Download log.jsonl from AdaptiveChunking/spec_D_s1: direct link, hf CLI and curl.
- Browser
- Download file 110 kB
-
https://huggingface.co/AdaptiveChunking/spec_D_s1/resolve/main/log.jsonl
- Command line
-
hf download hf://AdaptiveChunking/spec_D_s1/log.jsonl
-
curl -L -o log.jsonl https://huggingface.co/AdaptiveChunking/spec_D_s1/resolve/main/log.jsonl
110 kB
| {"event": "start", "run": "spec_D_s1", "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": 14124.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.811584, "cps": 110.34375, "mean_p": 0.52085, "boundary_entropy": 0.686717, "bpb": 8.06011}, "de": {"chunks": 18700.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.910729, "cps": 146.09375, "mean_p": 0.525595, "boundary_entropy": 0.687217, "bpb": 8.1963}, "vi": {"chunks": 21351.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.158177, "cps": 166.804688, "mean_p": 0.524018, "boundary_entropy": 0.687874, "bpb": 8.061647}, "tr": {"chunks": 17846.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.985803, "cps": 139.421875, "mean_p": 0.529939, "boundary_entropy": 0.686373, "bpb": 8.199}, "fi": {"chunks": 17574.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.934241, "cps": 137.296875, "mean_p": 0.526553, "boundary_entropy": 0.686826, "bpb": 8.038283}, "ru": {"chunks": 34128.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.80095, "cps": 266.625, "mean_p": 0.540146, "boundary_entropy": 0.687211, "bpb": 8.262097}, "el": {"chunks": 36913.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.798616, "cps": 288.382812, "mean_p": 0.535776, "boundary_entropy": 0.688142, "bpb": 8.178552}, "hi": {"chunks": 34684.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.061456, "cps": 270.96875, "mean_p": 0.509835, "boundary_entropy": 0.69075, "bpb": 8.397747}, "ta": {"chunks": 50678.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.710778, "cps": 395.921875, "mean_p": 0.522012, "boundary_entropy": 0.690366, "bpb": 8.106024}, "th": {"chunks": 45409.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.823065, "cps": 354.757812, "mean_p": 0.531604, "boundary_entropy": 0.689071, "bpb": 8.030889}, "zh": {"chunks": 4820.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.816119, "cps": 37.65625, "mean_p": 0.492076, "boundary_entropy": 0.68644, "bpb": 8.099598}, "my": {"chunks": 28215.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.489075, "cps": 220.429688, "mean_p": 0.506229, "boundary_entropy": 0.690851, "bpb": 8.204861}}, "bpb": {"en": 8.06011, "de": 8.1963, "vi": 8.06165, "tr": 8.199, "fi": 8.03828, "ru": 8.2621, "el": 8.17855, "hi": 8.39775, "ta": 8.10602, "th": 8.03089, "zh": 8.0996, "my": 8.20486}, "gini_cpc": 0.247425, "gini_cps": 0.271932, "bpic": 1.182896, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1221621360827407, "vi": 1.4270575576222595, "tr": 1.2146656456674263, "fi": 1.1511320610100555, "ru": 2.2190548395702145, "el": 2.2161793753498933, "hi": 2.540039766651812, "ta": 3.3401072258858213, "th": 3.478462165603346, "zh": 1.0055878258102513, "my": 1.8347761655608366}, "t": 9.4} | |
| {"event": "train", "step": 0, "phase": 1, "objective": "parity_a", "lm_loss": 5.63341, "bpb": 8.1273, "chunk_loss": 0.135106, "F": 0.86952, "G": 0.52249, "chunks": 890.39, "bytes_per_chunk": 1.15, "boundary_entropy": 0.69066, "router_noise": 0.0, "grad_norm": 9.429, "lr": 6e-07, "router_trainable": true, "sec_per_step": 0.005, "n_target_mean": 9.43, "alpha": 0.03, "t": 10.2} | |
| {"event": "probe", "step": 1, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 14061.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.807964, "cps": 109.851562, "mean_p": 0.520574, "boundary_entropy": 0.686738, "bpb": 8.024465}, "de": {"chunks": 18670.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.909268, "cps": 145.859375, "mean_p": 0.525335, "boundary_entropy": 0.687241, "bpb": 8.163719}, "vi": {"chunks": 21327.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.156876, "cps": 166.617188, "mean_p": 0.523894, "boundary_entropy": 0.687885, "bpb": 8.04658}, "tr": {"chunks": 17814.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.984036, "cps": 139.171875, "mean_p": 0.529735, "boundary_entropy": 0.686396, "bpb": 8.174371}, "fi": {"chunks": 17529.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.931848, "cps": 136.945312, "mean_p": 0.526268, "boundary_entropy": 0.686854, "bpb": 8.007937}, "ru": {"chunks": 34123.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.800686, "cps": 266.585938, "mean_p": 0.540002, "boundary_entropy": 0.687233, "bpb": 8.228421}, "el": {"chunks": 36911.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.798519, "cps": 288.367188, "mean_p": 0.535657, "boundary_entropy": 0.688159, "bpb": 8.166301}, "hi": {"chunks": 34326.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.040178, "cps": 268.171875, "mean_p": 0.509578, "boundary_entropy": 0.690758, "bpb": 8.350512}, "ta": {"chunks": 50656.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.709601, "cps": 395.75, "mean_p": 0.521735, "boundary_entropy": 0.690389, "bpb": 8.072011}, "th": {"chunks": 45388.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.821759, "cps": 354.59375, "mean_p": 0.531356, "boundary_entropy": 0.689101, "bpb": 7.98586}, "zh": {"chunks": 4804.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.81341, "cps": 37.53125, "mean_p": 0.49203, "boundary_entropy": 0.686437, "bpb": 8.094802}, "my": {"chunks": 28140.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.485117, "cps": 219.84375, "mean_p": 0.506185, "boundary_entropy": 0.690852, "bpb": 8.19073}}, "bpb": {"en": 8.02446, "de": 8.16372, "vi": 8.04658, "tr": 8.17437, "fi": 8.00794, "ru": 8.22842, "el": 8.1663, "hi": 8.35051, "ta": 8.07201, "th": 7.98586, "zh": 8.0948, "my": 8.19073}, "gini_cpc": 0.24759, "gini_cps": 0.272258, "bpic": 1.185428, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1253816326164388, "vi": 1.4318401589633083, "tr": 1.217920129676378, "fi": 1.153328886401543, "ru": 2.228670701479926, "el": 2.2259883042854365, "hi": 2.5250852045317727, "ta": 3.353616020548834, "th": 3.4924314695525873, "zh": 1.0067403329207583, "my": 1.8380978581561684}, "t": 17.2} | |
| {"event": "probe", "step": 2, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 13921.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.79992, "cps": 108.757812, "mean_p": 0.519961, "boundary_entropy": 0.686782, "bpb": 7.952918}, "de": {"chunks": 18576.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.90469, "cps": 145.125, "mean_p": 0.524739, "boundary_entropy": 0.687297, "bpb": 8.096459}, "vi": {"chunks": 21253.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.152861, "cps": 166.039062, "mean_p": 0.523612, "boundary_entropy": 0.68791, "bpb": 8.014652}, "tr": {"chunks": 17761.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.981108, "cps": 138.757812, "mean_p": 0.529283, "boundary_entropy": 0.686444, "bpb": 8.123395}, "fi": {"chunks": 17439.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.927064, "cps": 136.242188, "mean_p": 0.525686, "boundary_entropy": 0.686911, "bpb": 7.947504}, "ru": {"chunks": 34105.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.799736, "cps": 266.445312, "mean_p": 0.539704, "boundary_entropy": 0.687279, "bpb": 8.161407}, "el": {"chunks": 36881.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.797057, "cps": 288.132812, "mean_p": 0.535362, "boundary_entropy": 0.688199, "bpb": 8.135207}, "hi": {"chunks": 33694.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 2.002615, "cps": 263.234375, "mean_p": 0.509137, "boundary_entropy": 0.690772, "bpb": 8.266845}, "ta": {"chunks": 50607.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.70698, "cps": 395.367188, "mean_p": 0.521245, "boundary_entropy": 0.690429, "bpb": 8.009476}, "th": {"chunks": 45341.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.818837, "cps": 354.226562, "mean_p": 0.530922, "boundary_entropy": 0.689153, "bpb": 7.909967}, "zh": {"chunks": 4773.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.808161, "cps": 37.289062, "mean_p": 0.491937, "boundary_entropy": 0.686433, "bpb": 8.086223}, "my": {"chunks": 28005.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.477992, "cps": 218.789062, "mean_p": 0.506091, "boundary_entropy": 0.690854, "bpb": 8.160242}}, "bpb": {"en": 7.95292, "de": 8.09646, "vi": 8.01465, "tr": 8.12339, "fi": 7.9475, "ru": 8.16141, "el": 8.13521, "hi": 8.26684, "ta": 8.00948, "th": 7.90997, "zh": 8.08622, "my": 8.16024}, "gini_cpc": 0.248112, "gini_cps": 0.273216, "bpic": 1.19055, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1309762420041207, "vi": 1.441221681639665, "tr": 1.2265084640082995, "fi": 1.1589464805448055, "ru": 2.249896428375801, "el": 2.246547107484514, "hi": 2.503520692491797, "ta": 3.3840658878619676, "th": 3.5239011369654922, "zh": 1.010303083436351, "my": 1.8476762977953174}, "t": 25.2} | |
| {"event": "probe", "step": 4, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 13352.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.767224, "cps": 104.3125, "mean_p": 0.51777, "boundary_entropy": 0.68693, "bpb": 7.704497}, "de": {"chunks": 18154.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.884138, "cps": 141.828125, "mean_p": 0.522627, "boundary_entropy": 0.687485, "bpb": 7.858612}, "vi": {"chunks": 21022.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.140331, "cps": 164.234375, "mean_p": 0.522597, "boundary_entropy": 0.687997, "bpb": 7.903295}, "tr": {"chunks": 17546.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.969232, "cps": 137.078125, "mean_p": 0.527719, "boundary_entropy": 0.686607, "bpb": 7.943501}, "fi": {"chunks": 17149.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.911647, "cps": 133.976562, "mean_p": 0.523669, "boundary_entropy": 0.687099, "bpb": 7.74307}, "ru": {"chunks": 34065.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.797625, "cps": 266.132812, "mean_p": 0.538526, "boundary_entropy": 0.687457, "bpb": 7.918121}, "el": {"chunks": 36826.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.794377, "cps": 287.703125, "mean_p": 0.534454, "boundary_entropy": 0.688319, "bpb": 8.041995}, "hi": {"chunks": 31778.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.888737, "cps": 248.265625, "mean_p": 0.507915, "boundary_entropy": 0.690807, "bpb": 8.044763}, "ta": {"chunks": 50433.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.697673, "cps": 394.007812, "mean_p": 0.519697, "boundary_entropy": 0.690549, "bpb": 7.829479}, "th": {"chunks": 45160.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.807585, "cps": 352.8125, "mean_p": 0.52952, "boundary_entropy": 0.689316, "bpb": 7.678553}, "zh": {"chunks": 4687.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.7936, "cps": 36.617188, "mean_p": 0.491611, "boundary_entropy": 0.686417, "bpb": 8.054991}, "my": {"chunks": 27699.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.461843, "cps": 216.398438, "mean_p": 0.505898, "boundary_entropy": 0.690858, "bpb": 8.073512}}, "bpb": {"en": 7.7045, "de": 7.85861, "vi": 7.90329, "tr": 7.9435, "fi": 7.74307, "ru": 7.91812, "el": 8.04199, "hi": 8.04476, "ta": 7.82948, "th": 7.67855, "zh": 8.05499, "my": 8.07351}, "gini_cpc": 0.250454, "gini_cps": 0.277199, "bpic": 1.207348, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.152385328537904, "vi": 1.4863075583523992, "tr": 1.2632967245844768, "fi": 1.1882414922123221, "ru": 2.343025285708188, "el": 2.3387914644975765, "hi": 2.4617802565264495, "ta": 3.5161478623695213, "th": 3.6594065790020176, "zh": 1.0343780770923416, "my": 1.9053664364822749}, "t": 33.4} | |
| {"event": "probe", "step": 8, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 11238.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.645751, "cps": 87.796875, "mean_p": 0.510778, "boundary_entropy": 0.68727, "bpb": 7.083636}, "de": {"chunks": 16350.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.796279, "cps": 127.734375, "mean_p": 0.515628, "boundary_entropy": 0.687977, "bpb": 7.216359}, "vi": {"chunks": 19972.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 1.083374, "cps": 156.03125, "mean_p": 0.519099, "boundary_entropy": 0.688267, "bpb": 7.564801}, "tr": {"chunks": 16619.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.918025, "cps": 129.835938, "mean_p": 0.522619, "boundary_entropy": 0.687074, "bpb": 7.42407}, "fi": {"chunks": 15669.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.83297, "cps": 122.414062, "mean_p": 0.517365, "boundary_entropy": 0.687574, "bpb": 7.214434}, "ru": {"chunks": 33819.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.784644, "cps": 264.210938, "mean_p": 0.533915, "boundary_entropy": 0.688105, "bpb": 7.187379}, "el": {"chunks": 36559.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.781367, "cps": 285.617188, "mean_p": 0.5302, "boundary_entropy": 0.688842, "bpb": 7.596685}, "hi": {"chunks": 25072.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 1.490163, "cps": 195.875, "mean_p": 0.503504, "boundary_entropy": 0.69088, "bpb": 7.250302}, "ta": {"chunks": 48394.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 2.588607, "cps": 378.078125, "mean_p": 0.5131, "boundary_entropy": 0.690952, "bpb": 7.103722}, "th": {"chunks": 43830.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 2.724899, "cps": 342.421875, "mean_p": 0.523707, "boundary_entropy": 0.689911, "bpb": 6.80992}, "zh": {"chunks": 4322.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.731798, "cps": 33.765625, "mean_p": 0.490213, "boundary_entropy": 0.686344, "bpb": 7.92145}, "my": {"chunks": 25699.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.356291, "cps": 200.773438, "mean_p": 0.50468, "boundary_entropy": 0.690876, "bpb": 7.664091}}, "bpb": {"en": 7.08364, "de": 7.21636, "vi": 7.5648, "tr": 7.42407, "fi": 7.21443, "ru": 7.18738, "el": 7.59668, "hi": 7.2503, "ta": 7.10372, "th": 6.80992, "zh": 7.92145, "my": 7.66409}, "gini_cpc": 0.265085, "gini_cps": 0.292405, "bpic": 1.289834, "worst_lang": "th", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2331060889858678, "vi": 1.6776969224636105, "tr": 1.4216393267698548, "fi": 1.289925088133986, "ru": 2.763672899289585, "el": 2.758598877994996, "hi": 2.307644996664099, "ta": 4.008677727313164, "th": 4.219738107136059, "zh": 1.1332517864410379, "my": 2.100331959184442}, "t": 42.2} | |
