "chunks": 177.97, "bytes_per_chunk": 5.754, "boundary_entropy": 0.31337, "router_noise": 0.0, "grad_norm": 0.7965, "lr": 0.0002591, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 112.9} |
| {"event": "probe", "step": 1800, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2891.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166121, "cps": 22.585938, "mean_p": 0.230996, "boundary_entropy": 0.237012, "bpb": 2.332288}, "de": {"chunks": 3324.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161886, "cps": 25.96875, "mean_p": 0.253749, "boundary_entropy": 0.309053, "bpb": 2.601624}, "vi": {"chunks": 4314.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234011, "cps": 33.703125, "mean_p": 0.243814, "boundary_entropy": 0.257132, "bpb": 2.115327}, "tr": {"chunks": 3822.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.211125, "cps": 29.859375, "mean_p": 0.244205, "boundary_entropy": 0.272202, "bpb": 2.620403}, "fi": {"chunks": 2982.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158524, "cps": 23.296875, "mean_p": 0.247563, "boundary_entropy": 0.316322, "bpb": 2.853277}, "ru": {"chunks": 6060.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.319789, "cps": 47.34375, "mean_p": 0.26378, "boundary_entropy": 0.310428, "bpb": 1.61012}, "el": {"chunks": 6821.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.332359, "cps": 53.289062, "mean_p": 0.249803, "boundary_entropy": 0.299433, "bpb": 1.718239}, "hi": {"chunks": 7509.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.4463, "cps": 58.664062, "mean_p": 0.319053, "boundary_entropy": 0.465011, "bpb": 1.368327}, "ta": {"chunks": 9101.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.486815, "cps": 71.101562, "mean_p": 0.244052, "boundary_entropy": 0.339999, "bpb": 1.096539}, "th": {"chunks": 7822.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.486292, "cps": 61.109375, "mean_p": 0.354676, "boundary_entropy": 0.544142, "bpb": 1.304663}, "zh": {"chunks": 3249.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.550119, "cps": 25.382812, "mean_p": 0.319886, "boundary_entropy": 0.497511, "bpb": 3.481482}, "my": {"chunks": 8890.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.469179, "cps": 69.453125, "mean_p": 0.343836, "boundary_entropy": 0.534605, "bpb": 1.27602}}, "bpb": {"en": 2.33229, "de": 2.60162, "vi": 2.11533, "tr": 2.6204, "fi": 2.85328, "ru": 1.61012, "el": 1.71824, "hi": 1.36833, "ta": 1.09654, "th": 1.30466, "zh": 3.48148, "my": 1.27602}, "gini_cpc": 0.236436, "gini_cps": 0.231041, "bpic": 5.746515, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9745062671974514, "vi": 1.4086822013493236, "tr": 1.2709139884109477, "fi": 0.95427112800287, "ru": 1.925038583028578, "el": 2.0007058080789566, "hi": 2.6866003064282604, "ta": 2.9304861515142906, "th": 2.9273373563373495, "zh": 3.3115574766815468, "my": 2.82432332965818}, "t": 115.9} |
| {"event": "eval", "step": 1800, "phase": 1, "bpb": {"en": 2.34321, "de": 2.60426, "vi": 2.12082, "tr": 2.62701, "fi": 2.85545, "ru": 1.58817, "el": 1.70821, "hi": 1.36793, "ta": 1.10397, "th": 1.32246, "zh": 3.48626, "my": 1.28233}, "bytes_per_chunk": {"en": 5.733, "de": 6.247, "vi": 5.668, "tr": 5.115, "fi": 6.419, "ru": 5.865, "el": 5.561, "hi": 5.759, "ta": 5.516, "th": 5.918, "zh": 5.556, "my": 5.953}, "bpb_mean": 2.03417, "t": 117.2} |
| {"event": "train", "step": 1800, "phase": 1, "objective": "global", "lm_loss": 1.57716, "bpb": 2.27537, "chunk_loss": 0.030114, "F": 0.172, "G": 0.23904, "chunks": 176.12, "bytes_per_chunk": 5.814, "boundary_entropy": 0.26201, "router_noise": 0.0, "grad_norm": 0.8704, "lr": 0.0002564, "router_trainable": true, "sec_per_step": 0.087, "alpha": 0.03, "t": 117.2} |
| {"event": "train", "step": 1850, "phase": 1, "objective": "global", "lm_loss": 1.54935, "bpb": 2.23525, "chunk_loss": 0.03033, "F": 0.1767, "G": 0.25982, "chunks": 180.94, "bytes_per_chunk": 5.659, "boundary_entropy": 0.2916, "router_noise": 0.0, "grad_norm": 0.6873, "lr": 0.0002537, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 119.4} |
| {"event": "train", "step": 1900, "phase": 1, "objective": "global", "lm_loss": 1.59362, "bpb": 2.2991, "chunk_loss": 0.030022, "F": 0.16953, "G": 0.26707, "chunks": 173.59, "bytes_per_chunk": 5.899, "boundary_entropy": 0.337, "router_noise": 0.0, "grad_norm": 0.6585, "lr": 0.0002508, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 121.5} |
| {"event": "train", "step": 1950, "phase": 1, "objective": "global", "lm_loss": 1.60155, "bpb": 2.31054, "chunk_loss": 0.03015, "F": 0.17447, "G": 0.26196, "chunks": 178.66, "bytes_per_chunk": 5.732, "boundary_entropy": 0.30904, "router_noise": 0.0, "grad_norm": 0.6942, "lr": 0.0002479, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 123.7} |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2892.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166178, "cps": 22.59375, "mean_p": 0.234025, "boundary_entropy": 0.242214, "bpb": 2.283066}, "de": {"chunks": 3338.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162568, "cps": 26.078125, "mean_p": 0.265267, "boundary_entropy": 0.327101, "bpb": 2.549855}, "vi": {"chunks": 4273.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.231787, "cps": 33.382812, "mean_p": 0.239938, "boundary_entropy": 0.243618, "bpb": 2.085786}, "tr": {"chunks": 3802.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.21002, "cps": 29.703125, "mean_p": 0.237878, "boundary_entropy": 0.262075, "bpb": 2.556454}, "fi": {"chunks": 3127.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.166233, "cps": 24.429688, "mean_p": 0.268126, "boundary_entropy": 0.341992, "bpb": 2.780736}, "ru": {"chunks": 5986.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.315884, "cps": 46.765625, "mean_p": 0.27034, "boundary_entropy": 0.324054, "bpb": 1.575276}, "el": {"chunks": 6781.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.33041, "cps": 52.976562, "mean_p": 0.253151, "boundary_entropy": 0.298353, "bpb": 1.700122}, "hi": {"chunks": 6570.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.39049, "cps": 51.328125, "mean_p": 0.299128, "boundary_entropy": 0.450665, "bpb": 1.347888}, "ta": {"chunks": 8744.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.467719, "cps": 68.3125, "mean_p": 0.245297, "boundary_entropy": 0.353689, "bpb": 1.069495}, "th": {"chunks": 9791.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.608704, "cps": 76.492188, "mean_p": 0.411895, "boundary_entropy": 0.56959, "bpb": 1.252038}, "zh": {"chunks": 3170.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.536742, "cps": 24.765625, "mean_p": 0.319634, "boundary_entropy": 0.50269, "bpb": 3.46046}, "my": {"chunks": 9256.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.488495, "cps": 72.3125, "mean_p": 0.354715, "boundary_entropy": 0.550649, "bpb": 1.232017}}, "bpb": {"en": 2.28307, "de": 2.54985, "vi": 2.08579, "tr": 2.55645, "fi": 2.78074, "ru": 1.57528, "el": 1.70012, "hi": 1.34789, "ta": 1.0695, "th": 1.25204, "zh": 3.46046, "my": 1.23202}, "gini_cpc": 0.25106, "gini_cps": 0.242753, "bpic": 5.666337, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9782723004543037, "vi": 1.3948117024377453, "tr": 1.2638263115418618, "fi": 1.0003266275515055, "ru": 1.9008739968688075, "el": 1.9882854334262763, "hi": 2.349828290986664, "ta": 2.8145599984019514, "th": 3.6629569700385782, "zh": 3.2299191482605076, "my": 2.939583503048746}, "t": 126.7} |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.29318, "de": 2.55793, "vi": 2.08851, "tr": 2.56223, "fi": 2.78607, "ru": 1.55494, "el": 1.68881, "hi": 1.34565, "ta": 1.0779, "th": 1.27271, "zh": 3.46386, "my": 1.23915}, "bytes_per_chunk": {"en": 5.722, "de": 6.202, "vi": 5.713, "tr": 5.157, "fi": 6.119, "ru": 5.94, "el": 5.542, "hi": 6.709, "ta": 5.824, "th": 4.802, "zh": 5.638, "my": 5.847}, "bpb_mean": 1.99424, "t": 128.0} |
| {"event": "train", "step": 2000, "phase": 1, "objective": "global", "lm_loss": 1.51611, "bpb": 2.18729, "chunk_loss": 0.030239, "F": 0.17404, "G": 0.27223, "chunks": 178.22, "bytes_per_chunk": 5.746, "boundary_entropy": 0.33527, "router_noise": 0.0, "grad_norm": 0.6561, "lr": 0.000245, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 128.1} |
| {"event": "train", "step": 2050, "phase": 1, "objective": "global", "lm_loss": 1.39126, "bpb": 2.00716, "chunk_loss": 0.03018, "F": 0.17685, "G": 0.25162, "chunks": 181.09, "bytes_per_chunk": 5.655, "boundary_entropy": 0.28595, "router_noise": 0.0, "grad_norm": 0.6993, "lr": 0.0002419, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 130.2} |
| {"event": "train", "step": 2100, "phase": 1, "objective": "global", "lm_loss": 1.56349, "bpb": 2.25563, "chunk_loss": 0.029943, "F": 0.16681, "G": 0.25099, "chunks": 170.81, "bytes_per_chunk": 5.995, "boundary_entropy": 0.29346, "router_noise": 0.0, "grad_norm": 0.7015, "lr": 0.0002388, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 132.4} |
| {"event": "train", "step": 2150, "phase": 1, "objective": "global", "lm_loss": 1.48057, "bpb": 2.13601, "chunk_loss": 0.030134, "F": 0.17319, "G": 0.25395, "chunks": 177.34, "bytes_per_chunk": 5.774, "boundary_entropy": 0.29424, "router_noise": 0.0, "grad_norm": 0.6911, "lr": 0.0002357, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 134.5} |
| {"event": "probe", "step": 2200, "phase": 1, "objective": "global", "router_trainable": true, "per_lang": {"en": {"chunks": 2895.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166351, "cps": 22.617188, "mean_p": 0.228321, "boundary_entropy": 0.227306, "bpb": 2.248696}, "de": {"chunks": 3325.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161934, "cps": 25.976562, "mean_p": 0.25492, "boundary_entropy": 0.307243, "bpb": 2.511901}, "vi": {"chunks": 4327.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234717, "cps": 33.804688, "mean_p": 0.241402, "boundary_entropy": 0.243337, "bpb": 2.05245}, "tr": {"chunks": 3778.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.208695, "cps": 29.515625, "mean_p": 0.235571, "boundary_entropy": 0.256911, "bpb": 2.507472}, "fi": {"chunks": 3182.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.169156, "cps": 24.859375, "mean_p": 0.267958, "boundary_entropy": 0.34017, "bpb": 2.74608}, "ru": {"chunks": 6142.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.324116, "cps": 47.984375, "mean_p": 0.268557, "boundary_entropy": 0.30953, "bpb": 1.54037}, "el": {"chunks": 6909.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.336647, "cps": 53.976562, "mean_p": 0.247265, "boundary_entropy": 0.278979, "bpb": 1.642293}, "hi": {"chunks": 7615.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.4526, "cps": 59.492188, "mean_p": 0.297258, "boundary_entropy": 0.440355, "bpb": 1.319699}, "ta": {"chunks": 9350.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.500134, "cps": 73.046875, "mean_p": 0.256272, "boundary_entropy": 0.356078, "bpb": 1.06262}, "th": {"chunks": 9069.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.563817, "cps": 70.851562, "mean_p": 0.408564, "boundary_entropy": 0.579322, "bpb": 1.2367}, "zh": {"chunks": 3004.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.508635, "cps": 23.46875, "mean_p": 0.300711, "boundary_entropy": 0.46933, "bpb": 3.410343}, "my": {"chunks": 9599.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.506597, "cps": 74.992188, "mean_p": 0.356319, "boundary_entropy": 0.542198, "bpb": 1.192247}}, "bpb": {"en": 2.2487, "de": 2.5119, "vi": 2.05245, "tr": 2.50747, "fi": 2.74608, "ru": 1.54037, "el": 1.64229, "hi": 1.3197, "ta": 1.06262, "th": 1.2367, "zh": 3.41034, "my": 1.19225}, "gini_cpc": 0.241664, "gini_cps": 0.246556, "bpic": 5.546369, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9734525668145982, "vi": 1.4109749561194562, "tr": 1.254547052144622, "fi": 1.0168662751895525, "ru": 1.948391157532093, "el": 2.0237175578821085, "hi": 2.720760957432885, "ta": 3.006503359768966, "th": 3.389330258133201, "zh": 3.057609632076978, "my": 3.0453566947883544}, "t": 137.6} |
| {"event": "eval", "step": 2200, "phase": 1, "bpb": {"en": 2.25974, "de": 2.51784, "vi": 2.05703, "tr": 2.51356, "fi": 2.75107, "ru": 1.51925, "el": 1.63372, "hi": 1.31671, "ta": 1.06986, "th": 1.25644, "zh": 3.41575, "my": 1.19995}, "bytes_per_chunk": {"en": 5.723, "de": 6.234, "vi": 5.659, "tr": 5.182, "fi": 6.031, "ru": 5.783, "el": 5.531, "hi": 5.686, "ta": 5.285, "th": 5.288, "zh": 6.088, "my": 5.587}, "bpb_mean": 1.95925, "t": 138.9} |
| {"event": "train", "step": 2200, "phase": 1, "objective": "global", "lm_loss": 1.48574, "bpb": 2.14347, "chunk_loss": 0.030204, "F": 0.17365, "G": 0.26551, "chunks": 177.81, "bytes_per_chunk": 5.759, "boundary_entropy": 0.32488, "router_noise": 0.0, "grad_norm": 0.674, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.088, "alpha": 0.03, "t": 138.9} |
| {"event": "train", "step": 2250, "phase": 1, "objective": "global", "lm_loss": 1.71709, "bpb": 2.47724, "chunk_loss": 0.030246, "F": 0.17828, "G": 0.24998, "chunks": 182.56, "bytes_per_chunk": 5.609, "boundary_entropy": 0.27279, "router_noise": 0.0, "grad_norm": 0.7312, "lr": 0.0002293, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 141.1} |
