Download log.jsonl from AdaptiveChunking/spec_D_s2: direct link, hf CLI and curl.
- Browser
- Download file 110 kB
-
https://huggingface.co/AdaptiveChunking/spec_D_s2/resolve/main/log.jsonl
- Command line
-
hf download hf://AdaptiveChunking/spec_D_s2/log.jsonl
-
curl -L -o log.jsonl https://huggingface.co/AdaptiveChunking/spec_D_s2/resolve/main/log.jsonl
110 kB
| {"event": "start", "run": "spec_D_s2", "condition": "D", "params": 82470528, "langs": ["en", "de", "vi", "tr", "fi", "ru", "el", "hi", "ta", "th", "zh", "my"], "premium": {"en": 1.0, "de": 1.184, "vi": 1.39, "tr": 1.12, "fi": 1.104, "ru": 1.971, "el": 2.174, "hi": 2.552, "ta": 3.192, "th": 2.762, "zh": 0.915, "my": 3.524}, "schedule": {"tau1": 0.4, "tau2": 0.7, "p1_objective": "parity_a", "p2_objective": "parity_a", "p3_objective": "parity_a", "freeze_in_p3": false, "freeze_scope": "router"}, "corpus": "/local/scratch/sas245/beetle-hnet/scratch/data/tier1_corpus", "premium_kind": "flores_per_sent", "doc_filtered": {"my": 37511}, "weights": {"en": 0.2946, "de": 0.0909, "vi": 0.0576, "tr": 0.0612, "fi": 0.0424, "ru": 0.1476, "el": 0.0522, "hi": 0.042, "ta": 0.0313, "th": 0.0542, "zh": 0.1038, "my": 0.0222}, "t": 2.3} | |
| {"event": "probe_schedule", "n": 21, "steps": [0, 1, 2, 4, 8, 16, 32, 64, 128, 256, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 11000], "t": 2.3} | |
| {"event": "probe", "step": 0, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 10185.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.585244, "cps": 79.570312, "mean_p": 0.50663, "boundary_entropy": 0.687329, "bpb": 8.017475}, "de": {"chunks": 10966.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.534067, "cps": 85.671875, "mean_p": 0.503879, "boundary_entropy": 0.688251, "bpb": 8.111776}, "vi": {"chunks": 5741.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.311418, "cps": 44.851562, "mean_p": 0.489692, "boundary_entropy": 0.688475, "bpb": 7.989146}, "tr": {"chunks": 6370.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.351875, "cps": 49.765625, "mean_p": 0.495802, "boundary_entropy": 0.687878, "bpb": 8.107657}, "fi": {"chunks": 2115.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.112434, "cps": 16.523438, "mean_p": 0.479404, "boundary_entropy": 0.686744, "bpb": 8.107682}, "ru": {"chunks": 23602.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.245488, "cps": 184.390625, "mean_p": 0.507156, "boundary_entropy": 0.690197, "bpb": 8.281387}, "el": {"chunks": 1455.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.070896, "cps": 11.367188, "mean_p": 0.480429, "boundary_entropy": 0.689544, "bpb": 7.890102}, "hi": {"chunks": 7663.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.455453, "cps": 59.867188, "mean_p": 0.494285, "boundary_entropy": 0.690846, "bpb": 7.810752}, "ta": {"chunks": 357.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.019096, "cps": 2.789062, "mean_p": 0.475379, "boundary_entropy": 0.689908, "bpb": 7.609497}, "th": {"chunks": 27234.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.69313, "cps": 212.765625, "mean_p": 0.50318, "boundary_entropy": 0.691076, "bpb": 7.876604}, "zh": {"chunks": 14702.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.489333, "cps": 114.859375, "mean_p": 0.531852, "boundary_entropy": 0.685234, "bpb": 8.024454}, "my": {"chunks": 789.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.04164, "cps": 6.164062, "mean_p": 0.475663, "boundary_entropy": 0.690063, "bpb": 8.245358}}, "bpb": {"en": 8.01748, "de": 8.11178, "vi": 7.98915, "tr": 8.10766, "fi": 8.10768, "ru": 8.28139, "el": 7.8901, "hi": 7.81075, "ta": 7.6095, "th": 7.8766, "zh": 8.02445, "my": 8.24536}, "gini_cpc": 0.562508, "gini_cps": 0.489131, "bpic": 3.45192, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9125547315729478, "vi": 0.5321174384544991, "tr": 0.6012456783662331, "fi": 0.19211513271032077, "ru": 2.1281521716321925, "el": 0.1211393488838307, "hi": 0.7782279770749311, "ta": 0.032629155488401715, "th": 2.8930334475858417, "zh": 4.253496332079356, "my": 0.07115029646287195}, "t": 8.1} | |
| {"event": "train", "step": 0, "phase": 1, "objective": "parity_a", "lm_loss": 5.5908, "bpb": 8.06582, "chunk_loss": 0.082627, "F": 0.56091, "G": 0.50224, "chunks": 574.38, "bytes_per_chunk": 1.783, "boundary_entropy": 0.6917, "router_noise": 0.0, "grad_norm": 9.2842, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.006, "n_target_mean": 8.9638, "alpha": 0.03, "t": 9.0} | |
| {"event": "probe", "step": 1, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 10116.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.581279, "cps": 79.03125, "mean_p": 0.50646, "boundary_entropy": 0.687331, "bpb": 7.982813}, "de": {"chunks": 10916.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.531632, "cps": 85.28125, "mean_p": 0.503755, "boundary_entropy": 0.688251, "bpb": 8.080574}, "vi": {"chunks": 5695.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.308923, "cps": 44.492188, "mean_p": 0.489573, "boundary_entropy": 0.688469, "bpb": 7.968805}, "tr": {"chunks": 6319.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.349058, "cps": 49.367188, "mean_p": 0.495635, "boundary_entropy": 0.687872, "bpb": 8.083917}, "fi": {"chunks": 2105.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.111903, "cps": 16.445312, "mean_p": 0.479281, "boundary_entropy": 0.68673, "bpb": 8.087042}, "ru": {"chunks": 23485.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.239314, "cps": 183.476562, "mean_p": 0.50705, "boundary_entropy": 0.6902, "bpb": 8.229006}, "el": {"chunks": 1445.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.070409, "cps": 11.289062, "mean_p": 0.480398, "boundary_entropy": 0.689541, "bpb": 7.88016}, "hi": {"chunks": 7578.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.450401, "cps": 59.203125, "mean_p": 0.494214, "boundary_entropy": 0.690844, "bpb": 7.787942}, "ta": {"chunks": 355.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.018989, "cps": 2.773438, "mean_p": 0.475287, "boundary_entropy": 0.689899, "bpb": 7.59574}, "th": {"chunks": 26773.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.66447, "cps": 209.164062, "mean_p": 0.503002, "boundary_entropy": 0.691077, "bpb": 7.839513}, "zh": {"chunks": 14695.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.488148, "cps": 114.804688, "mean_p": 0.531768, "boundary_entropy": 0.685244, "bpb": 8.019147}, "my": {"chunks": 784.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.041376, "cps": 6.125, "mean_p": 0.47562, "boundary_entropy": 0.690059, "bpb": 8.246774}}, "bpb": {"en": 7.98281, "de": 8.08057, "vi": 7.96881, "tr": 8.08392, "fi": 8.08704, "ru": 8.22901, "el": 7.88016, "hi": 7.78794, "ta": 7.59574, "th": 7.83951, "zh": 8.01915, "my": 8.24677}, "gini_cpc": 0.562924, "gini_cps": 0.48839, "bpic": 3.480502, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9145899385180896, "vi": 0.5314542519145689, "tr": 0.6005001268279724, "fi": 0.1925109850606706, "ru": 2.132046388541989, "el": 0.12112737385922769, "hi": 0.7748449868597082, "ta": 0.032667671750142846, "th": 2.8634609988233475, "zh": 4.280469898316006, "my": 0.07118163940469151}, "t": 14.1} | |
| {"event": "probe", "step": 2, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 9980.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.573464, "cps": 77.96875, "mean_p": 0.506081, "boundary_entropy": 0.687335, "bpb": 7.915977}, "de": {"chunks": 10778.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.524911, "cps": 84.203125, "mean_p": 0.503468, "boundary_entropy": 0.68825, "bpb": 8.020592}, "vi": {"chunks": 5571.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.302197, "cps": 43.523438, "mean_p": 0.489288, "boundary_entropy": 0.688454, "bpb": 7.931178}, "tr": {"chunks": 6169.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.340772, "cps": 48.195312, "mean_p": 0.495276, "boundary_entropy": 0.687859, "bpb": 8.037788}, "fi": {"chunks": 2072.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.110148, "cps": 16.1875, "mean_p": 0.479, "boundary_entropy": 0.686699, "bpb": 8.047566}, "ru": {"chunks": 23243.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.226544, "cps": 181.585938, "mean_p": 0.506846, "boundary_entropy": 0.690205, "bpb": 8.121394}, "el": {"chunks": 1432.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.069775, "cps": 11.1875, "mean_p": 0.480342, "boundary_entropy": 0.689536, "bpb": 7.855022}, "hi": {"chunks": 7499.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.445706, "cps": 58.585938, "mean_p": 0.49414, "boundary_entropy": 0.690841, "bpb": 7.746181}, "ta": {"chunks": 350.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.018722, "cps": 2.734375, "mean_p": 0.47522, "boundary_entropy": 0.689892, "bpb": 7.565769}, "th": {"chunks": 26138.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.624992, "cps": 204.203125, "mean_p": 0.502773, "boundary_entropy": 0.691078, "bpb": 7.777491}, "zh": {"chunks": 14681.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.485777, "cps": 114.695312, "mean_p": 0.531567, "boundary_entropy": 0.685268, "bpb": 8.006239}, "my": {"chunks": 675.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.035624, "cps": 5.273438, "mean_p": 0.475469, "boundary_entropy": 0.690043, "bpb": 8.235705}}, "bpb": {"en": 7.91598, "de": 8.02059, "vi": 7.93118, "tr": 8.03779, "fi": 8.04757, "ru": 8.12139, "el": 7.85502, "hi": 7.74618, "ta": 7.56577, "th": 7.77749, "zh": 8.00624, "my": 8.2357}, "gini_cpc": 0.565273, "gini_cps": 0.489319, "bpic": 3.534286, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9153334868246209, "vi": 0.5269672135157215, "tr": 0.5942344114844061, "fi": 0.19207526743455547, "ru": 2.138831377795168, "el": 0.12167343018151491, "hi": 0.7772162274027996, "ta": 0.03264646436494278, "th": 2.8336412582703487, "zh": 4.334667353671108, "my": 0.06212036168600659}, "t": 19.9} | |
| {"event": "probe", "step": 4, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 9457.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.543412, "cps": 73.882812, "mean_p": 0.504636, "boundary_entropy": 0.687345, "bpb": 7.680886}, "de": {"chunks": 10236.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.498515, "cps": 79.96875, "mean_p": 0.502357, "boundary_entropy": 0.688245, "bpb": 7.806258}, "vi": {"chunks": 5232.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.283808, "cps": 40.875, "mean_p": 0.488352, "boundary_entropy": 0.688401, "bpb": 7.800581}, "tr": {"chunks": 5714.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.315638, "cps": 44.640625, "mean_p": 0.493982, "boundary_entropy": 0.687807, "bpb": 7.874044}, "fi": {"chunks": 1987.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.10563, "cps": 15.523438, "mean_p": 0.478024, "boundary_entropy": 0.686588, "bpb": 7.910087}, "ru": {"chunks": 22496.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.187124, "cps": 175.75, "mean_p": 0.506209, "boundary_entropy": 0.690223, "bpb": 7.802961}, "el": {"chunks": 1383.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.067388, "cps": 10.804688, "mean_p": 0.480023, "boundary_entropy": 0.689508, "bpb": 7.777461}, "hi": {"chunks": 7114.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.422823, "cps": 55.578125, "mean_p": 0.493763, "boundary_entropy": 0.690829, "bpb": 7.595523}, "ta": {"chunks": 341.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.01824, "cps": 2.664062, "mean_p": 0.47497, "boundary_entropy": 0.689868, "bpb": 7.469041}, "th": {"chunks": 23721.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.474728, "cps": 185.320312, "mean_p": 0.501956, "boundary_entropy": 0.691081, "bpb": 7.562664}, "zh": {"chunks": 14618.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.47511, "cps": 114.203125, "mean_p": 0.530921, "boundary_entropy": 0.685344, "bpb": 7.968462}, "my": {"chunks": 649.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.034252, "cps": 5.070312, "mean_p": 0.475101, "boundary_entropy": 0.690005, "bpb": 8.211103}}, "bpb": {"en": 7.68089, "de": 7.80626, "vi": 7.80058, "tr": 7.87404, "fi": 7.91009, "ru": 7.80296, "el": 7.77746, "hi": 7.59552, "ta": 7.46904, "th": 7.56266, "zh": 7.96846, "my": 8.2111}, "gini_cpc": 0.569661, "gini_cps": 0.487113, "bpic": 3.727911, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9173785920436606, "vi": 0.522270294054206, "tr": 0.5808452147195663, "fi": 0.19438229923437073, "ru": 2.184574300060571, "el": 0.12400867469010941, "hi": 0.7780894361763521, "ta": 0.0335660038965803, "th": 2.7138301272954175, "zh": 4.5547573577102565, "my": 0.0630306886214011}, "t": 26.1} | |