| {"event": "probe", "step": 16, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 6976.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.40085, "cps": 54.5, "mean_p": 0.499847, "boundary_entropy": 0.687496, "bpb": 6.269589}, "de": {"chunks": 11023.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.536843, "cps": 86.117188, "mean_p": 0.503715, "boundary_entropy": 0.688416, "bpb": 6.398104}, "vi": {"chunks": 16721.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.907025, "cps": 130.632812, "mean_p": 0.510315, "boundary_entropy": 0.688753, "bpb": 7.013452}, "tr": {"chunks": 13701.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.756836, "cps": 107.039062, "mean_p": 0.512074, "boundary_entropy": 0.687801, "bpb": 6.707258}, "fi": {"chunks": 11266.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.598905, "cps": 88.015625, "mean_p": 0.505693, "boundary_entropy": 0.688063, "bpb": 6.474274}, "ru": {"chunks": 19520.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 1.030079, "cps": 152.5, "mean_p": 0.504288, "boundary_entropy": 0.69026, "bpb": 5.778762}, "el": {"chunks": 31225.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 1.521464, "cps": 243.945312, "mean_p": 0.512469, "boundary_entropy": 0.690269, "bpb": 6.468739}, "hi": {"chunks": 4848.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.288143, "cps": 37.875, "mean_p": 0.489643, "boundary_entropy": 0.690612, "bpb": 5.900847}, "ta": {"chunks": 7995.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.427654, "cps": 62.460938, "mean_p": 0.49195, "boundary_entropy": 0.691075, "bpb": 6.039866}, "th": {"chunks": 22880.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 1.422443, "cps": 178.75, "mean_p": 0.500531, "boundary_entropy": 0.690961, "bpb": 5.554415}, "zh": {"chunks": 3232.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.54724, "cps": 25.25, "mean_p": 0.485584, "boundary_entropy": 0.686045, "bpb": 7.54408}, "my": {"chunks": 20136.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 1.062698, "cps": 157.3125, "mean_p": 0.498655, "boundary_entropy": 0.690879, "bpb": 6.39209}}, "bpb": {"en": 6.26959, "de": 6.3981, "vi": 7.01345, "tr": 6.70726, "fi": 6.47427, "ru": 5.77876, "el": 6.46874, "hi": 5.90085, "ta": 6.03987, "th": 5.55442, "zh": 7.54408, "my": 6.39209}, "gini_cpc": 0.271132, "gini_cps": 0.314355, "bpic": 2.263887, "worst_lang": "el", "cpc_parity_vs_en": {"en": 1.0, "de": 1.3392604646380384, "vi": 2.2627509359067193, "tr": 1.888075527739586, "fi": 1.4940857090497996, "ru": 2.5697344532933117, "el": 3.7955896190199674, "hi": 0.718828332674455, "ta": 1.0668678965945366, "th": 3.5485636797631845, "zh": 1.3651977308102163, "my": 2.6511083327038945}, "t": 51.8} | |
| {"event": "probe", "step": 32, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 1374.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.078952, "cps": 10.734375, "mean_p": 0.486119, "boundary_entropy": 0.687138, "bpb": 5.813134}, "de": {"chunks": 1927.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.093849, "cps": 15.054688, "mean_p": 0.487057, "boundary_entropy": 0.688061, "bpb": 5.952869}, "vi": {"chunks": 2570.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.139409, "cps": 20.078125, "mean_p": 0.483002, "boundary_entropy": 0.688204, "bpb": 6.727115}, "tr": {"chunks": 4318.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.238524, "cps": 33.734375, "mean_p": 0.492344, "boundary_entropy": 0.687996, "bpb": 6.316795}, "fi": {"chunks": 2814.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.149593, "cps": 21.984375, "mean_p": 0.48998, "boundary_entropy": 0.68792, "bpb": 6.012585}, "ru": {"chunks": 229.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.012084, "cps": 1.789062, "mean_p": 0.470251, "boundary_entropy": 0.688385, "bpb": 5.483753}, "el": {"chunks": 2794.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.13614, "cps": 21.828125, "mean_p": 0.489029, "boundary_entropy": 0.690278, "bpb": 5.791848}, "hi": {"chunks": 162.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.009629, "cps": 1.265625, "mean_p": 0.462271, "boundary_entropy": 0.687883, "bpb": 5.583871}, "ta": {"chunks": 212.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.01134, "cps": 1.65625, "mean_p": 0.464297, "boundary_entropy": 0.688551, "bpb": 5.521822}, "th": {"chunks": 384.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.023873, "cps": 3.0, "mean_p": 0.473061, "boundary_entropy": 0.689475, "bpb": 5.362568}, "zh": {"chunks": 4121.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.697765, "cps": 32.195312, "mean_p": 0.489851, "boundary_entropy": 0.686376, "bpb": 7.022408}, "my": {"chunks": 340.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.017944, "cps": 2.65625, "mean_p": 0.464574, "boundary_entropy": 0.688362, "bpb": 5.766663}}, "bpb": {"en": 5.81313, "de": 5.95287, "vi": 6.72712, "tr": 6.3168, "fi": 6.01258, "ru": 5.48375, "el": 5.79185, "hi": 5.58387, "ta": 5.52182, "th": 5.36257, "zh": 7.02241, "my": 5.76666}, "gini_cpc": 0.597421, "gini_cps": 0.464317, "bpic": 18.064533, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.1886847607334405, "vi": 1.7657424942824014, "tr": 3.0211304213371952, "fi": 1.8947398843796448, "ru": 0.15306068601583112, "el": 1.7243401640974965, "hi": 0.12195435935036367, "ta": 0.14363086596958063, "th": 0.30237615940425655, "zh": 8.837848638864775, "my": 0.2272756604377137}, "t": 61.9} | |
| {"event": "probe", "step": 64, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 1390.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.079871, "cps": 10.859375, "mean_p": 0.48957, "boundary_entropy": 0.687449, "bpb": 5.429005}, "de": {"chunks": 1416.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.068962, "cps": 11.0625, "mean_p": 0.488341, "boundary_entropy": 0.688234, "bpb": 5.607902}, "vi": {"chunks": 3907.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.211934, "cps": 30.523438, "mean_p": 0.489254, "boundary_entropy": 0.688753, "bpb": 6.093205}, "tr": {"chunks": 2615.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.144451, "cps": 20.429688, "mean_p": 0.490346, "boundary_entropy": 0.688007, "bpb": 5.981546}, "fi": {"chunks": 1617.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.08596, "cps": 12.632812, "mean_p": 0.488803, "boundary_entropy": 0.687964, "bpb": 5.630504}, "ru": {"chunks": 1575.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.083113, "cps": 12.304688, "mean_p": 0.491111, "boundary_entropy": 0.690257, "bpb": 4.888601}, "el": {"chunks": 590.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.028748, "cps": 4.609375, "mean_p": 0.488179, "boundary_entropy": 0.690332, "bpb": 5.327889}, "hi": {"chunks": 6865.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.408024, "cps": 53.632812, "mean_p": 0.496205, "boundary_entropy": 0.690979, "bpb": 4.514106}, "ta": {"chunks": 20684.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 1.106392, "cps": 161.59375, "mean_p": 0.500112, "boundary_entropy": 0.691326, "bpb": 4.57931}, "th": {"chunks": 7637.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.47479, "cps": 59.664062, "mean_p": 0.494915, "boundary_entropy": 0.691037, "bpb": 4.536395}, "zh": {"chunks": 3545.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.600237, "cps": 27.695312, "mean_p": 0.49168, "boundary_entropy": 0.686847, "bpb": 6.692137}, "my": {"chunks": 564.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.029766, "cps": 4.40625, "mean_p": 0.483422, "boundary_entropy": 0.690709, "bpb": 4.57366}}, "bpb": {"en": 5.429, "de": 5.6079, "vi": 6.0932, "tr": 5.98155, "fi": 5.6305, "ru": 4.8886, "el": 5.32789, "hi": 4.51411, "ta": 4.57931, "th": 4.5364, "zh": 6.69214, "my": 4.57366}, "gini_cpc": 0.549699, "gini_cps": 0.544527, "bpic": 7.323366, "worst_lang": "ta", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8634161467397454, "vi": 2.653441939694786, "tr": 1.8085497574431202, "fi": 1.0762358546526236, "ru": 1.0405924337047512, "el": 0.3599320222231804, "hi": 5.108516360759832, "ta": 13.852188077834068, "th": 5.944441333473476, "zh": 7.515054194368853, "my": 0.3726705272893781}, "t": 77.3} | |
| {"event": "train", "step": 100, "phase": 1, "objective": "parity_a", "lm_loss": 3.4139, "bpb": 4.92522, "chunk_loss": 0.033884, "F": 0.16879, "G": 0.48316, "chunks": 172.84, "bytes_per_chunk": 5.924, "boundary_entropy": 0.6911, "router_noise": 0.0, "grad_norm": 1.7833, "lr": 6.06e-05, "router_trainable": true, "sec_per_step": 0.783, "n_target_mean": 8.4315, "alpha": 0.03, "t": 88.4} | |
| {"event": "probe", "step": 128, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 2208.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.126875, "cps": 17.25, "mean_p": 0.463669, "boundary_entropy": 0.682609, "bpb": 4.367298}, "de": {"chunks": 3287.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.160084, "cps": 25.679688, "mean_p": 0.462109, "boundary_entropy": 0.682705, "bpb": 4.630838}, "vi": {"chunks": 4889.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.265202, "cps": 38.195312, "mean_p": 0.483762, "boundary_entropy": 0.687771, "bpb": 4.967063}, "tr": {"chunks": 3221.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.177926, "cps": 25.164062, "mean_p": 0.468584, "boundary_entropy": 0.683947, "bpb": 4.850718}, "fi": {"chunks": 2108.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.112062, "cps": 16.46875, "mean_p": 0.465787, "boundary_entropy": 0.68389, "bpb": 4.648756}, "ru": {"chunks": 4503.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.237625, "cps": 35.179688, "mean_p": 0.449741, "boundary_entropy": 0.681378, "bpb": 3.283858}, "el": {"chunks": 7331.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.357209, "cps": 57.273438, "mean_p": 0.494746, "boundary_entropy": 0.690535, "bpb": 4.492744}, "hi": {"chunks": 1721.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.102288, "cps": 13.445312, "mean_p": 0.473434, "boundary_entropy": 0.688857, "bpb": 2.993691}, "ta": {"chunks": 1548.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.082803, "cps": 12.09375, "mean_p": 0.482003, "boundary_entropy": 0.690382, "bpb": 3.259341}, "th": {"chunks": 2401.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.14927, "cps": 18.757812, "mean_p": 0.473266, "boundary_entropy": 0.688932, "bpb": 3.006675}, "zh": {"chunks": 3009.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.509482, "cps": 23.507812, "mean_p": 0.489284, "boundary_entropy": 0.686692, "bpb": 6.257684}, "my": {"chunks": 489.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.025807, "cps": 3.820312, "mean_p": 0.463841, "boundary_entropy": 0.687908, "bpb": 3.650685}}, "bpb": {"en": 4.3673, "de": 4.63084, "vi": 4.96706, "tr": 4.85072, "fi": 4.64876, "ru": 3.28386, "el": 4.49274, "hi": 2.99369, "ta": 3.25934, "th": 3.00668, "zh": 6.25768, "my": 3.65069}, "gini_cpc": 0.351243, "gini_cps": 0.304932, "bpic": 10.452976, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 1.2617471953948998, "vi": 2.090267877147675, "tr": 1.4023784341533592, "fi": 0.883250260601452, "ru": 1.8729137748078466, "el": 2.8154474486737042, "hi": 0.8062149525163124, "ta": 0.6526352663465004, "th": 1.176511416929086, "zh": 4.015630981389596, "my": 0.20340917304568107}, "t": 98.4} | |
| {"event": "train", "step": 200, "phase": 1, "objective": "parity_a", "lm_loss": 2.39426, "bpb": 3.45419, "chunk_loss": 0.029656, "F": 0.13551, "G": 0.40263, "chunks": 138.77, "bytes_per_chunk": 7.379, "boundary_entropy": 0.65328, "router_noise": 0.0, "grad_norm": 1.4827, "lr": 0.0001206, "router_trainable": true, "sec_per_step": 0.256, "n_target_mean": 9.5617, "alpha": 0.03, "t": 114.0} | |
| {"event": "probe", "step": 256, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3272.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.188014, "cps": 25.5625, "mean_p": 0.33201, "boundary_entropy": 0.56097, "bpb": 3.597863}, "de": {"chunks": 3211.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.156382, "cps": 25.085938, "mean_p": 0.323059, "boundary_entropy": 0.565409, "bpb": 3.787944}, "vi": {"chunks": 4322.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.234445, "cps": 33.765625, "mean_p": 0.348778, "boundary_entropy": 0.595882, "bpb": 3.063932}, "tr": {"chunks": 3930.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.217091, "cps": 30.703125, "mean_p": 0.348079, "boundary_entropy": 0.570433, "bpb": 3.788647}, "fi": {"chunks": 2345.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.124661, "cps": 18.320312, "mean_p": 0.333671, "boundary_entropy": 0.576568, "bpb": 3.795792}, "ru": {"chunks": 3227.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.17029, "cps": 25.210938, "mean_p": 0.341599, "boundary_entropy": 0.620241, "bpb": 2.625312}, "el": {"chunks": 3589.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.174877, "cps": 28.039062, "mean_p": 0.411216, "boundary_entropy": 0.665188, "bpb": 2.683246}, "hi": {"chunks": 3248.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.193046, "cps": 25.375, "mean_p": 0.415697, "boundary_entropy": 0.671803, "bpb": 2.091178}, "ta": {"chunks": 3060.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.16368, "cps": 23.90625, "mean_p": 0.44329, "boundary_entropy": 0.682124, "bpb": 1.956154}, "th": {"chunks": 4716.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.293192, "cps": 36.84375, "mean_p": 0.474839, "boundary_entropy": 0.688171, "bpb": 2.084164}, "zh": {"chunks": 2902.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.491365, "cps": 22.671875, "mean_p": 0.432943, "boundary_entropy": 0.662862, "bpb": 4.886808}, "my": {"chunks": 9722.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.513088, "cps": 75.953125, "mean_p": 0.451051, "boundary_entropy": 0.681853, "bpb": 2.179205}}, "bpb": {"en": 3.59786, "de": 3.78794, "vi": 3.06393, "tr": 3.78865, "fi": 3.79579, "ru": 2.62531, "el": 2.68325, "hi": 2.09118, "ta": 1.95615, "th": 2.08416, "zh": 4.88681, "my": 2.1792}, "gini_cpc": 0.246752, "gini_cps": 0.194578, "bpic": 8.072123, "worst_lang": "my", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8317613265401261, "vi": 1.246959780050571, "tr": 1.1546565516771932, "fi": 0.66304314413652, "ru": 0.9057338027623846, "el": 0.9301295421786303, "hi": 1.0267666943983258, "ta": 0.8705761840273442, "th": 1.5594216391678377, "zh": 2.613453580420889, "my": 2.7289970478979146}, "t": 128.2} | |