| {"event": "train", "step": 2300, "phase": 1, "objective": "global", "lm_loss": 1.48817, "bpb": 2.14698, "chunk_loss": 0.030052, "F": 0.16852, "G": 0.2491, "chunks": 172.56, "bytes_per_chunk": 5.934, "boundary_entropy": 0.29753, "router_noise": 0.0, "grad_norm": 0.6485, "lr": 0.000226, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 143.2} |
| {"event": "train", "step": 2350, "phase": 1, "objective": "global", "lm_loss": 1.49329, "bpb": 2.15437, "chunk_loss": 0.030057, "F": 0.172, "G": 0.26259, "chunks": 176.12, "bytes_per_chunk": 5.814, "boundary_entropy": 0.31323, "router_noise": 0.0, "grad_norm": 0.7092, "lr": 0.0002226, "router_trainable": true, "sec_per_step": 0.043, "alpha": 0.03, "t": 145.4} |
| {"event": "probe", "step": 2400, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2894.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166293, "cps": 22.609375, "mean_p": 0.224777, "boundary_entropy": 0.215751, "bpb": 2.224607}, "de": {"chunks": 3332.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.162275, "cps": 26.03125, "mean_p": 0.256415, "boundary_entropy": 0.305433, "bpb": 2.482869}, "vi": {"chunks": 4391.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.238188, "cps": 34.304688, "mean_p": 0.244084, "boundary_entropy": 0.24444, "bpb": 2.045137}, "tr": {"chunks": 3748.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.207038, "cps": 29.28125, "mean_p": 0.228013, "boundary_entropy": 0.244843, "bpb": 2.479253}, "fi": {"chunks": 3111.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.165382, "cps": 24.304688, "mean_p": 0.266446, "boundary_entropy": 0.332486, "bpb": 2.700383}, "ru": {"chunks": 6127.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.323325, "cps": 47.867188, "mean_p": 0.263336, "boundary_entropy": 0.303284, "bpb": 1.52074}, "el": {"chunks": 6807.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.331677, "cps": 53.179688, "mean_p": 0.236618, "boundary_entropy": 0.268823, "bpb": 1.622485}, "hi": {"chunks": 7154.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.425201, "cps": 55.890625, "mean_p": 0.277252, "boundary_entropy": 0.42124, "bpb": 1.295872}, "ta": {"chunks": 8852.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.473496, "cps": 69.15625, "mean_p": 0.242081, "boundary_entropy": 0.338491, "bpb": 1.046547}, "th": {"chunks": 7126.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.443021, "cps": 55.671875, "mean_p": 0.385578, "boundary_entropy": 0.566968, "bpb": 1.21282}, "zh": {"chunks": 3400.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.575686, "cps": 26.5625, "mean_p": 0.311785, "boundary_entropy": 0.458632, "bpb": 3.395884}, "my": {"chunks": 8591.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.453399, "cps": 67.117188, "mean_p": 0.333741, "boundary_entropy": 0.536887, "bpb": 1.224099}}, "bpb": {"en": 2.22461, "de": 2.48287, "vi": 2.04514, "tr": 2.47925, "fi": 2.70038, "ru": 1.52074, "el": 1.62248, "hi": 1.29587, "ta": 1.04655, "th": 1.21282, "zh": 3.39588, "my": 1.2241}, "gini_cpc": 0.232863, "gini_cps": 0.217645, "bpic": 5.856301, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9758390180204798, "vi": 1.4323392355778881, "tr": 1.2450151156472706, "fi": 0.9945204575084854, "ru": 1.9443044019744244, "el": 1.9945296596287658, "hi": 2.556933636329096, "ta": 2.8473544234707915, "th": 2.664098918173774, "zh": 3.4618724916574832, "my": 2.726502726919008}, "t": 148.5} |
| {"event": "eval", "step": 2400, "phase": 1, "bpb": {"en": 2.23624, "de": 2.48847, "vi": 2.04848, "tr": 2.48373, "fi": 2.70421, "ru": 1.49768, "el": 1.61284, "hi": 1.29286, "ta": 1.05285, "th": 1.23162, "zh": 3.39843, "my": 1.2313}, "bytes_per_chunk": {"en": 5.73, "de": 6.229, "vi": 5.572, "tr": 5.263, "fi": 6.114, "ru": 5.824, "el": 5.521, "hi": 6.119, "ta": 5.64, "th": 6.591, "zh": 5.191, "my": 6.105}, "bpb_mean": 1.93989, "t": 149.8} |
| {"event": "phase_transition", "step": 2400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 149.8} |
| {"event": "train", "step": 2400, "phase": 2, "objective": "parity_b", "lm_loss": 1.5076, "bpb": 2.17501, "chunk_loss": 0.0, "F": 0.17404, "G": 0.25548, "chunks": 178.22, "bytes_per_chunk": 5.746, "boundary_entropy": 0.30279, "router_noise": 0.0, "grad_norm": 0.7212, "lr": 0.0002192, "router_trainable": true, "sec_per_step": 0.088, "worst_lang": "my", "n_target_mean": 21.145, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 149.8} |
| {"event": "train", "step": 2450, "phase": 2, "objective": "parity_b", "lm_loss": 1.31909, "bpb": 1.90305, "chunk_loss": 0.034573, "F": 0.15518, "G": 0.22053, "chunks": 158.91, "bytes_per_chunk": 6.444, "boundary_entropy": 0.25788, "router_noise": 0.0, "grad_norm": 0.654, "lr": 0.0002158, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1875, "fallback": 0.0, "alpha": 0.03, "t": 152.0} |
| {"event": "train", "step": 2500, "phase": 2, "objective": "parity_b", "lm_loss": 1.421, "bpb": 2.05007, "chunk_loss": 0.032159, "F": 0.15686, "G": 0.2119, "chunks": 160.62, "bytes_per_chunk": 6.375, "boundary_entropy": 0.24576, "router_noise": 0.0, "grad_norm": 0.6861, "lr": 0.0002123, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 154.1} |
| {"event": "train", "step": 2550, "phase": 2, "objective": "parity_b", "lm_loss": 1.58452, "bpb": 2.28597, "chunk_loss": 0.032174, "F": 0.16061, "G": 0.20559, "chunks": 164.47, "bytes_per_chunk": 6.226, "boundary_entropy": 0.21921, "router_noise": 0.0, "grad_norm": 0.6466, "lr": 0.0002088, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 156.3} |
| {"event": "probe", "step": 2600, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2905.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166925, "cps": 22.695312, "mean_p": 0.202917, "boundary_entropy": 0.161376, "bpb": 2.199976}, "de": {"chunks": 3287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160084, "cps": 25.679688, "mean_p": 0.211945, "boundary_entropy": 0.22085, "bpb": 2.450253}, "vi": {"chunks": 4236.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22978, "cps": 33.09375, "mean_p": 0.219486, "boundary_entropy": 0.196709, "bpb": 2.025793}, "tr": {"chunks": 3856.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213003, "cps": 30.125, "mean_p": 0.215639, "boundary_entropy": 0.204044, "bpb": 2.460467}, "fi": {"chunks": 2966.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157674, "cps": 23.171875, "mean_p": 0.215962, "boundary_entropy": 0.245513, "bpb": 2.665324}, "ru": {"chunks": 5325.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.281003, "cps": 41.601562, "mean_p": 0.147303, "boundary_entropy": 0.167885, "bpb": 1.526286}, "el": {"chunks": 5433.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.264727, "cps": 42.445312, "mean_p": 0.154222, "boundary_entropy": 0.189919, "bpb": 1.63917}, "hi": {"chunks": 4520.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.268648, "cps": 35.3125, "mean_p": 0.158052, "boundary_entropy": 0.298992, "bpb": 1.319334}, "ta": {"chunks": 4272.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.22851, "cps": 33.375, "mean_p": 0.121988, "boundary_entropy": 0.234143, "bpb": 1.115987}, "th": {"chunks": 4230.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.262978, "cps": 33.046875, "mean_p": 0.250076, "boundary_entropy": 0.452422, "bpb": 1.235284}, "zh": {"chunks": 2956.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.500508, "cps": 23.09375, "mean_p": 0.271039, "boundary_entropy": 0.416214, "bpb": 3.368951}, "my": {"chunks": 2119.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.111832, "cps": 16.554688, "mean_p": 0.133602, "boundary_entropy": 0.313246, "bpb": 1.234409}}, "bpb": {"en": 2.19998, "de": 2.45025, "vi": 2.02579, "tr": 2.46047, "fi": 2.66532, "ru": 1.52629, "el": 1.63917, "hi": 1.31933, "ta": 1.11599, "th": 1.23528, "zh": 3.36895, "my": 1.23441}, "gini_cpc": 0.198709, "gini_cps": 0.142669, "bpic": 8.324065, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9590147357769153, "vi": 1.3765462034118108, "tr": 1.2760404961359473, "fi": 0.9445767394860242, "ru": 1.6834041026526008, "el": 1.5859038132189207, "hi": 1.6093901737282392, "ta": 1.3689379333710499, "th": 1.575423377421048, "zh": 2.998395866859631, "my": 0.6699548942172381}, "t": 159.3} |
| {"event": "eval", "step": 2600, "phase": 2, "bpb": {"en": 2.21129, "de": 2.46083, "vi": 2.02914, "tr": 2.46542, "fi": 2.6713, "ru": 1.50505, "el": 1.62898, "hi": 1.31582, "ta": 1.12342, "th": 1.25622, "zh": 3.37247, "my": 1.24231}, "bytes_per_chunk": {"en": 5.738, "de": 6.294, "vi": 5.771, "tr": 5.143, "fi": 6.425, "ru": 6.645, "el": 6.919, "hi": 9.211, "ta": 11.271, "th": 11.081, "zh": 5.894, "my": 24.212}, "bpb_mean": 1.94019, "t": 160.6} |
| {"event": "train", "step": 2600, "phase": 2, "objective": "parity_b", "lm_loss": 1.68202, "bpb": 2.42664, "chunk_loss": 0.031755, "F": 0.1637, "G": 0.21242, "chunks": 167.62, "bytes_per_chunk": 6.109, "boundary_entropy": 0.24943, "router_noise": 0.0, "grad_norm": 0.7, "lr": 0.0002053, "router_trainable": true, "sec_per_step": 0.087, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 160.6} |
| {"event": "train", "step": 2650, "phase": 2, "objective": "parity_b", "lm_loss": 1.43413, "bpb": 2.06901, "chunk_loss": 0.031509, "F": 0.15469, "G": 0.20078, "chunks": 158.41, "bytes_per_chunk": 6.464, "boundary_entropy": 0.2211, "router_noise": 0.0, "grad_norm": 0.6288, "lr": 0.0002017, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 162.8} |
| {"event": "train", "step": 2700, "phase": 2, "objective": "parity_b", "lm_loss": 1.37861, "bpb": 1.98891, "chunk_loss": 0.03126, "F": 0.14664, "G": 0.18174, "chunks": 150.16, "bytes_per_chunk": 6.82, "boundary_entropy": 0.2061, "router_noise": 0.0, "grad_norm": 0.6687, "lr": 0.0001981, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.2188, "fallback": 0.0, "alpha": 0.03, "t": 165.0} |
| {"event": "train", "step": 2750, "phase": 2, "objective": "parity_b", "lm_loss": 1.35162, "bpb": 1.94998, "chunk_loss": 0.031582, "F": 0.14621, "G": 0.18662, "chunks": 149.72, "bytes_per_chunk": 6.839, "boundary_entropy": 0.2011, "router_noise": 0.0, "grad_norm": 0.6747, "lr": 0.0001945, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "th", "n_target_mean": 16.5724, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 167.1} |
| {"event": "probe", "step": 2800, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2899.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.16658, "cps": 22.648438, "mean_p": 0.196058, "boundary_entropy": 0.138712, "bpb": 2.17504}, "de": {"chunks": 3292.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160327, "cps": 25.71875, "mean_p": 0.201246, "boundary_entropy": 0.185858, "bpb": 2.428241}, "vi": {"chunks": 4214.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.228587, "cps": 32.921875, "mean_p": 0.216684, "boundary_entropy": 0.190559, "bpb": 2.002048}, "tr": {"chunks": 3920.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216539, "cps": 30.625, "mean_p": 0.218159, "boundary_entropy": 0.19122, "bpb": 2.422018}, "fi": {"chunks": 2959.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157302, "cps": 23.117188, "mean_p": 0.204322, "boundary_entropy": 0.215811, "bpb": 2.642246}, "ru": {"chunks": 4782.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.252348, "cps": 37.359375, "mean_p": 0.13793, "boundary_entropy": 0.159323, "bpb": 1.5223}, "el": {"chunks": 4131.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.201286, "cps": 32.273438, "mean_p": 0.125221, "boundary_entropy": 0.188539, "bpb": 1.647743}, "hi": {"chunks": 1021.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.060684, "cps": 7.976562, "mean_p": 0.088511, "boundary_entropy": 0.235014, "bpb": 1.35791}, "ta": {"chunks": 1700.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.090933, "cps": 13.28125, "mean_p": 0.085973, "boundary_entropy": 0.214355, "bpb": 1.083913}, "th": {"chunks": 2433.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.151259, "cps": 19.007812, "mean_p": 0.150561, "boundary_entropy": 0.332394, "bpb": 1.258152}, "zh": {"chunks": 2817.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.476973, "cps": 22.007812, "mean_p": 0.263017, "boundary_entropy": 0.409101, "bpb": 3.344954}, "my": {"chunks": 3123.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.16482, "cps": 24.398438, "mean_p": 0.135495, "boundary_entropy": 0.317977, "bpb": 1.20551}}, "bpb": {"en": 2.17504, "de": 2.42824, "vi": 2.00205, "tr": 2.42202, "fi": 2.64225, "ru": 1.5223, "el": 1.64774, "hi": 1.35791, "ta": 1.08391, "th": 1.25815, "zh": 3.34495, "my": 1.20551}, "gini_cpc": 0.250398, "gini_cps": 0.184899, "bpic": 10.291518, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.96246140729013, "vi": 1.3722312146580666, "tr": 1.299904419913516, "fi": 0.9442978188893719, "ru": 1.514873129757236, "el": 1.208343067152029, "hi": 0.3642894332062037, "ta": 0.5458827323861619, "th": 0.9080232075995893, "zh": 2.8633161919164296, "my": 0.9894287213537634}, "t": 170.2} |