| {"event": "probe", "step": 8, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 7488.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.430271, "cps": 58.5, "mean_p": 0.499677, "boundary_entropy": 0.687325, "bpb": 7.071254}, "de": {"chunks": 8453.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.411679, "cps": 66.039062, "mean_p": 0.498656, "boundary_entropy": 0.688199, "bpb": 7.216092}, "vi": {"chunks": 4072.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.220884, "cps": 31.8125, "mean_p": 0.48515, "boundary_entropy": 0.688203, "bpb": 7.451103}, "tr": {"chunks": 4352.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.240402, "cps": 34.0, "mean_p": 0.489741, "boundary_entropy": 0.687593, "bpb": 7.394811}, "fi": {"chunks": 1727.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.091808, "cps": 13.492188, "mean_p": 0.475016, "boundary_entropy": 0.686224, "bpb": 7.50038}, "ru": {"chunks": 18746.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.989235, "cps": 146.453125, "mean_p": 0.503354, "boundary_entropy": 0.690278, "bpb": 6.819247}, "el": {"chunks": 1212.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.059056, "cps": 9.46875, "mean_p": 0.478824, "boundary_entropy": 0.689399, "bpb": 7.515131}, "hi": {"chunks": 6129.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.364279, "cps": 47.882812, "mean_p": 0.492637, "boundary_entropy": 0.690786, "bpb": 7.046534}, "ta": {"chunks": 317.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.016956, "cps": 2.476562, "mean_p": 0.474522, "boundary_entropy": 0.689829, "bpb": 7.127018}, "th": {"chunks": 16186.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.006279, "cps": 126.453125, "mean_p": 0.499004, "boundary_entropy": 0.691068, "bpb": 6.859555}, "zh": {"chunks": 14352.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.430071, "cps": 112.125, "mean_p": 0.528401, "boundary_entropy": 0.685624, "bpb": 7.848903}, "my": {"chunks": 573.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.030241, "cps": 4.476562, "mean_p": 0.473997, "boundary_entropy": 0.689888, "bpb": 8.107625}}, "bpb": {"en": 7.07125, "de": 7.21609, "vi": 7.4511, "tr": 7.39481, "fi": 7.50038, "ru": 6.81925, "el": 7.51513, "hi": 7.04653, "ta": 7.12702, "th": 6.85955, "zh": 7.8489, "my": 8.10763}, "gini_cpc": 0.585854, "gini_cps": 0.479297, "bpic": 4.590297, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.9567902614865126, "vi": 0.5133610467825276, "tr": 0.5587230902372039, "fi": 0.21337264390448032, "ru": 2.299099054529463, "el": 0.1372524353539556, "hi": 0.8466284003886159, "ta": 0.03940869713552918, "th": 2.338712060527805, "zh": 5.6477734507280735, "my": 0.07028287692226498}, "t": 32.8} | |
| {"event": "probe", "step": 16, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 4216.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.242257, "cps": 32.9375, "mean_p": 0.491603, "boundary_entropy": 0.687168, "bpb": 6.368322}, "de": {"chunks": 6082.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.296206, "cps": 47.515625, "mean_p": 0.494473, "boundary_entropy": 0.688154, "bpb": 6.435993}, "vi": {"chunks": 2179.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.118199, "cps": 17.023438, "mean_p": 0.478824, "boundary_entropy": 0.687752, "bpb": 6.992591}, "tr": {"chunks": 2789.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.154063, "cps": 21.789062, "mean_p": 0.484376, "boundary_entropy": 0.687288, "bpb": 6.724068}, "fi": {"chunks": 1371.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.072883, "cps": 10.710938, "mean_p": 0.473154, "boundary_entropy": 0.68605, "bpb": 6.73212}, "ru": {"chunks": 5032.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.265541, "cps": 39.3125, "mean_p": 0.490416, "boundary_entropy": 0.690066, "bpb": 5.735069}, "el": {"chunks": 1163.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.056668, "cps": 9.085938, "mean_p": 0.477426, "boundary_entropy": 0.689257, "bpb": 6.762285}, "hi": {"chunks": 4175.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.248143, "cps": 32.617188, "mean_p": 0.489957, "boundary_entropy": 0.690666, "bpb": 6.020743}, "ta": {"chunks": 291.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.015566, "cps": 2.273438, "mean_p": 0.475615, "boundary_entropy": 0.689969, "bpb": 6.500044}, "th": {"chunks": 1461.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.09083, "cps": 11.414062, "mean_p": 0.485367, "boundary_entropy": 0.690575, "bpb": 5.686266}, "zh": {"chunks": 12459.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 2.10955, "cps": 97.335938, "mean_p": 0.518082, "boundary_entropy": 0.686489, "bpb": 7.484754}, "my": {"chunks": 464.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.024488, "cps": 3.625, "mean_p": 0.471401, "boundary_entropy": 0.689586, "bpb": 7.448917}}, "bpb": {"en": 6.36832, "de": 6.43599, "vi": 6.99259, "tr": 6.72407, "fi": 6.73212, "ru": 5.73507, "el": 6.76229, "hi": 6.02074, "ta": 6.50004, "th": 5.68627, "zh": 7.48475, "my": 7.44892}, "gini_cpc": 0.638532, "gini_cps": 0.468767, "bpic": 9.207356, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2226932837600897, "vi": 0.487907626107216, "tr": 0.635947816945853, "fi": 0.3008493581967917, "ru": 1.0961120095327261, "el": 0.23391732430343112, "hi": 1.02429469846703, "ta": 0.06425264901015147, "th": 0.3749321382668803, "zh": 8.707896554239055, "my": 0.1010829999531325}, "t": 40.3} | |
| {"event": "probe", "step": 32, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3485.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.200253, "cps": 27.226562, "mean_p": 0.493267, "boundary_entropy": 0.687491, "bpb": 5.838237}, "de": {"chunks": 6112.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.297667, "cps": 47.75, "mean_p": 0.496699, "boundary_entropy": 0.688431, "bpb": 5.951871}, "vi": {"chunks": 1470.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.07974, "cps": 11.484375, "mean_p": 0.480194, "boundary_entropy": 0.688102, "bpb": 6.593182}, "tr": {"chunks": 3009.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.166216, "cps": 23.507812, "mean_p": 0.489698, "boundary_entropy": 0.68787, "bpb": 6.262496}, "fi": {"chunks": 2054.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.109191, "cps": 16.046875, "mean_p": 0.486971, "boundary_entropy": 0.687693, "bpb": 5.99411}, "ru": {"chunks": 4464.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.235567, "cps": 34.875, "mean_p": 0.491277, "boundary_entropy": 0.690174, "bpb": 5.174995}, "el": {"chunks": 7435.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.362276, "cps": 58.085938, "mean_p": 0.492289, "boundary_entropy": 0.69033, "bpb": 5.56601}, "hi": {"chunks": 4503.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.267637, "cps": 35.179688, "mean_p": 0.490401, "boundary_entropy": 0.690698, "bpb": 5.369144}, "ta": {"chunks": 5931.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.317251, "cps": 46.335938, "mean_p": 0.493851, "boundary_entropy": 0.691205, "bpb": 5.237689}, "th": {"chunks": 848.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.05272, "cps": 6.625, "mean_p": 0.484871, "boundary_entropy": 0.690582, "bpb": 5.167687}, "zh": {"chunks": 3685.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.623942, "cps": 28.789062, "mean_p": 0.49101, "boundary_entropy": 0.686704, "bpb": 7.019701}, "my": {"chunks": 2828.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.149251, "cps": 22.09375, "mean_p": 0.487731, "boundary_entropy": 0.691004, "bpb": 5.266751}}, "bpb": {"en": 5.83824, "de": 5.95187, "vi": 6.59318, "tr": 6.2625, "fi": 5.99411, "ru": 5.17499, "el": 5.56601, "hi": 5.36914, "ta": 5.23769, "th": 5.16769, "zh": 7.0197, "my": 5.26675}, "gini_cpc": 0.330931, "gini_cps": 0.280628, "bpic": 8.375109, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.48645674552236, "vi": 0.3981947507217378, "tr": 0.8300284415869624, "fi": 0.545267852450873, "ru": 1.176349329769877, "el": 1.8090953838332762, "hi": 1.3364976880325574, "ta": 1.5842502790586483, "th": 0.2632668182704062, "zh": 3.1157699705719595, "my": 0.7453107182414819}, "t": 51.1} | |
| {"event": "probe", "step": 64, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 8390.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.482101, "cps": 65.546875, "mean_p": 0.502863, "boundary_entropy": 0.68783, "bpb": 5.421319}, "de": {"chunks": 8812.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.429163, "cps": 68.84375, "mean_p": 0.500824, "boundary_entropy": 0.688628, "bpb": 5.568148}, "vi": {"chunks": 2453.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.133062, "cps": 19.164062, "mean_p": 0.488903, "boundary_entropy": 0.688862, "bpb": 5.980739}, "tr": {"chunks": 4788.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.264487, "cps": 37.40625, "mean_p": 0.495693, "boundary_entropy": 0.688269, "bpb": 5.908706}, "fi": {"chunks": 5748.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.305566, "cps": 44.90625, "mean_p": 0.497647, "boundary_entropy": 0.68829, "bpb": 5.585643}, "ru": {"chunks": 3351.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.176834, "cps": 26.179688, "mean_p": 0.4938, "boundary_entropy": 0.690366, "bpb": 4.879865}, "el": {"chunks": 6032.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.293914, "cps": 47.125, "mean_p": 0.493091, "boundary_entropy": 0.690453, "bpb": 5.073312}, "hi": {"chunks": 1018.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.060505, "cps": 7.953125, "mean_p": 0.483113, "boundary_entropy": 0.690254, "bpb": 5.032213}, "ta": {"chunks": 1767.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.094517, "cps": 13.804688, "mean_p": 0.49231, "boundary_entropy": 0.691184, "bpb": 4.567038}, "th": {"chunks": 2070.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.128691, "cps": 16.171875, "mean_p": 0.489435, "boundary_entropy": 0.690823, "bpb": 4.503394}, "zh": {"chunks": 5434.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.920081, "cps": 42.453125, "mean_p": 0.498794, "boundary_entropy": 0.687248, "bpb": 6.695624}, "my": {"chunks": 1107.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.058423, "cps": 8.648438, "mean_p": 0.485545, "boundary_entropy": 0.690918, "bpb": 4.620287}}, "bpb": {"en": 5.42132, "de": 5.56815, "vi": 5.98074, "tr": 5.90871, "fi": 5.58564, "ru": 4.87987, "el": 5.07331, "hi": 5.03221, "ta": 4.56704, "th": 4.50339, "zh": 6.69562, "my": 4.62029}, "gini_cpc": 0.424956, "gini_cps": 0.343186, "bpic": 7.529547, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8901931348397158, "vi": 0.27600475594274637, "tr": 0.5486125644298547, "fi": 0.6338215999577759, "ru": 0.36679834958692503, "el": 0.6096529043231069, "hi": 0.12550321882521293, "ta": 0.19605288584297015, "th": 0.26693863759386127, "zh": 1.908483241825549, "my": 0.12118431272791502}, "t": 65.7} | |
| {"event": "train", "step": 100, "phase": 1, "objective": "parity_a", "lm_loss": 3.30014, "bpb": 4.7611, "chunk_loss": 0.040338, "F": 0.24129, "G": 0.48732, "chunks": 247.08, "bytes_per_chunk": 4.144, "boundary_entropy": 0.69093, "router_noise": 0.0, "grad_norm": 1.8234, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.685, "n_target_mean": 9.1186, "alpha": 0.03, "t": 77.5} | |
| {"event": "probe", "step": 128, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 4993.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.286905, "cps": 39.007812, "mean_p": 0.474506, "boundary_entropy": 0.681905, "bpb": 4.362857}, "de": {"chunks": 6457.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.314469, "cps": 50.445312, "mean_p": 0.474523, "boundary_entropy": 0.682109, "bpb": 4.61974}, "vi": {"chunks": 4674.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.253539, "cps": 36.515625, "mean_p": 0.481206, "boundary_entropy": 0.687339, "bpb": 4.964723}, "tr": {"chunks": 4317.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.238469, "cps": 33.726562, "mean_p": 0.47295, "boundary_entropy": 0.684119, "bpb": 4.912408}, "fi": {"chunks": 5585.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.296901, "cps": 43.632812, "mean_p": 0.480877, "boundary_entropy": 0.684833, "bpb": 4.62991}, "ru": {"chunks": 3003.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.15847, "cps": 23.460938, "mean_p": 0.481138, "boundary_entropy": 0.689324, "bpb": 3.265759}, "el": {"chunks": 6068.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.295668, "cps": 47.40625, "mean_p": 0.490466, "boundary_entropy": 0.69028, "bpb": 4.278529}, "hi": {"chunks": 3037.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.180505, "cps": 23.726562, "mean_p": 0.476665, "boundary_entropy": 0.689227, "bpb": 3.200146}, "ta": {"chunks": 5769.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.308585, "cps": 45.070312, "mean_p": 0.492978, "boundary_entropy": 0.691152, "bpb": 3.202591}, "th": {"chunks": 3793.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.23581, "cps": 29.632812, "mean_p": 0.487145, "boundary_entropy": 0.690496, "bpb": 2.929234}, "zh": {"chunks": 4945.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.837284, "cps": 38.632812, "mean_p": 0.498563, "boundary_entropy": 0.687305, "bpb": 6.259082}, "my": {"chunks": 3134.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.1654, "cps": 24.484375, "mean_p": 0.488225, "boundary_entropy": 0.691002, "bpb": 3.645332}}, "bpb": {"en": 4.36286, "de": 4.61974, "vi": 4.96472, "tr": 4.91241, "fi": 4.62991, "ru": 3.26576, "el": 4.27853, "hi": 3.20015, "ta": 3.20259, "th": 2.92923, "zh": 6.25908, "my": 3.64533}, "gini_cpc": 0.238325, "gini_cps": 0.141905, "bpic": 6.880879, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0960766687349528, "vi": 0.8837066441431348, "tr": 0.8311780225398472, "fi": 1.034841527061388, "ru": 0.5523427680018517, "el": 1.0305457607029025, "hi": 0.6291472072788977, "ta": 1.0755673829962642, "th": 0.8219101270843526, "zh": 2.918336772056565, "my": 0.5764984848325337}, "t": 87.8} | |
| {"event": "train", "step": 200, "phase": 1, "objective": "parity_a", "lm_loss": 2.54562, "bpb": 3.67255, "chunk_loss": 0.035058, "F": 0.19495, "G": 0.40626, "chunks": 199.62, "bytes_per_chunk": 5.13, "boundary_entropy": 0.64382, "router_noise": 0.0, "grad_norm": 1.1399, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.275, "n_target_mean": 8.9885, "alpha": 0.03, "t": 105.0} | |