| {"event": "train", "step": 300, "phase": 1, "objective": "parity_a", "lm_loss": 2.13926, "bpb": 3.08631, "chunk_loss": 0.029964, "F": 0.13618, "G": 0.32924, "chunks": 139.45, "bytes_per_chunk": 7.343, "boundary_entropy": 0.57725, "router_noise": 0.0, "grad_norm": 1.1875, "lr": 0.0001806, "router_trainable": true, "sec_per_step": 0.266, "n_target_mean": 9.5002, "alpha": 0.03, "t": 140.6} | |
| {"event": "train", "step": 400, "phase": 1, "objective": "parity_a", "lm_loss": 2.00257, "bpb": 2.8891, "chunk_loss": 0.030627, "F": 0.15169, "G": 0.3008, "chunks": 155.33, "bytes_per_chunk": 6.592, "boundary_entropy": 0.51887, "router_noise": 0.0, "grad_norm": 1.0925, "lr": 0.0002406, "router_trainable": true, "sec_per_step": 0.189, "n_target_mean": 8.7646, "alpha": 0.03, "t": 159.6} | |
| {"event": "train", "step": 500, "phase": 1, "objective": "parity_a", "lm_loss": 1.7914, "bpb": 2.58444, "chunk_loss": 0.030762, "F": 0.13762, "G": 0.28692, "chunks": 140.92, "bytes_per_chunk": 7.266, "boundary_entropy": 0.49455, "router_noise": 0.0, "grad_norm": 1.1378, "lr": 0.0003, "router_trainable": true, "sec_per_step": 0.18, "n_target_mean": 9.6185, "alpha": 0.03, "t": 177.6} | |
| {"event": "train", "step": 600, "phase": 1, "objective": "parity_a", "lm_loss": 1.79223, "bpb": 2.58564, "chunk_loss": 0.030361, "F": 0.14359, "G": 0.27215, "chunks": 147.03, "bytes_per_chunk": 6.965, "boundary_entropy": 0.45324, "router_noise": 0.0, "grad_norm": 0.8667, "lr": 0.0002999, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 8.8282, "alpha": 0.03, "t": 195.5} | |
| {"event": "train", "step": 700, "phase": 1, "objective": "parity_a", "lm_loss": 1.76872, "bpb": 2.55173, "chunk_loss": 0.030425, "F": 0.14525, "G": 0.27031, "chunks": 148.73, "bytes_per_chunk": 6.885, "boundary_entropy": 0.43928, "router_noise": 0.0, "grad_norm": 0.8179, "lr": 0.0002998, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 8.8615, "alpha": 0.03, "t": 212.7} | |
| {"event": "train", "step": 800, "phase": 1, "objective": "parity_a", "lm_loss": 1.62745, "bpb": 2.34792, "chunk_loss": 0.030279, "F": 0.13554, "G": 0.26336, "chunks": 138.8, "bytes_per_chunk": 7.378, "boundary_entropy": 0.43657, "router_noise": 0.0, "grad_norm": 0.7477, "lr": 0.0002995, "router_trainable": true, "sec_per_step": 0.196, "n_target_mean": 9.5699, "alpha": 0.03, "t": 232.3} | |
| {"event": "train", "step": 900, "phase": 1, "objective": "parity_a", "lm_loss": 1.56483, "bpb": 2.25757, "chunk_loss": 0.030374, "F": 0.14677, "G": 0.25051, "chunks": 150.3, "bytes_per_chunk": 6.813, "boundary_entropy": 0.38908, "router_noise": 0.0, "grad_norm": 0.7444, "lr": 0.000299, "router_trainable": true, "sec_per_step": 0.18, "n_target_mean": 8.7456, "alpha": 0.03, "t": 250.3} | |
| {"event": "probe", "step": 1000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3141.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.180486, "cps": 24.539062, "mean_p": 0.225121, "boundary_entropy": 0.284046, "bpb": 2.322583}, "de": {"chunks": 3207.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.156188, "cps": 25.054688, "mean_p": 0.197765, "boundary_entropy": 0.284137, "bpb": 2.615987}, "vi": {"chunks": 3264.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.177055, "cps": 25.5, "mean_p": 0.211248, "boundary_entropy": 0.337223, "bpb": 2.124232}, "tr": {"chunks": 3101.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.171298, "cps": 24.226562, "mean_p": 0.215444, "boundary_entropy": 0.319634, "bpb": 2.61619}, "fi": {"chunks": 3179.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.168997, "cps": 24.835938, "mean_p": 0.211461, "boundary_entropy": 0.295083, "bpb": 2.828311}, "ru": {"chunks": 2599.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.13715, "cps": 20.304688, "mean_p": 0.252737, "boundary_entropy": 0.483759, "bpb": 1.76387}, "el": {"chunks": 3078.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.149978, "cps": 24.046875, "mean_p": 0.266422, "boundary_entropy": 0.504276, "bpb": 1.818023}, "hi": {"chunks": 2539.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.150906, "cps": 19.835938, "mean_p": 0.332048, "boundary_entropy": 0.590189, "bpb": 1.411933}, "ta": {"chunks": 2906.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.155443, "cps": 22.703125, "mean_p": 0.199272, "boundary_entropy": 0.418182, "bpb": 1.165239}, "th": {"chunks": 1648.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.102456, "cps": 12.875, "mean_p": 0.222893, "boundary_entropy": 0.462595, "bpb": 1.429053}, "zh": {"chunks": 2966.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.502201, "cps": 23.171875, "mean_p": 0.322675, "boundary_entropy": 0.506736, "bpb": 3.512967}, "my": {"chunks": 2360.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.124551, "cps": 18.4375, "mean_p": 0.223188, "boundary_entropy": 0.465415, "bpb": 1.308641}}, "bpb": {"en": 2.32258, "de": 2.61599, "vi": 2.12423, "tr": 2.61619, "fi": 2.82831, "ru": 1.76387, "el": 1.81802, "hi": 1.41193, "ta": 1.16524, "th": 1.42905, "zh": 3.51297, "my": 1.30864}, "gini_cpc": 0.204497, "gini_cps": 0.081612, "bpic": 11.291662, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8653717957967445, "vi": 0.9809868639368711, "tr": 0.9490900023905445, "fi": 0.9363426985480716, "ru": 0.7598944087013277, "el": 0.8309673387618643, "hi": 0.8361107586807847, "ta": 0.8612442278949803, "th": 0.5676652715590735, "zh": 2.7824917660993256, "my": 0.6900885326532838}, "t": 271.4} | |
| {"event": "eval", "step": 1000, "phase": 1, "bpb": {"en": 2.33557, "de": 2.60875, "vi": 2.12665, "tr": 2.60997, "fi": 2.84054, "ru": 1.76224, "el": 1.8125, "hi": 1.41888, "ta": 1.16486, "th": 1.42751, "zh": 3.51714, "my": 1.31931}, "bytes_per_chunk": {"en": 5.37, "de": 6.457, "vi": 7.289, "tr": 6.638, "fi": 6.137, "ru": 13.672, "el": 11.979, "hi": 16.531, "ta": 19.128, "th": 26.899, "zh": 5.289, "my": 23.766}, "bpb_mean": 2.07866, "t": 276.4} | |
| {"event": "train", "step": 1000, "phase": 1, "objective": "parity_a", "lm_loss": 1.55458, "bpb": 2.24279, "chunk_loss": 0.029865, "F": 0.13002, "G": 0.25013, "chunks": 133.14, "bytes_per_chunk": 7.691, "boundary_entropy": 0.40504, "router_noise": 0.0, "grad_norm": 0.6025, "lr": 0.0002985, "router_trainable": true, "sec_per_step": 0.262, "n_target_mean": 9.5287, "alpha": 0.03, "t": 276.6} | |
| {"event": "train", "step": 1100, "phase": 1, "objective": "parity_a", "lm_loss": 1.46976, "bpb": 2.12041, "chunk_loss": 0.030037, "F": 0.12657, "G": 0.2427, "chunks": 129.61, "bytes_per_chunk": 7.901, "boundary_entropy": 0.40065, "router_noise": 0.0, "grad_norm": 0.6525, "lr": 0.0002978, "router_trainable": true, "sec_per_step": 0.183, "n_target_mean": 9.6044, "alpha": 0.03, "t": 294.9} | |
| {"event": "train", "step": 1200, "phase": 1, "objective": "parity_a", "lm_loss": 1.5258, "bpb": 2.20127, "chunk_loss": 0.030439, "F": 0.13934, "G": 0.24349, "chunks": 142.69, "bytes_per_chunk": 7.177, "boundary_entropy": 0.38333, "router_noise": 0.0, "grad_norm": 0.6584, "lr": 0.000297, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 9.0775, "alpha": 0.03, "t": 312.8} | |
| {"event": "train", "step": 1300, "phase": 1, "objective": "parity_a", "lm_loss": 1.39388, "bpb": 2.01094, "chunk_loss": 0.030501, "F": 0.13501, "G": 0.23479, "chunks": 138.25, "bytes_per_chunk": 7.407, "boundary_entropy": 0.36692, "router_noise": 0.0, "grad_norm": 0.6348, "lr": 0.0002962, "router_trainable": true, "sec_per_step": 0.184, "n_target_mean": 9.6616, "alpha": 0.03, "t": 331.2} | |
| {"event": "train", "step": 1400, "phase": 1, "objective": "parity_a", "lm_loss": 1.35453, "bpb": 1.95417, "chunk_loss": 0.030179, "F": 0.13095, "G": 0.2236, "chunks": 134.09, "bytes_per_chunk": 7.636, "boundary_entropy": 0.35161, "router_noise": 0.0, "grad_norm": 0.6043, "lr": 0.0002951, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 9.9045, "alpha": 0.03, "t": 349.8} | |
| {"event": "train", "step": 1500, "phase": 1, "objective": "parity_a", "lm_loss": 1.43629, "bpb": 2.07214, "chunk_loss": 0.030223, "F": 0.13371, "G": 0.22292, "chunks": 136.92, "bytes_per_chunk": 7.479, "boundary_entropy": 0.35154, "router_noise": 0.0, "grad_norm": 0.5898, "lr": 0.000294, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.3047, "alpha": 0.03, "t": 367.6} | |
| {"event": "train", "step": 1600, "phase": 1, "objective": "parity_a", "lm_loss": 1.46703, "bpb": 2.11647, "chunk_loss": 0.030216, "F": 0.14027, "G": 0.22373, "chunks": 143.64, "bytes_per_chunk": 7.129, "boundary_entropy": 0.34069, "router_noise": 0.0, "grad_norm": 0.5555, "lr": 0.0002928, "router_trainable": true, "sec_per_step": 0.179, "n_target_mean": 8.9522, "alpha": 0.03, "t": 385.5} | |
| {"event": "train", "step": 1700, "phase": 1, "objective": "parity_a", "lm_loss": 1.49818, "bpb": 2.16141, "chunk_loss": 0.030468, "F": 0.15173, "G": 0.23732, "chunks": 155.38, "bytes_per_chunk": 6.591, "boundary_entropy": 0.34673, "router_noise": 0.0, "grad_norm": 0.5522, "lr": 0.0002914, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 8.7266, "alpha": 0.03, "t": 404.0} | |
| {"event": "train", "step": 1800, "phase": 1, "objective": "parity_a", "lm_loss": 1.30248, "bpb": 1.87909, "chunk_loss": 0.029911, "F": 0.12378, "G": 0.2121, "chunks": 126.75, "bytes_per_chunk": 8.079, "boundary_entropy": 0.33467, "router_noise": 0.0, "grad_norm": 0.4905, "lr": 0.0002899, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 9.8471, "alpha": 0.03, "t": 422.5} | |
| {"event": "train", "step": 1900, "phase": 1, "objective": "parity_a", "lm_loss": 1.29012, "bpb": 1.86124, "chunk_loss": 0.030111, "F": 0.1291, "G": 0.21909, "chunks": 132.2, "bytes_per_chunk": 7.746, "boundary_entropy": 0.33678, "router_noise": 0.0, "grad_norm": 0.4973, "lr": 0.0002883, "router_trainable": true, "sec_per_step": 0.178, "n_target_mean": 9.8798, "alpha": 0.03, "t": 440.3} | |
| {"event": "probe", "step": 2000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3124.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.179509, "cps": 24.40625, "mean_p": 0.205127, "boundary_entropy": 0.216634, "bpb": 2.096102}, "de": {"chunks": 3245.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.158038, "cps": 25.351562, "mean_p": 0.193311, "boundary_entropy": 0.238019, "bpb": 2.319988}, "vi": {"chunks": 3061.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.166043, "cps": 23.914062, "mean_p": 0.191057, "boundary_entropy": 0.287139, "bpb": 1.960217}, "tr": {"chunks": 3167.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.174943, "cps": 24.742188, "mean_p": 0.208184, "boundary_entropy": 0.271947, "bpb": 2.289836}, "fi": {"chunks": 3248.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.172665, "cps": 25.375, "mean_p": 0.210848, "boundary_entropy": 0.258843, "bpb": 2.468308}, "ru": {"chunks": 2846.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.150185, "cps": 22.234375, "mean_p": 0.240994, "boundary_entropy": 0.463492, "bpb": 1.459237}, "el": {"chunks": 2398.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.116845, "cps": 18.734375, "mean_p": 0.194033, "boundary_entropy": 0.396423, "bpb": 1.511739}, "hi": {"chunks": 2620.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.155721, "cps": 20.46875, "mean_p": 0.152815, "boundary_entropy": 0.342197, "bpb": 1.219373}, "ta": {"chunks": 1614.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.086333, "cps": 12.609375, "mean_p": 0.183044, "boundary_entropy": 0.413143, "bpb": 1.017082}, "th": {"chunks": 3472.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.215853, "cps": 27.125, "mean_p": 0.224352, "boundary_entropy": 0.445998, "bpb": 1.16812}, "zh": {"chunks": 3080.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.521504, "cps": 24.0625, "mean_p": 0.290169, "boundary_entropy": 0.412354, "bpb": 3.18456}, "my": {"chunks": 2682.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.141545, "cps": 20.953125, "mean_p": 0.358973, "boundary_entropy": 0.631928, "bpb": 1.16297}}, "bpb": {"en": 2.0961, "de": 2.31999, "vi": 1.96022, "tr": 2.28984, "fi": 2.46831, "ru": 1.45924, "el": 1.51174, "hi": 1.21937, "ta": 1.01708, "th": 1.16812, "zh": 3.18456, "my": 1.16297}, "gini_cpc": 0.227997, "gini_cps": 0.086189, "bpic": 11.105738, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8803905838968943, "vi": 0.9249820023774573, "tr": 0.9745645436346846, "fi": 0.9618718881009976, "ru": 0.8366402927036916, "el": 0.6509107267490339, "hi": 0.8674796856242141, "ta": 0.4809402844160198, "th": 1.2024630673236014, "zh": 2.905162093454735, "my": 0.7885123366139609}, "t": 461.3} | |