| {"event": "eval", "step": 2800, "phase": 2, "bpb": {"en": 2.18762, "de": 2.43747, "vi": 2.00532, "tr": 2.42633, "fi": 2.65109, "ru": 1.50259, "el": 1.63673, "hi": 1.35565, "ta": 1.09129, "th": 1.27528, "zh": 3.34757, "my": 1.2135}, "bytes_per_chunk": {"en": 5.738, "de": 6.283, "vi": 5.768, "tr": 5.06, "fi": 6.455, "ru": 7.417, "el": 9.214, "hi": 34.901, "ta": 26.525, "th": 19.755, "zh": 6.219, "my": 16.161}, "bpb_mean": 1.92754, "t": 171.4} |
| {"event": "train", "step": 2800, "phase": 2, "objective": "parity_b", "lm_loss": 1.47435, "bpb": 2.12703, "chunk_loss": 0.031064, "F": 0.1564, "G": 0.18904, "chunks": 160.16, "bytes_per_chunk": 6.394, "boundary_entropy": 0.19103, "router_noise": 0.0, "grad_norm": 0.7, "lr": 0.0001909, "router_trainable": true, "sec_per_step": 0.087, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 171.5} |
| {"event": "train", "step": 2850, "phase": 2, "objective": "parity_b", "lm_loss": 1.42169, "bpb": 2.05106, "chunk_loss": 0.030809, "F": 0.14447, "G": 0.1804, "chunks": 147.94, "bytes_per_chunk": 6.922, "boundary_entropy": 0.19434, "router_noise": 0.0, "grad_norm": 0.7392, "lr": 0.0001872, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 173.6} |
| {"event": "train", "step": 2900, "phase": 2, "objective": "parity_b", "lm_loss": 1.49039, "bpb": 2.15018, "chunk_loss": 0.0, "F": 0.15314, "G": 0.18924, "chunks": 156.81, "bytes_per_chunk": 6.53, "boundary_entropy": 0.19775, "router_noise": 0.0, "grad_norm": 0.6954, "lr": 0.0001835, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 175.8} |
| {"event": "train", "step": 2950, "phase": 2, "objective": "parity_b", "lm_loss": 1.3816, "bpb": 1.99323, "chunk_loss": 0.030792, "F": 0.14993, "G": 0.1884, "chunks": 153.53, "bytes_per_chunk": 6.67, "boundary_entropy": 0.20772, "router_noise": 0.0, "grad_norm": 0.698, "lr": 0.0001799, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.25, "fallback": 0.0, "alpha": 0.03, "t": 178.0} |
| {"event": "probe", "step": 3000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2897.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.166466, "cps": 22.632812, "mean_p": 0.1953, "boundary_entropy": 0.133829, "bpb": 2.1508}, "de": {"chunks": 3308.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.161107, "cps": 25.84375, "mean_p": 0.203169, "boundary_entropy": 0.185648, "bpb": 2.39482}, "vi": {"chunks": 4226.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.229238, "cps": 33.015625, "mean_p": 0.21462, "boundary_entropy": 0.173953, "bpb": 1.992577}, "tr": {"chunks": 3878.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.214219, "cps": 30.296875, "mean_p": 0.217972, "boundary_entropy": 0.194069, "bpb": 2.395368}, "fi": {"chunks": 2980.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.158418, "cps": 23.28125, "mean_p": 0.204928, "boundary_entropy": 0.209738, "bpb": 2.58727}, "ru": {"chunks": 4467.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.235726, "cps": 34.898438, "mean_p": 0.126384, "boundary_entropy": 0.138581, "bpb": 1.525563}, "el": {"chunks": 4337.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.211324, "cps": 33.882812, "mean_p": 0.1287, "boundary_entropy": 0.183717, "bpb": 1.612008}, "hi": {"chunks": 1640.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.097474, "cps": 12.8125, "mean_p": 0.10846, "boundary_entropy": 0.251531, "bpb": 1.307665}, "ta": {"chunks": 2646.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.141535, "cps": 20.671875, "mean_p": 0.098715, "boundary_entropy": 0.21387, "bpb": 1.058422}, "th": {"chunks": 3542.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.220205, "cps": 27.671875, "mean_p": 0.188921, "boundary_entropy": 0.367714, "bpb": 1.204963}, "zh": {"chunks": 3209.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.543346, "cps": 25.070312, "mean_p": 0.280697, "boundary_entropy": 0.40036, "bpb": 3.326124}, "my": {"chunks": 2967.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.156586, "cps": 23.179688, "mean_p": 0.145758, "boundary_entropy": 0.323074, "bpb": 1.195546}}, "bpb": {"en": 2.1508, "de": 2.39482, "vi": 1.99258, "tr": 2.39537, "fi": 2.58727, "ru": 1.52556, "el": 1.61201, "hi": 1.30766, "ta": 1.05842, "th": 1.20496, "zh": 3.32612, "my": 1.19555}, "gini_cpc": 0.222626, "gini_cps": 0.128769, "bpic": 9.571315, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9678069102868344, "vi": 1.3770888939007917, "tr": 1.2868646715718368, "fi": 0.9516560340697695, "ru": 1.4160623081105645, "el": 1.2694751442058045, "hi": 0.5855505592966234, "ta": 0.8502369901890556, "th": 1.3228272008956559, "zh": 3.264013147643539, "my": 0.9406537587856495}, "t": 181.0} |
| {"event": "eval", "step": 3000, "phase": 2, "bpb": {"en": 2.1643, "de": 2.40532, "vi": 1.99483, "tr": 2.39989, "fi": 2.595, "ru": 1.50381, "el": 1.60244, "hi": 1.30516, "ta": 1.06572, "th": 1.22157, "zh": 3.32786, "my": 1.20396}, "bytes_per_chunk": {"en": 5.736, "de": 6.253, "vi": 5.762, "tr": 5.094, "fi": 6.374, "ru": 8.008, "el": 8.805, "hi": 16.163, "ta": 15.708, "th": 13.012, "zh": 5.698, "my": 16.716}, "bpb_mean": 1.89916, "t": 182.3} |
| {"event": "train", "step": 3000, "phase": 2, "objective": "parity_b", "lm_loss": 1.48796, "bpb": 2.14668, "chunk_loss": 0.0, "F": 0.15381, "G": 0.20144, "chunks": 157.5, "bytes_per_chunk": 6.502, "boundary_entropy": 0.21492, "router_noise": 0.0, "grad_norm": 0.7183, "lr": 0.0001761, "router_trainable": true, "sec_per_step": 0.088, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 182.3} |
| {"event": "train", "step": 3050, "phase": 2, "objective": "parity_b", "lm_loss": 1.39036, "bpb": 2.00586, "chunk_loss": 0.030474, "F": 0.14871, "G": 0.18749, "chunks": 152.28, "bytes_per_chunk": 6.724, "boundary_entropy": 0.17787, "router_noise": 0.0, "grad_norm": 0.6712, "lr": 0.0001724, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0625, "fallback": 0.0, "alpha": 0.03, "t": 184.5} |
| {"event": "train", "step": 3100, "phase": 2, "objective": "parity_b", "lm_loss": 1.42967, "bpb": 2.06257, "chunk_loss": 0.030365, "F": 0.15784, "G": 0.18452, "chunks": 161.62, "bytes_per_chunk": 6.336, "boundary_entropy": 0.15676, "router_noise": 0.0, "grad_norm": 0.6989, "lr": 0.0001687, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 186.7} |
| {"event": "train", "step": 3150, "phase": 2, "objective": "parity_b", "lm_loss": 1.38064, "bpb": 1.99184, "chunk_loss": 0.030735, "F": 0.14316, "G": 0.18395, "chunks": 146.59, "bytes_per_chunk": 6.985, "boundary_entropy": 0.19768, "router_noise": 0.0, "grad_norm": 0.727, "lr": 0.000165, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1562, "fallback": 0.0, "alpha": 0.03, "t": 188.8} |
| {"event": "probe", "step": 3200, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2883.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165661, "cps": 22.523438, "mean_p": 0.189699, "boundary_entropy": 0.11904, "bpb": 2.149396}, "de": {"chunks": 3301.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160766, "cps": 25.789062, "mean_p": 0.197003, "boundary_entropy": 0.173032, "bpb": 2.376002}, "vi": {"chunks": 4168.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226092, "cps": 32.5625, "mean_p": 0.173928, "boundary_entropy": 0.156822, "bpb": 1.984835}, "tr": {"chunks": 3845.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.212396, "cps": 30.039062, "mean_p": 0.212792, "boundary_entropy": 0.181784, "bpb": 2.382649}, "fi": {"chunks": 2967.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157727, "cps": 23.179688, "mean_p": 0.196665, "boundary_entropy": 0.195017, "bpb": 2.569144}, "ru": {"chunks": 4459.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.235303, "cps": 34.835938, "mean_p": 0.127413, "boundary_entropy": 0.125787, "bpb": 1.492923}, "el": {"chunks": 4176.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.203479, "cps": 32.625, "mean_p": 0.126646, "boundary_entropy": 0.170555, "bpb": 1.579363}, "hi": {"chunks": 2786.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.165587, "cps": 21.765625, "mean_p": 0.121181, "boundary_entropy": 0.256222, "bpb": 1.283983}, "ta": {"chunks": 2922.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.156298, "cps": 22.828125, "mean_p": 0.098786, "boundary_entropy": 0.210715, "bpb": 1.057593}, "th": {"chunks": 3589.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.223127, "cps": 28.039062, "mean_p": 0.208139, "boundary_entropy": 0.379681, "bpb": 1.190376}, "zh": {"chunks": 3006.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.508974, "cps": 23.484375, "mean_p": 0.263771, "boundary_entropy": 0.374355, "bpb": 3.30255}, "my": {"chunks": 1163.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.061379, "cps": 9.085938, "mean_p": 0.075194, "boundary_entropy": 0.211912, "bpb": 1.195068}}, "bpb": {"en": 2.1494, "de": 2.376, "vi": 1.98483, "tr": 2.38265, "fi": 2.56914, "ru": 1.49292, "el": 1.57936, "hi": 1.28398, "ta": 1.05759, "th": 1.19038, "zh": 3.30255, "my": 1.19507}, "gini_cpc": 0.222169, "gini_cps": 0.136969, "bpic": 9.774125, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9704487276718531, "vi": 1.3647843888319255, "tr": 1.2821099497539314, "fi": 0.9521056463354101, "ru": 1.420390424442437, "el": 1.2282849277045227, "hi": 0.9995522865761735, "ta": 0.9434833054518018, "th": 1.3468892010673008, "zh": 3.0723805805345132, "my": 0.37050648722161333}, "t": 191.9} |
| {"event": "eval", "step": 3200, "phase": 2, "bpb": {"en": 2.15989, "de": 2.38537, "vi": 1.98885, "tr": 2.38665, "fi": 2.57657, "ru": 1.47276, "el": 1.56935, "hi": 1.28057, "ta": 1.06364, "th": 1.20758, "zh": 3.3028, "my": 1.20368}, "bytes_per_chunk": {"en": 5.738, "de": 6.266, "vi": 5.796, "tr": 5.098, "fi": 6.405, "ru": 8.086, "el": 9.129, "hi": 11.088, "ta": 14.558, "th": 13.093, "zh": 5.845, "my": 34.049}, "bpb_mean": 1.88314, "t": 193.2} |
| {"event": "train", "step": 3200, "phase": 2, "objective": "parity_b", "lm_loss": 1.32959, "bpb": 1.9182, "chunk_loss": 0.030849, "F": 0.14853, "G": 0.17555, "chunks": 152.09, "bytes_per_chunk": 6.733, "boundary_entropy": 0.1676, "router_noise": 0.0, "grad_norm": 0.7268, "lr": 0.0001613, "router_trainable": true, "sec_per_step": 0.088, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 193.2} |
| {"event": "train", "step": 3250, "phase": 2, "objective": "parity_b", "lm_loss": 1.5581, "bpb": 2.24787, "chunk_loss": 0.0, "F": 0.15793, "G": 0.19424, "chunks": 161.72, "bytes_per_chunk": 6.332, "boundary_entropy": 0.18108, "router_noise": 0.0, "grad_norm": 0.7097, "lr": 0.0001576, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "hi", "n_target_mean": 15.3104, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 195.4} |
| {"event": "train", "step": 3300, "phase": 2, "objective": "parity_b", "lm_loss": 1.5175, "bpb": 2.18928, "chunk_loss": 0.030595, "F": 0.16193, "G": 0.19131, "chunks": 165.81, "bytes_per_chunk": 6.176, "boundary_entropy": 0.1598, "router_noise": 0.0, "grad_norm": 0.7183, "lr": 0.0001539, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 197.5} |
| {"event": "train", "step": 3350, "phase": 2, "objective": "parity_b", "lm_loss": 1.34751, "bpb": 1.94405, "chunk_loss": 0.030498, "F": 0.142, "G": 0.17205, "chunks": 145.41, "bytes_per_chunk": 7.042, "boundary_entropy": 0.19037, "router_noise": 0.0, "grad_norm": 0.6172, "lr": 0.0001501, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.25, "fallback": 0.0, "alpha": 0.03, "t": 199.7} |
| {"event": "probe", "step": 3400, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2885.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165776, "cps": 22.539062, "mean_p": 0.187541, "boundary_entropy": 0.113829, "bpb": 2.123867}, "de": {"chunks": 3284.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159938, "cps": 25.65625, "mean_p": 0.18858, "boundary_entropy": 0.152402, "bpb": 2.353068}, "vi": {"chunks": 4154.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225332, "cps": 32.453125, "mean_p": 0.146555, "boundary_entropy": 0.166586, "bpb": 1.960237}, "tr": {"chunks": 3870.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213777, "cps": 30.234375, "mean_p": 0.212026, "boundary_entropy": 0.17494, "bpb": 2.360343}, "fi": {"chunks": 2949.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15677, "cps": 23.039062, "mean_p": 0.191248, "boundary_entropy": 0.18058, "bpb": 2.549226}, "ru": {"chunks": 4448.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.234723, "cps": 34.75, "mean_p": 0.12783, "boundary_entropy": 0.114521, "bpb": 1.470627}, "el": {"chunks": 4093.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.199435, "cps": 31.976562, "mean_p": 0.126654, "boundary_entropy": 0.166995, "bpb": 1.564763}, "hi": {"chunks": 1475.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.087667, "cps": 11.523438, "mean_p": 0.100506, "boundary_entropy": 0.237541, "bpb": 1.27933}, "ta": {"chunks": 2387.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.127681, "cps": 18.648438, "mean_p": 0.09659, "boundary_entropy": 0.211868, "bpb": 1.045074}, "th": {"chunks": 3463.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.215294, "cps": 27.054688, "mean_p": 0.182117, "boundary_entropy": 0.345043, "bpb": 1.174812}, "zh": {"chunks": 3244.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.549272, "cps": 25.34375, "mean_p": 0.281977, "boundary_entropy": 0.385009, "bpb": 3.278084}, "my": {"chunks": 2764.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.145873, "cps": 21.59375, "mean_p": 0.093796, "boundary_entropy": 0.224327, "bpb": 1.170406}}, "bpb": {"en": 2.12387, "de": 2.35307, "vi": 1.96024, "tr": 2.36034, "fi": 2.54923, "ru": 1.47063, "el": 1.56476, "hi": 1.27933, "ta": 1.04507, "th": 1.17481, "zh": 3.27808, "my": 1.17041}, "gini_cpc": 0.237925, "gini_cps": 0.137585, "bpic": 9.836503, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9647816707281281, "vi": 1.3592572338334274, "tr": 1.2895515752293991, "fi": 0.9456734402701616, "ru": 1.4159041900833627, "el": 1.2030376030747427, "hi": 0.5288289842681698, "ta": 0.7702030821316185, "th": 1.2987026568667643, "zh": 3.3133377272239075, "my": 0.879939807898951}, "t": 202.8} |