| {"event": "probe", "step": 256, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2939.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.168879, "cps": 22.960938, "mean_p": 0.303975, "boundary_entropy": 0.52425, "bpb": 3.593538}, "de": {"chunks": 3482.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.169581, "cps": 27.203125, "mean_p": 0.301637, "boundary_entropy": 0.527817, "bpb": 3.80867}, "vi": {"chunks": 3592.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.194847, "cps": 28.0625, "mean_p": 0.366347, "boundary_entropy": 0.617383, "bpb": 3.084622}, "tr": {"chunks": 3743.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.206761, "cps": 29.242188, "mean_p": 0.366757, "boundary_entropy": 0.57374, "bpb": 3.771788}, "fi": {"chunks": 2924.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155441, "cps": 22.84375, "mean_p": 0.332638, "boundary_entropy": 0.572247, "bpb": 3.82666}, "ru": {"chunks": 4261.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.224855, "cps": 33.289062, "mean_p": 0.312898, "boundary_entropy": 0.577087, "bpb": 2.653677}, "el": {"chunks": 1907.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.09292, "cps": 14.898438, "mean_p": 0.34468, "boundary_entropy": 0.625841, "bpb": 2.697072}, "hi": {"chunks": 2639.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.15685, "cps": 20.617188, "mean_p": 0.444128, "boundary_entropy": 0.682358, "bpb": 2.101937}, "ta": {"chunks": 1448.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.077454, "cps": 11.3125, "mean_p": 0.452235, "boundary_entropy": 0.685185, "bpb": 1.987384}, "th": {"chunks": 557.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.034629, "cps": 4.351562, "mean_p": 0.412748, "boundary_entropy": 0.671807, "bpb": 2.090565}, "zh": {"chunks": 3147.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.532848, "cps": 24.585938, "mean_p": 0.405483, "boundary_entropy": 0.644024, "bpb": 4.915075}, "my": {"chunks": 3351.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.176852, "cps": 26.179688, "mean_p": 0.446532, "boundary_entropy": 0.681479, "bpb": 2.179344}}, "bpb": {"en": 3.59354, "de": 3.80867, "vi": 3.08462, "tr": 3.77179, "fi": 3.82666, "ru": 2.65368, "el": 2.69707, "hi": 2.10194, "ta": 1.98738, "th": 2.09056, "zh": 4.91508, "my": 2.17934}, "gini_cpc": 0.298888, "gini_cps": 0.193434, "bpic": 11.290997, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0041553205905756, "vi": 1.1537659560507647, "tr": 1.2243168025872593, "fi": 0.9204284180766363, "ru": 1.3314561070706834, "el": 0.5502174781389098, "hi": 0.9287714382640182, "ta": 0.45863545656575383, "th": 0.20504947612118887, "zh": 3.155206837482358, "my": 1.0472143528072713}, "t": 118.3} | |
| {"event": "train", "step": 300, "phase": 1, "objective": "parity_a", "lm_loss": 2.19705, "bpb": 3.16967, "chunk_loss": 0.033136, "F": 0.18738, "G": 0.33785, "chunks": 191.88, "bytes_per_chunk": 5.337, "boundary_entropy": 0.53545, "router_noise": 0.0, "grad_norm": 1.1457, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.252, "n_target_mean": 8.5838, "alpha": 0.03, "t": 130.2} | |
| {"event": "train", "step": 400, "phase": 1, "objective": "parity_a", "lm_loss": 2.06412, "bpb": 2.9779, "chunk_loss": 0.031871, "F": 0.17604, "G": 0.29773, "chunks": 180.27, "bytes_per_chunk": 5.681, "boundary_entropy": 0.45833, "router_noise": 0.0, "grad_norm": 1.3873, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.193, "n_target_mean": 9.1878, "alpha": 0.03, "t": 149.5} | |
| {"event": "train", "step": 500, "phase": 1, "objective": "parity_a", "lm_loss": 1.80824, "bpb": 2.60875, "chunk_loss": 0.03082, "F": 0.15561, "G": 0.26511, "chunks": 159.34, "bytes_per_chunk": 6.426, "boundary_entropy": 0.41058, "router_noise": 0.0, "grad_norm": 0.9005, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.192, "n_target_mean": 9.429, "alpha": 0.03, "t": 168.7} | |
| {"event": "train", "step": 600, "phase": 1, "objective": "parity_a", "lm_loss": 1.78464, "bpb": 2.57469, "chunk_loss": 0.030835, "F": 0.168, "G": 0.23697, "chunks": 172.03, "bytes_per_chunk": 5.952, "boundary_entropy": 0.3296, "router_noise": 0.0, "grad_norm": 0.8498, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 8.8883, "alpha": 0.03, "t": 186.5} | |
| {"event": "train", "step": 700, "phase": 1, "objective": "parity_a", "lm_loss": 1.6838, "bpb": 2.42921, "chunk_loss": 0.030152, "F": 0.14951, "G": 0.23521, "chunks": 153.09, "bytes_per_chunk": 6.689, "boundary_entropy": 0.33938, "router_noise": 0.0, "grad_norm": 0.7665, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 9.6257, "alpha": 0.03, "t": 203.9} | |
| {"event": "train", "step": 800, "phase": 1, "objective": "parity_a", "lm_loss": 1.756, "bpb": 2.53338, "chunk_loss": 0.030554, "F": 0.16457, "G": 0.22509, "chunks": 168.52, "bytes_per_chunk": 6.077, "boundary_entropy": 0.28843, "router_noise": 0.0, "grad_norm": 0.8546, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 8.2854, "alpha": 0.03, "t": 222.5} | |
| {"event": "train", "step": 900, "phase": 1, "objective": "parity_a", "lm_loss": 1.64437, "bpb": 2.37233, "chunk_loss": 0.030663, "F": 0.15866, "G": 0.22863, "chunks": 162.47, "bytes_per_chunk": 6.303, "boundary_entropy": 0.30923, "router_noise": 0.0, "grad_norm": 0.7336, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.2171, "alpha": 0.03, "t": 239.5} | |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3717.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.213584, "cps": 29.039062, "mean_p": 0.219134, "boundary_entropy": 0.137637, "bpb": 2.300779}, "de": {"chunks": 4667.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.227293, "cps": 36.460938, "mean_p": 0.22488, "boundary_entropy": 0.13991, "bpb": 2.572519}, "vi": {"chunks": 3692.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.200271, "cps": 28.84375, "mean_p": 0.179616, "boundary_entropy": 0.202467, "bpb": 2.131658}, "tr": {"chunks": 3831.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.211622, "cps": 29.929688, "mean_p": 0.190388, "boundary_entropy": 0.176178, "bpb": 2.58969}, "fi": {"chunks": 3999.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.212588, "cps": 31.242188, "mean_p": 0.206819, "boundary_entropy": 0.169906, "bpb": 2.857201}, "ru": {"chunks": 2435.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.128496, "cps": 19.023438, "mean_p": 0.161246, "boundary_entropy": 0.33803, "bpb": 1.784217}, "el": {"chunks": 2077.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.101204, "cps": 16.226562, "mean_p": 0.198635, "boundary_entropy": 0.430262, "bpb": 1.827104}, "hi": {"chunks": 2379.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.141397, "cps": 18.585938, "mean_p": 0.236038, "boundary_entropy": 0.493841, "bpb": 1.420334}, "ta": {"chunks": 1523.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.081466, "cps": 11.898438, "mean_p": 0.156703, "boundary_entropy": 0.374751, "bpb": 1.161588}, "th": {"chunks": 2840.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.176562, "cps": 22.1875, "mean_p": 0.191333, "boundary_entropy": 0.395473, "bpb": 1.384389}, "zh": {"chunks": 2811.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.475957, "cps": 21.960938, "mean_p": 0.273033, "boundary_entropy": 0.445947, "bpb": 3.491655}, "my": {"chunks": 4562.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.240764, "cps": 35.640625, "mean_p": 0.291609, "boundary_entropy": 0.55606, "bpb": 1.285904}}, "bpb": {"en": 2.30078, "de": 2.57252, "vi": 2.13166, "tr": 2.58969, "fi": 2.8572, "ru": 1.78422, "el": 1.8271, "hi": 1.42033, "ta": 1.16159, "th": 1.38439, "zh": 3.49166, "my": 1.2859}, "gini_cpc": 0.233887, "gini_cps": 0.170501, "bpic": 9.959801, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0641845581987412, "vi": 0.9376701903794344, "tr": 0.9908163389504154, "fi": 0.9953391342940732, "ru": 0.6016186770759464, "el": 0.47383508028571814, "hi": 0.6620197230427165, "ta": 0.3814222229705387, "th": 0.8266636359553753, "zh": 2.2284298234897166, "my": 1.1272583578215762}, "t": 260.8} | |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.3135, "de": 2.56302, "vi": 2.13325, "tr": 2.58497, "fi": 2.86362, "ru": 1.78318, "el": 1.82084, "hi": 1.42723, "ta": 1.16093, "th": 1.37832, "zh": 3.49575, "my": 1.29831}, "bytes_per_chunk": {"en": 4.779, "de": 4.547, "vi": 7.199, "tr": 5.336, "fi": 5.04, "ru": 15.109, "el": 17.47, "hi": 16.331, "ta": 30.135, "th": 17.526, "zh": 5.595, "my": 12.115}, "bpb_mean": 2.06858, "t": 265.8} | |
| {"event": "train", "step": 1000, "phase": 1, "objective": "parity_a", "lm_loss": 1.57545, "bpb": 2.27289, "chunk_loss": 0.030302, "F": 0.15396, "G": 0.20569, "chunks": 157.66, "bytes_per_chunk": 6.495, "boundary_entropy": 0.253, "router_noise": 0.0, "grad_norm": 0.6986, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.265, "n_target_mean": 8.9577, "alpha": 0.03, "t": 265.9} | |
| {"event": "train", "step": 1100, "phase": 1, "objective": "parity_a", "lm_loss": 1.53584, "bpb": 2.21575, "chunk_loss": 0.03051, "F": 0.1535, "G": 0.20269, "chunks": 157.19, "bytes_per_chunk": 6.515, "boundary_entropy": 0.24841, "router_noise": 0.0, "grad_norm": 0.6038, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 8.9558, "alpha": 0.03, "t": 282.9} | |
| {"event": "train", "step": 1200, "phase": 1, "objective": "parity_a", "lm_loss": 1.53003, "bpb": 2.20736, "chunk_loss": 0.030516, "F": 0.15668, "G": 0.21686, "chunks": 160.44, "bytes_per_chunk": 6.383, "boundary_entropy": 0.25789, "router_noise": 0.0, "grad_norm": 0.6306, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.1676, "alpha": 0.03, "t": 300.4} | |
| {"event": "train", "step": 1300, "phase": 1, "objective": "parity_a", "lm_loss": 1.57089, "bpb": 2.26631, "chunk_loss": 0.030871, "F": 0.17302, "G": 0.21334, "chunks": 177.17, "bytes_per_chunk": 5.78, "boundary_entropy": 0.22067, "router_noise": 0.0, "grad_norm": 0.6099, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 8.5049, "alpha": 0.03, "t": 318.3} | |
| {"event": "train", "step": 1400, "phase": 1, "objective": "parity_a", "lm_loss": 1.4001, "bpb": 2.01992, "chunk_loss": 0.030157, "F": 0.15002, "G": 0.19712, "chunks": 153.62, "bytes_per_chunk": 6.666, "boundary_entropy": 0.22736, "router_noise": 0.0, "grad_norm": 0.5401, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.2869, "alpha": 0.03, "t": 335.2} | |
| {"event": "train", "step": 1500, "phase": 1, "objective": "parity_a", "lm_loss": 1.51181, "bpb": 2.18109, "chunk_loss": 0.030395, "F": 0.16385, "G": 0.20113, "chunks": 167.78, "bytes_per_chunk": 6.103, "boundary_entropy": 0.20613, "router_noise": 0.0, "grad_norm": 0.5478, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.19, "n_target_mean": 8.4518, "alpha": 0.03, "t": 354.2} | |
| {"event": "train", "step": 1600, "phase": 1, "objective": "parity_a", "lm_loss": 1.53348, "bpb": 2.21235, "chunk_loss": 0.030466, "F": 0.1691, "G": 0.20975, "chunks": 173.16, "bytes_per_chunk": 5.914, "boundary_entropy": 0.21188, "router_noise": 0.0, "grad_norm": 0.6052, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 8.32, "alpha": 0.03, "t": 372.7} | |
| {"event": "train", "step": 1700, "phase": 1, "objective": "parity_a", "lm_loss": 1.40874, "bpb": 2.03238, "chunk_loss": 0.030264, "F": 0.14629, "G": 0.19873, "chunks": 149.8, "bytes_per_chunk": 6.836, "boundary_entropy": 0.23694, "router_noise": 0.0, "grad_norm": 0.524, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.166, "n_target_mean": 9.7805, "alpha": 0.03, "t": 389.4} | |
| {"event": "train", "step": 1800, "phase": 1, "objective": "parity_a", "lm_loss": 1.42547, "bpb": 2.05652, "chunk_loss": 0.030319, "F": 0.14882, "G": 0.2013, "chunks": 152.39, "bytes_per_chunk": 6.72, "boundary_entropy": 0.23396, "router_noise": 0.0, "grad_norm": 0.511, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.311, "alpha": 0.03, "t": 406.4} | |
| {"event": "train", "step": 1900, "phase": 1, "objective": "parity_a", "lm_loss": 1.40912, "bpb": 2.03293, "chunk_loss": 0.030314, "F": 0.15134, "G": 0.20023, "chunks": 154.97, "bytes_per_chunk": 6.608, "boundary_entropy": 0.22353, "router_noise": 0.0, "grad_norm": 0.5163, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.182, "n_target_mean": 9.4579, "alpha": 0.03, "t": 424.6} | |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3686.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.211803, "cps": 28.796875, "mean_p": 0.21117, "boundary_entropy": 0.092812, "bpb": 2.062595}, "de": {"chunks": 4549.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.221546, "cps": 35.539062, "mean_p": 0.216106, "boundary_entropy": 0.106945, "bpb": 2.2605}, "vi": {"chunks": 3711.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.201302, "cps": 28.992188, "mean_p": 0.17258, "boundary_entropy": 0.150299, "bpb": 1.958868}, "tr": {"chunks": 3328.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.183837, "cps": 26.0, "mean_p": 0.176878, "boundary_entropy": 0.125951, "bpb": 2.252244}, "fi": {"chunks": 3614.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.192122, "cps": 28.234375, "mean_p": 0.191978, "boundary_entropy": 0.119651, "bpb": 2.441742}, "ru": {"chunks": 2845.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.150132, "cps": 22.226562, "mean_p": 0.175678, "boundary_entropy": 0.330109, "bpb": 1.455685}, "el": {"chunks": 4043.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.196998, "cps": 31.585938, "mean_p": 0.235169, "boundary_entropy": 0.448515, "bpb": 1.534799}, "hi": {"chunks": 3623.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.215334, "cps": 28.304688, "mean_p": 0.179869, "boundary_entropy": 0.361998, "bpb": 1.242504}, "ta": {"chunks": 2409.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.128858, "cps": 18.820312, "mean_p": 0.15488, "boundary_entropy": 0.356676, "bpb": 1.024699}, "th": {"chunks": 2537.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.157725, "cps": 19.820312, "mean_p": 0.186693, "boundary_entropy": 0.385788, "bpb": 1.167659}, "zh": {"chunks": 3113.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.527091, "cps": 24.320312, "mean_p": 0.254915, "boundary_entropy": 0.332176, "bpb": 3.179263}, "my": {"chunks": 2767.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.146031, "cps": 21.617188, "mean_p": 0.210539, "boundary_entropy": 0.444556, "bpb": 1.145293}}, "bpb": {"en": 2.06259, "de": 2.2605, "vi": 1.95887, "tr": 2.25224, "fi": 2.44174, "ru": 1.45568, "el": 1.5348, "hi": 1.2425, "ta": 1.0247, "th": 1.16766, "zh": 3.17926, "my": 1.14529}, "gini_cpc": 0.192288, "gini_cps": 0.10301, "bpic": 9.540858, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0460015280776, "vi": 0.9504222649684533, "tr": 0.867963686062532, "fi": 0.9070788778099468, "ru": 0.7088296012724463, "el": 0.9301043713405568, "hi": 1.016674780039973, "ta": 0.6083872722256946, "th": 0.7446774300459897, "zh": 2.48859639323171, "my": 0.689468727331082}, "t": 443.8} | |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.07206, "de": 2.25125, "vi": 1.96066, "tr": 2.24504, "fi": 2.4542, "ru": 1.45614, "el": 1.52945, "hi": 1.2499, "ta": 1.0257, "th": 1.16503, "zh": 3.18792, "my": 1.15466}, "bytes_per_chunk": {"en": 4.863, "de": 4.726, "vi": 7.275, "tr": 6.051, "fi": 5.415, "ru": 12.357, "el": 9.442, "hi": 11.729, "ta": 21.259, "th": 18.17, "zh": 5.477, "my": 19.94}, "bpb_mean": 1.81267, "t": 448.6} | |