| {"event": "eval", "step": 2000, "phase": 1, "bpb": {"en": 2.1043, "de": 2.3139, "vi": 1.96533, "tr": 2.28102, "fi": 2.4832, "ru": 1.45814, "el": 1.50752, "hi": 1.22644, "ta": 1.01738, "th": 1.16421, "zh": 3.19279, "my": 1.17125}, "bytes_per_chunk": {"en": 5.351, "de": 6.294, "vi": 7.382, "tr": 6.109, "fi": 5.703, "ru": 11.963, "el": 15.444, "hi": 15.411, "ta": 29.826, "th": 13.828, "zh": 5.349, "my": 17.213}, "bpb_mean": 1.82379, "t": 466.3} | |
| {"event": "train", "step": 2000, "phase": 1, "objective": "parity_a", "lm_loss": 1.38488, "bpb": 1.99796, "chunk_loss": 0.030208, "F": 0.14616, "G": 0.21982, "chunks": 149.67, "bytes_per_chunk": 6.842, "boundary_entropy": 0.31246, "router_noise": 0.0, "grad_norm": 0.5025, "lr": 0.0002866, "router_trainable": true, "sec_per_step": 0.262, "n_target_mean": 8.7556, "alpha": 0.03, "t": 466.5} | |
| {"event": "train", "step": 2100, "phase": 1, "objective": "parity_a", "lm_loss": 1.40609, "bpb": 2.02856, "chunk_loss": 0.029961, "F": 0.14003, "G": 0.21501, "chunks": 143.39, "bytes_per_chunk": 7.141, "boundary_entropy": 0.31059, "router_noise": 0.0, "grad_norm": 0.5111, "lr": 0.0002848, "router_trainable": true, "sec_per_step": 0.188, "n_target_mean": 8.4312, "alpha": 0.03, "t": 485.3} | |
| {"event": "train", "step": 2200, "phase": 1, "objective": "parity_a", "lm_loss": 1.26569, "bpb": 1.82601, "chunk_loss": 0.030271, "F": 0.13295, "G": 0.2121, "chunks": 136.14, "bytes_per_chunk": 7.522, "boundary_entropy": 0.31283, "router_noise": 0.0, "grad_norm": 0.4737, "lr": 0.0002829, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 9.4538, "alpha": 0.03, "t": 503.8} | |
| {"event": "train", "step": 2300, "phase": 1, "objective": "parity_a", "lm_loss": 1.40658, "bpb": 2.02927, "chunk_loss": 0.030472, "F": 0.15121, "G": 0.22789, "chunks": 154.84, "bytes_per_chunk": 6.613, "boundary_entropy": 0.31576, "router_noise": 0.0, "grad_norm": 0.4846, "lr": 0.0002809, "router_trainable": true, "sec_per_step": 0.189, "n_target_mean": 8.852, "alpha": 0.03, "t": 522.8} | |
| {"event": "train", "step": 2400, "phase": 1, "objective": "parity_a", "lm_loss": 1.3116, "bpb": 1.89224, "chunk_loss": 0.030108, "F": 0.13823, "G": 0.20522, "chunks": 141.55, "bytes_per_chunk": 7.234, "boundary_entropy": 0.2867, "router_noise": 0.0, "grad_norm": 0.4805, "lr": 0.0002788, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 8.6967, "alpha": 0.03, "t": 541.4} | |
| {"event": "train", "step": 2500, "phase": 1, "objective": "parity_a", "lm_loss": 1.28248, "bpb": 1.85023, "chunk_loss": 0.030239, "F": 0.1394, "G": 0.21312, "chunks": 142.75, "bytes_per_chunk": 7.173, "boundary_entropy": 0.30021, "router_noise": 0.0, "grad_norm": 0.4728, "lr": 0.0002765, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 9.137, "alpha": 0.03, "t": 560.0} | |
| {"event": "train", "step": 2600, "phase": 1, "objective": "parity_a", "lm_loss": 1.38399, "bpb": 1.99668, "chunk_loss": 0.029898, "F": 0.14558, "G": 0.2153, "chunks": 149.08, "bytes_per_chunk": 6.869, "boundary_entropy": 0.29006, "router_noise": 0.0, "grad_norm": 0.45, "lr": 0.0002742, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 8.5157, "alpha": 0.03, "t": 578.6} | |
| {"event": "train", "step": 2700, "phase": 1, "objective": "parity_a", "lm_loss": 1.24402, "bpb": 1.79475, "chunk_loss": 0.030087, "F": 0.13216, "G": 0.2111, "chunks": 135.33, "bytes_per_chunk": 7.567, "boundary_entropy": 0.30085, "router_noise": 0.0, "grad_norm": 0.4688, "lr": 0.0002718, "router_trainable": true, "sec_per_step": 0.189, "n_target_mean": 9.8013, "alpha": 0.03, "t": 597.5} | |
| {"event": "train", "step": 2800, "phase": 1, "objective": "parity_a", "lm_loss": 1.3096, "bpb": 1.88935, "chunk_loss": 0.03012, "F": 0.13977, "G": 0.20926, "chunks": 143.12, "bytes_per_chunk": 7.155, "boundary_entropy": 0.28594, "router_noise": 0.0, "grad_norm": 0.4288, "lr": 0.0002693, "router_trainable": true, "sec_per_step": 0.189, "n_target_mean": 8.995, "alpha": 0.03, "t": 616.4} | |
| {"event": "train", "step": 2900, "phase": 1, "objective": "parity_a", "lm_loss": 1.37962, "bpb": 1.99037, "chunk_loss": 0.030216, "F": 0.13853, "G": 0.20333, "chunks": 141.86, "bytes_per_chunk": 7.218, "boundary_entropy": 0.27329, "router_noise": 0.0, "grad_norm": 0.4924, "lr": 0.0002667, "router_trainable": true, "sec_per_step": 0.187, "n_target_mean": 8.9685, "alpha": 0.03, "t": 635.1} | |
| {"event": "probe", "step": 3000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3066.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.176177, "cps": 23.953125, "mean_p": 0.196764, "boundary_entropy": 0.178943, "bpb": 1.99039}, "de": {"chunks": 3095.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.150733, "cps": 24.179688, "mean_p": 0.182558, "boundary_entropy": 0.201161, "bpb": 2.187277}, "vi": {"chunks": 2837.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.153892, "cps": 22.164062, "mean_p": 0.178599, "boundary_entropy": 0.264109, "bpb": 1.859927}, "tr": {"chunks": 2868.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.158427, "cps": 22.40625, "mean_p": 0.183235, "boundary_entropy": 0.230231, "bpb": 2.162659}, "fi": {"chunks": 2817.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.149753, "cps": 22.007812, "mean_p": 0.178731, "boundary_entropy": 0.201864, "bpb": 2.326774}, "ru": {"chunks": 2918.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.153984, "cps": 22.796875, "mean_p": 0.244452, "boundary_entropy": 0.459353, "bpb": 1.337054}, "el": {"chunks": 2649.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.129075, "cps": 20.695312, "mean_p": 0.177217, "boundary_entropy": 0.357612, "bpb": 1.379896}, "hi": {"chunks": 2396.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.142407, "cps": 18.71875, "mean_p": 0.149416, "boundary_entropy": 0.329617, "bpb": 1.128084}, "ta": {"chunks": 2388.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.127735, "cps": 18.65625, "mean_p": 0.209763, "boundary_entropy": 0.440542, "bpb": 0.946685}, "th": {"chunks": 3127.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.194405, "cps": 24.429688, "mean_p": 0.192659, "boundary_entropy": 0.39608, "bpb": 1.083054}, "zh": {"chunks": 3223.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.545716, "cps": 25.179688, "mean_p": 0.266635, "boundary_entropy": 0.331423, "bpb": 2.99385}, "my": {"chunks": 2591.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.136743, "cps": 20.242188, "mean_p": 0.34091, "boundary_entropy": 0.61062, "bpb": 1.082755}}, "bpb": {"en": 1.99039, "de": 2.18728, "vi": 1.85993, "tr": 2.16266, "fi": 2.32677, "ru": 1.33705, "el": 1.3799, "hi": 1.12808, "ta": 0.94669, "th": 1.08305, "zh": 2.99385, "my": 1.08275}, "gini_cpc": 0.208714, "gini_cps": 0.053252, "bpic": 11.295982, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8555791960304842, "vi": 0.8735105678862156, "tr": 0.8992502459020911, "fi": 0.8500156724304938, "ru": 0.8740334281686796, "el": 0.7326441442768664, "hi": 0.8083207183420905, "ta": 0.725038089550614, "th": 1.1034655666721618, "zh": 3.0975536172375695, "my": 0.7761684879040669}, "t": 656.9} | |
| {"event": "eval", "step": 3000, "phase": 1, "bpb": {"en": 1.99919, "de": 2.17975, "vi": 1.86597, "tr": 2.15501, "fi": 2.34236, "ru": 1.33775, "el": 1.3765, "hi": 1.135, "ta": 0.94676, "th": 1.07939, "zh": 3.0019, "my": 1.09267}, "bytes_per_chunk": {"en": 5.42, "de": 6.404, "vi": 7.885, "tr": 6.687, "fi": 6.343, "ru": 11.648, "el": 13.33, "hi": 16.626, "ta": 20.607, "th": 14.303, "zh": 5.08, "my": 20.652}, "bpb_mean": 1.70935, "t": 661.9} | |
| {"event": "train", "step": 3000, "phase": 1, "objective": "parity_a", "lm_loss": 1.41096, "bpb": 2.03558, "chunk_loss": 0.030238, "F": 0.15308, "G": 0.21746, "chunks": 156.75, "bytes_per_chunk": 6.533, "boundary_entropy": 0.27951, "router_noise": 0.0, "grad_norm": 0.4602, "lr": 0.000264, "router_trainable": true, "sec_per_step": 0.27, "n_target_mean": 8.276, "alpha": 0.03, "t": 662.1} | |
| {"event": "train", "step": 3100, "phase": 1, "objective": "parity_a", "lm_loss": 1.28748, "bpb": 1.85745, "chunk_loss": 0.029929, "F": 0.13408, "G": 0.20687, "chunks": 137.3, "bytes_per_chunk": 7.458, "boundary_entropy": 0.28661, "router_noise": 0.0, "grad_norm": 0.4421, "lr": 0.0002612, "router_trainable": true, "sec_per_step": 0.182, "n_target_mean": 9.1918, "alpha": 0.03, "t": 680.3} | |
| {"event": "train", "step": 3200, "phase": 1, "objective": "parity_a", "lm_loss": 1.32247, "bpb": 1.90792, "chunk_loss": 0.030459, "F": 0.1436, "G": 0.21007, "chunks": 147.05, "bytes_per_chunk": 6.964, "boundary_entropy": 0.28232, "router_noise": 0.0, "grad_norm": 0.4364, "lr": 0.0002583, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 8.8996, "alpha": 0.03, "t": 698.4} | |
| {"event": "train", "step": 3300, "phase": 1, "objective": "parity_a", "lm_loss": 1.19274, "bpb": 1.72076, "chunk_loss": 0.030126, "F": 0.12474, "G": 0.21553, "chunks": 127.73, "bytes_per_chunk": 8.017, "boundary_entropy": 0.32143, "router_noise": 0.0, "grad_norm": 0.4077, "lr": 0.0002553, "router_trainable": true, "sec_per_step": 0.182, "n_target_mean": 10.211, "alpha": 0.03, "t": 716.6} | |
| {"event": "train", "step": 3400, "phase": 1, "objective": "parity_a", "lm_loss": 1.2422, "bpb": 1.79212, "chunk_loss": 0.030109, "F": 0.14059, "G": 0.20628, "chunks": 143.97, "bytes_per_chunk": 7.113, "boundary_entropy": 0.27308, "router_noise": 0.0, "grad_norm": 0.4354, "lr": 0.0002523, "router_trainable": true, "sec_per_step": 0.18, "n_target_mean": 9.1142, "alpha": 0.03, "t": 734.6} | |
| {"event": "train", "step": 3500, "phase": 1, "objective": "parity_a", "lm_loss": 1.22547, "bpb": 1.76798, "chunk_loss": 0.030531, "F": 0.14574, "G": 0.21853, "chunks": 149.23, "bytes_per_chunk": 6.862, "boundary_entropy": 0.28675, "router_noise": 0.0, "grad_norm": 0.4313, "lr": 0.0002492, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 9.3465, "alpha": 0.03, "t": 752.7} | |
| {"event": "train", "step": 3600, "phase": 1, "objective": "parity_a", "lm_loss": 1.28655, "bpb": 1.8561, "chunk_loss": 0.030139, "F": 0.14052, "G": 0.19686, "chunks": 143.89, "bytes_per_chunk": 7.117, "boundary_entropy": 0.24906, "router_noise": 0.0, "grad_norm": 0.445, "lr": 0.000246, "router_trainable": true, "sec_per_step": 0.187, "n_target_mean": 8.6273, "alpha": 0.03, "t": 771.4} | |
| {"event": "train", "step": 3700, "phase": 1, "objective": "parity_a", "lm_loss": 1.19695, "bpb": 1.72684, "chunk_loss": 0.02995, "F": 0.12209, "G": 0.20454, "chunks": 125.02, "bytes_per_chunk": 8.191, "boundary_entropy": 0.29685, "router_noise": 0.0, "grad_norm": 0.4455, "lr": 0.0002427, "router_trainable": true, "sec_per_step": 0.177, "n_target_mean": 9.8774, "alpha": 0.03, "t": 789.0} | |
| {"event": "train", "step": 3800, "phase": 1, "objective": "parity_a", "lm_loss": 1.34556, "bpb": 1.94123, "chunk_loss": 0.030084, "F": 0.14299, "G": 0.21753, "chunks": 146.42, "bytes_per_chunk": 6.993, "boundary_entropy": 0.28356, "router_noise": 0.0, "grad_norm": 0.4491, "lr": 0.0002394, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 8.9776, "alpha": 0.03, "t": 807.7} | |
| {"event": "train", "step": 3900, "phase": 1, "objective": "parity_a", "lm_loss": 1.19659, "bpb": 1.72632, "chunk_loss": 0.030251, "F": 0.13849, "G": 0.19887, "chunks": 141.81, "bytes_per_chunk": 7.221, "boundary_entropy": 0.25817, "router_noise": 0.0, "grad_norm": 0.398, "lr": 0.000236, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 9.1074, "alpha": 0.03, "t": 824.8} | |
| {"event": "probe", "step": 4000, "phase": 1, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3100.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.17813, "cps": 24.21875, "mean_p": 0.196593, "boundary_entropy": 0.152592, "bpb": 1.903259}, "de": {"chunks": 3107.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151317, "cps": 24.273438, "mean_p": 0.177651, "boundary_entropy": 0.17943, "bpb": 2.092656}, "vi": {"chunks": 2934.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.159154, "cps": 22.921875, "mean_p": 0.177661, "boundary_entropy": 0.252828, "bpb": 1.788013}, "tr": {"chunks": 2997.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.165553, "cps": 23.414062, "mean_p": 0.187302, "boundary_entropy": 0.215315, "bpb": 2.053317}, "fi": {"chunks": 2967.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157727, "cps": 23.179688, "mean_p": 0.185318, "boundary_entropy": 0.195301, "bpb": 2.194177}, "ru": {"chunks": 3029.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.159842, "cps": 23.664062, "mean_p": 0.243381, "boundary_entropy": 0.459654, "bpb": 1.285555}, "el": {"chunks": 3054.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.148809, "cps": 23.859375, "mean_p": 0.176859, "boundary_entropy": 0.346014, "bpb": 1.302451}, "hi": {"chunks": 2925.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.173848, "cps": 22.851562, "mean_p": 0.150353, "boundary_entropy": 0.323337, "bpb": 1.07021}, "ta": {"chunks": 2866.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.153303, "cps": 22.390625, "mean_p": 0.236201, "boundary_entropy": 0.479946, "bpb": 0.911851}, "th": {"chunks": 3215.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.199876, "cps": 25.117188, "mean_p": 0.18434, "boundary_entropy": 0.380946, "bpb": 1.032818}, "zh": {"chunks": 2973.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.503386, "cps": 23.226562, "mean_p": 0.244862, "boundary_entropy": 0.297276, "bpb": 2.885964}, "my": {"chunks": 1890.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.099747, "cps": 14.765625, "mean_p": 0.307348, "boundary_entropy": 0.577291, "bpb": 1.035514}}, "bpb": {"en": 1.90326, "de": 2.09266, "vi": 1.78801, "tr": 2.05332, "fi": 2.19418, "ru": 1.28556, "el": 1.30245, "hi": 1.07021, "ta": 0.91185, "th": 1.03282, "zh": 2.88596, "my": 1.03551}, "gini_cpc": 0.19306, "gini_cps": 0.044639, "bpic": 10.947343, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8494763111466515, "vi": 0.8934688049555107, "tr": 0.9293913323936684, "fi": 0.8854582510919314, "ru": 0.8973306153715209, "el": 0.8353925807866234, "hi": 0.9759627091022384, "ta": 0.8606233855869692, "th": 1.122076167938472, "zh": 2.8259462219940357, "my": 0.5599649635334737}, "t": 846.8} | |