| {"event": "eval", "step": 3400, "phase": 2, "bpb": {"en": 2.13399, "de": 2.36514, "vi": 1.96345, "tr": 2.3652, "fi": 2.5568, "ru": 1.45098, "el": 1.55535, "hi": 1.27683, "ta": 1.05253, "th": 1.19099, "zh": 3.27612, "my": 1.17854}, "bytes_per_chunk": {"en": 5.736, "de": 6.283, "vi": 5.798, "tr": 5.093, "fi": 6.43, "ru": 8.087, "el": 9.309, "hi": 22.047, "ta": 19.658, "th": 12.942, "zh": 5.516, "my": 18.297}, "bpb_mean": 1.86383, "t": 204.1} |
| {"event": "train", "step": 3400, "phase": 2, "objective": "parity_b", "lm_loss": 1.37511, "bpb": 1.98386, "chunk_loss": 0.03053, "F": 0.14243, "G": 0.1778, "chunks": 145.84, "bytes_per_chunk": 7.021, "boundary_entropy": 0.17753, "router_noise": 0.0, "grad_norm": 0.6869, "lr": 0.0001465, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 204.1} |
| {"event": "train", "step": 3450, "phase": 2, "objective": "parity_b", "lm_loss": 1.3767, "bpb": 1.98616, "chunk_loss": 0.030348, "F": 0.14395, "G": 0.18027, "chunks": 147.41, "bytes_per_chunk": 6.947, "boundary_entropy": 0.17928, "router_noise": 0.0, "grad_norm": 0.764, "lr": 0.0001428, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 206.3} |
| {"event": "train", "step": 3500, "phase": 2, "objective": "parity_b", "lm_loss": 1.45458, "bpb": 2.09852, "chunk_loss": 0.03054, "F": 0.14584, "G": 0.17344, "chunks": 149.34, "bytes_per_chunk": 6.857, "boundary_entropy": 0.17815, "router_noise": 0.0, "grad_norm": 0.705, "lr": 0.0001391, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 208.4} |
| {"event": "train", "step": 3550, "phase": 2, "objective": "parity_b", "lm_loss": 1.4012, "bpb": 2.0215, "chunk_loss": 0.030394, "F": 0.14954, "G": 0.18153, "chunks": 153.12, "bytes_per_chunk": 6.687, "boundary_entropy": 0.16521, "router_noise": 0.0, "grad_norm": 0.6906, "lr": 0.0001355, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 210.6} |
| {"event": "probe", "step": 3600, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2883.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165661, "cps": 22.523438, "mean_p": 0.186027, "boundary_entropy": 0.103771, "bpb": 2.109896}, "de": {"chunks": 3283.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159889, "cps": 25.648438, "mean_p": 0.188024, "boundary_entropy": 0.144411, "bpb": 2.336155}, "vi": {"chunks": 4173.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226363, "cps": 32.601562, "mean_p": 0.156811, "boundary_entropy": 0.143067, "bpb": 1.950569}, "tr": {"chunks": 3856.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213003, "cps": 30.125, "mean_p": 0.21206, "boundary_entropy": 0.174196, "bpb": 2.335416}, "fi": {"chunks": 2956.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157142, "cps": 23.09375, "mean_p": 0.188791, "boundary_entropy": 0.16986, "bpb": 2.518523}, "ru": {"chunks": 4337.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.228865, "cps": 33.882812, "mean_p": 0.122713, "boundary_entropy": 0.110433, "bpb": 1.469693}, "el": {"chunks": 4125.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.200994, "cps": 32.226562, "mean_p": 0.126298, "boundary_entropy": 0.160534, "bpb": 1.542255}, "hi": {"chunks": 1878.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.11162, "cps": 14.671875, "mean_p": 0.106469, "boundary_entropy": 0.241376, "bpb": 1.264218}, "ta": {"chunks": 1892.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.101204, "cps": 14.78125, "mean_p": 0.079481, "boundary_entropy": 0.186043, "bpb": 1.02931}, "th": {"chunks": 3477.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.216164, "cps": 27.164062, "mean_p": 0.182428, "boundary_entropy": 0.339356, "bpb": 1.152073}, "zh": {"chunks": 3323.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.562648, "cps": 25.960938, "mean_p": 0.279818, "boundary_entropy": 0.369056, "bpb": 3.253451}, "my": {"chunks": 3463.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.182763, "cps": 27.054688, "mean_p": 0.108178, "boundary_entropy": 0.238144, "bpb": 1.161708}}, "bpb": {"en": 2.1099, "de": 2.33615, "vi": 1.95057, "tr": 2.33542, "fi": 2.51852, "ru": 1.46969, "el": 1.54226, "hi": 1.26422, "ta": 1.02931, "th": 1.15207, "zh": 3.25345, "my": 1.16171}, "gini_cpc": 0.232792, "gini_cps": 0.129256, "bpic": 9.680195, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9651569745370172, "vi": 1.3664216061889694, "tr": 1.2857778845906787, "fi": 0.9485757635886324, "ru": 1.381527981791175, "el": 1.2132843215472116, "hi": 0.6737829124874566, "ta": 0.6109070547278608, "th": 1.3048575514380063, "zh": 3.3963807947824978, "my": 1.1032364275567044}, "t": 213.7} |
| {"event": "eval", "step": 3600, "phase": 2, "bpb": {"en": 2.12051, "de": 2.34508, "vi": 1.95279, "tr": 2.34298, "fi": 2.52661, "ru": 1.45017, "el": 1.53271, "hi": 1.26171, "ta": 1.03602, "th": 1.16986, "zh": 3.2523, "my": 1.16989}, "bytes_per_chunk": {"en": 5.736, "de": 6.282, "vi": 5.791, "tr": 5.073, "fi": 6.403, "ru": 8.319, "el": 9.324, "hi": 14.722, "ta": 24.614, "th": 13.143, "zh": 5.413, "my": 14.942}, "bpb_mean": 1.84672, "t": 215.0} |
| {"event": "train", "step": 3600, "phase": 2, "objective": "parity_b", "lm_loss": 1.48033, "bpb": 2.13566, "chunk_loss": 0.030817, "F": 0.15796, "G": 0.1799, "chunks": 161.75, "bytes_per_chunk": 6.331, "boundary_entropy": 0.16378, "router_noise": 0.0, "grad_norm": 1.4183, "lr": 0.0001319, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 215.1} |
| {"event": "train", "step": 3650, "phase": 2, "objective": "parity_b", "lm_loss": 1.33746, "bpb": 1.92955, "chunk_loss": 0.030761, "F": 0.15201, "G": 0.16834, "chunks": 155.66, "bytes_per_chunk": 6.579, "boundary_entropy": 0.14364, "router_noise": 0.0, "grad_norm": 0.6991, "lr": 0.0001283, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.25, "fallback": 0.0, "alpha": 0.03, "t": 217.2} |
| {"event": "train", "step": 3700, "phase": 2, "objective": "parity_b", "lm_loss": 1.58667, "bpb": 2.28909, "chunk_loss": 0.0, "F": 0.16647, "G": 0.19949, "chunks": 170.47, "bytes_per_chunk": 6.007, "boundary_entropy": 0.18117, "router_noise": 0.0, "grad_norm": 0.7591, "lr": 0.0001247, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 219.4} |
| {"event": "train", "step": 3750, "phase": 2, "objective": "parity_b", "lm_loss": 1.31733, "bpb": 1.90051, "chunk_loss": 0.030445, "F": 0.13898, "G": 0.16481, "chunks": 142.31, "bytes_per_chunk": 7.195, "boundary_entropy": 0.15885, "router_noise": 0.0, "grad_norm": 0.7104, "lr": 0.0001212, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.0938, "fallback": 0.0, "alpha": 0.03, "t": 221.6} |
| {"event": "probe", "step": 3800, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2883.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165661, "cps": 22.523438, "mean_p": 0.185112, "boundary_entropy": 0.099952, "bpb": 2.091763}, "de": {"chunks": 3284.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159938, "cps": 25.65625, "mean_p": 0.187028, "boundary_entropy": 0.140305, "bpb": 2.314454}, "vi": {"chunks": 4159.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225603, "cps": 32.492188, "mean_p": 0.154237, "boundary_entropy": 0.137962, "bpb": 1.939326}, "tr": {"chunks": 3864.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213445, "cps": 30.1875, "mean_p": 0.211612, "boundary_entropy": 0.169609, "bpb": 2.317446}, "fi": {"chunks": 2990.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15895, "cps": 23.359375, "mean_p": 0.192044, "boundary_entropy": 0.175896, "bpb": 2.518392}, "ru": {"chunks": 4416.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.233034, "cps": 34.5, "mean_p": 0.126855, "boundary_entropy": 0.103058, "bpb": 1.440069}, "el": {"chunks": 3859.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.188033, "cps": 30.148438, "mean_p": 0.116791, "boundary_entropy": 0.153197, "bpb": 1.525561}, "hi": {"chunks": 4110.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.244279, "cps": 32.109375, "mean_p": 0.129034, "boundary_entropy": 0.246506, "bpb": 1.243661}, "ta": {"chunks": 3274.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.175127, "cps": 25.578125, "mean_p": 0.093185, "boundary_entropy": 0.188024, "bpb": 1.018955}, "th": {"chunks": 3725.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.231582, "cps": 29.101562, "mean_p": 0.195476, "boundary_entropy": 0.350084, "bpb": 1.134884}, "zh": {"chunks": 3178.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.538097, "cps": 24.828125, "mean_p": 0.275396, "boundary_entropy": 0.365176, "bpb": 3.24725}, "my": {"chunks": 2949.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.155636, "cps": 23.039062, "mean_p": 0.098094, "boundary_entropy": 0.22285, "bpb": 1.149827}}, "bpb": {"en": 2.09176, "de": 2.31445, "vi": 1.93933, "tr": 2.31745, "fi": 2.51839, "ru": 1.44007, "el": 1.52556, "hi": 1.24366, "ta": 1.01895, "th": 1.13488, "zh": 3.24725, "my": 1.14983}, "gini_cpc": 0.185704, "gini_cps": 0.08046, "bpic": 8.98974, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9654509608222858, "vi": 1.361837397589246, "tr": 1.2884454735628585, "fi": 0.9594863102604909, "ru": 1.4066930061309266, "el": 1.1350458659031974, "hi": 1.4745728276482675, "ta": 1.057140431912799, "th": 1.3979276327600154, "zh": 3.2481788040381514, "my": 0.9394872148035578}, "t": 224.7} |
| {"event": "eval", "step": 3800, "phase": 2, "bpb": {"en": 2.10326, "de": 2.32432, "vi": 1.94156, "tr": 2.32388, "fi": 2.52676, "ru": 1.41922, "el": 1.51534, "hi": 1.24059, "ta": 1.02815, "th": 1.15277, "zh": 3.24578, "my": 1.15778}, "bytes_per_chunk": {"en": 5.738, "de": 6.28, "vi": 5.791, "tr": 5.09, "fi": 6.34, "ru": 8.142, "el": 9.955, "hi": 9.584, "ta": 13.888, "th": 12.371, "zh": 5.592, "my": 16.573}, "bpb_mean": 1.83162, "t": 226.0} |
| {"event": "train", "step": 3800, "phase": 2, "objective": "parity_b", "lm_loss": 1.37922, "bpb": 1.98979, "chunk_loss": 0.030396, "F": 0.15433, "G": 0.17637, "chunks": 158.03, "bytes_per_chunk": 6.48, "boundary_entropy": 0.14014, "router_noise": 0.0, "grad_norm": 0.7051, "lr": 0.0001177, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0312, "fallback": 0.0, "alpha": 0.03, "t": 226.0} |
| {"event": "train", "step": 3850, "phase": 2, "objective": "parity_b", "lm_loss": 1.48678, "bpb": 2.14497, "chunk_loss": 0.030556, "F": 0.15518, "G": 0.19178, "chunks": 158.91, "bytes_per_chunk": 6.444, "boundary_entropy": 0.19925, "router_noise": 0.0, "grad_norm": 0.6896, "lr": 0.0001142, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 228.2} |
| {"event": "train", "step": 3900, "phase": 2, "objective": "parity_b", "lm_loss": 1.49547, "bpb": 2.1575, "chunk_loss": 0.0, "F": 0.14987, "G": 0.18436, "chunks": 153.47, "bytes_per_chunk": 6.672, "boundary_entropy": 0.1853, "router_noise": 0.0, "grad_norm": 0.7025, "lr": 0.0001108, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "el", "n_target_mean": 13.046, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 230.3} |
| {"event": "train", "step": 3950, "phase": 2, "objective": "parity_b", "lm_loss": 1.43556, "bpb": 2.07108, "chunk_loss": 0.0, "F": 0.14597, "G": 0.17724, "chunks": 149.47, "bytes_per_chunk": 6.851, "boundary_entropy": 0.16598, "router_noise": 0.0, "grad_norm": 0.7276, "lr": 0.0001074, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "vi", "n_target_mean": 8.3396, "frac_in_batch": 0.0, "fallback": 0.0, "alpha": 0.03, "t": 232.5} |