| {"event": "train", "step": 2000, "phase": 1, "objective": "parity_a", "lm_loss": 1.40191, "bpb": 2.02253, "chunk_loss": 0.030318, "F": 0.15926, "G": 0.19886, "chunks": 163.08, "bytes_per_chunk": 6.279, "boundary_entropy": 0.19789, "router_noise": 0.0, "grad_norm": 0.4718, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.242, "n_target_mean": 8.8903, "alpha": 0.03, "t": 448.8} | |
| {"event": "train", "step": 2100, "phase": 1, "objective": "parity_a", "lm_loss": 1.46138, "bpb": 2.10832, "chunk_loss": 0.030065, "F": 0.15311, "G": 0.19648, "chunks": 156.78, "bytes_per_chunk": 6.531, "boundary_entropy": 0.21957, "router_noise": 0.0, "grad_norm": 0.4478, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.166, "n_target_mean": 8.7438, "alpha": 0.03, "t": 465.4} | |
| {"event": "train", "step": 2200, "phase": 1, "objective": "parity_a", "lm_loss": 1.41106, "bpb": 2.03574, "chunk_loss": 0.030728, "F": 0.15964, "G": 0.19727, "chunks": 163.47, "bytes_per_chunk": 6.264, "boundary_entropy": 0.2004, "router_noise": 0.0, "grad_norm": 0.4363, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.165, "n_target_mean": 8.9385, "alpha": 0.03, "t": 481.9} | |
| {"event": "train", "step": 2300, "phase": 1, "objective": "parity_a", "lm_loss": 1.35093, "bpb": 1.94899, "chunk_loss": 0.030288, "F": 0.15067, "G": 0.19263, "chunks": 154.28, "bytes_per_chunk": 6.637, "boundary_entropy": 0.20722, "router_noise": 0.0, "grad_norm": 0.4495, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.183, "n_target_mean": 8.9615, "alpha": 0.03, "t": 500.2} | |
| {"event": "train", "step": 2400, "phase": 1, "objective": "parity_a", "lm_loss": 1.36849, "bpb": 1.97432, "chunk_loss": 0.029947, "F": 0.14279, "G": 0.18714, "chunks": 146.22, "bytes_per_chunk": 7.003, "boundary_entropy": 0.21039, "router_noise": 0.0, "grad_norm": 0.4402, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.188, "n_target_mean": 9.3689, "alpha": 0.03, "t": 519.0} | |
| {"event": "train", "step": 2500, "phase": 1, "objective": "parity_a", "lm_loss": 1.2293, "bpb": 1.77351, "chunk_loss": 0.030227, "F": 0.13748, "G": 0.19171, "chunks": 140.78, "bytes_per_chunk": 7.274, "boundary_entropy": 0.23424, "router_noise": 0.0, "grad_norm": 0.4345, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.183, "n_target_mean": 10.1802, "alpha": 0.03, "t": 537.3} | |
| {"event": "train", "step": 2600, "phase": 1, "objective": "parity_a", "lm_loss": 1.22027, "bpb": 1.76047, "chunk_loss": 0.030247, "F": 0.13623, "G": 0.18788, "chunks": 139.5, "bytes_per_chunk": 7.341, "boundary_entropy": 0.22923, "router_noise": 0.0, "grad_norm": 0.4211, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.183, "n_target_mean": 9.7041, "alpha": 0.03, "t": 555.6} | |
| {"event": "train", "step": 2700, "phase": 1, "objective": "parity_a", "lm_loss": 1.36409, "bpb": 1.96796, "chunk_loss": 0.030362, "F": 0.1619, "G": 0.19788, "chunks": 165.78, "bytes_per_chunk": 6.177, "boundary_entropy": 0.19068, "router_noise": 0.0, "grad_norm": 0.4311, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 8.6479, "alpha": 0.03, "t": 574.0} | |
| {"event": "train", "step": 2800, "phase": 1, "objective": "parity_a", "lm_loss": 1.18505, "bpb": 1.70967, "chunk_loss": 0.030257, "F": 0.13876, "G": 0.18885, "chunks": 142.09, "bytes_per_chunk": 7.207, "boundary_entropy": 0.22272, "router_noise": 0.0, "grad_norm": 0.4014, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.184, "n_target_mean": 10.0068, "alpha": 0.03, "t": 592.4} | |
| {"event": "train", "step": 2900, "phase": 1, "objective": "parity_a", "lm_loss": 1.2326, "bpb": 1.77827, "chunk_loss": 0.030325, "F": 0.1438, "G": 0.19189, "chunks": 147.25, "bytes_per_chunk": 6.954, "boundary_entropy": 0.21792, "router_noise": 0.0, "grad_norm": 0.419, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.188, "n_target_mean": 9.7749, "alpha": 0.03, "t": 611.3} | |
| {"event": "probe", "step": 3000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3554.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.204218, "cps": 27.765625, "mean_p": 0.203517, "boundary_entropy": 0.083503, "bpb": 1.898053}, "de": {"chunks": 4500.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.219159, "cps": 35.15625, "mean_p": 0.2128, "boundary_entropy": 0.094601, "bpb": 2.065671}, "vi": {"chunks": 3365.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.182533, "cps": 26.289062, "mean_p": 0.154908, "boundary_entropy": 0.137921, "bpb": 1.816377}, "tr": {"chunks": 3155.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174281, "cps": 24.648438, "mean_p": 0.168201, "boundary_entropy": 0.112224, "bpb": 2.017235}, "fi": {"chunks": 3530.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.187656, "cps": 27.578125, "mean_p": 0.185355, "boundary_entropy": 0.105401, "bpb": 2.237592}, "ru": {"chunks": 2515.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.132718, "cps": 19.648438, "mean_p": 0.159723, "boundary_entropy": 0.304195, "bpb": 1.337604}, "el": {"chunks": 2616.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.127467, "cps": 20.4375, "mean_p": 0.183796, "boundary_entropy": 0.387104, "bpb": 1.397711}, "hi": {"chunks": 2775.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.164933, "cps": 21.679688, "mean_p": 0.154532, "boundary_entropy": 0.331217, "bpb": 1.1543}, "ta": {"chunks": 2532.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.135437, "cps": 19.78125, "mean_p": 0.169339, "boundary_entropy": 0.373576, "bpb": 0.959149}, "th": {"chunks": 2511.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.156108, "cps": 19.617188, "mean_p": 0.174815, "boundary_entropy": 0.366591, "bpb": 1.085798}, "zh": {"chunks": 3484.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.589909, "cps": 27.21875, "mean_p": 0.255411, "boundary_entropy": 0.279645, "bpb": 2.968507}, "my": {"chunks": 2138.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.112835, "cps": 16.703125, "mean_p": 0.159295, "boundary_entropy": 0.357585, "bpb": 1.07846}}, "bpb": {"en": 1.89805, "de": 2.06567, "vi": 1.81638, "tr": 2.01723, "fi": 2.23759, "ru": 1.3376, "el": 1.39771, "hi": 1.1543, "ta": 0.95915, "th": 1.0858, "zh": 2.96851, "my": 1.07846}, "gini_cpc": 0.241674, "gini_cps": 0.113295, "bpic": 10.464376, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0731657489963382, "vi": 0.8938170264380821, "tr": 0.8534056403501865, "fi": 0.9189026811052705, "ru": 0.6498834417498289, "el": 0.6241710064042508, "hi": 0.8076340326827608, "ta": 0.6632006276749822, "th": 0.7644205332217054, "zh": 2.8886265619637306, "my": 0.5525238400888419}, "t": 632.8} | |
| {"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 1.9008, "de": 2.06753, "vi": 1.82528, "tr": 2.01064, "fi": 2.24808, "ru": 1.33695, "el": 1.39276, "hi": 1.16125, "ta": 0.96008, "th": 1.08369, "zh": 2.9773, "my": 1.08769}, "bytes_per_chunk": {"en": 5.067, "de": 4.803, "vi": 8.137, "tr": 6.46, "fi": 5.563, "ru": 14.695, "el": 15.159, "hi": 16.337, "ta": 20.017, "th": 19.645, "zh": 4.83, "my": 23.803}, "bpb_mean": 1.671, "t": 637.7} | |
| {"event": "train", "step": 3000, "phase": 1, "objective": "parity_a", "lm_loss": 1.26282, "bpb": 1.82186, "chunk_loss": 0.030129, "F": 0.14668, "G": 0.17928, "chunks": 150.2, "bytes_per_chunk": 6.817, "boundary_entropy": 0.17655, "router_noise": 0.0, "grad_norm": 0.4148, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.265, "n_target_mean": 8.6784, "alpha": 0.03, "t": 637.8} | |
| {"event": "train", "step": 3100, "phase": 1, "objective": "parity_a", "lm_loss": 1.19035, "bpb": 1.71732, "chunk_loss": 0.030307, "F": 0.1474, "G": 0.18682, "chunks": 150.94, "bytes_per_chunk": 6.784, "boundary_entropy": 0.19099, "router_noise": 0.0, "grad_norm": 0.4224, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.6548, "alpha": 0.03, "t": 655.1} | |
| {"event": "train", "step": 3200, "phase": 1, "objective": "parity_a", "lm_loss": 1.25931, "bpb": 1.8168, "chunk_loss": 0.030089, "F": 0.13693, "G": 0.18625, "chunks": 140.22, "bytes_per_chunk": 7.303, "boundary_entropy": 0.21228, "router_noise": 0.0, "grad_norm": 0.4392, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 10.1171, "alpha": 0.03, "t": 673.0} | |
| {"event": "train", "step": 3300, "phase": 1, "objective": "parity_a", "lm_loss": 1.25749, "bpb": 1.81418, "chunk_loss": 0.030196, "F": 0.14819, "G": 0.18622, "chunks": 151.75, "bytes_per_chunk": 6.748, "boundary_entropy": 0.19004, "router_noise": 0.0, "grad_norm": 0.4221, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 8.7235, "alpha": 0.03, "t": 690.8} | |
| {"event": "train", "step": 3400, "phase": 1, "objective": "parity_a", "lm_loss": 1.31838, "bpb": 1.90202, "chunk_loss": 0.030336, "F": 0.15807, "G": 0.19216, "chunks": 161.86, "bytes_per_chunk": 6.326, "boundary_entropy": 0.18467, "router_noise": 0.0, "grad_norm": 0.4382, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 8.3404, "alpha": 0.03, "t": 708.6} | |
| {"event": "train", "step": 3500, "phase": 1, "objective": "parity_a", "lm_loss": 1.17157, "bpb": 1.69022, "chunk_loss": 0.030203, "F": 0.14633, "G": 0.19063, "chunks": 149.84, "bytes_per_chunk": 6.834, "boundary_entropy": 0.20413, "router_noise": 0.0, "grad_norm": 0.405, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 9.2986, "alpha": 0.03, "t": 726.5} | |
| {"event": "train", "step": 3600, "phase": 1, "objective": "parity_a", "lm_loss": 1.21892, "bpb": 1.75853, "chunk_loss": 0.030612, "F": 0.15559, "G": 0.19736, "chunks": 159.33, "bytes_per_chunk": 6.427, "boundary_entropy": 0.19987, "router_noise": 0.0, "grad_norm": 0.4134, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.112, "alpha": 0.03, "t": 744.4} | |
| {"event": "train", "step": 3700, "phase": 1, "objective": "parity_a", "lm_loss": 1.21422, "bpb": 1.75175, "chunk_loss": 0.030331, "F": 0.14072, "G": 0.18708, "chunks": 144.09, "bytes_per_chunk": 7.106, "boundary_entropy": 0.21349, "router_noise": 0.0, "grad_norm": 0.4113, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 9.5465, "alpha": 0.03, "t": 762.3} | |
| {"event": "train", "step": 3800, "phase": 1, "objective": "parity_a", "lm_loss": 1.23402, "bpb": 1.78031, "chunk_loss": 0.03025, "F": 0.1512, "G": 0.19193, "chunks": 154.83, "bytes_per_chunk": 6.614, "boundary_entropy": 0.19431, "router_noise": 0.0, "grad_norm": 0.399, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 8.8211, "alpha": 0.03, "t": 780.2} | |
| {"event": "train", "step": 3900, "phase": 1, "objective": "parity_a", "lm_loss": 1.2109, "bpb": 1.74695, "chunk_loss": 0.030193, "F": 0.15163, "G": 0.18701, "chunks": 155.27, "bytes_per_chunk": 6.595, "boundary_entropy": 0.18404, "router_noise": 0.0, "grad_norm": 0.3985, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 8.7584, "alpha": 0.03, "t": 798.0} | |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3542.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203528, "cps": 27.671875, "mean_p": 0.200665, "boundary_entropy": 0.071999, "bpb": 1.81734}, "de": {"chunks": 4410.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.214776, "cps": 34.453125, "mean_p": 0.208383, "boundary_entropy": 0.085521, "bpb": 1.96826}, "vi": {"chunks": 3408.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.184866, "cps": 26.625, "mean_p": 0.161169, "boundary_entropy": 0.131372, "bpb": 1.711786}, "tr": {"chunks": 3150.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174004, "cps": 24.609375, "mean_p": 0.169602, "boundary_entropy": 0.106248, "bpb": 1.907607}, "fi": {"chunks": 3467.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.184307, "cps": 27.085938, "mean_p": 0.184965, "boundary_entropy": 0.094874, "bpb": 2.104507}, "ru": {"chunks": 3010.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.158839, "cps": 23.515625, "mean_p": 0.185551, "boundary_entropy": 0.336423, "bpb": 1.276826}, "el": {"chunks": 3193.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.155582, "cps": 24.945312, "mean_p": 0.199835, "boundary_entropy": 0.398499, "bpb": 1.323021}, "hi": {"chunks": 3165.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.188113, "cps": 24.726562, "mean_p": 0.173596, "boundary_entropy": 0.354134, "bpb": 1.093107}, "ta": {"chunks": 3105.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.166087, "cps": 24.257812, "mean_p": 0.171384, "boundary_entropy": 0.37297, "bpb": 0.942319}, "th": {"chunks": 3414.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.212247, "cps": 26.671875, "mean_p": 0.195711, "boundary_entropy": 0.38458, "bpb": 1.040706}, "zh": {"chunks": 3472.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.587877, "cps": 27.125, "mean_p": 0.253486, "boundary_entropy": 0.242945, "bpb": 2.842761}, "my": {"chunks": 2506.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.132257, "cps": 19.578125, "mean_p": 0.155803, "boundary_entropy": 0.346956, "bpb": 1.037298}}, "bpb": {"en": 1.81734, "de": 1.96826, "vi": 1.71179, "tr": 1.90761, "fi": 2.10451, "ru": 1.27683, "el": 1.32302, "hi": 1.09311, "ta": 0.94232, "th": 1.04071, "zh": 2.84276, "my": 1.0373}, "gini_cpc": 0.199953, "gini_cps": 0.064522, "bpic": 9.632574, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0552655139735534, "vi": 0.908305632881994, "tr": 0.8549398599450355, "fi": 0.905560606275729, "ru": 0.780428003796135, "el": 0.7644228082728683, "hi": 0.9242600993553897, "ta": 0.8160404719645158, "th": 1.0428408018033783, "zh": 2.8884299345878577, "my": 0.6498202694956282}, "t": 819.3} | |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.8188, "de": 1.97129, "vi": 1.72258, "tr": 1.90035, "fi": 2.11503, "ru": 1.27703, "el": 1.32164, "hi": 1.0997, "ta": 0.94472, "th": 1.03865, "zh": 2.85154, "my": 1.04577}, "bytes_per_chunk": {"en": 5.098, "de": 4.906, "vi": 8.025, "tr": 6.416, "fi": 5.648, "ru": 11.93, "el": 11.968, "hi": 13.822, "ta": 15.997, "th": 14.085, "zh": 4.912, "my": 21.363}, "bpb_mean": 1.59226, "t": 824.5} | |