| {"event": "eval", "step": 4000, "phase": 1, "bpb": {"en": 1.90854, "de": 2.087, "vi": 1.79643, "tr": 2.04616, "fi": 2.21292, "ru": 1.28603, "el": 1.30076, "hi": 1.07703, "ta": 0.91248, "th": 1.03168, "zh": 2.89324, "my": 1.04352}, "bytes_per_chunk": {"en": 5.406, "de": 6.467, "vi": 7.974, "tr": 6.432, "fi": 6.078, "ru": 11.537, "el": 11.922, "hi": 14.627, "ta": 17.672, "th": 14.11, "zh": 5.52, "my": 26.811}, "bpb_mean": 1.63298, "t": 851.8} | |
| {"event": "train", "step": 4000, "phase": 1, "objective": "parity_a", "lm_loss": 1.21067, "bpb": 1.74663, "chunk_loss": 0.030304, "F": 0.14119, "G": 0.21315, "chunks": 144.58, "bytes_per_chunk": 7.083, "boundary_entropy": 0.27228, "router_noise": 0.0, "grad_norm": 0.4164, "lr": 0.0002325, "router_trainable": true, "sec_per_step": 0.272, "n_target_mean": 9.2596, "alpha": 0.03, "t": 852.0} | |
| {"event": "train", "step": 4100, "phase": 1, "objective": "parity_a", "lm_loss": 1.27335, "bpb": 1.83706, "chunk_loss": 0.029964, "F": 0.1385, "G": 0.2033, "chunks": 141.83, "bytes_per_chunk": 7.22, "boundary_entropy": 0.26509, "router_noise": 0.0, "grad_norm": 0.4252, "lr": 0.000229, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 8.9573, "alpha": 0.03, "t": 868.9} | |
| {"event": "train", "step": 4200, "phase": 1, "objective": "parity_a", "lm_loss": 1.09735, "bpb": 1.58315, "chunk_loss": 0.030119, "F": 0.12064, "G": 0.20606, "chunks": 123.53, "bytes_per_chunk": 8.289, "boundary_entropy": 0.29464, "router_noise": 0.0, "grad_norm": 0.4252, "lr": 0.0002254, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 10.9544, "alpha": 0.03, "t": 886.3} | |
| {"event": "train", "step": 4300, "phase": 1, "objective": "parity_a", "lm_loss": 1.24898, "bpb": 1.8019, "chunk_loss": 0.030294, "F": 0.14655, "G": 0.20452, "chunks": 150.06, "bytes_per_chunk": 6.824, "boundary_entropy": 0.24796, "router_noise": 0.0, "grad_norm": 0.3908, "lr": 0.0002217, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 8.6079, "alpha": 0.03, "t": 904.4} | |
| {"event": "phase_transition", "step": 4400, "phase": 2, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 921.2} | |
| {"event": "train", "step": 4400, "phase": 2, "objective": "parity_a", "lm_loss": 1.2651, "bpb": 1.82515, "chunk_loss": 0.030079, "F": 0.14131, "G": 0.20033, "chunks": 144.7, "bytes_per_chunk": 7.077, "boundary_entropy": 0.24532, "router_noise": 0.0, "grad_norm": 0.4312, "lr": 0.0002181, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.7204, "alpha": 0.03, "t": 921.3} | |
| {"event": "train", "step": 4500, "phase": 2, "objective": "parity_a", "lm_loss": 1.2503, "bpb": 1.8038, "chunk_loss": 0.030257, "F": 0.14395, "G": 0.1992, "chunks": 147.41, "bytes_per_chunk": 6.947, "boundary_entropy": 0.23835, "router_noise": 0.0, "grad_norm": 0.4371, "lr": 0.0002143, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 8.6056, "alpha": 0.03, "t": 939.4} | |
| {"event": "train", "step": 4600, "phase": 2, "objective": "parity_a", "lm_loss": 1.19228, "bpb": 1.7201, "chunk_loss": 0.030234, "F": 0.1295, "G": 0.20244, "chunks": 132.61, "bytes_per_chunk": 7.722, "boundary_entropy": 0.27439, "router_noise": 0.0, "grad_norm": 0.4147, "lr": 0.0002105, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.6816, "alpha": 0.03, "t": 956.4} | |
| {"event": "train", "step": 4700, "phase": 2, "objective": "parity_a", "lm_loss": 1.21058, "bpb": 1.7465, "chunk_loss": 0.029999, "F": 0.12659, "G": 0.19581, "chunks": 129.62, "bytes_per_chunk": 7.9, "boundary_entropy": 0.27158, "router_noise": 0.0, "grad_norm": 0.4201, "lr": 0.0002067, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 9.4602, "alpha": 0.03, "t": 974.5} | |
| {"event": "train", "step": 4800, "phase": 2, "objective": "parity_a", "lm_loss": 1.2181, "bpb": 1.75735, "chunk_loss": 0.03006, "F": 0.13091, "G": 0.19476, "chunks": 134.05, "bytes_per_chunk": 7.639, "boundary_entropy": 0.25575, "router_noise": 0.0, "grad_norm": 0.4154, "lr": 0.0002029, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.0803, "alpha": 0.03, "t": 991.4} | |
| {"event": "train", "step": 4900, "phase": 2, "objective": "parity_a", "lm_loss": 1.18639, "bpb": 1.7116, "chunk_loss": 0.030143, "F": 0.14053, "G": 0.19147, "chunks": 143.91, "bytes_per_chunk": 7.116, "boundary_entropy": 0.22941, "router_noise": 0.0, "grad_norm": 0.4119, "lr": 0.000199, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 8.5379, "alpha": 0.03, "t": 1009.5} | |
| {"event": "probe", "step": 5000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3068.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.176291, "cps": 23.96875, "mean_p": 0.190157, "boundary_entropy": 0.129851, "bpb": 1.816241}, "de": {"chunks": 3088.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.150392, "cps": 24.125, "mean_p": 0.170599, "boundary_entropy": 0.153644, "bpb": 1.983447}, "vi": {"chunks": 2706.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.146786, "cps": 21.140625, "mean_p": 0.159495, "boundary_entropy": 0.226295, "bpb": 1.703681}, "tr": {"chunks": 2823.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.155941, "cps": 22.054688, "mean_p": 0.173137, "boundary_entropy": 0.189806, "bpb": 1.923912}, "fi": {"chunks": 2931.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.155813, "cps": 22.898438, "mean_p": 0.174094, "boundary_entropy": 0.162617, "bpb": 2.094673}, "ru": {"chunks": 2987.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.157625, "cps": 23.335938, "mean_p": 0.216352, "boundary_entropy": 0.415107, "bpb": 1.239447}, "el": {"chunks": 2974.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.144911, "cps": 23.234375, "mean_p": 0.160047, "boundary_entropy": 0.316835, "bpb": 1.247663}, "hi": {"chunks": 2623.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.155899, "cps": 20.492188, "mean_p": 0.153312, "boundary_entropy": 0.332778, "bpb": 1.036811}, "ta": {"chunks": 2189.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.11709, "cps": 17.101562, "mean_p": 0.193844, "boundary_entropy": 0.42524, "bpb": 0.878133}, "th": {"chunks": 2259.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.140441, "cps": 17.648438, "mean_p": 0.141077, "boundary_entropy": 0.315441, "bpb": 0.99453}, "zh": {"chunks": 2992.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.506603, "cps": 23.375, "mean_p": 0.228242, "boundary_entropy": 0.249656, "bpb": 2.802323}, "my": {"chunks": 1793.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.094627, "cps": 14.007812, "mean_p": 0.282442, "boundary_entropy": 0.549793, "bpb": 1.002722}}, "bpb": {"en": 1.81624, "de": 1.98345, "vi": 1.70368, "tr": 1.92391, "fi": 2.09467, "ru": 1.23945, "el": 1.24766, "hi": 1.03681, "ta": 0.87813, "th": 0.99453, "zh": 2.80232, "my": 1.00272}, "gini_cpc": 0.208995, "gini_cps": 0.077131, "bpic": 11.83304, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8530876394414791, "vi": 0.8326326085273004, "tr": 0.8845636561351347, "fi": 0.8838380612691804, "ru": 0.8941178666152951, "el": 0.8219944475435422, "hi": 0.8843251625851672, "ta": 0.6641849904439376, "th": 0.7966433415375214, "zh": 2.8736701147177133, "my": 0.5367668399536617}, "t": 1029.1} | |
| {"event": "eval", "step": 5000, "phase": 2, "bpb": {"en": 1.81899, "de": 1.98405, "vi": 1.71501, "tr": 1.91657, "fi": 2.10959, "ru": 1.23937, "el": 1.24577, "hi": 1.04334, "ta": 0.88105, "th": 0.99333, "zh": 2.81164, "my": 1.01177}, "bytes_per_chunk": {"en": 5.449, "de": 6.498, "vi": 8.615, "tr": 6.748, "fi": 6.078, "ru": 11.784, "el": 12.157, "hi": 16.188, "ta": 21.959, "th": 20.016, "zh": 5.459, "my": 28.645}, "bpb_mean": 1.56421, "t": 1034.1} | |
| {"event": "train", "step": 5000, "phase": 2, "objective": "parity_a", "lm_loss": 1.14127, "bpb": 1.64651, "chunk_loss": 0.029971, "F": 0.12802, "G": 0.19008, "chunks": 131.09, "bytes_per_chunk": 7.811, "boundary_entropy": 0.25149, "router_noise": 0.0, "grad_norm": 0.4221, "lr": 0.000195, "router_trainable": true, "sec_per_step": 0.248, "n_target_mean": 9.5372, "alpha": 0.03, "t": 1034.3} | |
| {"event": "train", "step": 5100, "phase": 2, "objective": "parity_a", "lm_loss": 1.13126, "bpb": 1.63206, "chunk_loss": 0.030037, "F": 0.12675, "G": 0.19668, "chunks": 129.8, "bytes_per_chunk": 7.889, "boundary_entropy": 0.26858, "router_noise": 0.0, "grad_norm": 0.4066, "lr": 0.0001911, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 10.2808, "alpha": 0.03, "t": 1051.8} | |
| {"event": "train", "step": 5200, "phase": 2, "objective": "parity_a", "lm_loss": 1.25098, "bpb": 1.80478, "chunk_loss": 0.030235, "F": 0.14204, "G": 0.19909, "chunks": 145.45, "bytes_per_chunk": 7.04, "boundary_entropy": 0.23959, "router_noise": 0.0, "grad_norm": 0.4166, "lr": 0.0001871, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 8.7255, "alpha": 0.03, "t": 1069.3} | |
| {"event": "train", "step": 5300, "phase": 2, "objective": "parity_a", "lm_loss": 1.15884, "bpb": 1.67186, "chunk_loss": 0.030012, "F": 0.13426, "G": 0.18937, "chunks": 137.48, "bytes_per_chunk": 7.448, "boundary_entropy": 0.23336, "router_noise": 0.0, "grad_norm": 0.3985, "lr": 0.0001831, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 9.0671, "alpha": 0.03, "t": 1086.5} | |
| {"event": "train", "step": 5400, "phase": 2, "objective": "parity_a", "lm_loss": 1.09707, "bpb": 1.58274, "chunk_loss": 0.029954, "F": 0.12701, "G": 0.18971, "chunks": 130.06, "bytes_per_chunk": 7.873, "boundary_entropy": 0.24616, "router_noise": 0.0, "grad_norm": 0.4044, "lr": 0.0001791, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 9.3868, "alpha": 0.03, "t": 1104.7} | |
| {"event": "train", "step": 5500, "phase": 2, "objective": "parity_a", "lm_loss": 1.11567, "bpb": 1.60956, "chunk_loss": 0.030132, "F": 0.13193, "G": 0.1943, "chunks": 135.09, "bytes_per_chunk": 7.58, "boundary_entropy": 0.2463, "router_noise": 0.0, "grad_norm": 0.3871, "lr": 0.0001751, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.5342, "alpha": 0.03, "t": 1121.9} | |
| {"event": "train", "step": 5600, "phase": 2, "objective": "parity_a", "lm_loss": 1.13784, "bpb": 1.64156, "chunk_loss": 0.030099, "F": 0.13429, "G": 0.18387, "chunks": 137.52, "bytes_per_chunk": 7.446, "boundary_entropy": 0.21942, "router_noise": 0.0, "grad_norm": 0.4176, "lr": 0.0001711, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 9.2622, "alpha": 0.03, "t": 1140.0} | |
| {"event": "train", "step": 5700, "phase": 2, "objective": "parity_a", "lm_loss": 1.23281, "bpb": 1.77856, "chunk_loss": 0.030145, "F": 0.14647, "G": 0.19463, "chunks": 149.98, "bytes_per_chunk": 6.827, "boundary_entropy": 0.21806, "router_noise": 0.0, "grad_norm": 0.4102, "lr": 0.000167, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.2312, "alpha": 0.03, "t": 1157.0} | |
| {"event": "train", "step": 5800, "phase": 2, "objective": "parity_a", "lm_loss": 1.23194, "bpb": 1.77731, "chunk_loss": 0.030193, "F": 0.15263, "G": 0.19458, "chunks": 156.3, "bytes_per_chunk": 6.552, "boundary_entropy": 0.21034, "router_noise": 0.0, "grad_norm": 0.4272, "lr": 0.000163, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.0077, "alpha": 0.03, "t": 1174.1} | |
| {"event": "train", "step": 5900, "phase": 2, "objective": "parity_a", "lm_loss": 1.16723, "bpb": 1.68396, "chunk_loss": 0.030253, "F": 0.14507, "G": 0.19125, "chunks": 148.55, "bytes_per_chunk": 6.893, "boundary_entropy": 0.21154, "router_noise": 0.0, "grad_norm": 0.4158, "lr": 0.0001589, "router_trainable": true, "sec_per_step": 0.182, "n_target_mean": 8.8393, "alpha": 0.03, "t": 1192.4} | |
| {"event": "probe", "step": 6000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3083.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.177153, "cps": 24.085938, "mean_p": 0.190233, "boundary_entropy": 0.123128, "bpb": 1.764466}, "de": {"chunks": 3180.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154873, "cps": 24.84375, "mean_p": 0.173765, "boundary_entropy": 0.146126, "bpb": 1.916874}, "vi": {"chunks": 2893.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.15693, "cps": 22.601562, "mean_p": 0.167365, "boundary_entropy": 0.222769, "bpb": 1.642539}, "tr": {"chunks": 3027.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16721, "cps": 23.648438, "mean_p": 0.182384, "boundary_entropy": 0.187985, "bpb": 1.847895}, "fi": {"chunks": 2961.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.157408, "cps": 23.132812, "mean_p": 0.174873, "boundary_entropy": 0.156462, "bpb": 2.017591}, "ru": {"chunks": 2938.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.15504, "cps": 22.953125, "mean_p": 0.197654, "boundary_entropy": 0.382498, "bpb": 1.206789}, "el": {"chunks": 3157.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.153827, "cps": 24.664062, "mean_p": 0.162745, "boundary_entropy": 0.310135, "bpb": 1.215171}, "hi": {"chunks": 2749.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.163388, "cps": 21.476562, "mean_p": 0.137182, "boundary_entropy": 0.296437, "bpb": 1.010722}, "ta": {"chunks": 2647.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.141589, "cps": 20.679688, "mean_p": 0.187247, "boundary_entropy": 0.404822, "bpb": 0.859924}, "th": {"chunks": 2562.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.159279, "cps": 20.015625, "mean_p": 0.144328, "boundary_entropy": 0.313263, "bpb": 0.969464}, "zh": {"chunks": 3057.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.517609, "cps": 23.882812, "mean_p": 0.238618, "boundary_entropy": 0.254768, "bpb": 2.752055}, "my": {"chunks": 2489.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.13136, "cps": 19.445312, "mean_p": 0.286364, "boundary_entropy": 0.544396, "bpb": 0.96628}}, "bpb": {"en": 1.76447, "de": 1.91687, "vi": 1.64254, "tr": 1.84789, "fi": 2.01759, "ru": 1.20679, "el": 1.21517, "hi": 1.01072, "ta": 0.85992, "th": 0.96946, "zh": 2.75206, "my": 0.96628}, "gini_cpc": 0.175872, "gini_cps": 0.043109, "bpic": 11.046283, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.87422920018358, "vi": 0.885841224363006, "tr": 0.943870613589313, "fi": 0.8885402780786454, "ru": 0.8751715125160789, "el": 0.8683290531896771, "hi": 0.9222958668516753, "ta": 0.7992434158904177, "th": 0.8991013621581989, "zh": 2.9218141740330372, "my": 0.7415016401821097}, "t": 1212.0} | |