| {"event": "probe", "step": 4000, "phase": 2, "objective": "parity_b", "router_trainable": true, "per_lang": {"en": {"chunks": 2883.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165661, "cps": 22.523438, "mean_p": 0.183583, "boundary_entropy": 0.095226, "bpb": 2.077356}, "de": {"chunks": 3285.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.159986, "cps": 25.664062, "mean_p": 0.186068, "boundary_entropy": 0.137014, "bpb": 2.299596}, "vi": {"chunks": 4180.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226743, "cps": 32.65625, "mean_p": 0.154038, "boundary_entropy": 0.134599, "bpb": 1.926246}, "tr": {"chunks": 3884.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.21455, "cps": 30.34375, "mean_p": 0.210182, "boundary_entropy": 0.160884, "bpb": 2.293307}, "fi": {"chunks": 2952.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.156929, "cps": 23.0625, "mean_p": 0.18299, "boundary_entropy": 0.15311, "bpb": 2.486304}, "ru": {"chunks": 4260.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.224802, "cps": 33.28125, "mean_p": 0.118747, "boundary_entropy": 0.099985, "bpb": 1.433674}, "el": {"chunks": 3896.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.189836, "cps": 30.4375, "mean_p": 0.119577, "boundary_entropy": 0.15175, "bpb": 1.508192}, "hi": {"chunks": 2567.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.152571, "cps": 20.054688, "mean_p": 0.11042, "boundary_entropy": 0.227585, "bpb": 1.245211}, "ta": {"chunks": 3252.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.17395, "cps": 25.40625, "mean_p": 0.099143, "boundary_entropy": 0.194506, "bpb": 1.017161}, "th": {"chunks": 3889.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.241778, "cps": 30.382812, "mean_p": 0.203787, "boundary_entropy": 0.352649, "bpb": 1.131488}, "zh": {"chunks": 3328.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.563495, "cps": 26.0, "mean_p": 0.282014, "boundary_entropy": 0.356408, "bpb": 3.212256}, "my": {"chunks": 3264.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.172261, "cps": 25.5, "mean_p": 0.099037, "boundary_entropy": 0.221441, "bpb": 1.146514}}, "bpb": {"en": 2.07736, "de": 2.2996, "vi": 1.92625, "tr": 2.29331, "fi": 2.4863, "ru": 1.43367, "el": 1.50819, "hi": 1.24521, "ta": 1.01716, "th": 1.13149, "zh": 3.21226, "my": 1.14651}, "gini_cpc": 0.195188, "gini_cps": 0.084038, "bpic": 9.216643, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9657449471075545, "vi": 1.3687137104888312, "tr": 1.2951144459933082, "fi": 0.9472921698625314, "ru": 1.3570000466752146, "el": 1.1459286586055601, "hi": 0.9209801578036747, "ta": 1.050036861508987, "th": 1.4594739768600538, "zh": 3.4014912082564406, "my": 1.0398393588059724}, "t": 235.6} |
| {"event": "eval", "step": 4000, "phase": 2, "bpb": {"en": 2.09032, "de": 2.30904, "vi": 1.928, "tr": 2.29913, "fi": 2.49555, "ru": 1.41205, "el": 1.49918, "hi": 1.24213, "ta": 1.02434, "th": 1.14882, "zh": 3.21009, "my": 1.15451}, "bytes_per_chunk": {"en": 5.738, "de": 6.272, "vi": 5.786, "tr": 5.059, "fi": 6.415, "ru": 8.492, "el": 9.816, "hi": 11.222, "ta": 13.783, "th": 11.957, "zh": 5.325, "my": 15.81}, "bpb_mean": 1.81776, "t": 236.9} |
| {"event": "train", "step": 4000, "phase": 2, "objective": "parity_b", "lm_loss": 1.41496, "bpb": 2.04135, "chunk_loss": 0.030366, "F": 0.15216, "G": 0.18633, "chunks": 155.81, "bytes_per_chunk": 6.572, "boundary_entropy": 0.17931, "router_noise": 0.0, "grad_norm": 0.7666, "lr": 0.000104, "router_trainable": true, "sec_per_step": 0.089, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1875, "fallback": 0.0, "alpha": 0.03, "t": 237.0} |
| {"event": "train", "step": 4050, "phase": 2, "objective": "parity_b", "lm_loss": 1.30848, "bpb": 1.88774, "chunk_loss": 0.030486, "F": 0.14374, "G": 0.16801, "chunks": 147.19, "bytes_per_chunk": 6.957, "boundary_entropy": 0.15174, "router_noise": 0.0, "grad_norm": 0.7006, "lr": 0.0001007, "router_trainable": true, "sec_per_step": 0.044, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.1875, "fallback": 0.0, "alpha": 0.03, "t": 239.1} |
| {"event": "train", "step": 4100, "phase": 2, "objective": "parity_b", "lm_loss": 1.45281, "bpb": 2.09596, "chunk_loss": 0.030451, "F": 0.15707, "G": 0.18317, "chunks": 160.84, "bytes_per_chunk": 6.366, "boundary_entropy": 0.16309, "router_noise": 0.0, "grad_norm": 0.7053, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.125, "fallback": 0.0, "alpha": 0.03, "t": 241.3} |
| {"event": "train", "step": 4150, "phase": 2, "objective": "parity_b", "lm_loss": 1.25191, "bpb": 1.80613, "chunk_loss": 0.030415, "F": 0.1377, "G": 0.16656, "chunks": 141.0, "bytes_per_chunk": 7.262, "boundary_entropy": 0.15969, "router_noise": 0.0, "grad_norm": 0.6864, "lr": 9.43e-05, "router_trainable": true, "sec_per_step": 0.043, "worst_lang": "ru", "n_target_mean": 11.8251, "frac_in_batch": 0.2188, "fallback": 0.0, "alpha": 0.03, "t": 243.5} |
| {"event": "probe", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2884.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165719, "cps": 22.53125, "mean_p": 0.183934, "boundary_entropy": 0.094976, "bpb": 2.063697}, "de": {"chunks": 3287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160084, "cps": 25.679688, "mean_p": 0.185372, "boundary_entropy": 0.13396, "bpb": 2.288246}, "vi": {"chunks": 4160.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.225658, "cps": 32.5, "mean_p": 0.151007, "boundary_entropy": 0.135373, "bpb": 1.919393}, "tr": {"chunks": 3887.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.214716, "cps": 30.367188, "mean_p": 0.211087, "boundary_entropy": 0.161503, "bpb": 2.286247}, "fi": {"chunks": 2947.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.156664, "cps": 23.023438, "mean_p": 0.182762, "boundary_entropy": 0.15478, "bpb": 2.465539}, "ru": {"chunks": 4297.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.226755, "cps": 33.570312, "mean_p": 0.120584, "boundary_entropy": 0.098486, "bpb": 1.414066}, "el": {"chunks": 3906.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.190323, "cps": 30.515625, "mean_p": 0.119481, "boundary_entropy": 0.151098, "bpb": 1.494107}, "hi": {"chunks": 1701.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.1011, "cps": 13.289062, "mean_p": 0.102095, "boundary_entropy": 0.231062, "bpb": 1.236097}, "ta": {"chunks": 2040.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.10912, "cps": 15.9375, "mean_p": 0.078471, "boundary_entropy": 0.177592, "bpb": 1.012147}, "th": {"chunks": 3495.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.217283, "cps": 27.304688, "mean_p": 0.184311, "boundary_entropy": 0.336911, "bpb": 1.125115}, "zh": {"chunks": 3282.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.555706, "cps": 25.640625, "mean_p": 0.277739, "boundary_entropy": 0.354192, "bpb": 3.200451}, "my": {"chunks": 2976.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.157061, "cps": 23.25, "mean_p": 0.09182, "boundary_entropy": 0.204664, "bpb": 1.135411}}, "bpb": {"en": 2.0637, "de": 2.28825, "vi": 1.91939, "tr": 2.28625, "fi": 2.46554, "ru": 1.41407, "el": 1.49411, "hi": 1.2361, "ta": 1.01215, "th": 1.12511, "zh": 3.20045, "my": 1.13541}, "gini_cpc": 0.236271, "gini_cps": 0.130916, "bpic": 9.875482, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9659978527849996, "vi": 1.3616925231546007, "tr": 1.295665376188022, "fi": 0.9453597693561865, "ru": 1.368311583516005, "el": 1.1484715940297152, "hi": 0.610067802478397, "ta": 0.6584663139476352, "th": 1.311157848582664, "zh": 3.353312271354315, "my": 0.9477600862218138}, "t": 246.6} |
| {"event": "eval", "step": 4200, "phase": 2, "bpb": {"en": 2.07725, "de": 2.2986, "vi": 1.92267, "tr": 2.29101, "fi": 2.47559, "ru": 1.39428, "el": 1.4846, "hi": 1.2343, "ta": 1.01856, "th": 1.14159, "zh": 3.19908, "my": 1.14469}, "bytes_per_chunk": {"en": 5.738, "de": 6.274, "vi": 5.798, "tr": 5.055, "fi": 6.422, "ru": 8.461, "el": 9.8, "hi": 17.912, "ta": 22.675, "th": 13.151, "zh": 5.39, "my": 16.622}, "bpb_mean": 1.80685, "t": 247.9} |
| {"event": "phase_transition", "step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584, "t": 247.9} |
| {"event": "train", "step": 4200, "phase": 3, "objective": "none", "lm_loss": 1.26703, "bpb": 1.82794, "chunk_loss": 0.0, "F": 0.14066, "G": 0.16573, "chunks": 144.03, "bytes_per_chunk": 7.11, "boundary_entropy": 0.13989, "router_noise": 0.0, "grad_norm": 0.713, "lr": 9.12e-05, "router_trainable": false, "sec_per_step": 0.089, "alpha": 0.0, "t": 247.9} |
| {"event": "train", "step": 4250, "phase": 3, "objective": "none", "lm_loss": 1.29071, "bpb": 1.8621, "chunk_loss": 0.0, "F": 0.14594, "G": 0.16067, "chunks": 149.44, "bytes_per_chunk": 6.852, "boundary_entropy": 0.12832, "router_noise": 0.0, "grad_norm": 0.7228, "lr": 8.81e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 250.0} |
| {"event": "train", "step": 4300, "phase": 3, "objective": "none", "lm_loss": 1.36565, "bpb": 1.97022, "chunk_loss": 0.0, "F": 0.14566, "G": 0.17608, "chunks": 149.16, "bytes_per_chunk": 6.865, "boundary_entropy": 0.16273, "router_noise": 0.0, "grad_norm": 0.7301, "lr": 8.5e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 252.1} |
| {"event": "train", "step": 4350, "phase": 3, "objective": "none", "lm_loss": 1.33211, "bpb": 1.92184, "chunk_loss": 0.0, "F": 0.14691, "G": 0.16973, "chunks": 150.44, "bytes_per_chunk": 6.807, "boundary_entropy": 0.14834, "router_noise": 0.0, "grad_norm": 0.715, "lr": 8.21e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 254.2} |
| {"event": "probe", "step": 4400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2884.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165719, "cps": 22.53125, "mean_p": 0.184804, "boundary_entropy": 0.096813, "bpb": 2.054077}, "de": {"chunks": 3287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160084, "cps": 25.679688, "mean_p": 0.186697, "boundary_entropy": 0.136692, "bpb": 2.273728}, "vi": {"chunks": 4188.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227177, "cps": 32.71875, "mean_p": 0.165462, "boundary_entropy": 0.12589, "bpb": 1.910008}, "tr": {"chunks": 3914.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216207, "cps": 30.578125, "mean_p": 0.213458, "boundary_entropy": 0.163033, "bpb": 2.262875}, "fi": {"chunks": 2963.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157514, "cps": 23.148438, "mean_p": 0.18535, "boundary_entropy": 0.159236, "bpb": 2.443037}, "ru": {"chunks": 4490.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.236939, "cps": 35.078125, "mean_p": 0.131719, "boundary_entropy": 0.099589, "bpb": 1.396276}, "el": {"chunks": 4167.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.20304, "cps": 32.554688, "mean_p": 0.128449, "boundary_entropy": 0.155185, "bpb": 1.482384}, "hi": {"chunks": 3468.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.206122, "cps": 27.09375, "mean_p": 0.116226, "boundary_entropy": 0.224123, "bpb": 1.218974}, "ta": {"chunks": 3296.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.176304, "cps": 25.75, "mean_p": 0.093601, "boundary_entropy": 0.184815, "bpb": 0.994792}, "th": {"chunks": 3897.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.242275, "cps": 30.445312, "mean_p": 0.191777, "boundary_entropy": 0.338259, "bpb": 1.107304}, "zh": {"chunks": 3483.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.589739, "cps": 27.210938, "mean_p": 0.284933, "boundary_entropy": 0.352055, "bpb": 3.185184}, "my": {"chunks": 3232.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.170572, "cps": 25.25, "mean_p": 0.103627, "boundary_entropy": 0.203997, "bpb": 1.125055}}, "bpb": {"en": 2.05408, "de": 2.27373, "vi": 1.91001, "tr": 2.26287, "fi": 2.44304, "ru": 1.39628, "el": 1.48238, "hi": 1.21897, "ta": 0.99479, "th": 1.1073, "zh": 3.18518, "my": 1.12505}, "gini_cpc": 0.193776, "gini_cps": 0.077401, "bpic": 8.869653, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9659978527849996, "vi": 1.370857761291218, "tr": 1.3046653672240591, "fi": 0.9504923639641603, "ru": 1.4297693763059955, "el": 1.2252127834925304, "hi": 1.2438066660758853, "ta": 1.0638749856722578, "th": 1.4619691376041892, "zh": 3.5586796590880803, "my": 1.029287835574228}, "t": 257.3} |
| {"event": "eval", "step": 4400, "phase": 3, "bpb": {"en": 2.06728, "de": 2.28366, "vi": 1.91193, "tr": 2.26866, "fi": 2.45379, "ru": 1.37529, "el": 1.47397, "hi": 1.21485, "ta": 1.0019, "th": 1.12363, "zh": 3.1826, "my": 1.13307}, "bytes_per_chunk": {"en": 5.736, "de": 6.271, "vi": 5.78, "tr": 5.034, "fi": 6.394, "ru": 7.997, "el": 9.084, "hi": 9.938, "ta": 13.187, "th": 11.842, "zh": 5.099, "my": 15.157}, "bpb_mean": 1.79089, "t": 258.6} |
| {"event": "train", "step": 4400, "phase": 3, "objective": "none", "lm_loss": 1.33453, "bpb": 1.92532, "chunk_loss": 0.0, "F": 0.14908, "G": 0.17638, "chunks": 152.66, "bytes_per_chunk": 6.708, "boundary_entropy": 0.16364, "router_noise": 0.0, "grad_norm": 0.7056, "lr": 7.92e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 258.6} |
| {"event": "train", "step": 4450, "phase": 3, "objective": "none", "lm_loss": 1.17549, "bpb": 1.69587, "chunk_loss": 0.0, "F": 0.14023, "G": 0.16107, "chunks": 143.59, "bytes_per_chunk": 7.131, "boundary_entropy": 0.14454, "router_noise": 0.0, "grad_norm": 0.8342, "lr": 7.63e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 260.7} |
| {"event": "train", "step": 4500, "phase": 3, "objective": "none", "lm_loss": 1.30094, "bpb": 1.87685, "chunk_loss": 0.0, "F": 0.1467, "G": 0.16987, "chunks": 150.22, "bytes_per_chunk": 6.817, "boundary_entropy": 0.13901, "router_noise": 0.0, "grad_norm": 0.7314, "lr": 7.36e-05, "router_trainable": false, "sec_per_step": 0.043, "alpha": 0.0, "t": 262.8} |
| {"event": "train", "step": 4550, "phase": 3, "objective": "none", "lm_loss": 1.2995, "bpb": 1.87479, "chunk_loss": 0.0, "F": 0.15744, "G": 0.17657, "chunks": 161.22, "bytes_per_chunk": 6.352, "boundary_entropy": 0.12909, "router_noise": 0.0, "grad_norm": 0.6936, "lr": 7.09e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 264.9} |