| {"event": "train", "step": 4000, "phase": 1, "objective": "parity_a", "lm_loss": 1.14457, "bpb": 1.65126, "chunk_loss": 0.030309, "F": 0.14484, "G": 0.18779, "chunks": 148.31, "bytes_per_chunk": 6.904, "boundary_entropy": 0.19546, "router_noise": 0.0, "grad_norm": 0.3796, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.268, "n_target_mean": 9.4453, "alpha": 0.03, "t": 824.8} | |
| {"event": "train", "step": 4100, "phase": 1, "objective": "parity_a", "lm_loss": 1.22211, "bpb": 1.76313, "chunk_loss": 0.030259, "F": 0.15314, "G": 0.19143, "chunks": 156.81, "bytes_per_chunk": 6.53, "boundary_entropy": 0.18931, "router_noise": 0.0, "grad_norm": 0.399, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 8.5221, "alpha": 0.03, "t": 841.6} | |
| {"event": "train", "step": 4200, "phase": 1, "objective": "parity_a", "lm_loss": 1.20162, "bpb": 1.73357, "chunk_loss": 0.030105, "F": 0.13861, "G": 0.18504, "chunks": 141.94, "bytes_per_chunk": 7.214, "boundary_entropy": 0.20764, "router_noise": 0.0, "grad_norm": 0.415, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 9.4361, "alpha": 0.03, "t": 859.7} | |
| {"event": "train", "step": 4300, "phase": 1, "objective": "parity_a", "lm_loss": 1.13132, "bpb": 1.63215, "chunk_loss": 0.030165, "F": 0.13875, "G": 0.18584, "chunks": 142.08, "bytes_per_chunk": 7.207, "boundary_entropy": 0.20725, "router_noise": 0.0, "grad_norm": 0.3799, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.8096, "alpha": 0.03, "t": 876.9} | |
| {"event": "phase_transition", "step": 4400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 894.7} | |
| {"event": "train", "step": 4400, "phase": 2, "objective": "parity_a", "lm_loss": 1.17081, "bpb": 1.68912, "chunk_loss": 0.030226, "F": 0.14688, "G": 0.18388, "chunks": 150.41, "bytes_per_chunk": 6.808, "boundary_entropy": 0.18041, "router_noise": 0.0, "grad_norm": 0.3961, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.18, "n_target_mean": 8.9436, "alpha": 0.03, "t": 894.8} | |
| {"event": "train", "step": 4500, "phase": 2, "objective": "parity_a", "lm_loss": 1.26438, "bpb": 1.82411, "chunk_loss": 0.030229, "F": 0.16602, "G": 0.19062, "chunks": 170.0, "bytes_per_chunk": 6.024, "boundary_entropy": 0.14655, "router_noise": 0.0, "grad_norm": 0.385, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 7.6421, "alpha": 0.03, "t": 912.9} | |
| {"event": "train", "step": 4600, "phase": 2, "objective": "parity_a", "lm_loss": 1.113, "bpb": 1.60572, "chunk_loss": 0.030176, "F": 0.14267, "G": 0.18281, "chunks": 146.09, "bytes_per_chunk": 7.009, "boundary_entropy": 0.18726, "router_noise": 0.0, "grad_norm": 0.3877, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.3412, "alpha": 0.03, "t": 930.1} | |
| {"event": "train", "step": 4700, "phase": 2, "objective": "parity_a", "lm_loss": 1.18474, "bpb": 1.70921, "chunk_loss": 0.030174, "F": 0.14845, "G": 0.18456, "chunks": 152.02, "bytes_per_chunk": 6.736, "boundary_entropy": 0.17676, "router_noise": 0.0, "grad_norm": 0.3798, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 8.9439, "alpha": 0.03, "t": 948.0} | |
| {"event": "train", "step": 4800, "phase": 2, "objective": "parity_a", "lm_loss": 1.16099, "bpb": 1.67495, "chunk_loss": 0.030151, "F": 0.1525, "G": 0.18716, "chunks": 156.16, "bytes_per_chunk": 6.558, "boundary_entropy": 0.16805, "router_noise": 0.0, "grad_norm": 0.3991, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.9799, "alpha": 0.03, "t": 965.0} | |
| {"event": "train", "step": 4900, "phase": 2, "objective": "parity_a", "lm_loss": 1.22876, "bpb": 1.77273, "chunk_loss": 0.030192, "F": 0.1615, "G": 0.18739, "chunks": 165.38, "bytes_per_chunk": 6.192, "boundary_entropy": 0.15144, "router_noise": 0.0, "grad_norm": 0.3924, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 8.1677, "alpha": 0.03, "t": 982.6} | |
| {"event": "probe", "step": 5000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3555.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.204275, "cps": 27.773438, "mean_p": 0.201055, "boundary_entropy": 0.064499, "bpb": 1.767945}, "de": {"chunks": 4389.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.213753, "cps": 34.289062, "mean_p": 0.208011, "boundary_entropy": 0.073582, "bpb": 1.909305}, "vi": {"chunks": 3450.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.187144, "cps": 26.953125, "mean_p": 0.160613, "boundary_entropy": 0.119597, "bpb": 1.635}, "tr": {"chunks": 3163.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174722, "cps": 24.710938, "mean_p": 0.167359, "boundary_entropy": 0.090315, "bpb": 1.848685}, "fi": {"chunks": 3471.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18452, "cps": 27.117188, "mean_p": 0.182853, "boundary_entropy": 0.082417, "bpb": 2.042029}, "ru": {"chunks": 2944.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.155356, "cps": 23.0, "mean_p": 0.175902, "boundary_entropy": 0.318158, "bpb": 1.233105}, "el": {"chunks": 2778.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.13536, "cps": 21.703125, "mean_p": 0.182476, "boundary_entropy": 0.374711, "bpb": 1.268238}, "hi": {"chunks": 3261.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.193819, "cps": 25.476562, "mean_p": 0.16271, "boundary_entropy": 0.325194, "bpb": 1.048621}, "ta": {"chunks": 2495.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.133458, "cps": 19.492188, "mean_p": 0.157409, "boundary_entropy": 0.351781, "bpb": 0.881349}, "th": {"chunks": 3347.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.208082, "cps": 26.148438, "mean_p": 0.189001, "boundary_entropy": 0.370984, "bpb": 0.991534}, "zh": {"chunks": 3275.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.554521, "cps": 25.585938, "mean_p": 0.235865, "boundary_entropy": 0.216981, "bpb": 2.712603}, "my": {"chunks": 3052.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.161072, "cps": 23.84375, "mean_p": 0.167399, "boundary_entropy": 0.355489, "bpb": 1.002592}}, "bpb": {"en": 1.76795, "de": 1.9093, "vi": 1.635, "tr": 1.84869, "fi": 2.04203, "ru": 1.23311, "el": 1.26824, "hi": 1.04862, "ta": 0.88135, "th": 0.99153, "zh": 2.7126, "my": 1.00259}, "gini_cpc": 0.198924, "gini_cps": 0.072469, "bpic": 9.795329, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0463998984037428, "vi": 0.9161370947530324, "tr": 0.8553289183716898, "fi": 0.903290089491134, "ru": 0.7605243200516572, "el": 0.6626373497594857, "hi": 0.9488121579175053, "ta": 0.6533254790705337, "th": 1.0186363320387866, "zh": 2.7145783234099343, "my": 0.7885072091596009}, "t": 1003.0} | |
| {"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.76731, "de": 1.91239, "vi": 1.64797, "tr": 1.84124, "fi": 2.05261, "ru": 1.23271, "el": 1.26613, "hi": 1.05404, "ta": 0.88471, "th": 0.99004, "zh": 2.71995, "my": 1.01068}, "bytes_per_chunk": {"en": 5.101, "de": 4.932, "vi": 7.926, "tr": 6.399, "fi": 5.633, "ru": 12.302, "el": 13.719, "hi": 13.095, "ta": 19.532, "th": 14.693, "zh": 5.159, "my": 17.597}, "bpb_mean": 1.53165, "t": 1007.7} | |
| {"event": "train", "step": 5000, "phase": 2, "objective": "parity_a", "lm_loss": 1.16461, "bpb": 1.68018, "chunk_loss": 0.030206, "F": 0.15224, "G": 0.1852, "chunks": 155.89, "bytes_per_chunk": 6.569, "boundary_entropy": 0.16762, "router_noise": 0.0, "grad_norm": 0.3712, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.253, "n_target_mean": 9.2014, "alpha": 0.03, "t": 1007.9} | |
| {"event": "train", "step": 5100, "phase": 2, "objective": "parity_a", "lm_loss": 1.17204, "bpb": 1.69089, "chunk_loss": 0.030132, "F": 0.15303, "G": 0.18052, "chunks": 156.7, "bytes_per_chunk": 6.535, "boundary_entropy": 0.15261, "router_noise": 0.0, "grad_norm": 0.4114, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 8.5306, "alpha": 0.03, "t": 1025.7} | |
| {"event": "train", "step": 5200, "phase": 2, "objective": "parity_a", "lm_loss": 1.17337, "bpb": 1.69281, "chunk_loss": 0.030133, "F": 0.14958, "G": 0.18195, "chunks": 153.17, "bytes_per_chunk": 6.685, "boundary_entropy": 0.16444, "router_noise": 0.0, "grad_norm": 0.4018, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.802, "alpha": 0.03, "t": 1042.7} | |
| {"event": "train", "step": 5300, "phase": 2, "objective": "parity_a", "lm_loss": 1.13597, "bpb": 1.63886, "chunk_loss": 0.030131, "F": 0.14499, "G": 0.17804, "chunks": 148.47, "bytes_per_chunk": 6.897, "boundary_entropy": 0.16409, "router_noise": 0.0, "grad_norm": 0.3897, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 8.9067, "alpha": 0.03, "t": 1060.3} | |
| {"event": "train", "step": 5400, "phase": 2, "objective": "parity_a", "lm_loss": 1.10365, "bpb": 1.59223, "chunk_loss": 0.030067, "F": 0.14574, "G": 0.17773, "chunks": 149.23, "bytes_per_chunk": 6.862, "boundary_entropy": 0.1594, "router_noise": 0.0, "grad_norm": 0.3643, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.9318, "alpha": 0.03, "t": 1077.3} | |
| {"event": "train", "step": 5500, "phase": 2, "objective": "parity_a", "lm_loss": 1.10726, "bpb": 1.59744, "chunk_loss": 0.030092, "F": 0.13808, "G": 0.17782, "chunks": 141.39, "bytes_per_chunk": 7.242, "boundary_entropy": 0.181, "router_noise": 0.0, "grad_norm": 0.3667, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.165, "n_target_mean": 9.5417, "alpha": 0.03, "t": 1093.9} | |
| {"event": "train", "step": 5600, "phase": 2, "objective": "parity_a", "lm_loss": 1.12092, "bpb": 1.61714, "chunk_loss": 0.030104, "F": 0.14236, "G": 0.1814, "chunks": 145.78, "bytes_per_chunk": 7.024, "boundary_entropy": 0.1815, "router_noise": 0.0, "grad_norm": 0.3834, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.1849, "alpha": 0.03, "t": 1111.1} | |
| {"event": "train", "step": 5700, "phase": 2, "objective": "parity_a", "lm_loss": 1.15581, "bpb": 1.66749, "chunk_loss": 0.030195, "F": 0.15543, "G": 0.1844, "chunks": 159.16, "bytes_per_chunk": 6.434, "boundary_entropy": 0.15772, "router_noise": 0.0, "grad_norm": 0.3842, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.6518, "alpha": 0.03, "t": 1128.2} | |
| {"event": "train", "step": 5800, "phase": 2, "objective": "parity_a", "lm_loss": 1.08193, "bpb": 1.5609, "chunk_loss": 0.030114, "F": 0.13359, "G": 0.17502, "chunks": 136.8, "bytes_per_chunk": 7.486, "boundary_entropy": 0.18379, "router_noise": 0.0, "grad_norm": 0.3539, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.9423, "alpha": 0.03, "t": 1145.7} | |
| {"event": "train", "step": 5900, "phase": 2, "objective": "parity_a", "lm_loss": 1.11287, "bpb": 1.60552, "chunk_loss": 0.030128, "F": 0.13882, "G": 0.17778, "chunks": 142.16, "bytes_per_chunk": 7.203, "boundary_entropy": 0.17832, "router_noise": 0.0, "grad_norm": 0.372, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.2794, "alpha": 0.03, "t": 1163.0} | |
| {"event": "probe", "step": 6000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3511.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.201747, "cps": 27.429688, "mean_p": 0.198465, "boundary_entropy": 0.061628, "bpb": 1.729447}, "de": {"chunks": 4363.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.212487, "cps": 34.085938, "mean_p": 0.204529, "boundary_entropy": 0.075074, "bpb": 1.863749}, "vi": {"chunks": 3393.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.184052, "cps": 26.507812, "mean_p": 0.154769, "boundary_entropy": 0.105878, "bpb": 1.578779}, "tr": {"chunks": 3078.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.170027, "cps": 24.046875, "mean_p": 0.1633, "boundary_entropy": 0.085148, "bpb": 1.78947}, "fi": {"chunks": 3277.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.174207, "cps": 25.601562, "mean_p": 0.172324, "boundary_entropy": 0.075366, "bpb": 1.987413}, "ru": {"chunks": 2944.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.155356, "cps": 23.0, "mean_p": 0.178577, "boundary_entropy": 0.320169, "bpb": 1.196665}, "el": {"chunks": 3101.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.151099, "cps": 24.226562, "mean_p": 0.191059, "boundary_entropy": 0.386246, "bpb": 1.226307}, "hi": {"chunks": 2846.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.169153, "cps": 22.234375, "mean_p": 0.144344, "boundary_entropy": 0.299987, "bpb": 1.003707}, "ta": {"chunks": 3123.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.16705, "cps": 24.398438, "mean_p": 0.15748, "boundary_entropy": 0.339597, "bpb": 0.858344}, "th": {"chunks": 2668.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.165869, "cps": 20.84375, "mean_p": 0.163752, "boundary_entropy": 0.340506, "bpb": 0.955435}, "zh": {"chunks": 3359.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.568744, "cps": 26.242188, "mean_p": 0.234474, "boundary_entropy": 0.193126, "bpb": 2.618839}, "my": {"chunks": 2872.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.151573, "cps": 22.4375, "mean_p": 0.159385, "boundary_entropy": 0.338456, "bpb": 0.964074}}, "bpb": {"en": 1.72945, "de": 1.86375, "vi": 1.57878, "tr": 1.78947, "fi": 1.98741, "ru": 1.19666, "el": 1.22631, "hi": 1.00371, "ta": 0.85834, "th": 0.95543, "zh": 2.61884, "my": 0.96407}, "gini_cpc": 0.188166, "gini_cps": 0.066976, "bpic": 9.959284, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0532369737550946, "vi": 0.9122922981335655, "tr": 0.8427744382979682, "fi": 0.8634910662466667, "ru": 0.7700552428891032, "el": 0.7489523914045322, "hi": 0.8384421704995931, "ta": 0.8280180526125471, "th": 0.8221632319119261, "zh": 2.819095912869456, "my": 0.7513016631525745}, "t": 1182.9} | |