| {"event": "eval", "step": 6000, "phase": 2, "bpb": {"en": 1.76621, "de": 1.92005, "vi": 1.65615, "tr": 1.8406, "fi": 2.03218, "ru": 1.20527, "el": 1.21473, "hi": 1.01707, "ta": 0.86264, "th": 0.96792, "zh": 2.75983, "my": 0.97493}, "bytes_per_chunk": {"en": 5.454, "de": 6.357, "vi": 8.04, "tr": 6.389, "fi": 6.068, "ru": 11.87, "el": 11.518, "hi": 15.756, "ta": 18.981, "th": 17.034, "zh": 5.388, "my": 21.553}, "bpb_mean": 1.51813, "t": 1217.1} | |
| {"event": "train", "step": 6000, "phase": 2, "objective": "parity_a", "lm_loss": 1.19502, "bpb": 1.72405, "chunk_loss": 0.03002, "F": 0.1344, "G": 0.19584, "chunks": 137.62, "bytes_per_chunk": 7.441, "boundary_entropy": 0.23943, "router_noise": 0.0, "grad_norm": 0.4093, "lr": 0.0001549, "router_trainable": true, "sec_per_step": 0.249, "n_target_mean": 9.5562, "alpha": 0.03, "t": 1217.2} | |
| {"event": "train", "step": 6100, "phase": 2, "objective": "parity_a", "lm_loss": 1.14234, "bpb": 1.64805, "chunk_loss": 0.030114, "F": 0.13225, "G": 0.1849, "chunks": 135.42, "bytes_per_chunk": 7.562, "boundary_entropy": 0.22721, "router_noise": 0.0, "grad_norm": 0.4439, "lr": 0.0001509, "router_trainable": true, "sec_per_step": 0.181, "n_target_mean": 9.1355, "alpha": 0.03, "t": 1235.3} | |
| {"event": "train", "step": 6200, "phase": 2, "objective": "parity_a", "lm_loss": 1.08319, "bpb": 1.56272, "chunk_loss": 0.030183, "F": 0.1295, "G": 0.1953, "chunks": 132.61, "bytes_per_chunk": 7.722, "boundary_entropy": 0.25461, "router_noise": 0.0, "grad_norm": 0.4077, "lr": 0.0001469, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.8141, "alpha": 0.03, "t": 1252.3} | |
| {"event": "train", "step": 6300, "phase": 2, "objective": "parity_a", "lm_loss": 1.15786, "bpb": 1.67043, "chunk_loss": 0.030543, "F": 0.14188, "G": 0.19571, "chunks": 145.28, "bytes_per_chunk": 7.048, "boundary_entropy": 0.23379, "router_noise": 0.0, "grad_norm": 0.4271, "lr": 0.0001429, "router_trainable": true, "sec_per_step": 0.172, "n_target_mean": 9.3066, "alpha": 0.03, "t": 1269.5} | |
| {"event": "train", "step": 6400, "phase": 2, "objective": "parity_a", "lm_loss": 1.11177, "bpb": 1.60394, "chunk_loss": 0.030023, "F": 0.12744, "G": 0.18409, "chunks": 130.5, "bytes_per_chunk": 7.847, "boundary_entropy": 0.23252, "router_noise": 0.0, "grad_norm": 0.3946, "lr": 0.0001389, "router_trainable": true, "sec_per_step": 0.184, "n_target_mean": 9.6866, "alpha": 0.03, "t": 1287.9} | |
| {"event": "train", "step": 6500, "phase": 2, "objective": "parity_a", "lm_loss": 1.09712, "bpb": 1.58281, "chunk_loss": 0.03008, "F": 0.13431, "G": 0.18448, "chunks": 137.53, "bytes_per_chunk": 7.446, "boundary_entropy": 0.21882, "router_noise": 0.0, "grad_norm": 0.4121, "lr": 0.000135, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.3935, "alpha": 0.03, "t": 1304.8} | |
| {"event": "train", "step": 6600, "phase": 2, "objective": "parity_a", "lm_loss": 1.14488, "bpb": 1.65172, "chunk_loss": 0.030062, "F": 0.13853, "G": 0.19143, "chunks": 141.86, "bytes_per_chunk": 7.218, "boundary_entropy": 0.2268, "router_noise": 0.0, "grad_norm": 0.389, "lr": 0.000131, "router_trainable": true, "sec_per_step": 0.185, "n_target_mean": 8.955, "alpha": 0.03, "t": 1323.3} | |
| {"event": "train", "step": 6700, "phase": 2, "objective": "parity_a", "lm_loss": 1.20755, "bpb": 1.74213, "chunk_loss": 0.03064, "F": 0.15649, "G": 0.19252, "chunks": 160.25, "bytes_per_chunk": 6.39, "boundary_entropy": 0.18736, "router_noise": 0.0, "grad_norm": 0.4345, "lr": 0.0001271, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.2156, "alpha": 0.03, "t": 1340.4} | |
| {"event": "train", "step": 6800, "phase": 2, "objective": "parity_a", "lm_loss": 1.08025, "bpb": 1.55848, "chunk_loss": 0.030009, "F": 0.1333, "G": 0.17661, "chunks": 136.5, "bytes_per_chunk": 7.502, "boundary_entropy": 0.19979, "router_noise": 0.0, "grad_norm": 0.3887, "lr": 0.0001233, "router_trainable": true, "sec_per_step": 0.183, "n_target_mean": 8.8796, "alpha": 0.03, "t": 1358.7} | |
| {"event": "train", "step": 6900, "phase": 2, "objective": "parity_a", "lm_loss": 1.14399, "bpb": 1.65043, "chunk_loss": 0.03013, "F": 0.13525, "G": 0.19158, "chunks": 138.5, "bytes_per_chunk": 7.394, "boundary_entropy": 0.22941, "router_noise": 0.0, "grad_norm": 0.4196, "lr": 0.0001195, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.5822, "alpha": 0.03, "t": 1375.8} | |
| {"event": "probe", "step": 7000, "phase": 2, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3034.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.174338, "cps": 23.703125, "mean_p": 0.184189, "boundary_entropy": 0.105018, "bpb": 1.730127}, "de": {"chunks": 3107.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.151317, "cps": 24.273438, "mean_p": 0.167086, "boundary_entropy": 0.130361, "bpb": 1.877249}, "vi": {"chunks": 2899.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.157255, "cps": 22.648438, "mean_p": 0.164127, "boundary_entropy": 0.211913, "bpb": 1.59003}, "tr": {"chunks": 2973.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.164227, "cps": 23.226562, "mean_p": 0.17658, "boundary_entropy": 0.169024, "bpb": 1.805483}, "fi": {"chunks": 3021.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.160598, "cps": 23.601562, "mean_p": 0.173465, "boundary_entropy": 0.140686, "bpb": 1.963757}, "ru": {"chunks": 3255.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.171768, "cps": 25.429688, "mean_p": 0.200836, "boundary_entropy": 0.381954, "bpb": 1.182728}, "el": {"chunks": 2782.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.135555, "cps": 21.734375, "mean_p": 0.150821, "boundary_entropy": 0.295622, "bpb": 1.182279}, "hi": {"chunks": 3039.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.180624, "cps": 23.742188, "mean_p": 0.137077, "boundary_entropy": 0.292639, "bpb": 0.980309}, "ta": {"chunks": 2571.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.137523, "cps": 20.085938, "mean_p": 0.188092, "boundary_entropy": 0.409199, "bpb": 0.841485}, "th": {"chunks": 2745.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170656, "cps": 21.445312, "mean_p": 0.150498, "boundary_entropy": 0.320939, "bpb": 0.947321}, "zh": {"chunks": 2983.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.50508, "cps": 23.304688, "mean_p": 0.226134, "boundary_entropy": 0.227208, "bpb": 2.70525}, "my": {"chunks": 2784.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.146928, "cps": 21.75, "mean_p": 0.271429, "boundary_entropy": 0.517649, "bpb": 0.936342}}, "bpb": {"en": 1.73013, "de": 1.87725, "vi": 1.59003, "tr": 1.80548, "fi": 1.96376, "ru": 1.18273, "el": 1.18228, "hi": 0.98031, "ta": 0.84148, "th": 0.94732, "zh": 2.70525, "my": 0.93634}, "gini_cpc": 0.176969, "gini_cps": 0.033641, "bpic": 10.905038, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8679553607628937, "vi": 0.9020147039814029, "tr": 0.9420043353328902, "fi": 0.9211861199741928, "ru": 0.985258799567957, "el": 0.7775437311020826, "hi": 1.0360582443059883, "ta": 0.7888331458941835, "th": 0.9788808384265445, "zh": 2.897132476029092, "my": 0.8427803450278311}, "t": 1396.8} | |
| {"event": "eval", "step": 7000, "phase": 2, "bpb": {"en": 1.73087, "de": 1.88192, "vi": 1.60391, "tr": 1.79735, "fi": 1.97831, "ru": 1.18131, "el": 1.18131, "hi": 0.98701, "ta": 0.84612, "th": 0.94513, "zh": 2.71628, "my": 0.94607}, "bytes_per_chunk": {"en": 5.554, "de": 6.455, "vi": 8.188, "tr": 6.518, "fi": 6.043, "ru": 10.56, "el": 12.865, "hi": 13.972, "ta": 19.103, "th": 16.07, "zh": 5.533, "my": 18.999}, "bpb_mean": 1.48297, "t": 1402.0} | |
| {"event": "train", "step": 7000, "phase": 2, "objective": "parity_a", "lm_loss": 1.15599, "bpb": 1.66774, "chunk_loss": 0.030206, "F": 0.13625, "G": 0.18585, "chunks": 139.52, "bytes_per_chunk": 7.34, "boundary_entropy": 0.21867, "router_noise": 0.0, "grad_norm": 0.4356, "lr": 0.0001157, "router_trainable": true, "sec_per_step": 0.264, "n_target_mean": 9.0446, "alpha": 0.03, "t": 1402.2} | |
| {"event": "train", "step": 7100, "phase": 2, "objective": "parity_a", "lm_loss": 1.05288, "bpb": 1.51898, "chunk_loss": 0.030156, "F": 0.12924, "G": 0.18745, "chunks": 132.34, "bytes_per_chunk": 7.737, "boundary_entropy": 0.23158, "router_noise": 0.0, "grad_norm": 0.399, "lr": 0.0001119, "router_trainable": true, "sec_per_step": 0.182, "n_target_mean": 10.0028, "alpha": 0.03, "t": 1420.4} | |
| {"event": "train", "step": 7200, "phase": 2, "objective": "parity_a", "lm_loss": 1.09963, "bpb": 1.58643, "chunk_loss": 0.029978, "F": 0.13211, "G": 0.17934, "chunks": 135.28, "bytes_per_chunk": 7.569, "boundary_entropy": 0.20726, "router_noise": 0.0, "grad_norm": 0.4205, "lr": 0.0001083, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.1147, "alpha": 0.03, "t": 1437.3} | |
| {"event": "train", "step": 7300, "phase": 2, "objective": "parity_a", "lm_loss": 1.14301, "bpb": 1.64902, "chunk_loss": 0.030123, "F": 0.14084, "G": 0.17682, "chunks": 144.22, "bytes_per_chunk": 7.1, "boundary_entropy": 0.18316, "router_noise": 0.0, "grad_norm": 0.4178, "lr": 0.0001046, "router_trainable": true, "sec_per_step": 0.183, "n_target_mean": 8.6182, "alpha": 0.03, "t": 1455.6} | |
| {"event": "train", "step": 7400, "phase": 2, "objective": "parity_a", "lm_loss": 1.09555, "bpb": 1.58055, "chunk_loss": 0.030205, "F": 0.13095, "G": 0.17724, "chunks": 134.09, "bytes_per_chunk": 7.636, "boundary_entropy": 0.20729, "router_noise": 0.0, "grad_norm": 0.4112, "lr": 0.000101, "router_trainable": true, "sec_per_step": 0.184, "n_target_mean": 9.7931, "alpha": 0.03, "t": 1474.0} | |
| {"event": "train", "step": 7500, "phase": 2, "objective": "parity_a", "lm_loss": 1.14917, "bpb": 1.65791, "chunk_loss": 0.030194, "F": 0.14355, "G": 0.18601, "chunks": 147.0, "bytes_per_chunk": 6.966, "boundary_entropy": 0.20313, "router_noise": 0.0, "grad_norm": 0.4126, "lr": 9.75e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 8.8162, "alpha": 0.03, "t": 1490.9} | |
| {"event": "train", "step": 7600, "phase": 2, "objective": "parity_a", "lm_loss": 1.09623, "bpb": 1.58152, "chunk_loss": 0.030273, "F": 0.13086, "G": 0.18583, "chunks": 134.0, "bytes_per_chunk": 7.642, "boundary_entropy": 0.2275, "router_noise": 0.0, "grad_norm": 0.4199, "lr": 9.4e-05, "router_trainable": true, "sec_per_step": 0.184, "n_target_mean": 9.4461, "alpha": 0.03, "t": 1509.3} | |
| {"event": "phase_transition", "step": 7700, "phase": 3, "objective": "parity_a", "router_trainable": true, "encoder_trainable": true, "frozen_router_params": 0, "t": 1527.8} | |
| {"event": "train", "step": 7700, "phase": 3, "objective": "parity_a", "lm_loss": 1.1028, "bpb": 1.59101, "chunk_loss": 0.029995, "F": 0.12978, "G": 0.17892, "chunks": 132.89, "bytes_per_chunk": 7.706, "boundary_entropy": 0.20677, "router_noise": 0.0, "grad_norm": 0.4117, "lr": 9.06e-05, "router_trainable": true, "sec_per_step": 0.187, "n_target_mean": 9.5412, "alpha": 0.03, "t": 1528.0} | |
| {"event": "train", "step": 7800, "phase": 3, "objective": "parity_a", "lm_loss": 1.12606, "bpb": 1.62457, "chunk_loss": 0.030053, "F": 0.14137, "G": 0.18285, "chunks": 144.77, "bytes_per_chunk": 7.074, "boundary_entropy": 0.18796, "router_noise": 0.0, "grad_norm": 0.4066, "lr": 8.73e-05, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 8.4053, "alpha": 0.03, "t": 1544.8} | |
| {"event": "train", "step": 7900, "phase": 3, "objective": "parity_a", "lm_loss": 1.03819, "bpb": 1.4978, "chunk_loss": 0.030039, "F": 0.12154, "G": 0.17388, "chunks": 124.45, "bytes_per_chunk": 8.228, "boundary_entropy": 0.21819, "router_noise": 0.0, "grad_norm": 0.4011, "lr": 8.4e-05, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 10.2831, "alpha": 0.03, "t": 1561.6} | |
| {"event": "probe", "step": 8000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3057.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.175659, "cps": 23.882812, "mean_p": 0.184797, "boundary_entropy": 0.095518, "bpb": 1.701603}, "de": {"chunks": 3231.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.157356, "cps": 25.242188, "mean_p": 0.171403, "boundary_entropy": 0.126438, "bpb": 1.841142}, "vi": {"chunks": 2929.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.158883, "cps": 22.882812, "mean_p": 0.16462, "boundary_entropy": 0.201995, "bpb": 1.557675}, "tr": {"chunks": 2922.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.16141, "cps": 22.828125, "mean_p": 0.175099, "boundary_entropy": 0.162405, "bpb": 1.768056}, "fi": {"chunks": 3052.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162245, "cps": 23.84375, "mean_p": 0.174074, "boundary_entropy": 0.133547, "bpb": 1.925071}, "ru": {"chunks": 2906.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.153351, "cps": 22.703125, "mean_p": 0.172758, "boundary_entropy": 0.3325, "bpb": 1.161012}, "el": {"chunks": 2953.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.143887, "cps": 23.070312, "mean_p": 0.139224, "boundary_entropy": 0.267445, "bpb": 1.159757}, "hi": {"chunks": 2800.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.166419, "cps": 21.875, "mean_p": 0.125491, "boundary_entropy": 0.270301, "bpb": 0.961258}, "ta": {"chunks": 2501.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.133779, "cps": 19.539062, "mean_p": 0.178432, "boundary_entropy": 0.389056, "bpb": 0.82441}, "th": {"chunks": 2742.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.170469, "cps": 21.421875, "mean_p": 0.143062, "boundary_entropy": 0.305745, "bpb": 0.928917}, "zh": {"chunks": 3006.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.508974, "cps": 23.484375, "mean_p": 0.221686, "boundary_entropy": 0.205973, "bpb": 2.670151}, "my": {"chunks": 2922.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.154212, "cps": 22.828125, "mean_p": 0.270005, "boundary_entropy": 0.512305, "bpb": 0.91646}}, "bpb": {"en": 1.7016, "de": 1.84114, "vi": 1.55768, "tr": 1.76806, "fi": 1.92507, "ru": 1.16101, "el": 1.15976, "hi": 0.96126, "ta": 0.82441, "th": 0.92892, "zh": 2.67015, "my": 0.91646}, "gini_cpc": 0.171038, "gini_cps": 0.031479, "bpic": 10.958596, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8958044846738255, "vi": 0.9044923774032467, "tr": 0.9188790324604758, "fi": 0.9236370112929039, "ru": 0.8730016752934353, "el": 0.8191270790898476, "hi": 0.9473962031942863, "ta": 0.7615824081683111, "th": 0.9704542508014496, "zh": 2.8975051402293106, "my": 0.877900945990711}, "t": 1582.7} | |