| {"event": "probe", "step": 4600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2884.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165719, "cps": 22.53125, "mean_p": 0.184343, "boundary_entropy": 0.094848, "bpb": 2.041876}, "de": {"chunks": 3282.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15984, "cps": 25.640625, "mean_p": 0.183968, "boundary_entropy": 0.129318, "bpb": 2.269242}, "vi": {"chunks": 4183.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226905, "cps": 32.679688, "mean_p": 0.167691, "boundary_entropy": 0.115227, "bpb": 1.89862}, "tr": {"chunks": 3869.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213721, "cps": 30.226562, "mean_p": 0.209603, "boundary_entropy": 0.157705, "bpb": 2.25571}, "fi": {"chunks": 2946.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15661, "cps": 23.015625, "mean_p": 0.183239, "boundary_entropy": 0.155136, "bpb": 2.432221}, "ru": {"chunks": 4548.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.24, "cps": 35.53125, "mean_p": 0.13783, "boundary_entropy": 0.10321, "bpb": 1.383144}, "el": {"chunks": 4272.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.208157, "cps": 33.375, "mean_p": 0.131038, "boundary_entropy": 0.15512, "bpb": 1.463193}, "hi": {"chunks": 4189.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.248975, "cps": 32.726562, "mean_p": 0.128966, "boundary_entropy": 0.239588, "bpb": 1.208505}, "ta": {"chunks": 4082.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.218347, "cps": 31.890625, "mean_p": 0.100508, "boundary_entropy": 0.190379, "bpb": 0.989884}, "th": {"chunks": 3832.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.238234, "cps": 29.9375, "mean_p": 0.201842, "boundary_entropy": 0.350041, "bpb": 1.102697}, "zh": {"chunks": 3424.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.579749, "cps": 26.75, "mean_p": 0.284618, "boundary_entropy": 0.355134, "bpb": 3.179916}, "my": {"chunks": 3099.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.163553, "cps": 24.210938, "mean_p": 0.101895, "boundary_entropy": 0.203568, "bpb": 1.116774}}, "bpb": {"en": 2.04188, "de": 2.26924, "vi": 1.89862, "tr": 2.25571, "fi": 2.43222, "ru": 1.38314, "el": 1.46319, "hi": 1.20851, "ta": 0.98988, "th": 1.1027, "zh": 3.17992, "my": 1.11677}, "gini_cpc": 0.189591, "gini_cps": 0.082911, "bpic": 8.603026, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9645284310436169, "vi": 1.369221111623965, "tr": 1.2896653821639972, "fi": 0.9450389821931882, "ru": 1.4482385575589458, "el": 1.256085675805157, "hi": 1.5023950761798972, "ta": 1.317578183105023, "th": 1.4375842276877735, "zh": 3.4983976895542885, "my": 0.9869316220434816}, "t": 268.0} |
| {"event": "eval", "step": 4600, "phase": 3, "bpb": {"en": 2.05384, "de": 2.27762, "vi": 1.90063, "tr": 2.25901, "fi": 2.44295, "ru": 1.36271, "el": 1.45574, "hi": 1.20444, "ta": 0.99786, "th": 1.11924, "zh": 3.17624, "my": 1.12543}, "bytes_per_chunk": {"en": 5.735, "de": 6.281, "vi": 5.783, "tr": 5.082, "fi": 6.411, "ru": 7.85, "el": 8.882, "hi": 9.342, "ta": 11.801, "th": 11.981, "zh": 5.2, "my": 15.66}, "bpb_mean": 1.78131, "t": 269.3} |
| {"event": "train", "step": 4600, "phase": 3, "objective": "none", "lm_loss": 1.45406, "bpb": 2.09776, "chunk_loss": 0.0, "F": 0.16226, "G": 0.20401, "chunks": 166.16, "bytes_per_chunk": 6.163, "boundary_entropy": 0.19693, "router_noise": 0.0, "grad_norm": 0.6748, "lr": 6.82e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 269.3} |
| {"event": "train", "step": 4650, "phase": 3, "objective": "none", "lm_loss": 1.34331, "bpb": 1.93798, "chunk_loss": 0.0, "F": 0.15024, "G": 0.17742, "chunks": 153.84, "bytes_per_chunk": 6.656, "boundary_entropy": 0.16763, "router_noise": 0.0, "grad_norm": 0.7507, "lr": 6.57e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 271.4} |
| {"event": "train", "step": 4700, "phase": 3, "objective": "none", "lm_loss": 1.40219, "bpb": 2.02293, "chunk_loss": 0.0, "F": 0.16367, "G": 0.19451, "chunks": 167.59, "bytes_per_chunk": 6.11, "boundary_entropy": 0.16833, "router_noise": 0.0, "grad_norm": 0.7549, "lr": 6.32e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 273.5} |
| {"event": "train", "step": 4750, "phase": 3, "objective": "none", "lm_loss": 1.37674, "bpb": 1.98621, "chunk_loss": 0.0, "F": 0.15744, "G": 0.18529, "chunks": 161.22, "bytes_per_chunk": 6.352, "boundary_entropy": 0.15127, "router_noise": 0.0, "grad_norm": 0.6983, "lr": 6.08e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 275.6} |
| {"event": "probe", "step": 4800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2886.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165833, "cps": 22.546875, "mean_p": 0.185698, "boundary_entropy": 0.099074, "bpb": 2.035083}, "de": {"chunks": 3293.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160376, "cps": 25.726562, "mean_p": 0.187437, "boundary_entropy": 0.138314, "bpb": 2.253976}, "vi": {"chunks": 4183.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226905, "cps": 32.679688, "mean_p": 0.169468, "boundary_entropy": 0.116372, "bpb": 1.893456}, "tr": {"chunks": 3873.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.213942, "cps": 30.257812, "mean_p": 0.212167, "boundary_entropy": 0.166269, "bpb": 2.238434}, "fi": {"chunks": 2962.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157461, "cps": 23.140625, "mean_p": 0.187681, "boundary_entropy": 0.163325, "bpb": 2.411909}, "ru": {"chunks": 4606.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.243061, "cps": 35.984375, "mean_p": 0.141979, "boundary_entropy": 0.11387, "bpb": 1.37331}, "el": {"chunks": 4361.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.212493, "cps": 34.070312, "mean_p": 0.134685, "boundary_entropy": 0.157926, "bpb": 1.449944}, "hi": {"chunks": 4556.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.270788, "cps": 35.59375, "mean_p": 0.132493, "boundary_entropy": 0.223876, "bpb": 1.197244}, "ta": {"chunks": 4345.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.232415, "cps": 33.945312, "mean_p": 0.106159, "boundary_entropy": 0.186231, "bpb": 0.980494}, "th": {"chunks": 3715.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.230961, "cps": 29.023438, "mean_p": 0.193376, "boundary_entropy": 0.343923, "bpb": 1.095313}, "zh": {"chunks": 3496.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.59194, "cps": 27.3125, "mean_p": 0.286633, "boundary_entropy": 0.351014, "bpb": 3.158805}, "my": {"chunks": 3185.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.168092, "cps": 24.882812, "mean_p": 0.104915, "boundary_entropy": 0.207015, "bpb": 1.112539}}, "bpb": {"en": 2.03508, "de": 2.25398, "vi": 1.89346, "tr": 2.23843, "fi": 2.41191, "ru": 1.37331, "el": 1.44994, "hi": 1.19724, "ta": 0.98049, "th": 1.09531, "zh": 3.1588, "my": 1.11254}, "gini_cpc": 0.194355, "gini_cps": 0.089778, "bpic": 8.441983, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9670904997208996, "vi": 1.3682722404447383, "tr": 1.2901040511657538, "fi": 0.9495131072399692, "ru": 1.4656913093324704, "el": 1.2813655243855506, "hi": 1.6328881447455, "ta": 1.401496781411197, "th": 1.392725560739238, "zh": 3.5694867547221087, "my": 1.013616925982347}, "t": 278.7} |
| {"event": "eval", "step": 4800, "phase": 3, "bpb": {"en": 2.04669, "de": 2.2642, "vi": 1.89424, "tr": 2.24299, "fi": 2.42268, "ru": 1.35149, "el": 1.44138, "hi": 1.19253, "ta": 0.98902, "th": 1.11245, "zh": 3.15573, "my": 1.12158}, "bytes_per_chunk": {"en": 5.733, "de": 6.262, "vi": 5.784, "tr": 5.05, "fi": 6.379, "ru": 7.731, "el": 8.752, "hi": 9.111, "ta": 11.381, "th": 12.352, "zh": 5.103, "my": 15.564}, "bpb_mean": 1.76958, "t": 279.9} |
| {"event": "train", "step": 4800, "phase": 3, "objective": "none", "lm_loss": 1.25683, "bpb": 1.81323, "chunk_loss": 0.0, "F": 0.15189, "G": 0.179, "chunks": 155.53, "bytes_per_chunk": 6.584, "boundary_entropy": 0.14824, "router_noise": 0.0, "grad_norm": 0.7176, "lr": 5.85e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 280.0} |
| {"event": "train", "step": 4850, "phase": 3, "objective": "none", "lm_loss": 1.34645, "bpb": 1.94251, "chunk_loss": 0.0, "F": 0.15216, "G": 0.1813, "chunks": 155.81, "bytes_per_chunk": 6.572, "boundary_entropy": 0.16667, "router_noise": 0.0, "grad_norm": 0.7343, "lr": 5.62e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 282.1} |
| {"event": "train", "step": 4900, "phase": 3, "objective": "none", "lm_loss": 1.37688, "bpb": 1.98642, "chunk_loss": 0.0, "F": 0.16165, "G": 0.19211, "chunks": 165.53, "bytes_per_chunk": 6.186, "boundary_entropy": 0.16908, "router_noise": 0.0, "grad_norm": 0.7349, "lr": 5.41e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 284.2} |
| {"event": "train", "step": 4950, "phase": 3, "objective": "none", "lm_loss": 1.26801, "bpb": 1.82935, "chunk_loss": 0.0, "F": 0.14523, "G": 0.18717, "chunks": 148.72, "bytes_per_chunk": 6.885, "boundary_entropy": 0.19355, "router_noise": 0.0, "grad_norm": 0.7675, "lr": 5.2e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 286.3} |
| {"event": "probe", "step": 5000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2885.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165776, "cps": 22.539062, "mean_p": 0.18557, "boundary_entropy": 0.09877, "bpb": 2.027692}, "de": {"chunks": 3296.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160522, "cps": 25.75, "mean_p": 0.187581, "boundary_entropy": 0.139282, "bpb": 2.243567}, "vi": {"chunks": 4184.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.22696, "cps": 32.6875, "mean_p": 0.169476, "boundary_entropy": 0.114767, "bpb": 1.885458}, "tr": {"chunks": 3922.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216649, "cps": 30.640625, "mean_p": 0.213996, "boundary_entropy": 0.161353, "bpb": 2.231241}, "fi": {"chunks": 2956.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157142, "cps": 23.09375, "mean_p": 0.186755, "boundary_entropy": 0.162591, "bpb": 2.402199}, "ru": {"chunks": 4695.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.247757, "cps": 36.679688, "mean_p": 0.148786, "boundary_entropy": 0.121381, "bpb": 1.362403}, "el": {"chunks": 4412.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.214978, "cps": 34.46875, "mean_p": 0.137752, "boundary_entropy": 0.159005, "bpb": 1.441871}, "hi": {"chunks": 4672.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.277682, "cps": 36.5, "mean_p": 0.141372, "boundary_entropy": 0.233042, "bpb": 1.191236}, "ta": {"chunks": 4595.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.245788, "cps": 35.898438, "mean_p": 0.113748, "boundary_entropy": 0.190481, "bpb": 0.976048}, "th": {"chunks": 3671.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.228225, "cps": 28.679688, "mean_p": 0.195517, "boundary_entropy": 0.345881, "bpb": 1.091679}, "zh": {"chunks": 3500.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.592618, "cps": 27.34375, "mean_p": 0.286411, "boundary_entropy": 0.348452, "bpb": 3.150742}, "my": {"chunks": 3291.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.173686, "cps": 25.710938, "mean_p": 0.103876, "boundary_entropy": 0.202281, "bpb": 1.112499}}, "bpb": {"en": 2.02769, "de": 2.24357, "vi": 1.88546, "tr": 2.23124, "fi": 2.4022, "ru": 1.3624, "el": 1.44187, "hi": 1.19124, "ta": 0.97605, "th": 1.09168, "zh": 3.15074, "my": 1.1125}, "gini_cpc": 0.195713, "gini_cps": 0.094701, "bpic": 8.328761, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.968307060511544, "vi": 1.3690737280594754, "tr": 1.3068788832169778, "fi": 0.9479181720714132, "ru": 1.4945301646675782, "el": 1.2967998790045845, "hi": 1.67504339966162, "ta": 1.4826489997464543, "th": 1.3767073212122125, "zh": 3.574809508410505, "my": 1.0477141489853283}, "t": 289.3} |
| {"event": "eval", "step": 5000, "phase": 3, "bpb": {"en": 2.03955, "de": 2.25344, "vi": 1.88761, "tr": 2.2361, "fi": 2.41323, "ru": 1.34137, "el": 1.43342, "hi": 1.18762, "ta": 0.9838, "th": 1.10822, "zh": 3.14853, "my": 1.12161}, "bytes_per_chunk": {"en": 5.734, "de": 6.266, "vi": 5.783, "tr": 5.019, "fi": 6.388, "ru": 7.575, "el": 8.588, "hi": 8.861, "ta": 10.82, "th": 12.608, "zh": 5.112, "my": 15.222}, "bpb_mean": 1.76287, "t": 290.6} |
| {"event": "train", "step": 5000, "phase": 3, "objective": "none", "lm_loss": 1.45956, "bpb": 2.1057, "chunk_loss": 0.0, "F": 0.17383, "G": 0.19818, "chunks": 178.0, "bytes_per_chunk": 5.753, "boundary_entropy": 0.14669, "router_noise": 0.0, "grad_norm": 0.7648, "lr": 5e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 290.7} |
| {"event": "train", "step": 5050, "phase": 3, "objective": "none", "lm_loss": 1.27798, "bpb": 1.84373, "chunk_loss": 0.0, "F": 0.14755, "G": 0.18252, "chunks": 151.09, "bytes_per_chunk": 6.777, "boundary_entropy": 0.18192, "router_noise": 0.0, "grad_norm": 0.7286, "lr": 4.81e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 292.7} |
| {"event": "train", "step": 5100, "phase": 3, "objective": "none", "lm_loss": 1.45016, "bpb": 2.09214, "chunk_loss": 0.0, "F": 0.16272, "G": 0.19404, "chunks": 166.62, "bytes_per_chunk": 6.146, "boundary_entropy": 0.1741, "router_noise": 0.0, "grad_norm": 0.7415, "lr": 4.63e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 294.9} |