| {"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.72877, "de": 1.86809, "vi": 1.59382, "tr": 1.78281, "fi": 1.99977, "ru": 1.19646, "el": 1.22333, "hi": 1.00901, "ta": 0.86143, "th": 0.9541, "zh": 2.625, "my": 0.97152}, "bytes_per_chunk": {"en": 5.14, "de": 4.973, "vi": 8.015, "tr": 6.515, "fi": 5.893, "ru": 12.021, "el": 12.374, "hi": 14.607, "ta": 16.287, "th": 17.703, "zh": 5.086, "my": 17.747}, "bpb_mean": 1.48451, "t": 1187.9} | |
| {"event": "train", "step": 6000, "phase": 2, "objective": "parity_a", "lm_loss": 1.11835, "bpb": 1.61344, "chunk_loss": 0.030249, "F": 0.1456, "G": 0.18283, "chunks": 149.09, "bytes_per_chunk": 6.868, "boundary_entropy": 0.18025, "router_noise": 0.0, "grad_norm": 0.3763, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.25, "n_target_mean": 9.5014, "alpha": 0.03, "t": 1188.0} | |
| {"event": "train", "step": 6100, "phase": 2, "objective": "parity_a", "lm_loss": 1.18319, "bpb": 1.70698, "chunk_loss": 0.030178, "F": 0.1505, "G": 0.18444, "chunks": 154.11, "bytes_per_chunk": 6.645, "boundary_entropy": 0.17022, "router_noise": 0.0, "grad_norm": 0.3893, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.0707, "alpha": 0.03, "t": 1205.2} | |
| {"event": "train", "step": 6200, "phase": 2, "objective": "parity_a", "lm_loss": 1.16658, "bpb": 1.68302, "chunk_loss": 0.030182, "F": 0.14664, "G": 0.17723, "chunks": 150.16, "bytes_per_chunk": 6.82, "boundary_entropy": 0.16281, "router_noise": 0.0, "grad_norm": 0.3814, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.6046, "alpha": 0.03, "t": 1222.5} | |
| {"event": "train", "step": 6300, "phase": 2, "objective": "parity_a", "lm_loss": 1.194, "bpb": 1.72258, "chunk_loss": 0.030391, "F": 0.16043, "G": 0.191, "chunks": 164.28, "bytes_per_chunk": 6.233, "boundary_entropy": 0.16253, "router_noise": 0.0, "grad_norm": 0.3807, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.3698, "alpha": 0.03, "t": 1239.7} | |
| {"event": "train", "step": 6400, "phase": 2, "objective": "parity_a", "lm_loss": 1.10546, "bpb": 1.59484, "chunk_loss": 0.030224, "F": 0.14815, "G": 0.1817, "chunks": 151.7, "bytes_per_chunk": 6.75, "boundary_entropy": 0.16892, "router_noise": 0.0, "grad_norm": 0.3688, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.9315, "alpha": 0.03, "t": 1256.8} | |
| {"event": "train", "step": 6500, "phase": 2, "objective": "parity_a", "lm_loss": 1.12591, "bpb": 1.62435, "chunk_loss": 0.0301, "F": 0.14603, "G": 0.17923, "chunks": 149.53, "bytes_per_chunk": 6.848, "boundary_entropy": 0.1619, "router_noise": 0.0, "grad_norm": 0.3705, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.0889, "alpha": 0.03, "t": 1274.1} | |
| {"event": "train", "step": 6600, "phase": 2, "objective": "parity_a", "lm_loss": 1.07955, "bpb": 1.55746, "chunk_loss": 0.030086, "F": 0.13789, "G": 0.17645, "chunks": 141.2, "bytes_per_chunk": 7.252, "boundary_entropy": 0.17777, "router_noise": 0.0, "grad_norm": 0.3709, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 9.6297, "alpha": 0.03, "t": 1291.6} | |
| {"event": "train", "step": 6700, "phase": 2, "objective": "parity_a", "lm_loss": 1.11057, "bpb": 1.60222, "chunk_loss": 0.030211, "F": 0.14499, "G": 0.18429, "chunks": 148.47, "bytes_per_chunk": 6.897, "boundary_entropy": 0.18013, "router_noise": 0.0, "grad_norm": 0.3863, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.4445, "alpha": 0.03, "t": 1308.8} | |
| {"event": "train", "step": 6800, "phase": 2, "objective": "parity_a", "lm_loss": 1.15236, "bpb": 1.66251, "chunk_loss": 0.030171, "F": 0.15804, "G": 0.18813, "chunks": 161.83, "bytes_per_chunk": 6.328, "boundary_entropy": 0.15671, "router_noise": 0.0, "grad_norm": 0.3676, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.2605, "alpha": 0.03, "t": 1326.0} | |
| {"event": "train", "step": 6900, "phase": 2, "objective": "parity_a", "lm_loss": 1.1282, "bpb": 1.62765, "chunk_loss": 0.030132, "F": 0.15636, "G": 0.17964, "chunks": 160.11, "bytes_per_chunk": 6.396, "boundary_entropy": 0.13497, "router_noise": 0.0, "grad_norm": 0.383, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.174, "n_target_mean": 8.0676, "alpha": 0.03, "t": 1343.4} | |
| {"event": "probe", "step": 7000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3567.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.204965, "cps": 27.867188, "mean_p": 0.201303, "boundary_entropy": 0.055809, "bpb": 1.698393}, "de": {"chunks": 4341.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.211416, "cps": 33.914062, "mean_p": 0.204217, "boundary_entropy": 0.066658, "bpb": 1.827607}, "vi": {"chunks": 3358.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.182154, "cps": 26.234375, "mean_p": 0.153015, "boundary_entropy": 0.100541, "bpb": 1.533639}, "tr": {"chunks": 3152.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174115, "cps": 24.625, "mean_p": 0.164641, "boundary_entropy": 0.074704, "bpb": 1.750836}, "fi": {"chunks": 3383.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179842, "cps": 26.429688, "mean_p": 0.177171, "boundary_entropy": 0.069672, "bpb": 1.935808}, "ru": {"chunks": 2828.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.149235, "cps": 22.09375, "mean_p": 0.164376, "boundary_entropy": 0.295886, "bpb": 1.16382}, "el": {"chunks": 2514.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.122497, "cps": 19.640625, "mean_p": 0.161521, "boundary_entropy": 0.342135, "bpb": 1.19}, "hi": {"chunks": 2596.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.154294, "cps": 20.28125, "mean_p": 0.135099, "boundary_entropy": 0.286531, "bpb": 0.975381}, "ta": {"chunks": 2500.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.133726, "cps": 19.53125, "mean_p": 0.142675, "boundary_entropy": 0.31728, "bpb": 0.839261}, "th": {"chunks": 2720.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.169102, "cps": 21.25, "mean_p": 0.153942, "boundary_entropy": 0.323065, "bpb": 0.932721}, "zh": {"chunks": 3289.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.556891, "cps": 25.695312, "mean_p": 0.227898, "boundary_entropy": 0.168383, "bpb": 2.572692}, "my": {"chunks": 2251.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.118799, "cps": 17.585938, "mean_p": 0.134781, "boundary_entropy": 0.303812, "bpb": 0.940861}}, "bpb": {"en": 1.69839, "de": 1.82761, "vi": 1.53364, "tr": 1.75084, "fi": 1.93581, "ru": 1.16382, "el": 1.19, "hi": 0.97538, "ta": 0.83926, "th": 0.93272, "zh": 2.57269, "my": 0.94086}, "gini_cpc": 0.225008, "gini_cps": 0.10164, "bpic": 10.514836, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.031474247469775, "vi": 0.8887069176519323, "tr": 0.849486865342589, "fi": 0.8774272644993832, "ru": 0.7281002860433482, "el": 0.5976479567304885, "hi": 0.7527844268348919, "ta": 0.6524324449560034, "th": 0.8250283070933353, "zh": 2.7170112815950027, "my": 0.5796063432779509}, "t": 1364.4} | |
| {"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.69664, "de": 1.83166, "vi": 1.54914, "tr": 1.74484, "fi": 1.94733, "ru": 1.16319, "el": 1.18748, "hi": 0.98044, "ta": 0.84233, "th": 0.9323, "zh": 2.57752, "my": 0.94764}, "bytes_per_chunk": {"en": 5.098, "de": 4.987, "vi": 8.109, "tr": 6.387, "fi": 5.758, "ru": 12.943, "el": 15.202, "hi": 15.847, "ta": 20.511, "th": 17.739, "zh": 5.159, "my": 23.03}, "bpb_mean": 1.45004, "t": 1369.3} | |
| {"event": "train", "step": 7000, "phase": 2, "objective": "parity_a", "lm_loss": 1.22528, "bpb": 1.76771, "chunk_loss": 0.030159, "F": 0.15616, "G": 0.17995, "chunks": 159.91, "bytes_per_chunk": 6.404, "boundary_entropy": 0.13571, "router_noise": 0.0, "grad_norm": 0.4043, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.261, "n_target_mean": 7.9992, "alpha": 0.03, "t": 1369.5} | |
| {"event": "train", "step": 7100, "phase": 2, "objective": "parity_a", "lm_loss": 1.14315, "bpb": 1.64922, "chunk_loss": 0.0302, "F": 0.15224, "G": 0.18513, "chunks": 155.89, "bytes_per_chunk": 6.569, "boundary_entropy": 0.16424, "router_noise": 0.0, "grad_norm": 0.394, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.8248, "alpha": 0.03, "t": 1386.6} | |
| {"event": "train", "step": 7200, "phase": 2, "objective": "parity_a", "lm_loss": 1.03926, "bpb": 1.49934, "chunk_loss": 0.030162, "F": 0.14464, "G": 0.17982, "chunks": 148.11, "bytes_per_chunk": 6.914, "boundary_entropy": 0.16948, "router_noise": 0.0, "grad_norm": 0.3675, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 9.0753, "alpha": 0.03, "t": 1405.1} | |
| {"event": "train", "step": 7300, "phase": 2, "objective": "parity_a", "lm_loss": 1.12309, "bpb": 1.62027, "chunk_loss": 0.03016, "F": 0.15083, "G": 0.17627, "chunks": 154.45, "bytes_per_chunk": 6.63, "boundary_entropy": 0.13868, "router_noise": 0.0, "grad_norm": 0.4007, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 8.9955, "alpha": 0.03, "t": 1422.0} | |
| {"event": "train", "step": 7400, "phase": 2, "objective": "parity_a", "lm_loss": 1.04543, "bpb": 1.50824, "chunk_loss": 0.030262, "F": 0.14117, "G": 0.17742, "chunks": 144.56, "bytes_per_chunk": 7.083, "boundary_entropy": 0.17434, "router_noise": 0.0, "grad_norm": 0.3754, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.8241, "alpha": 0.03, "t": 1439.1} | |
| {"event": "train", "step": 7500, "phase": 2, "objective": "parity_a", "lm_loss": 1.06306, "bpb": 1.53367, "chunk_loss": 0.030275, "F": 0.15392, "G": 0.18149, "chunks": 157.61, "bytes_per_chunk": 6.497, "boundary_entropy": 0.1539, "router_noise": 0.0, "grad_norm": 0.3782, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.3011, "alpha": 0.03, "t": 1456.2} | |
| {"event": "train", "step": 7600, "phase": 2, "objective": "parity_a", "lm_loss": 1.05916, "bpb": 1.52805, "chunk_loss": 0.03016, "F": 0.13849, "G": 0.17098, "chunks": 141.81, "bytes_per_chunk": 7.221, "boundary_entropy": 0.16064, "router_noise": 0.0, "grad_norm": 0.3843, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.7149, "alpha": 0.03, "t": 1473.5} | |
| {"event": "phase_transition", "step": 7700, "phase": 3, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 1490.5} | |
| {"event": "train", "step": 7700, "phase": 3, "objective": "parity_a", "lm_loss": 1.04435, "bpb": 1.50667, "chunk_loss": 0.030249, "F": 0.14922, "G": 0.17839, "chunks": 152.8, "bytes_per_chunk": 6.702, "boundary_entropy": 0.15037, "router_noise": 0.0, "grad_norm": 0.3738, "lr": 9.06e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.9762, "alpha": 0.03, "t": 1490.6} | |
| {"event": "train", "step": 7800, "phase": 3, "objective": "parity_a", "lm_loss": 1.10349, "bpb": 1.592, "chunk_loss": 0.03015, "F": 0.15591, "G": 0.18165, "chunks": 159.66, "bytes_per_chunk": 6.414, "boundary_entropy": 0.14302, "router_noise": 0.0, "grad_norm": 0.3978, "lr": 8.73e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.3835, "alpha": 0.03, "t": 1507.8} | |
| {"event": "train", "step": 7900, "phase": 3, "objective": "parity_a", "lm_loss": 1.09306, "bpb": 1.57695, "chunk_loss": 0.030132, "F": 0.14682, "G": 0.17787, "chunks": 150.34, "bytes_per_chunk": 6.811, "boundary_entropy": 0.15389, "router_noise": 0.0, "grad_norm": 0.3756, "lr": 8.4e-05, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 8.7313, "alpha": 0.03, "t": 1525.8} | |
| {"event": "probe", "step": 8000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3539.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203356, "cps": 27.648438, "mean_p": 0.198231, "boundary_entropy": 0.054015, "bpb": 1.674262}, "de": {"chunks": 4324.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.210588, "cps": 33.78125, "mean_p": 0.201001, "boundary_entropy": 0.06571, "bpb": 1.797155}, "vi": {"chunks": 3366.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.182587, "cps": 26.296875, "mean_p": 0.151274, "boundary_entropy": 0.091353, "bpb": 1.50373}, "tr": {"chunks": 3147.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.173839, "cps": 24.585938, "mean_p": 0.164973, "boundary_entropy": 0.072628, "bpb": 1.719823}, "fi": {"chunks": 3381.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.179735, "cps": 26.414062, "mean_p": 0.176506, "boundary_entropy": 0.066048, "bpb": 1.895568}, "ru": {"chunks": 3019.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.159314, "cps": 23.585938, "mean_p": 0.170931, "boundary_entropy": 0.295576, "bpb": 1.13906}, "el": {"chunks": 2701.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.131608, "cps": 21.101562, "mean_p": 0.165241, "boundary_entropy": 0.343355, "bpb": 1.163142}, "hi": {"chunks": 2718.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.161545, "cps": 21.234375, "mean_p": 0.134963, "boundary_entropy": 0.280553, "bpb": 0.956074}, "ta": {"chunks": 2684.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.143568, "cps": 20.96875, "mean_p": 0.13099, "boundary_entropy": 0.290426, "bpb": 0.814864}, "th": {"chunks": 2561.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.159217, "cps": 20.007812, "mean_p": 0.146662, "boundary_entropy": 0.31197, "bpb": 0.911354}, "zh": {"chunks": 3390.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.573993, "cps": 26.484375, "mean_p": 0.230733, "boundary_entropy": 0.156006, "bpb": 2.529143}, "my": {"chunks": 2460.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.129829, "cps": 19.21875, "mean_p": 0.126188, "boundary_entropy": 0.284363, "bpb": 0.919145}}, "bpb": {"en": 1.67426, "de": 1.79716, "vi": 1.50373, "tr": 1.71982, "fi": 1.89557, "ru": 1.13906, "el": 1.16314, "hi": 0.95607, "ta": 0.81486, "th": 0.91135, "zh": 2.52914, "my": 0.91914}, "gini_cpc": 0.214649, "gini_cps": 0.088764, "bpic": 10.291794, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0355637408582608, "vi": 0.8978722044331776, "tr": 0.8548496713189512, "fi": 0.8838464967824599, "ru": 0.7834250541087214, "el": 0.6471832917057323, "hi": 0.7943976249366536, "ta": 0.7059933325378065, "th": 0.7829464707087761, "zh": 2.8226030931805597, "my": 0.6384329432590544}, "t": 1545.8} | |