| {"event": "eval", "step": 8000, "phase": 3, "bpb": {"en": 1.70195, "de": 1.84579, "vi": 1.57261, "tr": 1.76001, "fi": 1.93892, "ru": 1.15985, "el": 1.15789, "hi": 0.96711, "ta": 0.82874, "th": 0.92673, "zh": 2.6818, "my": 0.92663}, "bytes_per_chunk": {"en": 5.551, "de": 6.34, "vi": 8.075, "tr": 6.66, "fi": 6.004, "ru": 11.872, "el": 12.411, "hi": 15.173, "ta": 19.417, "th": 16.464, "zh": 5.501, "my": 18.167}, "bpb_mean": 1.45567, "t": 1587.7} | |
| {"event": "train", "step": 8000, "phase": 3, "objective": "parity_a", "lm_loss": 1.05705, "bpb": 1.525, "chunk_loss": 0.030148, "F": 0.12978, "G": 0.1803, "chunks": 132.89, "bytes_per_chunk": 7.706, "boundary_entropy": 0.21506, "router_noise": 0.0, "grad_norm": 0.4167, "lr": 8.08e-05, "router_trainable": true, "sec_per_step": 0.263, "n_target_mean": 9.7647, "alpha": 0.03, "t": 1587.9} | |
| {"event": "train", "step": 8100, "phase": 3, "objective": "parity_a", "lm_loss": 1.12314, "bpb": 1.62035, "chunk_loss": 0.030083, "F": 0.13786, "G": 0.18523, "chunks": 141.17, "bytes_per_chunk": 7.254, "boundary_entropy": 0.20897, "router_noise": 0.0, "grad_norm": 0.4357, "lr": 7.77e-05, "router_trainable": true, "sec_per_step": 0.182, "n_target_mean": 9.0705, "alpha": 0.03, "t": 1606.1} | |
| {"event": "train", "step": 8200, "phase": 3, "objective": "parity_a", "lm_loss": 1.11516, "bpb": 1.60883, "chunk_loss": 0.030081, "F": 0.1429, "G": 0.17909, "chunks": 146.33, "bytes_per_chunk": 6.998, "boundary_entropy": 0.17978, "router_noise": 0.0, "grad_norm": 0.3987, "lr": 7.47e-05, "router_trainable": true, "sec_per_step": 0.183, "n_target_mean": 8.5059, "alpha": 0.03, "t": 1624.4} | |
| {"event": "train", "step": 8300, "phase": 3, "objective": "parity_a", "lm_loss": 1.07245, "bpb": 1.54722, "chunk_loss": 0.030153, "F": 0.12996, "G": 0.1772, "chunks": 133.08, "bytes_per_chunk": 7.695, "boundary_entropy": 0.20447, "router_noise": 0.0, "grad_norm": 0.4145, "lr": 7.17e-05, "router_trainable": true, "sec_per_step": 0.168, "n_target_mean": 9.785, "alpha": 0.03, "t": 1641.2} | |
| {"event": "train", "step": 8400, "phase": 3, "objective": "parity_a", "lm_loss": 1.21635, "bpb": 1.75483, "chunk_loss": 0.030064, "F": 0.14703, "G": 0.18504, "chunks": 150.56, "bytes_per_chunk": 6.801, "boundary_entropy": 0.18677, "router_noise": 0.0, "grad_norm": 0.44, "lr": 6.88e-05, "router_trainable": true, "sec_per_step": 0.175, "n_target_mean": 8.204, "alpha": 0.03, "t": 1658.7} | |
| {"event": "train", "step": 8500, "phase": 3, "objective": "parity_a", "lm_loss": 1.10204, "bpb": 1.58991, "chunk_loss": 0.030093, "F": 0.134, "G": 0.16767, "chunks": 137.22, "bytes_per_chunk": 7.463, "boundary_entropy": 0.17092, "router_noise": 0.0, "grad_norm": 0.4111, "lr": 6.6e-05, "router_trainable": true, "sec_per_step": 0.182, "n_target_mean": 9.0722, "alpha": 0.03, "t": 1676.9} | |
| {"event": "train", "step": 8600, "phase": 3, "objective": "parity_a", "lm_loss": 1.12214, "bpb": 1.61891, "chunk_loss": 0.030096, "F": 0.13881, "G": 0.18058, "chunks": 142.14, "bytes_per_chunk": 7.204, "boundary_entropy": 0.19251, "router_noise": 0.0, "grad_norm": 0.4318, "lr": 6.33e-05, "router_trainable": true, "sec_per_step": 0.191, "n_target_mean": 8.9134, "alpha": 0.03, "t": 1696.1} | |
| {"event": "train", "step": 8700, "phase": 3, "objective": "parity_a", "lm_loss": 1.07341, "bpb": 1.54861, "chunk_loss": 0.030101, "F": 0.13359, "G": 0.1756, "chunks": 136.8, "bytes_per_chunk": 7.486, "boundary_entropy": 0.18802, "router_noise": 0.0, "grad_norm": 0.4279, "lr": 6.07e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.3731, "alpha": 0.03, "t": 1712.9} | |
| {"event": "train", "step": 8800, "phase": 3, "objective": "parity_a", "lm_loss": 1.14578, "bpb": 1.65301, "chunk_loss": 0.030116, "F": 0.13824, "G": 0.17997, "chunks": 141.56, "bytes_per_chunk": 7.234, "boundary_entropy": 0.19102, "router_noise": 0.0, "grad_norm": 0.4274, "lr": 5.82e-05, "router_trainable": true, "sec_per_step": 0.189, "n_target_mean": 8.9276, "alpha": 0.03, "t": 1731.9} | |
| {"event": "train", "step": 8900, "phase": 3, "objective": "parity_a", "lm_loss": 1.09809, "bpb": 1.58421, "chunk_loss": 0.030159, "F": 0.12932, "G": 0.18337, "chunks": 132.42, "bytes_per_chunk": 7.733, "boundary_entropy": 0.22306, "router_noise": 0.0, "grad_norm": 0.4425, "lr": 5.58e-05, "router_trainable": true, "sec_per_step": 0.166, "n_target_mean": 10.0472, "alpha": 0.03, "t": 1748.5} | |
| {"event": "probe", "step": 9000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3040.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.174683, "cps": 23.75, "mean_p": 0.183038, "boundary_entropy": 0.089456, "bpb": 1.681546}, "de": {"chunks": 3133.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.152584, "cps": 24.476562, "mean_p": 0.16441, "boundary_entropy": 0.112636, "bpb": 1.820291}, "vi": {"chunks": 2924.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.158611, "cps": 22.84375, "mean_p": 0.159702, "boundary_entropy": 0.188016, "bpb": 1.526694}, "tr": {"chunks": 2905.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.160471, "cps": 22.695312, "mean_p": 0.171823, "boundary_entropy": 0.149783, "bpb": 1.742624}, "fi": {"chunks": 3052.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.162245, "cps": 23.84375, "mean_p": 0.17242, "boundary_entropy": 0.125148, "bpb": 1.890728}, "ru": {"chunks": 2977.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.157098, "cps": 23.257812, "mean_p": 0.16474, "boundary_entropy": 0.311639, "bpb": 1.145497}, "el": {"chunks": 2879.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.140282, "cps": 22.492188, "mean_p": 0.129615, "boundary_entropy": 0.245438, "bpb": 1.137594}, "hi": {"chunks": 2740.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.162853, "cps": 21.40625, "mean_p": 0.117765, "boundary_entropy": 0.251058, "bpb": 0.94682}, "ta": {"chunks": 2590.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.13854, "cps": 20.234375, "mean_p": 0.177327, "boundary_entropy": 0.385396, "bpb": 0.808184}, "th": {"chunks": 2681.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.166677, "cps": 20.945312, "mean_p": 0.133873, "boundary_entropy": 0.28692, "bpb": 0.916104}, "zh": {"chunks": 3000.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.507958, "cps": 23.4375, "mean_p": 0.222784, "boundary_entropy": 0.201857, "bpb": 2.639735}, "my": {"chunks": 2770.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.14619, "cps": 21.640625, "mean_p": 0.257586, "boundary_entropy": 0.494937, "bpb": 0.899539}}, "bpb": {"en": 1.68155, "de": 1.82029, "vi": 1.52669, "tr": 1.74262, "fi": 1.89073, "ru": 1.1455, "el": 1.13759, "hi": 0.94682, "ta": 0.80818, "th": 0.9161, "zh": 2.63974, "my": 0.89954}, "gini_cpc": 0.171487, "gini_cps": 0.030928, "bpic": 11.062841, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8734911804361144, "vi": 0.9079977302899233, "tr": 0.9186416288954724, "fi": 0.9288020866850022, "ru": 0.8993322281627552, "el": 0.8030662164657367, "hi": 0.9322792680065691, "ta": 0.7930943046972875, "th": 0.9541711589745266, "zh": 2.9078925089561016, "my": 0.8368872079264024}, "t": 1770.3} | |
| {"event": "eval", "step": 9000, "phase": 3, "bpb": {"en": 1.68124, "de": 1.82567, "vi": 1.54131, "tr": 1.73578, "fi": 1.90454, "ru": 1.14418, "el": 1.1369, "hi": 0.95319, "ta": 0.81282, "th": 0.91426, "zh": 2.65122, "my": 0.90746}, "bytes_per_chunk": {"en": 5.573, "de": 6.452, "vi": 8.052, "tr": 6.716, "fi": 6.013, "ru": 11.493, "el": 12.542, "hi": 15.678, "ta": 19.939, "th": 16.893, "zh": 5.483, "my": 18.613}, "bpb_mean": 1.43405, "t": 1775.2} | |
| {"event": "train", "step": 9000, "phase": 3, "objective": "parity_a", "lm_loss": 1.12281, "bpb": 1.61987, "chunk_loss": 0.030193, "F": 0.14941, "G": 0.18312, "chunks": 153.0, "bytes_per_chunk": 6.693, "boundary_entropy": 0.16542, "router_noise": 0.0, "grad_norm": 0.4309, "lr": 5.35e-05, "router_trainable": true, "sec_per_step": 0.269, "n_target_mean": 8.2969, "alpha": 0.03, "t": 1775.4} | |
| {"event": "train", "step": 9100, "phase": 3, "objective": "parity_a", "lm_loss": 1.10316, "bpb": 1.59152, "chunk_loss": 0.030128, "F": 0.14264, "G": 0.17533, "chunks": 146.06, "bytes_per_chunk": 7.011, "boundary_entropy": 0.16709, "router_noise": 0.0, "grad_norm": 0.4276, "lr": 5.12e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 8.5607, "alpha": 0.03, "t": 1792.3} | |
| {"event": "train", "step": 9200, "phase": 3, "objective": "parity_a", "lm_loss": 1.03131, "bpb": 1.48787, "chunk_loss": 0.030137, "F": 0.13174, "G": 0.17583, "chunks": 134.91, "bytes_per_chunk": 7.59, "boundary_entropy": 0.19478, "router_noise": 0.0, "grad_norm": 0.421, "lr": 4.91e-05, "router_trainable": true, "sec_per_step": 0.187, "n_target_mean": 9.4012, "alpha": 0.03, "t": 1811.0} | |
| {"event": "train", "step": 9300, "phase": 3, "objective": "parity_a", "lm_loss": 1.07279, "bpb": 1.54771, "chunk_loss": 0.030193, "F": 0.12833, "G": 0.17293, "chunks": 131.41, "bytes_per_chunk": 7.793, "boundary_entropy": 0.19997, "router_noise": 0.0, "grad_norm": 0.4461, "lr": 4.71e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.6588, "alpha": 0.03, "t": 1828.0} | |
| {"event": "train", "step": 9400, "phase": 3, "objective": "parity_a", "lm_loss": 1.03535, "bpb": 1.4937, "chunk_loss": 0.030114, "F": 0.12419, "G": 0.17133, "chunks": 127.17, "bytes_per_chunk": 8.052, "boundary_entropy": 0.21017, "router_noise": 0.0, "grad_norm": 0.4127, "lr": 4.52e-05, "router_trainable": true, "sec_per_step": 0.188, "n_target_mean": 10.2413, "alpha": 0.03, "t": 1846.8} | |
| {"event": "train", "step": 9500, "phase": 3, "objective": "parity_a", "lm_loss": 1.11252, "bpb": 1.60502, "chunk_loss": 0.030119, "F": 0.14305, "G": 0.17845, "chunks": 146.48, "bytes_per_chunk": 6.991, "boundary_entropy": 0.17319, "router_noise": 0.0, "grad_norm": 0.462, "lr": 4.34e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.649, "alpha": 0.03, "t": 1863.8} | |
| {"event": "train", "step": 9600, "phase": 3, "objective": "parity_a", "lm_loss": 1.07041, "bpb": 1.54428, "chunk_loss": 0.030205, "F": 0.13789, "G": 0.17532, "chunks": 141.2, "bytes_per_chunk": 7.252, "boundary_entropy": 0.18075, "router_noise": 0.0, "grad_norm": 0.425, "lr": 4.17e-05, "router_trainable": true, "sec_per_step": 0.189, "n_target_mean": 9.0764, "alpha": 0.03, "t": 1882.7} | |
| {"event": "train", "step": 9700, "phase": 3, "objective": "parity_a", "lm_loss": 1.0905, "bpb": 1.57326, "chunk_loss": 0.030027, "F": 0.13148, "G": 0.17273, "chunks": 134.64, "bytes_per_chunk": 7.605, "boundary_entropy": 0.18328, "router_noise": 0.0, "grad_norm": 0.4414, "lr": 4.01e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 9.2508, "alpha": 0.03, "t": 1899.6} | |
| {"event": "train", "step": 9800, "phase": 3, "objective": "parity_a", "lm_loss": 1.01452, "bpb": 1.46364, "chunk_loss": 0.030208, "F": 0.12791, "G": 0.17253, "chunks": 130.98, "bytes_per_chunk": 7.818, "boundary_entropy": 0.19611, "router_noise": 0.0, "grad_norm": 0.422, "lr": 3.86e-05, "router_trainable": true, "sec_per_step": 0.19, "n_target_mean": 9.8501, "alpha": 0.03, "t": 1918.6} | |
| {"event": "train", "step": 9900, "phase": 3, "objective": "parity_a", "lm_loss": 1.07581, "bpb": 1.55206, "chunk_loss": 0.030096, "F": 0.14171, "G": 0.16883, "chunks": 145.11, "bytes_per_chunk": 7.057, "boundary_entropy": 0.15154, "router_noise": 0.0, "grad_norm": 0.4219, "lr": 3.72e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 8.3862, "alpha": 0.03, "t": 1935.5} | |