| {"event": "train", "step": 5150, "phase": 3, "objective": "none", "lm_loss": 1.24152, "bpb": 1.79113, "chunk_loss": 0.0, "F": 0.15079, "G": 0.18379, "chunks": 154.41, "bytes_per_chunk": 6.632, "boundary_entropy": 0.18527, "router_noise": 0.0, "grad_norm": 0.6694, "lr": 4.45e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 297.0} |
| {"event": "probe", "step": 5200, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.185811, "boundary_entropy": 0.099614, "bpb": 2.022004}, "de": {"chunks": 3296.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160522, "cps": 25.75, "mean_p": 0.187444, "boundary_entropy": 0.138244, "bpb": 2.234938}, "vi": {"chunks": 4182.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226851, "cps": 32.671875, "mean_p": 0.169787, "boundary_entropy": 0.116329, "bpb": 1.879646}, "tr": {"chunks": 3907.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215821, "cps": 30.523438, "mean_p": 0.213055, "boundary_entropy": 0.159886, "bpb": 2.219573}, "fi": {"chunks": 2953.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.156983, "cps": 23.070312, "mean_p": 0.185763, "boundary_entropy": 0.159795, "bpb": 2.395493}, "ru": {"chunks": 4758.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.251082, "cps": 37.171875, "mean_p": 0.148775, "boundary_entropy": 0.11992, "bpb": 1.352724}, "el": {"chunks": 4463.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.217463, "cps": 34.867188, "mean_p": 0.139912, "boundary_entropy": 0.161163, "bpb": 1.431522}, "hi": {"chunks": 4661.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.277028, "cps": 36.414062, "mean_p": 0.139908, "boundary_entropy": 0.233997, "bpb": 1.183681}, "ta": {"chunks": 4508.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.241134, "cps": 35.21875, "mean_p": 0.110741, "boundary_entropy": 0.188392, "bpb": 0.971196}, "th": {"chunks": 3894.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.242089, "cps": 30.421875, "mean_p": 0.201923, "boundary_entropy": 0.353457, "bpb": 1.08178}, "zh": {"chunks": 3448.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.583813, "cps": 26.9375, "mean_p": 0.286262, "boundary_entropy": 0.349417, "bpb": 3.138358}, "my": {"chunks": 3168.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.167194, "cps": 24.75, "mean_p": 0.10425, "boundary_entropy": 0.204956, "bpb": 1.107771}}, "bpb": {"en": 2.022, "de": 2.23494, "vi": 1.87965, "tr": 2.21957, "fi": 2.39549, "ru": 1.35272, "el": 1.43152, "hi": 1.18368, "ta": 0.9712, "th": 1.08178, "zh": 3.13836, "my": 1.10777}, "gini_cpc": 0.194987, "gini_cps": 0.096424, "bpic": 8.320275, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9673012013766636, "vi": 1.36699780692363, "tr": 1.3005282522310393, "fi": 0.9459724639514423, "ru": 1.5130112411287906, "el": 1.3104274232963138, "hi": 1.6693636835108023, "ta": 1.453066094523622, "th": 1.4588203510264461, "zh": 3.5180397792563647, "my": 1.0075085216457695}, "t": 300.0} |
| {"event": "eval", "step": 5200, "phase": 3, "bpb": {"en": 2.03386, "de": 2.24564, "vi": 1.8824, "tr": 2.22575, "fi": 2.40562, "ru": 1.33039, "el": 1.42148, "hi": 1.17982, "ta": 0.97949, "th": 1.09823, "zh": 3.13551, "my": 1.11601}, "bytes_per_chunk": {"en": 5.735, "de": 6.266, "vi": 5.783, "tr": 5.04, "fi": 6.399, "ru": 7.473, "el": 8.48, "hi": 8.895, "ta": 11.119, "th": 11.854, "zh": 5.094, "my": 15.548}, "bpb_mean": 1.75452, "t": 301.3} |
| {"event": "train", "step": 5200, "phase": 3, "objective": "none", "lm_loss": 1.35155, "bpb": 1.94987, "chunk_loss": 0.0, "F": 0.15833, "G": 0.18934, "chunks": 162.12, "bytes_per_chunk": 6.316, "boundary_entropy": 0.16208, "router_noise": 0.0, "grad_norm": 0.7627, "lr": 4.29e-05, "router_trainable": false, "sec_per_step": 0.087, "alpha": 0.0, "t": 301.3} |
| {"event": "train", "step": 5250, "phase": 3, "objective": "none", "lm_loss": 1.31682, "bpb": 1.89976, "chunk_loss": 0.0, "F": 0.15384, "G": 0.1853, "chunks": 157.53, "bytes_per_chunk": 6.5, "boundary_entropy": 0.15717, "router_noise": 0.0, "grad_norm": 0.751, "lr": 4.14e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 303.4} |
| {"event": "train", "step": 5300, "phase": 3, "objective": "none", "lm_loss": 1.28503, "bpb": 1.85391, "chunk_loss": 0.0, "F": 0.15131, "G": 0.18019, "chunks": 154.94, "bytes_per_chunk": 6.609, "boundary_entropy": 0.16234, "router_noise": 0.0, "grad_norm": 0.7151, "lr": 3.99e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 305.6} |
| {"event": "train", "step": 5350, "phase": 3, "objective": "none", "lm_loss": 1.3513, "bpb": 1.94952, "chunk_loss": 0.0, "F": 0.16025, "G": 0.18742, "chunks": 164.09, "bytes_per_chunk": 6.24, "boundary_entropy": 0.15182, "router_noise": 0.0, "grad_norm": 0.7252, "lr": 3.86e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 307.7} |
| {"event": "probe", "step": 5400, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2886.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165833, "cps": 22.546875, "mean_p": 0.185453, "boundary_entropy": 0.097741, "bpb": 2.013677}, "de": {"chunks": 3291.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160279, "cps": 25.710938, "mean_p": 0.186331, "boundary_entropy": 0.134863, "bpb": 2.225742}, "vi": {"chunks": 4185.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.227014, "cps": 32.695312, "mean_p": 0.171068, "boundary_entropy": 0.113127, "bpb": 1.87324}, "tr": {"chunks": 3911.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216042, "cps": 30.554688, "mean_p": 0.212351, "boundary_entropy": 0.160309, "bpb": 2.21246}, "fi": {"chunks": 2956.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157142, "cps": 23.09375, "mean_p": 0.185337, "boundary_entropy": 0.157056, "bpb": 2.388145}, "ru": {"chunks": 4759.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.251135, "cps": 37.179688, "mean_p": 0.151724, "boundary_entropy": 0.125575, "bpb": 1.346784}, "el": {"chunks": 4448.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.216732, "cps": 34.75, "mean_p": 0.141174, "boundary_entropy": 0.163632, "bpb": 1.423861}, "hi": {"chunks": 4691.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.278811, "cps": 36.648438, "mean_p": 0.142257, "boundary_entropy": 0.22556, "bpb": 1.178396}, "ta": {"chunks": 4605.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.246323, "cps": 35.976562, "mean_p": 0.115634, "boundary_entropy": 0.18965, "bpb": 0.966638}, "th": {"chunks": 3914.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.243332, "cps": 30.578125, "mean_p": 0.204306, "boundary_entropy": 0.352449, "bpb": 1.079353}, "zh": {"chunks": 3463.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.586353, "cps": 27.054688, "mean_p": 0.287449, "boundary_entropy": 0.349618, "bpb": 3.132103}, "my": {"chunks": 3285.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.173369, "cps": 25.664062, "mean_p": 0.107666, "boundary_entropy": 0.209914, "bpb": 1.10052}}, "bpb": {"en": 2.01368, "de": 2.22574, "vi": 1.87324, "tr": 2.21246, "fi": 2.38814, "ru": 1.34678, "el": 1.42386, "hi": 1.1784, "ta": 0.96664, "th": 1.07935, "zh": 3.1321, "my": 1.10052}, "gini_cpc": 0.195109, "gini_cps": 0.095939, "bpic": 8.272212, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9665031383484606, "vi": 1.3689264466318982, "tr": 1.3027619272164375, "fi": 0.9475897180963364, "ru": 1.514377972451851, "el": 1.3069281936406625, "hi": 1.681272670544587, "ta": 1.4853608005520282, "th": 1.4673291641274233, "zh": 3.5357930868428666, "my": 1.0454416332345402}, "t": 310.7} |
| {"event": "eval", "step": 5400, "phase": 3, "bpb": {"en": 2.02571, "de": 2.23598, "vi": 1.8754, "tr": 2.2172, "fi": 2.39663, "ru": 1.32528, "el": 1.41482, "hi": 1.17365, "ta": 0.97492, "th": 1.09662, "zh": 3.12811, "my": 1.10934}, "bytes_per_chunk": {"en": 5.735, "de": 6.27, "vi": 5.783, "tr": 5.035, "fi": 6.398, "ru": 7.463, "el": 8.482, "hi": 8.805, "ta": 10.882, "th": 11.781, "zh": 5.096, "my": 15.114}, "bpb_mean": 1.74781, "t": 312.0} |
| {"event": "train", "step": 5400, "phase": 3, "objective": "none", "lm_loss": 1.23209, "bpb": 1.77753, "chunk_loss": 0.0, "F": 0.14517, "G": 0.17948, "chunks": 148.66, "bytes_per_chunk": 6.888, "boundary_entropy": 0.17036, "router_noise": 0.0, "grad_norm": 0.7281, "lr": 3.73e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 312.1} |
| {"event": "train", "step": 5450, "phase": 3, "objective": "none", "lm_loss": 1.28764, "bpb": 1.85768, "chunk_loss": 0.0, "F": 0.14981, "G": 0.18486, "chunks": 153.41, "bytes_per_chunk": 6.675, "boundary_entropy": 0.18377, "router_noise": 0.0, "grad_norm": 0.7532, "lr": 3.62e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 314.2} |
| {"event": "train", "step": 5500, "phase": 3, "objective": "none", "lm_loss": 1.45519, "bpb": 2.0994, "chunk_loss": 0.0, "F": 0.17142, "G": 0.19943, "chunks": 175.53, "bytes_per_chunk": 5.834, "boundary_entropy": 0.15563, "router_noise": 0.0, "grad_norm": 0.7423, "lr": 3.51e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 316.3} |
| {"event": "train", "step": 5550, "phase": 3, "objective": "none", "lm_loss": 1.30583, "bpb": 1.88392, "chunk_loss": 0.0, "F": 0.15811, "G": 0.18403, "chunks": 161.91, "bytes_per_chunk": 6.325, "boundary_entropy": 0.14818, "router_noise": 0.0, "grad_norm": 0.749, "lr": 3.41e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 318.4} |
| {"event": "probe", "step": 5600, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.185526, "boundary_entropy": 0.098037, "bpb": 2.009265}, "de": {"chunks": 3288.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160132, "cps": 25.6875, "mean_p": 0.186245, "boundary_entropy": 0.134952, "bpb": 2.219948}, "vi": {"chunks": 4183.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226905, "cps": 32.679688, "mean_p": 0.172057, "boundary_entropy": 0.114812, "bpb": 1.868581}, "tr": {"chunks": 3909.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215931, "cps": 30.539062, "mean_p": 0.212968, "boundary_entropy": 0.159487, "bpb": 2.207811}, "fi": {"chunks": 2955.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157089, "cps": 23.085938, "mean_p": 0.184242, "boundary_entropy": 0.154374, "bpb": 2.381109}, "ru": {"chunks": 4769.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.251662, "cps": 37.257812, "mean_p": 0.155919, "boundary_entropy": 0.129046, "bpb": 1.340178}, "el": {"chunks": 4412.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.214978, "cps": 34.46875, "mean_p": 0.140837, "boundary_entropy": 0.164016, "bpb": 1.416269}, "hi": {"chunks": 4722.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.280654, "cps": 36.890625, "mean_p": 0.143781, "boundary_entropy": 0.225347, "bpb": 1.172249}, "ta": {"chunks": 4640.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.248195, "cps": 36.25, "mean_p": 0.116866, "boundary_entropy": 0.189514, "bpb": 0.96219}, "th": {"chunks": 3731.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.231955, "cps": 29.148438, "mean_p": 0.200092, "boundary_entropy": 0.350185, "bpb": 1.074543}, "zh": {"chunks": 3490.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.590924, "cps": 27.265625, "mean_p": 0.287569, "boundary_entropy": 0.344825, "bpb": 3.117063}, "my": {"chunks": 3223.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.170097, "cps": 25.179688, "mean_p": 0.105944, "boundary_entropy": 0.204889, "bpb": 1.102806}}, "bpb": {"en": 2.00927, "de": 2.21995, "vi": 1.86858, "tr": 2.20781, "fi": 2.38111, "ru": 1.34018, "el": 1.41627, "hi": 1.17225, "ta": 0.96219, "th": 1.07454, "zh": 3.11706, "my": 1.10281}, "gini_cpc": 0.197572, "gini_cps": 0.098107, "bpic": 8.30533, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9652876237936849, "vi": 1.3677982978605872, "tr": 1.301644702873204, "fi": 0.946941037841337, "ru": 1.5170344543321759, "el": 1.2959015070759359, "hi": 1.6917969830428046, "ta": 1.496131777564727, "th": 1.3982393499893784, "zh": 3.562126355273139, "my": 1.0253550288653928}, "t": 321.5} |
| {"event": "eval", "step": 5600, "phase": 3, "bpb": {"en": 2.02218, "de": 2.23079, "vi": 1.86987, "tr": 2.21283, "fi": 2.39126, "ru": 1.31849, "el": 1.40739, "hi": 1.16779, "ta": 0.9706, "th": 1.09253, "zh": 3.11385, "my": 1.11165}, "bytes_per_chunk": {"en": 5.734, "de": 6.277, "vi": 5.781, "tr": 5.037, "fi": 6.404, "ru": 7.452, "el": 8.552, "hi": 8.787, "ta": 10.751, "th": 12.206, "zh": 5.026, "my": 15.36}, "bpb_mean": 1.74244, "t": 322.7} |
| {"event": "train", "step": 5600, "phase": 3, "objective": "none", "lm_loss": 1.31485, "bpb": 1.89693, "chunk_loss": 0.0, "F": 0.1619, "G": 0.19136, "chunks": 165.78, "bytes_per_chunk": 6.177, "boundary_entropy": 0.15707, "router_noise": 0.0, "grad_norm": 0.7335, "lr": 3.33e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 322.8} |
| {"event": "train", "step": 5650, "phase": 3, "objective": "none", "lm_loss": 1.22395, "bpb": 1.76578, "chunk_loss": 0.0, "F": 0.15234, "G": 0.17992, "chunks": 156.0, "bytes_per_chunk": 6.564, "boundary_entropy": 0.15371, "router_noise": 0.0, "grad_norm": 0.7501, "lr": 3.25e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 324.9} |
| {"event": "train", "step": 5700, "phase": 3, "objective": "none", "lm_loss": 1.28317, "bpb": 1.85122, "chunk_loss": 0.0, "F": 0.15695, "G": 0.18342, "chunks": 160.72, "bytes_per_chunk": 6.371, "boundary_entropy": 0.14165, "router_noise": 0.0, "grad_norm": 0.7701, "lr": 3.18e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 327.0} |