| {"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.67187, "de": 1.80129, "vi": 1.52043, "tr": 1.71346, "fi": 1.908, "ru": 1.1384, "el": 1.1609, "hi": 0.96148, "ta": 0.8182, "th": 0.91102, "zh": 2.53488, "my": 0.92726}, "bytes_per_chunk": {"en": 5.135, "de": 5.023, "vi": 8.043, "tr": 6.396, "fi": 5.781, "ru": 12.144, "el": 13.848, "hi": 15.26, "ta": 18.54, "th": 18.679, "zh": 5.061, "my": 20.961}, "bpb_mean": 1.42227, "t": 1551.1} | |
| {"event": "train", "step": 8000, "phase": 3, "objective": "parity_a", "lm_loss": 1.0729, "bpb": 1.54786, "chunk_loss": 0.030163, "F": 0.1465, "G": 0.17622, "chunks": 150.02, "bytes_per_chunk": 6.826, "boundary_entropy": 0.15271, "router_noise": 0.0, "grad_norm": 0.4071, "lr": 8.08e-05, "router_trainable": true, "sec_per_step": 0.255, "n_target_mean": 9.3646, "alpha": 0.03, "t": 1551.4} | |
| {"event": "train", "step": 8100, "phase": 3, "objective": "parity_a", "lm_loss": 1.07639, "bpb": 1.55291, "chunk_loss": 0.030095, "F": 0.13396, "G": 0.17071, "chunks": 137.17, "bytes_per_chunk": 7.465, "boundary_entropy": 0.17208, "router_noise": 0.0, "grad_norm": 0.4081, "lr": 7.77e-05, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.4465, "alpha": 0.03, "t": 1568.7} | |
| {"event": "train", "step": 8200, "phase": 3, "objective": "parity_a", "lm_loss": 1.12538, "bpb": 1.62359, "chunk_loss": 0.030185, "F": 0.15668, "G": 0.17875, "chunks": 160.44, "bytes_per_chunk": 6.383, "boundary_entropy": 0.12729, "router_noise": 0.0, "grad_norm": 0.4027, "lr": 7.47e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.5428, "alpha": 0.03, "t": 1585.7} | |
| {"event": "train", "step": 8300, "phase": 3, "objective": "parity_a", "lm_loss": 1.05654, "bpb": 1.52427, "chunk_loss": 0.030471, "F": 0.15463, "G": 0.1778, "chunks": 158.34, "bytes_per_chunk": 6.467, "boundary_entropy": 0.13201, "router_noise": 0.0, "grad_norm": 0.3882, "lr": 7.17e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.1281, "alpha": 0.03, "t": 1602.7} | |
| {"event": "train", "step": 8400, "phase": 3, "objective": "parity_a", "lm_loss": 1.08245, "bpb": 1.56164, "chunk_loss": 0.03015, "F": 0.14029, "G": 0.17238, "chunks": 143.66, "bytes_per_chunk": 7.128, "boundary_entropy": 0.15468, "router_noise": 0.0, "grad_norm": 0.3991, "lr": 6.88e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.1518, "alpha": 0.03, "t": 1619.7} | |
| {"event": "train", "step": 8500, "phase": 3, "objective": "parity_a", "lm_loss": 1.07253, "bpb": 1.54733, "chunk_loss": 0.030151, "F": 0.14394, "G": 0.17532, "chunks": 147.39, "bytes_per_chunk": 6.948, "boundary_entropy": 0.15493, "router_noise": 0.0, "grad_norm": 0.3966, "lr": 6.6e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.307, "alpha": 0.03, "t": 1636.8} | |
| {"event": "train", "step": 8600, "phase": 3, "objective": "parity_a", "lm_loss": 1.03676, "bpb": 1.49574, "chunk_loss": 0.030214, "F": 0.13847, "G": 0.17241, "chunks": 141.8, "bytes_per_chunk": 7.222, "boundary_entropy": 0.16232, "router_noise": 0.0, "grad_norm": 0.3865, "lr": 6.33e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.6289, "alpha": 0.03, "t": 1654.0} | |
| {"event": "train", "step": 8700, "phase": 3, "objective": "parity_a", "lm_loss": 1.05296, "bpb": 1.51911, "chunk_loss": 0.030242, "F": 0.1404, "G": 0.17771, "chunks": 143.77, "bytes_per_chunk": 7.123, "boundary_entropy": 0.17331, "router_noise": 0.0, "grad_norm": 0.4099, "lr": 6.07e-05, "router_trainable": true, "sec_per_step": 0.176, "n_target_mean": 9.8233, "alpha": 0.03, "t": 1671.6} | |
| {"event": "train", "step": 8800, "phase": 3, "objective": "parity_a", "lm_loss": 1.0499, "bpb": 1.51468, "chunk_loss": 0.03021, "F": 0.14687, "G": 0.17675, "chunks": 150.39, "bytes_per_chunk": 6.809, "boundary_entropy": 0.14919, "router_noise": 0.0, "grad_norm": 0.3979, "lr": 5.82e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.4516, "alpha": 0.03, "t": 1688.6} | |
| {"event": "train", "step": 8900, "phase": 3, "objective": "parity_a", "lm_loss": 1.03992, "bpb": 1.50029, "chunk_loss": 0.030197, "F": 0.14491, "G": 0.17461, "chunks": 148.39, "bytes_per_chunk": 6.901, "boundary_entropy": 0.14892, "router_noise": 0.0, "grad_norm": 0.3848, "lr": 5.58e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.0423, "alpha": 0.03, "t": 1705.6} | |
| {"event": "probe", "step": 9000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3549.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.20393, "cps": 27.726562, "mean_p": 0.199607, "boundary_entropy": 0.050727, "bpb": 1.654712}, "de": {"chunks": 4334.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.211075, "cps": 33.859375, "mean_p": 0.20219, "boundary_entropy": 0.062928, "bpb": 1.772114}, "vi": {"chunks": 3370.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.182804, "cps": 26.328125, "mean_p": 0.150601, "boundary_entropy": 0.084958, "bpb": 1.478046}, "tr": {"chunks": 3096.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.171021, "cps": 24.1875, "mean_p": 0.16264, "boundary_entropy": 0.070682, "bpb": 1.699479}, "fi": {"chunks": 3347.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.177928, "cps": 26.148438, "mean_p": 0.174983, "boundary_entropy": 0.062315, "bpb": 1.86472}, "ru": {"chunks": 2981.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.157309, "cps": 23.289062, "mean_p": 0.162524, "boundary_entropy": 0.277339, "bpb": 1.119121}, "el": {"chunks": 2702.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.131657, "cps": 21.109375, "mean_p": 0.167327, "boundary_entropy": 0.347331, "bpb": 1.143281}, "hi": {"chunks": 2767.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.164458, "cps": 21.617188, "mean_p": 0.129023, "boundary_entropy": 0.267976, "bpb": 0.934984}, "ta": {"chunks": 2778.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.148596, "cps": 21.703125, "mean_p": 0.136269, "boundary_entropy": 0.297477, "bpb": 0.799873}, "th": {"chunks": 2884.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.179297, "cps": 22.53125, "mean_p": 0.146326, "boundary_entropy": 0.30527, "bpb": 0.896771}, "zh": {"chunks": 3365.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.56976, "cps": 26.289062, "mean_p": 0.2297, "boundary_entropy": 0.147659, "bpb": 2.498842}, "my": {"chunks": 2424.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.127929, "cps": 18.9375, "mean_p": 0.123027, "boundary_entropy": 0.275197, "bpb": 0.893003}}, "bpb": {"en": 1.65471, "de": 1.77211, "vi": 1.47805, "tr": 1.69948, "fi": 1.86472, "ru": 1.11912, "el": 1.14328, "hi": 0.93498, "ta": 0.79987, "th": 0.89677, "zh": 2.49884, "my": 0.893}, "gini_cpc": 0.210969, "gini_cps": 0.08266, "bpic": 10.207756, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.03503401056683, "vi": 0.8964062579885325, "tr": 0.8386263836656673, "fi": 0.8724929969169223, "ru": 0.7713844547982969, "el": 0.6455986601264507, "hi": 0.8064402713528327, "ta": 0.7286599384116202, "th": 0.879209377695784, "zh": 2.7938928533499894, "my": 0.6273174382493816}, "t": 1726.0} | |
| {"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.65211, "de": 1.77761, "vi": 1.49481, "tr": 1.6938, "fi": 1.87839, "ru": 1.11909, "el": 1.14098, "hi": 0.93991, "ta": 0.80376, "th": 0.89656, "zh": 2.50314, "my": 0.90111}, "bytes_per_chunk": {"en": 5.118, "de": 5.004, "vi": 8.049, "tr": 6.462, "fi": 5.795, "ru": 12.193, "el": 13.828, "hi": 14.829, "ta": 18.112, "th": 16.579, "zh": 5.013, "my": 20.78}, "bpb_mean": 1.40011, "t": 1730.9} | |
| {"event": "train", "step": 9000, "phase": 3, "objective": "parity_a", "lm_loss": 1.08607, "bpb": 1.56687, "chunk_loss": 0.030134, "F": 0.14783, "G": 0.17418, "chunks": 151.38, "bytes_per_chunk": 6.765, "boundary_entropy": 0.13972, "router_noise": 0.0, "grad_norm": 0.4019, "lr": 5.35e-05, "router_trainable": true, "sec_per_step": 0.254, "n_target_mean": 8.7035, "alpha": 0.03, "t": 1731.0} | |
| {"event": "train", "step": 9100, "phase": 3, "objective": "parity_a", "lm_loss": 1.02176, "bpb": 1.47409, "chunk_loss": 0.030193, "F": 0.14204, "G": 0.17399, "chunks": 145.45, "bytes_per_chunk": 7.04, "boundary_entropy": 0.15426, "router_noise": 0.0, "grad_norm": 0.3811, "lr": 5.12e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.4355, "alpha": 0.03, "t": 1748.1} | |
| {"event": "train", "step": 9200, "phase": 3, "objective": "parity_a", "lm_loss": 1.10789, "bpb": 1.59834, "chunk_loss": 0.030158, "F": 0.15109, "G": 0.17773, "chunks": 154.72, "bytes_per_chunk": 6.618, "boundary_entropy": 0.14143, "router_noise": 0.0, "grad_norm": 0.423, "lr": 4.91e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.5071, "alpha": 0.03, "t": 1765.2} | |
| {"event": "train", "step": 9300, "phase": 3, "objective": "parity_a", "lm_loss": 0.97316, "bpb": 1.40397, "chunk_loss": 0.030152, "F": 0.13452, "G": 0.1685, "chunks": 137.75, "bytes_per_chunk": 7.434, "boundary_entropy": 0.16349, "router_noise": 0.0, "grad_norm": 0.3952, "lr": 4.71e-05, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 10.0961, "alpha": 0.03, "t": 1782.9} | |
| {"event": "train", "step": 9400, "phase": 3, "objective": "parity_a", "lm_loss": 1.08132, "bpb": 1.56002, "chunk_loss": 0.030252, "F": 0.15581, "G": 0.17588, "chunks": 159.55, "bytes_per_chunk": 6.418, "boundary_entropy": 0.12122, "router_noise": 0.0, "grad_norm": 0.3888, "lr": 4.52e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.4801, "alpha": 0.03, "t": 1800.1} | |
| {"event": "train", "step": 9500, "phase": 3, "objective": "parity_a", "lm_loss": 1.01795, "bpb": 1.4686, "chunk_loss": 0.030178, "F": 0.13979, "G": 0.17023, "chunks": 143.14, "bytes_per_chunk": 7.154, "boundary_entropy": 0.1532, "router_noise": 0.0, "grad_norm": 0.3953, "lr": 4.34e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.6217, "alpha": 0.03, "t": 1817.2} | |
| {"event": "train", "step": 9600, "phase": 3, "objective": "parity_a", "lm_loss": 1.08613, "bpb": 1.56695, "chunk_loss": 0.030137, "F": 0.15169, "G": 0.17389, "chunks": 155.33, "bytes_per_chunk": 6.592, "boundary_entropy": 0.12711, "router_noise": 0.0, "grad_norm": 0.4149, "lr": 4.17e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.5189, "alpha": 0.03, "t": 1834.3} | |
| {"event": "train", "step": 9700, "phase": 3, "objective": "parity_a", "lm_loss": 1.06381, "bpb": 1.53475, "chunk_loss": 0.030212, "F": 0.14807, "G": 0.16932, "chunks": 151.62, "bytes_per_chunk": 6.754, "boundary_entropy": 0.12563, "router_noise": 0.0, "grad_norm": 0.3853, "lr": 4.01e-05, "router_trainable": true, "sec_per_step": 0.18, "n_target_mean": 9.1209, "alpha": 0.03, "t": 1852.3} | |
| {"event": "train", "step": 9800, "phase": 3, "objective": "parity_a", "lm_loss": 1.0316, "bpb": 1.48828, "chunk_loss": 0.030105, "F": 0.14194, "G": 0.16889, "chunks": 145.34, "bytes_per_chunk": 7.045, "boundary_entropy": 0.1388, "router_noise": 0.0, "grad_norm": 0.3992, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.3257, "alpha": 0.03, "t": 1869.5} | |
| {"event": "train", "step": 9900, "phase": 3, "objective": "parity_a", "lm_loss": 1.04339, "bpb": 1.5053, "chunk_loss": 0.030179, "F": 0.14781, "G": 0.17434, "chunks": 151.36, "bytes_per_chunk": 6.765, "boundary_entropy": 0.14143, "router_noise": 0.0, "grad_norm": 0.3981, "lr": 3.72e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.0365, "alpha": 0.03, "t": 1886.7} | |