| {"event": "probe", "step": 10000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3067.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.176234, "cps": 23.960938, "mean_p": 0.183951, "boundary_entropy": 0.085302, "bpb": 1.668106}, "de": {"chunks": 3171.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.154434, "cps": 24.773438, "mean_p": 0.166586, "boundary_entropy": 0.109503, "bpb": 1.801496}, "vi": {"chunks": 2901.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.157364, "cps": 22.664062, "mean_p": 0.15852, "boundary_entropy": 0.18419, "bpb": 1.509315}, "tr": {"chunks": 2908.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.160636, "cps": 22.71875, "mean_p": 0.169483, "boundary_entropy": 0.142902, "bpb": 1.723413}, "fi": {"chunks": 2987.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.15879, "cps": 23.335938, "mean_p": 0.169118, "boundary_entropy": 0.117594, "bpb": 1.870859}, "ru": {"chunks": 2860.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.150923, "cps": 22.34375, "mean_p": 0.160195, "boundary_entropy": 0.304761, "bpb": 1.13446}, "el": {"chunks": 2875.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.140087, "cps": 22.460938, "mean_p": 0.126871, "boundary_entropy": 0.238801, "bpb": 1.125595}, "hi": {"chunks": 2948.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.175215, "cps": 23.03125, "mean_p": 0.114486, "boundary_entropy": 0.240982, "bpb": 0.935163}, "ta": {"chunks": 2446.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.130837, "cps": 19.109375, "mean_p": 0.176975, "boundary_entropy": 0.38558, "bpb": 0.800803}, "th": {"chunks": 2775.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.172521, "cps": 21.679688, "mean_p": 0.134287, "boundary_entropy": 0.285749, "bpb": 0.905348}, "zh": {"chunks": 2970.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.502878, "cps": 23.203125, "mean_p": 0.219409, "boundary_entropy": 0.192502, "bpb": 2.61938}, "my": {"chunks": 2669.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.140859, "cps": 20.851562, "mean_p": 0.255705, "boundary_entropy": 0.494129, "bpb": 0.889613}}, "bpb": {"en": 1.66811, "de": 1.8015, "vi": 1.50932, "tr": 1.72341, "fi": 1.87086, "ru": 1.13446, "el": 1.1256, "hi": 0.93516, "ta": 0.8008, "th": 0.90535, "zh": 2.61938, "my": 0.88961}, "gini_cpc": 0.179603, "gini_cps": 0.0333, "bpic": 11.099315, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8763027590551389, "vi": 0.8929248943383502, "tr": 0.911494799142414, "fi": 0.9010184486295185, "ru": 0.8563812755238547, "el": 0.7948905759005844, "hi": 0.9942205842668758, "ta": 0.7424057509243721, "th": 0.9789314165788023, "zh": 2.8534702624565647, "my": 0.799273735218499}, "t": 1957.2} | |
| {"event": "eval", "step": 10000, "phase": 3, "bpb": {"en": 1.66689, "de": 1.80734, "vi": 1.52497, "tr": 1.71594, "fi": 1.88464, "ru": 1.13344, "el": 1.12433, "hi": 0.94151, "ta": 0.80552, "th": 0.90372, "zh": 2.632, "my": 0.89838}, "bytes_per_chunk": {"en": 5.541, "de": 6.379, "vi": 8.065, "tr": 6.746, "fi": 6.116, "ru": 11.851, "el": 12.726, "hi": 14.291, "ta": 20.682, "th": 16.514, "zh": 5.546, "my": 19.616}, "bpb_mean": 1.41989, "t": 1962.0} | |
| {"event": "train", "step": 10000, "phase": 3, "objective": "parity_a", "lm_loss": 1.09641, "bpb": 1.58179, "chunk_loss": 0.030074, "F": 0.14185, "G": 0.17741, "chunks": 145.25, "bytes_per_chunk": 7.05, "boundary_entropy": 0.16646, "router_noise": 0.0, "grad_norm": 0.4253, "lr": 3.6e-05, "router_trainable": true, "sec_per_step": 0.267, "n_target_mean": 8.546, "alpha": 0.03, "t": 1962.2} | |
| {"event": "train", "step": 10100, "phase": 3, "objective": "parity_a", "lm_loss": 1.15262, "bpb": 1.66288, "chunk_loss": 0.030142, "F": 0.14592, "G": 0.17975, "chunks": 149.42, "bytes_per_chunk": 6.853, "boundary_entropy": 0.1636, "router_noise": 0.0, "grad_norm": 0.4393, "lr": 3.49e-05, "router_trainable": true, "sec_per_step": 0.171, "n_target_mean": 8.3526, "alpha": 0.03, "t": 1979.3} | |
| {"event": "train", "step": 10200, "phase": 3, "objective": "parity_a", "lm_loss": 1.06536, "bpb": 1.53698, "chunk_loss": 0.030279, "F": 0.13916, "G": 0.17973, "chunks": 142.5, "bytes_per_chunk": 7.186, "boundary_entropy": 0.18618, "router_noise": 0.0, "grad_norm": 0.4413, "lr": 3.38e-05, "router_trainable": true, "sec_per_step": 0.188, "n_target_mean": 9.2727, "alpha": 0.03, "t": 1998.1} | |
| {"event": "train", "step": 10300, "phase": 3, "objective": "parity_a", "lm_loss": 1.05656, "bpb": 1.52429, "chunk_loss": 0.030091, "F": 0.13632, "G": 0.17247, "chunks": 139.59, "bytes_per_chunk": 7.336, "boundary_entropy": 0.17608, "router_noise": 0.0, "grad_norm": 0.4293, "lr": 3.3e-05, "router_trainable": true, "sec_per_step": 0.169, "n_target_mean": 8.8635, "alpha": 0.03, "t": 2015.0} | |
| {"event": "train", "step": 10400, "phase": 3, "objective": "parity_a", "lm_loss": 1.05895, "bpb": 1.52774, "chunk_loss": 0.030153, "F": 0.13399, "G": 0.17372, "chunks": 137.2, "bytes_per_chunk": 7.463, "boundary_entropy": 0.18289, "router_noise": 0.0, "grad_norm": 0.4392, "lr": 3.22e-05, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 9.198, "alpha": 0.03, "t": 2033.6} | |
| {"event": "train", "step": 10500, "phase": 3, "objective": "parity_a", "lm_loss": 1.01933, "bpb": 1.47059, "chunk_loss": 0.030011, "F": 0.12509, "G": 0.1732, "chunks": 128.09, "bytes_per_chunk": 7.994, "boundary_entropy": 0.19454, "router_noise": 0.0, "grad_norm": 0.4444, "lr": 3.15e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 10.3096, "alpha": 0.03, "t": 2050.6} | |
| {"event": "train", "step": 10600, "phase": 3, "objective": "parity_a", "lm_loss": 1.1063, "bpb": 1.59605, "chunk_loss": 0.030233, "F": 0.14684, "G": 0.18148, "chunks": 150.36, "bytes_per_chunk": 6.81, "boundary_entropy": 0.16473, "router_noise": 0.0, "grad_norm": 0.4222, "lr": 3.1e-05, "router_trainable": true, "sec_per_step": 0.173, "n_target_mean": 8.7263, "alpha": 0.03, "t": 2067.8} | |
| {"event": "train", "step": 10700, "phase": 3, "objective": "parity_a", "lm_loss": 1.08388, "bpb": 1.56371, "chunk_loss": 0.030285, "F": 0.14525, "G": 0.17959, "chunks": 148.73, "bytes_per_chunk": 6.885, "boundary_entropy": 0.1729, "router_noise": 0.0, "grad_norm": 0.4375, "lr": 3.05e-05, "router_trainable": true, "sec_per_step": 0.186, "n_target_mean": 9.0641, "alpha": 0.03, "t": 2086.5} | |
| {"event": "train", "step": 10800, "phase": 3, "objective": "parity_a", "lm_loss": 1.05675, "bpb": 1.52456, "chunk_loss": 0.030139, "F": 0.13504, "G": 0.17156, "chunks": 138.28, "bytes_per_chunk": 7.405, "boundary_entropy": 0.17271, "router_noise": 0.0, "grad_norm": 0.4503, "lr": 3.02e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 9.3092, "alpha": 0.03, "t": 2103.5} | |
| {"event": "train", "step": 10900, "phase": 3, "objective": "parity_a", "lm_loss": 1.13675, "bpb": 1.63998, "chunk_loss": 0.030148, "F": 0.14174, "G": 0.17698, "chunks": 145.14, "bytes_per_chunk": 7.055, "boundary_entropy": 0.17041, "router_noise": 0.0, "grad_norm": 0.4587, "lr": 3.01e-05, "router_trainable": true, "sec_per_step": 0.17, "n_target_mean": 8.8232, "alpha": 0.03, "t": 2120.5} | |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3053.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.17543, "cps": 23.851562, "mean_p": 0.183234, "boundary_entropy": 0.081847, "bpb": 1.658958}, "de": {"chunks": 3167.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15424, "cps": 24.742188, "mean_p": 0.165551, "boundary_entropy": 0.105892, "bpb": 1.791832}, "vi": {"chunks": 2970.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.161107, "cps": 23.203125, "mean_p": 0.159657, "boundary_entropy": 0.18213, "bpb": 1.497273}, "tr": {"chunks": 2931.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.161907, "cps": 22.898438, "mean_p": 0.170281, "boundary_entropy": 0.139809, "bpb": 1.712201}, "fi": {"chunks": 3000.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159481, "cps": 23.4375, "mean_p": 0.169485, "boundary_entropy": 0.116328, "bpb": 1.857881}, "ru": {"chunks": 2817.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.148654, "cps": 22.007812, "mean_p": 0.152513, "boundary_entropy": 0.285473, "bpb": 1.128818}, "el": {"chunks": 2972.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.144813, "cps": 23.21875, "mean_p": 0.127073, "boundary_entropy": 0.23444, "bpb": 1.118332}, "hi": {"chunks": 2829.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.168143, "cps": 22.101562, "mean_p": 0.113159, "boundary_entropy": 0.238989, "bpb": 0.928794}, "ta": {"chunks": 2958.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.158224, "cps": 23.109375, "mean_p": 0.179361, "boundary_entropy": 0.383231, "bpb": 0.795286}, "th": {"chunks": 2771.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.172272, "cps": 21.648438, "mean_p": 0.133314, "boundary_entropy": 0.281502, "bpb": 0.898668}, "zh": {"chunks": 2988.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.505926, "cps": 23.34375, "mean_p": 0.219981, "boundary_entropy": 0.189173, "bpb": 2.608566}, "my": {"chunks": 2824.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.149039, "cps": 22.0625, "mean_p": 0.254877, "boundary_entropy": 0.491595, "bpb": 0.881951}}, "bpb": {"en": 1.65896, "de": 1.79183, "vi": 1.49727, "tr": 1.7122, "fi": 1.85788, "ru": 1.12882, "el": 1.11833, "hi": 0.92879, "ta": 0.79529, "th": 0.89867, "zh": 2.60857, "my": 0.88195}, "gini_cpc": 0.164438, "gini_cps": 0.020413, "bpic": 10.878146, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8792107147659327, "vi": 0.9183550600631054, "tr": 0.9229168665621743, "fi": 0.9090895952380832, "ru": 0.8473736374187939, "el": 0.8254775676970892, "hi": 0.9584626428879786, "ta": 0.9019241505621425, "th": 0.9820029138112393, "zh": 2.883928350578986, "my": 0.8495689514793224}, "t": 2140.2} | |
| {"event": "eval", "step": 11000, "phase": 3, "bpb": {"en": 1.65791, "de": 1.79778, "vi": 1.51281, "tr": 1.70498, "fi": 1.87172, "ru": 1.12756, "el": 1.11742, "hi": 0.93502, "ta": 0.79978, "th": 0.89659, "zh": 2.61915, "my": 0.89061}, "bytes_per_chunk": {"en": 5.541, "de": 6.375, "vi": 7.997, "tr": 6.642, "fi": 6.054, "ru": 11.979, "el": 12.315, "hi": 15.011, "ta": 17.858, "th": 16.571, "zh": 5.547, "my": 18.297}, "bpb_mean": 1.41094, "t": 2145.8} | |
| {"event": "probe", "step": 11000, "phase": 3, "objective": "parity_a", "router_trainable": true, "per_lang": {"en": {"chunks": 3053.0, "chars": 17403.0, "bytes": 17418.0, "cpc": 0.17543, "cps": 23.851562, "mean_p": 0.183234, "boundary_entropy": 0.081847, "bpb": 1.658958}, "de": {"chunks": 3167.0, "chars": 20533.0, "bytes": 20862.0, "cpc": 0.15424, "cps": 24.742188, "mean_p": 0.165551, "boundary_entropy": 0.105892, "bpb": 1.791832}, "vi": {"chunks": 2970.0, "chars": 18435.0, "bytes": 24274.0, "cpc": 0.161107, "cps": 23.203125, "mean_p": 0.159657, "boundary_entropy": 0.18213, "bpb": 1.497273}, "tr": {"chunks": 2931.0, "chars": 18103.0, "bytes": 19742.0, "cpc": 0.161907, "cps": 22.898438, "mean_p": 0.170281, "boundary_entropy": 0.139809, "bpb": 1.712201}, "fi": {"chunks": 3000.0, "chars": 18811.0, "bytes": 19572.0, "cpc": 0.159481, "cps": 23.4375, "mean_p": 0.169485, "boundary_entropy": 0.116328, "bpb": 1.857881}, "ru": {"chunks": 2817.0, "chars": 18950.0, "bytes": 34560.0, "cpc": 0.148654, "cps": 22.007812, "mean_p": 0.152513, "boundary_entropy": 0.285473, "bpb": 1.128818}, "el": {"chunks": 2972.0, "chars": 20523.0, "bytes": 37264.0, "cpc": 0.144813, "cps": 23.21875, "mean_p": 0.127073, "boundary_entropy": 0.23444, "bpb": 1.118332}, "hi": {"chunks": 2829.0, "chars": 16825.0, "bytes": 43173.0, "cpc": 0.168143, "cps": 22.101562, "mean_p": 0.113159, "boundary_entropy": 0.238989, "bpb": 0.928794}, "ta": {"chunks": 2958.0, "chars": 18695.0, "bytes": 50915.0, "cpc": 0.158224, "cps": 23.109375, "mean_p": 0.179361, "boundary_entropy": 0.383231, "bpb": 0.795286}, "th": {"chunks": 2771.0, "chars": 16085.0, "bytes": 46085.0, "cpc": 0.172272, "cps": 21.648438, "mean_p": 0.133314, "boundary_entropy": 0.281502, "bpb": 0.898668}, "zh": {"chunks": 2988.0, "chars": 5906.0, "bytes": 16173.0, "cpc": 0.505926, "cps": 23.34375, "mean_p": 0.219981, "boundary_entropy": 0.189173, "bpb": 2.608566}, "my": {"chunks": 2824.0, "chars": 18948.0, "bytes": 53743.0, "cpc": 0.149039, "cps": 22.0625, "mean_p": 0.254877, "boundary_entropy": 0.491595, "bpb": 0.881951}}, "bpb": {"en": 1.65896, "de": 1.79183, "vi": 1.49727, "tr": 1.7122, "fi": 1.85788, "ru": 1.12882, "el": 1.11833, "hi": 0.92879, "ta": 0.79529, "th": 0.89867, "zh": 2.60857, "my": 0.88195}, "gini_cpc": 0.164438, "gini_cps": 0.020413, "bpic": 10.878146, "worst_lang": "zh", "cpc_parity_vs_en": {"en": 1.0, "de": 0.8792107147659327, "vi": 0.9183550600631054, "tr": 0.9229168665621743, "fi": 0.9090895952380832, "ru": 0.8473736374187939, "el": 0.8254775676970892, "hi": 0.9584626428879786, "ta": 0.9019241505621425, "th": 0.9820029138112393, "zh": 2.883928350578986, "my": 0.8495689514793224}, "t": 2149.5} | |
| {"event": "reps_dumped", "dirs": ["/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s1/th0.30", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s1/th0.40", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s1/th0.50", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s1/th0.60", "/local/scratch/sas245/beetle-hnet/scratch/reps/spec_D_s1/th0.70"], "t": 2152.4} | |
| {"event": "final", "run": "spec_D_s1", "condition": "D", "step": 11000, "params": 82470528, "per_lang": {"en": {"cpc": 0.1754295236453485, "cps": 23.8515625, "bpb": 1.6540044270787866, "bytes_per_chunk": 5.543768008670597}, "de": {"cpc": 0.15423951687527396, "cps": 24.7421875, "bpb": 1.7966894356572356, "bytes_per_chunk": 6.326212209881087}, "vi": {"cpc": 0.161106590724166, "cps": 23.203125, "bpb": 1.5140610046866572, "bytes_per_chunk": 7.978451752316893}, "tr": {"cpc": 0.1619068662652599, "cps": 22.8984375, "bpb": 1.7045004681769294, "bytes_per_chunk": 6.631059058628606}, "fi": {"cpc": 0.1594811546435596, "cps": 23.4375, "bpb": 1.8760725104608031, "bytes_per_chunk": 6.034969784172662}, "ru": {"cpc": 0.14865435356200526, "cps": 22.0078125, "bpb": 1.1289955739041115, "bytes_per_chunk": 11.96430934939869}, "el": {"cpc": 0.1448131364810213, "cps": 23.21875, "bpb": 1.1143450598077123, "bytes_per_chunk": 12.268495009886626}, "hi": {"cpc": 0.16814264487369984, "cps": 22.1015625, "bpb": 0.9351596626631545, "bytes_per_chunk": 15.01870577787963}, "ta": {"cpc": 0.15822412409735223, "cps": 23.109375, "bpb": 0.7979500119956463, "bytes_per_chunk": 17.62994098559107}, "th": {"cpc": 0.17227230338824992, "cps": 21.6484375, "bpb": 0.8991655856467282, "bytes_per_chunk": 16.481083884758657}, "zh": {"cpc": 0.5059261767693871, "cps": 23.34375, "bpb": 2.6232984753589252, "bytes_per_chunk": 5.545996435163882}, "my": {"cpc": 0.14903947646189572, "cps": 22.0625, "bpb": 0.8888833233930875, "bytes_per_chunk": 18.21203279144088}}, "gini_cpc": 0.16443756967222667, "gini_cps": 0.020412887377173172, "bpic": 10.878146258503401, "seed": 1, "mean_boundary_entropy": 0.1963818371295929, "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": 2161.9} | |