| {"event": "train", "step": 5750, "phase": 3, "objective": "none", "lm_loss": 1.30935, "bpb": 1.88899, "chunk_loss": 0.0, "F": 0.15167, "G": 0.1809, "chunks": 155.31, "bytes_per_chunk": 6.593, "boundary_entropy": 0.16378, "router_noise": 0.0, "grad_norm": 0.756, "lr": 3.13e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 329.1} |
| {"event": "probe", "step": 5800, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2887.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165891, "cps": 22.554688, "mean_p": 0.185811, "boundary_entropy": 0.098862, "bpb": 2.007699}, "de": {"chunks": 3294.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160425, "cps": 25.734375, "mean_p": 0.186593, "boundary_entropy": 0.135897, "bpb": 2.214984}, "vi": {"chunks": 4181.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226797, "cps": 32.664062, "mean_p": 0.172629, "boundary_entropy": 0.112591, "bpb": 1.866187}, "tr": {"chunks": 3919.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.216483, "cps": 30.617188, "mean_p": 0.213567, "boundary_entropy": 0.15902, "bpb": 2.201426}, "fi": {"chunks": 2956.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157142, "cps": 23.09375, "mean_p": 0.185302, "boundary_entropy": 0.157495, "bpb": 2.374371}, "ru": {"chunks": 4816.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.254142, "cps": 37.625, "mean_p": 0.157571, "boundary_entropy": 0.128747, "bpb": 1.337156}, "el": {"chunks": 4742.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.231058, "cps": 37.046875, "mean_p": 0.149925, "boundary_entropy": 0.166858, "bpb": 1.413063}, "hi": {"chunks": 4829.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.287013, "cps": 37.726562, "mean_p": 0.148241, "boundary_entropy": 0.225732, "bpb": 1.17083}, "ta": {"chunks": 4807.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.257128, "cps": 37.554688, "mean_p": 0.122039, "boundary_entropy": 0.192339, "bpb": 0.960448}, "th": {"chunks": 3878.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.241094, "cps": 30.296875, "mean_p": 0.203178, "boundary_entropy": 0.352459, "bpb": 1.071895}, "zh": {"chunks": 3464.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.586522, "cps": 27.0625, "mean_p": 0.287155, "boundary_entropy": 0.34627, "bpb": 3.113793}, "my": {"chunks": 3311.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.174741, "cps": 25.867188, "mean_p": 0.109521, "boundary_entropy": 0.212751, "bpb": 1.096469}}, "bpb": {"en": 2.0077, "de": 2.21498, "vi": 1.86619, "tr": 2.20143, "fi": 2.37437, "ru": 1.33716, "el": 1.41306, "hi": 1.17083, "ta": 0.96045, "th": 1.07189, "zh": 3.11379, "my": 1.09647}, "gini_cpc": 0.196507, "gini_cps": 0.102724, "bpic": 8.150985, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9670490975597318, "vi": 1.3671443182775795, "tr": 1.3049745690867451, "fi": 0.9472614916612495, "ru": 1.531985307625028, "el": 1.392829770297844, "hi": 1.730132916373084, "ta": 1.5499796238693193, "th": 1.4533294557112861, "zh": 3.5355890242596435, "my": 1.0533510706091578}, "t": 332.2} |
| {"event": "eval", "step": 5800, "phase": 3, "bpb": {"en": 2.02038, "de": 2.22657, "vi": 1.86749, "tr": 2.20579, "fi": 2.38507, "ru": 1.31561, "el": 1.4027, "hi": 1.16629, "ta": 0.96875, "th": 1.08843, "zh": 3.10986, "my": 1.10439}, "bytes_per_chunk": {"en": 5.735, "de": 6.265, "vi": 5.783, "tr": 5.028, "fi": 6.408, "ru": 7.386, "el": 7.964, "hi": 8.662, "ta": 10.441, "th": 11.813, "zh": 5.104, "my": 15.031}, "bpb_mean": 1.73844, "t": 333.5} |
| {"event": "train", "step": 5800, "phase": 3, "objective": "none", "lm_loss": 1.34258, "bpb": 1.93693, "chunk_loss": 0.0, "F": 0.16089, "G": 0.18891, "chunks": 164.75, "bytes_per_chunk": 6.215, "boundary_entropy": 0.16267, "router_noise": 0.0, "grad_norm": 0.7346, "lr": 3.08e-05, "router_trainable": false, "sec_per_step": 0.088, "alpha": 0.0, "t": 333.5} |
| {"event": "train", "step": 5850, "phase": 3, "objective": "none", "lm_loss": 1.30206, "bpb": 1.87847, "chunk_loss": 0.0, "F": 0.15378, "G": 0.18688, "chunks": 157.47, "bytes_per_chunk": 6.503, "boundary_entropy": 0.17546, "router_noise": 0.0, "grad_norm": 0.735, "lr": 3.05e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 335.6} |
| {"event": "train", "step": 5900, "phase": 3, "objective": "none", "lm_loss": 1.21587, "bpb": 1.75413, "chunk_loss": 0.0, "F": 0.14218, "G": 0.17599, "chunks": 145.59, "bytes_per_chunk": 7.033, "boundary_entropy": 0.16426, "router_noise": 0.0, "grad_norm": 0.7368, "lr": 3.02e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 337.7} |
| {"event": "train", "step": 5950, "phase": 3, "objective": "none", "lm_loss": 1.24186, "bpb": 1.79163, "chunk_loss": 0.0, "F": 0.14233, "G": 0.17685, "chunks": 145.75, "bytes_per_chunk": 7.026, "boundary_entropy": 0.16389, "router_noise": 0.0, "grad_norm": 0.7744, "lr": 3.01e-05, "router_trainable": false, "sec_per_step": 0.042, "alpha": 0.0, "t": 339.9} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.186422, "boundary_entropy": 0.101047, "bpb": 2.006422}, "de": {"chunks": 3290.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.16023, "cps": 25.703125, "mean_p": 0.188192, "boundary_entropy": 0.140551, "bpb": 2.210822}, "vi": {"chunks": 4183.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226905, "cps": 32.679688, "mean_p": 0.172638, "boundary_entropy": 0.11429, "bpb": 1.863597}, "tr": {"chunks": 3908.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215876, "cps": 30.53125, "mean_p": 0.213314, "boundary_entropy": 0.160166, "bpb": 2.198945}, "fi": {"chunks": 2961.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157408, "cps": 23.132812, "mean_p": 0.186766, "boundary_entropy": 0.161715, "bpb": 2.371395}, "ru": {"chunks": 4872.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.257098, "cps": 38.0625, "mean_p": 0.15879, "boundary_entropy": 0.127265, "bpb": 1.333898}, "el": {"chunks": 4584.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.223359, "cps": 35.8125, "mean_p": 0.144112, "boundary_entropy": 0.162914, "bpb": 1.408863}, "hi": {"chunks": 4760.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.282912, "cps": 37.1875, "mean_p": 0.144234, "boundary_entropy": 0.228428, "bpb": 1.168054}, "ta": {"chunks": 4739.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.25349, "cps": 37.023438, "mean_p": 0.119103, "boundary_entropy": 0.192822, "bpb": 0.958125}, "th": {"chunks": 3873.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.240783, "cps": 30.257812, "mean_p": 0.202586, "boundary_entropy": 0.353434, "bpb": 1.069157}, "zh": {"chunks": 3468.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.587199, "cps": 27.09375, "mean_p": 0.28739, "boundary_entropy": 0.34543, "bpb": 3.109874}, "my": {"chunks": 3302.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.174266, "cps": 25.796875, "mean_p": 0.108037, "boundary_entropy": 0.20892, "bpb": 1.096973}}, "bpb": {"en": 2.00642, "de": 2.21082, "vi": 1.8636, "tr": 2.19895, "fi": 2.37139, "ru": 1.3339, "el": 1.40886, "hi": 1.16805, "ta": 0.95813, "th": 1.06916, "zh": 3.10987, "my": 1.09697}, "gini_cpc": 0.196467, "gini_cps": 0.100987, "bpic": 8.195545, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9655403375392061, "vi": 1.3673246834915216, "tr": 1.3008611235523166, "fi": 0.9485352068270303, "ru": 1.549262456237803, "el": 1.345955480257742, "hi": 1.7048210970846212, "ta": 1.5275244502989007, "th": 1.450953060997798, "zh": 3.5384460424771085, "my": 1.050124096740635}, "t": 342.9} |
| {"event": "eval", "step": 6000, "phase": 3, "bpb": {"en": 2.01909, "de": 2.22078, "vi": 1.86459, "tr": 2.20322, "fi": 2.38245, "ru": 1.31293, "el": 1.39926, "hi": 1.16296, "ta": 0.96592, "th": 1.08636, "zh": 3.10477, "my": 1.10415}, "bytes_per_chunk": {"en": 5.734, "de": 6.265, "vi": 5.782, "tr": 5.038, "fi": 6.395, "ru": 7.303, "el": 8.27, "hi": 8.764, "ta": 10.587, "th": 11.886, "zh": 5.093, "my": 14.972}, "bpb_mean": 1.73554, "t": 344.2} |
| {"event": "probe", "step": 6000, "phase": 3, "objective": "none", "router_trainable": false, "per_lang": {"en": {"chunks": 2888.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.165948, "cps": 22.5625, "mean_p": 0.186422, "boundary_entropy": 0.101047, "bpb": 2.006422}, "de": {"chunks": 3290.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.16023, "cps": 25.703125, "mean_p": 0.188192, "boundary_entropy": 0.140551, "bpb": 2.210822}, "vi": {"chunks": 4183.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.226905, "cps": 32.679688, "mean_p": 0.172638, "boundary_entropy": 0.11429, "bpb": 1.863597}, "tr": {"chunks": 3908.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.215876, "cps": 30.53125, "mean_p": 0.213314, "boundary_entropy": 0.160166, "bpb": 2.198945}, "fi": {"chunks": 2961.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157408, "cps": 23.132812, "mean_p": 0.186766, "boundary_entropy": 0.161715, "bpb": 2.371395}, "ru": {"chunks": 4872.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.257098, "cps": 38.0625, "mean_p": 0.15879, "boundary_entropy": 0.127265, "bpb": 1.333898}, "el": {"chunks": 4584.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.223359, "cps": 35.8125, "mean_p": 0.144112, "boundary_entropy": 0.162914, "bpb": 1.408863}, "hi": {"chunks": 4760.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.282912, "cps": 37.1875, "mean_p": 0.144234, "boundary_entropy": 0.228428, "bpb": 1.168054}, "ta": {"chunks": 4739.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.25349, "cps": 37.023438, "mean_p": 0.119103, "boundary_entropy": 0.192822, "bpb": 0.958125}, "th": {"chunks": 3873.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.240783, "cps": 30.257812, "mean_p": 0.202586, "boundary_entropy": 0.353434, "bpb": 1.069157}, "zh": {"chunks": 3468.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.587199, "cps": 27.09375, "mean_p": 0.28739, "boundary_entropy": 0.34543, "bpb": 3.109874}, "my": {"chunks": 3302.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.174266, "cps": 25.796875, "mean_p": 0.108037, "boundary_entropy": 0.20892, "bpb": 1.096973}}, "bpb": {"en": 2.00642, "de": 2.21082, "vi": 1.8636, "tr": 2.19895, "fi": 2.37139, "ru": 1.3339, "el": 1.40886, "hi": 1.16805, "ta": 0.95813, "th": 1.06916, "zh": 3.10987, "my": 1.09697}, "gini_cpc": 0.196467, "gini_cps": 0.100987, "bpic": 8.195545, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9655403375392061, "vi": 1.3673246834915216, "tr": 1.3008611235523166, "fi": 0.9485352068270303, "ru": 1.549262456237803, "el": 1.345955480257742, "hi": 1.7048210970846212, "ta": 1.5275244502989007, "th": 1.450953060997798, "zh": 3.5384460424771085, "my": 1.050124096740635}, "t": 345.2} |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s1/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s1/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s1/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s1/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/pilot_E_s1/th0.70"], "t": 347.3} |
| {"event": "final", "run": "pilot_E_s1", "condition": "E", "step": 6000, "params": 22516992, "per_lang": {"en": {"cpc": 0.16594839970120095, "cps": 22.5625, "bpb": 2.027052285286028, "bytes_per_chunk": 5.732116109987427}, "de": {"cpc": 0.16022987386158866, "cps": 25.703125, "bpb": 2.2108564431276996, "bytes_per_chunk": 6.252465624366451}, "vi": {"cpc": 0.22690534309736912, "cps": 32.6796875, "bpb": 1.8681790764787312, "bytes_per_chunk": 5.773397495898074}, "tr": {"cpc": 0.2158758216870132, "cps": 30.53125, "bpb": 2.1935315020701425, "bytes_per_chunk": 5.056887671444279}, "fi": {"cpc": 0.15740789963319335, "cps": 23.1328125, "bpb": 2.3985133563139995, "bytes_per_chunk": 6.384335310091207}, "ru": {"cpc": 0.2570976253298153, "cps": 38.0625, "bpb": 1.3130465053519396, "bytes_per_chunk": 7.278947076137058}, "el": {"cpc": 0.22335915801783365, "cps": 35.8125, "bpb": 1.3963111640919865, "bytes_per_chunk": 8.278928751894897}, "hi": {"cpc": 0.2829123328380386, "cps": 37.1875, "bpb": 1.1711474198667287, "bytes_per_chunk": 8.791320824320472}, "ta": {"cpc": 0.25349023803155923, "cps": 37.0234375, "bpb": 0.9658224011153641, "bytes_per_chunk": 10.643928781696003}, "th": {"cpc": 0.2407833385141436, "cps": 30.2578125, "bpb": 1.0832735723389801, "bytes_per_chunk": 11.81680490443563}, "zh": {"cpc": 0.5871994581781239, "cps": 27.09375, "bpb": 3.112535372485221, "bytes_per_chunk": 5.089038370072703}, "my": {"cpc": 0.1742664133417775, "cps": 25.796875, "bpb": 1.1134285373889752, "bytes_per_chunk": 14.984366515190489}}, "gini_cpc": 0.1964674510925677, "gini_cps": 0.10098658922012471, "bpic": 8.195545400187921, "seed": 1, "mean_boundary_entropy": 0.17916804552078247, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "global", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 2400, "phase": 2, "objective": "parity_b", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4200, "phase": 3, "objective": "none", "router_trainable": false, "encoder_trainable": true, "frozen_router_params": 131584}], "t": 349.6} |
|
|