| {"event": "probe", "step": 10000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3541.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203471, "cps": 27.664062, "mean_p": 0.198161, "boundary_entropy": 0.050506, "bpb": 1.641832}, "de": {"chunks": 4335.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.211124, "cps": 33.867188, "mean_p": 0.201991, "boundary_entropy": 0.06184, "bpb": 1.76024}, "vi": {"chunks": 3370.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.182804, "cps": 26.328125, "mean_p": 0.150235, "boundary_entropy": 0.082479, "bpb": 1.461981}, "tr": {"chunks": 3101.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.171298, "cps": 24.226562, "mean_p": 0.161126, "boundary_entropy": 0.06786, "bpb": 1.683356}, "fi": {"chunks": 3414.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.18149, "cps": 26.671875, "mean_p": 0.176966, "boundary_entropy": 0.061462, "bpb": 1.847224}, "ru": {"chunks": 3030.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.159894, "cps": 23.671875, "mean_p": 0.159773, "boundary_entropy": 0.268506, "bpb": 1.105076}, "el": {"chunks": 2735.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.133265, "cps": 21.367188, "mean_p": 0.160024, "boundary_entropy": 0.329919, "bpb": 1.129529}, "hi": {"chunks": 2793.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.166003, "cps": 21.820312, "mean_p": 0.12919, "boundary_entropy": 0.263398, "bpb": 0.924335}, "ta": {"chunks": 2697.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.144263, "cps": 21.070312, "mean_p": 0.126253, "boundary_entropy": 0.278674, "bpb": 0.789848}, "th": {"chunks": 2833.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.176127, "cps": 22.132812, "mean_p": 0.144179, "boundary_entropy": 0.299721, "bpb": 0.884688}, "zh": {"chunks": 3303.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.559262, "cps": 25.804688, "mean_p": 0.225415, "boundary_entropy": 0.139385, "bpb": 2.486951}, "my": {"chunks": 2373.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.125237, "cps": 18.539062, "mean_p": 0.123249, "boundary_entropy": 0.27579, "bpb": 0.882025}}, "bpb": {"en": 1.64183, "de": 1.76024, "vi": 1.46198, "tr": 1.68336, "fi": 1.84722, "ru": 1.10508, "el": 1.12953, "hi": 0.92434, "ta": 0.78985, "th": 0.88469, "zh": 2.48695, "my": 0.88203}, "gini_cpc": 0.209013, "gini_cps": 0.085048, "bpic": 10.227342, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.0376117667218612, "vi": 0.898431462751003, "tr": 0.8418784799516239, "fi": 0.8919691352696996, "ru": 0.7858354339926038, "el": 0.6549598515619708, "hi": 0.8158570227850949, "ta": 0.7090121383262692, "th": 0.8656128655686393, "zh": 2.748611280196676, "my": 0.6155063752416952}, "t": 1907.3} | |
| {"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.63988, "de": 1.76548, "vi": 1.47841, "tr": 1.67733, "fi": 1.86049, "ru": 1.10462, "el": 1.12734, "hi": 0.92914, "ta": 0.7936, "th": 0.88471, "zh": 2.49221, "my": 0.88962}, "bytes_per_chunk": {"en": 5.134, "de": 5.005, "vi": 8.077, "tr": 6.515, "fi": 5.736, "ru": 11.961, "el": 13.899, "hi": 15.222, "ta": 18.84, "th": 16.509, "zh": 5.113, "my": 21.457}, "bpb_mean": 1.3869, "t": 1912.1} | |
| {"event": "train", "step": 10000, "phase": 3, "objective": "parity_a", "lm_loss": 1.05206, "bpb": 1.5178, "chunk_loss": 0.030153, "F": 0.1517, "G": 0.17145, "chunks": 155.34, "bytes_per_chunk": 6.592, "boundary_entropy": 0.12146, "router_noise": 0.0, "grad_norm": 0.382, "lr": 3.6e-05, "router_trainable": true, "sec_per_step": 0.256, "n_target_mean": 8.75, "alpha": 0.03, "t": 1912.3} | |
| {"event": "train", "step": 10100, "phase": 3, "objective": "parity_a", "lm_loss": 1.06554, "bpb": 1.53725, "chunk_loss": 0.030153, "F": 0.15263, "G": 0.17817, "chunks": 156.3, "bytes_per_chunk": 6.552, "boundary_entropy": 0.14156, "router_noise": 0.0, "grad_norm": 0.4211, "lr": 3.49e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.1007, "alpha": 0.03, "t": 1929.3} | |
| {"event": "train", "step": 10200, "phase": 3, "objective": "parity_a", "lm_loss": 1.02199, "bpb": 1.47442, "chunk_loss": 0.030167, "F": 0.14275, "G": 0.16997, "chunks": 146.17, "bytes_per_chunk": 7.005, "boundary_entropy": 0.14211, "router_noise": 0.0, "grad_norm": 0.4071, "lr": 3.38e-05, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 9.3349, "alpha": 0.03, "t": 1947.4} | |
| {"event": "train", "step": 10300, "phase": 3, "objective": "parity_a", "lm_loss": 0.96825, "bpb": 1.39689, "chunk_loss": 0.030282, "F": 0.13705, "G": 0.16974, "chunks": 140.34, "bytes_per_chunk": 7.296, "boundary_entropy": 0.15955, "router_noise": 0.0, "grad_norm": 0.4027, "lr": 3.3e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 10.2951, "alpha": 0.03, "t": 1964.6} | |
| {"event": "train", "step": 10400, "phase": 3, "objective": "parity_a", "lm_loss": 1.06433, "bpb": 1.5355, "chunk_loss": 0.030201, "F": 0.14839, "G": 0.17255, "chunks": 151.95, "bytes_per_chunk": 6.739, "boundary_entropy": 0.13487, "router_noise": 0.0, "grad_norm": 0.3971, "lr": 3.22e-05, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 9.215, "alpha": 0.03, "t": 1982.3} | |
| {"event": "train", "step": 10500, "phase": 3, "objective": "parity_a", "lm_loss": 1.0118, "bpb": 1.45971, "chunk_loss": 0.030218, "F": 0.13773, "G": 0.17103, "chunks": 141.03, "bytes_per_chunk": 7.261, "boundary_entropy": 0.16394, "router_noise": 0.0, "grad_norm": 0.4143, "lr": 3.15e-05, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.8085, "alpha": 0.03, "t": 1999.7} | |
| {"event": "train", "step": 10600, "phase": 3, "objective": "parity_a", "lm_loss": 0.97392, "bpb": 1.40508, "chunk_loss": 0.030145, "F": 0.12859, "G": 0.16579, "chunks": 131.67, "bytes_per_chunk": 7.777, "boundary_entropy": 0.1743, "router_noise": 0.0, "grad_norm": 0.403, "lr": 3.1e-05, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 10.3034, "alpha": 0.03, "t": 2016.9} | |
| {"event": "train", "step": 10700, "phase": 3, "objective": "parity_a", "lm_loss": 1.03469, "bpb": 1.49274, "chunk_loss": 0.03019, "F": 0.14044, "G": 0.17094, "chunks": 143.81, "bytes_per_chunk": 7.12, "boundary_entropy": 0.15486, "router_noise": 0.0, "grad_norm": 0.4191, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.6897, "alpha": 0.03, "t": 2034.2} | |
| {"event": "train", "step": 10800, "phase": 3, "objective": "parity_a", "lm_loss": 1.01024, "bpb": 1.45747, "chunk_loss": 0.030362, "F": 0.15787, "G": 0.17943, "chunks": 161.66, "bytes_per_chunk": 6.334, "boundary_entropy": 0.12666, "router_noise": 0.0, "grad_norm": 0.3959, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 8.6008, "alpha": 0.03, "t": 2052.1} | |
| {"event": "train", "step": 10900, "phase": 3, "objective": "parity_a", "lm_loss": 1.02444, "bpb": 1.47795, "chunk_loss": 0.030127, "F": 0.1429, "G": 0.16849, "chunks": 146.33, "bytes_per_chunk": 6.998, "boundary_entropy": 0.1355, "router_noise": 0.0, "grad_norm": 0.4121, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.8131, "alpha": 0.03, "t": 2069.9} | |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3536.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203183, "cps": 27.625, "mean_p": 0.198671, "boundary_entropy": 0.048888, "bpb": 1.632832}, "de": {"chunks": 4344.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.211562, "cps": 33.9375, "mean_p": 0.202152, "boundary_entropy": 0.060472, "bpb": 1.750652}, "vi": {"chunks": 3374.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.183021, "cps": 26.359375, "mean_p": 0.150486, "boundary_entropy": 0.080802, "bpb": 1.453067}, "tr": {"chunks": 3114.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172016, "cps": 24.328125, "mean_p": 0.16298, "boundary_entropy": 0.066365, "bpb": 1.673197}, "fi": {"chunks": 3389.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180161, "cps": 26.476562, "mean_p": 0.176796, "boundary_entropy": 0.05844, "bpb": 1.833903}, "ru": {"chunks": 3061.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.16153, "cps": 23.914062, "mean_p": 0.161285, "boundary_entropy": 0.268362, "bpb": 1.098899}, "el": {"chunks": 3205.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.156166, "cps": 25.039062, "mean_p": 0.168455, "boundary_entropy": 0.33501, "bpb": 1.121959}, "hi": {"chunks": 2992.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.177831, "cps": 23.375, "mean_p": 0.131998, "boundary_entropy": 0.264578, "bpb": 0.916888}, "ta": {"chunks": 3054.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.163359, "cps": 23.859375, "mean_p": 0.132666, "boundary_entropy": 0.280856, "bpb": 0.785605}, "th": {"chunks": 3041.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.189058, "cps": 23.757812, "mean_p": 0.14506, "boundary_entropy": 0.296782, "bpb": 0.879317}, "zh": {"chunks": 3376.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.571622, "cps": 26.375, "mean_p": 0.228546, "boundary_entropy": 0.135208, "bpb": 2.466209}, "my": {"chunks": 2765.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.145926, "cps": 21.601562, "mean_p": 0.12731, "boundary_entropy": 0.273794, "bpb": 0.87347}}, "bpb": {"en": 1.63283, "de": 1.75065, "vi": 1.45307, "tr": 1.6732, "fi": 1.8339, "ru": 1.0989, "el": 1.12196, "hi": 0.91689, "ta": 0.78561, "th": 0.87932, "zh": 2.46621, "my": 0.87347}, "gini_cpc": 0.186735, "gini_cps": 0.058047, "bpic": 9.777611, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.041236235324088, "vi": 0.9007697646248836, "tr": 0.8466032291840331, "fi": 0.8866894664982533, "ru": 0.7949978957485165, "el": 0.7685976505115218, "hi": 0.8752223111212709, "ta": 0.8039988491127518, "th": 0.9304803771536112, "zh": 2.8133311778956287, "my": 0.718197060757854}, "t": 2090.8} | |
| {"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.63137, "de": 1.7563, "vi": 1.46992, "tr": 1.66721, "fi": 1.84708, "ru": 1.09853, "el": 1.11984, "hi": 0.92134, "ta": 0.78966, "th": 0.87964, "zh": 2.47107, "my": 0.88076}, "bytes_per_chunk": {"en": 5.128, "de": 5.007, "vi": 8.045, "tr": 6.461, "fi": 5.748, "ru": 11.806, "el": 11.724, "hi": 14.303, "ta": 16.665, "th": 15.086, "zh": 5.002, "my": 18.506}, "bpb_mean": 1.37773, "t": 2095.5} | |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3536.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.203183, "cps": 27.625, "mean_p": 0.198671, "boundary_entropy": 0.048888, "bpb": 1.632832}, "de": {"chunks": 4344.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.211562, "cps": 33.9375, "mean_p": 0.202152, "boundary_entropy": 0.060472, "bpb": 1.750652}, "vi": {"chunks": 3374.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.183021, "cps": 26.359375, "mean_p": 0.150486, "boundary_entropy": 0.080802, "bpb": 1.453067}, "tr": {"chunks": 3114.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.172016, "cps": 24.328125, "mean_p": 0.16298, "boundary_entropy": 0.066365, "bpb": 1.673197}, "fi": {"chunks": 3389.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.180161, "cps": 26.476562, "mean_p": 0.176796, "boundary_entropy": 0.05844, "bpb": 1.833903}, "ru": {"chunks": 3061.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.16153, "cps": 23.914062, "mean_p": 0.161285, "boundary_entropy": 0.268362, "bpb": 1.098899}, "el": {"chunks": 3205.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.156166, "cps": 25.039062, "mean_p": 0.168455, "boundary_entropy": 0.33501, "bpb": 1.121959}, "hi": {"chunks": 2992.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.177831, "cps": 23.375, "mean_p": 0.131998, "boundary_entropy": 0.264578, "bpb": 0.916888}, "ta": {"chunks": 3054.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.163359, "cps": 23.859375, "mean_p": 0.132666, "boundary_entropy": 0.280856, "bpb": 0.785605}, "th": {"chunks": 3041.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.189058, "cps": 23.757812, "mean_p": 0.14506, "boundary_entropy": 0.296782, "bpb": 0.879317}, "zh": {"chunks": 3376.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.571622, "cps": 26.375, "mean_p": 0.228546, "boundary_entropy": 0.135208, "bpb": 2.466209}, "my": {"chunks": 2765.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.145926, "cps": 21.601562, "mean_p": 0.12731, "boundary_entropy": 0.273794, "bpb": 0.87347}}, "bpb": {"en": 1.63283, "de": 1.75065, "vi": 1.45307, "tr": 1.6732, "fi": 1.8339, "ru": 1.0989, "el": 1.12196, "hi": 0.91689, "ta": 0.78561, "th": 0.87932, "zh": 2.46621, "my": 0.87347}, "gini_cpc": 0.186735, "gini_cps": 0.058047, "bpic": 9.777611, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.041236235324088, "vi": 0.9007697646248836, "tr": 0.8466032291840331, "fi": 0.8866894664982533, "ru": 0.7949978957485165, "el": 0.7685976505115218, "hi": 0.8752223111212709, "ta": 0.8039988491127518, "th": 0.9304803771536112, "zh": 2.8133311778956287, "my": 0.718197060757854}, "t": 2099.0} | |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s2/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s2/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s2/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s2/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s2/th0.70"], "t": 2101.9} | |
| {"event": "final", "run": "spec_D_s2", "condition": "D", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 0.2031833591909441, "cps": 27.625, "bpb": 1.6271199808578398, "bytes_per_chunk": 5.119625027463809}, "de": {"cpc": 0.2115618760044806, "cps": 33.9375, "bpb": 1.7529729184093576, "bytes_per_chunk": 5.011379331768934}, "vi": {"cpc": 0.18302142663411988, "cps": 26.359375, "bpb": 1.4687836266441268, "bytes_per_chunk": 8.034449467473756}, "tr": {"cpc": 0.17201568800751257, "cps": 24.328125, "bpb": 1.666459152680295, "bytes_per_chunk": 6.452537137090323}, "fi": {"cpc": 0.1801605443623412, "cps": 26.4765625, "bpb": 1.8534615536419896, "bytes_per_chunk": 5.737165493054074}, "ru": {"cpc": 0.16153034300791558, "cps": 23.9140625, "bpb": 1.0995795699347535, "bytes_per_chunk": 11.814940845070423}, "el": {"cpc": 0.15616625249719826, "cps": 25.0390625, "bpb": 1.1165690373339712, "bytes_per_chunk": 11.628105039034777}, "hi": {"cpc": 0.17783060921248142, "cps": 23.375, "bpb": 0.9228888026387921, "bytes_per_chunk": 14.321091519960666}, "ta": {"cpc": 0.16335918694838192, "cps": 23.859375, "bpb": 0.7887196977493816, "bytes_per_chunk": 16.72423362786692}, "th": {"cpc": 0.18905812869132732, "cps": 23.7578125, "bpb": 0.8820690214261435, "bytes_per_chunk": 15.041326581842698}, "zh": {"cpc": 0.5716220792414494, "cps": 26.375, "bpb": 2.4713110971779577, "bytes_per_chunk": 5.006737238162085}, "my": {"cpc": 0.14592569136584335, "cps": 21.6015625, "bpb": 0.8804997711028036, "bytes_per_chunk": 18.62314181689015}}, "gini_cpc": 0.18673488006063543, "gini_cps": 0.058047353358301024, "bpic": 9.77761076150926, "seed": 2, "mean_boundary_entropy": 0.13966628909111023, "dechunk": "ema", "router_noise": 0.0, "transitions": [{"step": 0, "phase": 1, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 4400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}, {"step": 7700, "phase": 3, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0}], "t": 2111.0} | |