Upload add_sub_sorl_v1_abs30_K1_25K

add_sub_sorl_v1_abs30_K1_25K/config.json

@@ -0,0 +1,37 @@
+{
+  "architectures": [
+    "SorlModelWrapper"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": null,
+  "dtype": "float32",
+  "eos_token_id": null,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 510,
+  "initializer_range": 0.02,
+  "intermediate_size": 2040,
+  "layer_types": [
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 128,
+  "max_window_layers": 28,
+  "model_type": "qwen3",
+  "num_attention_heads": 3,
+  "num_hidden_layers": 2,
+  "num_key_value_heads": 3,
+  "pad_token_id": null,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 10000.0,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": false,
+  "transformers_version": "5.5.0",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 151674
+}

add_sub_sorl_v1_abs30_K1_25K/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "output_attentions": false,
+  "output_hidden_states": false,
+  "transformers_version": "5.5.0",
+  "use_cache": true
+}

add_sub_sorl_v1_abs30_K1_25K/metrics.json

@@ -0,0 +1,2197 @@
+{
+  "history": {
+    "step": [
+      50,
+      100,
+      150,
+      200,
+      250,
+      300,
+      350,
+      441,
+      491,
+      541,
+      591,
+      641,
+      691,
+      741,
+      832,
+      882,
+      932,
+      982,
+      1032,
+      1082,
+      1132,
+      1223,
+      1273,
+      1323,
+      1373,
+      1423,
+      1473,
+      1523,
+      1614,
+      1664,
+      1714,
+      1764,
+      1814,
+      1864,
+      1914,
+      2005,
+      2055,
+      2105,
+      2155,
+      2205,
+      2255,
+      2305,
+      2396,
+      2446,
+      2496,
+      2546,
+      2596,
+      2646,
+      2696,
+      2787,
+      2837,
+      2887,
+      2937,
+      2987,
+      3037,
+      3087,
+      3178,
+      3228,
+      3278,
+      3328,
+      3378,
+      3428,
+      3478,
+      3569,
+      3619,
+      3669,
+      3719,
+      3769,
+      3819,
+      3869,
+      3960,
+      4010,
+      4060,
+      4110,
+      4160,
+      4210,
+      4260,
+      4351,
+      4401,
+      4451,
+      4501,
+      4551,
+      4601,
+      4651,
+      4742,
+      4792,
+      4842,
+      4892,
+      4942,
+      4992,
+      5042,
+      5133,
+      5183,
+      5233,
+      5283,
+      5333,
+      5383,
+      5433,
+      5524,
+      5574,
+      5624,
+      5674,
+      5724,
+      5774,
+      5824,
+      5915,
+      5965,
+      6015,
+      6065,
+      6115,
+      6165,
+      6215,
+      6306,
+      6356,
+      6406,
+      6456,
+      6506,
+      6556,
+      6606,
+      6697,
+      6747,
+      6797,
+      6847,
+      6897,
+      6947,
+      6997,
+      7088,
+      7138,
+      7188,
+      7238,
+      7288,
+      7338,
+      7388,
+      7479,
+      7529,
+      7579,
+      7629,
+      7679,
+      7729,
+      7779
+    ],
+    "loss": [
+      1.7119760513305664,
+      6.14704704284668,
+      4.015985488891602,
+      2.931351661682129,
+      2.8381450176239014,
+      2.7390925884246826,
+      1.9953590631484985,
+      0.8327422142028809,
+      -3.8749427795410156,
+      -4.478009223937988,
+      -0.9891107082366943,
+      -0.2836305499076843,
+      -0.561787486076355,
+      -1.141134262084961,
+      -1.0379124879837036,
+      -0.8155723810195923,
+      -0.6522830128669739,
+      -0.969704270362854,
+      -0.6387378573417664,
+      -0.42737025022506714,
+      -0.7698933482170105,
+      -0.8871013522148132,
+      -0.723418116569519,
+      -1.4676158428192139,
+      -1.118342638015747,
+      -0.9780423045158386,
+      -1.3972547054290771,
+      -1.118117332458496,
+      -1.2728338241577148,
+      -1.3038815259933472,
+      -1.1130985021591187,
+      -1.2439594268798828,
+      -1.6075873374938965,
+      -1.2939362525939941,
+      -1.123915672302246,
+      -0.8386991024017334,
+      -1.2796339988708496,
+      -0.9666016101837158,
+      -1.2052180767059326,
+      -0.9123711585998535,
+      -0.7841464281082153,
+      -1.115072250366211,
+      -0.770645260810852,
+      -0.7825030088424683,
+      -0.9296966195106506,
+      -0.7025026679039001,
+      -0.9686148762702942,
+      -0.6636331677436829,
+      -0.8582434058189392,
+      -0.8699550628662109,
+      -1.1216692924499512,
+      -0.5576128363609314,
+      -0.5983526706695557,
+      -0.5066444873809814,
+      -0.5169333219528198,
+      -0.5374507308006287,
+      -0.5922688245773315,
+      -0.29379820823669434,
+      -0.7311683297157288,
+      -0.22624635696411133,
+      -0.3172416388988495,
+      -0.32080110907554626,
+      -0.4887930750846863,
+      -0.25087282061576843,
+      -0.5092266798019409,
+      -0.7776860594749451,
+      -0.15283048152923584,
+      -0.3975360095500946,
+      -0.3694022297859192,
+      -0.40873661637306213,
+      -0.33530089259147644,
+      -0.20308193564414978,
+      -0.2026340812444687,
+      -0.37627139687538147,
+      -0.2508956789970398,
+      -0.2133498638868332,
+      -0.1757795363664627,
+      -0.22155731916427612,
+      -0.12377417832612991,
+      -0.15623070299625397,
+      -0.15215444564819336,
+      -0.1778583824634552,
+      -0.3856639266014099,
+      -0.19510412216186523,
+      -0.45289498567581177,
+      -0.07906970381736755,
+      -0.20047199726104736,
+      -0.24871760606765747,
+      -0.15301105380058289,
+      -0.20984536409378052,
+      -0.15769784152507782,
+      -0.18541622161865234,
+      -0.0805714875459671,
+      -0.07824444025754929,
+      -0.06095310300588608,
+      -0.08044923841953278,
+      0.003586065024137497,
+      0.10273022204637527,
+      -0.10450196266174316,
+      -0.1762467324733734,
+      -0.08977089822292328,
+      -0.06519266963005066,
+      -0.041904985904693604,
+      -0.13669879734516144,
+      -0.018732935190200806,
+      -0.13292600214481354,
+      -0.11485083401203156,
+      -0.002781972289085388,
+      0.011605948209762573,
+      -0.0375908724963665,
+      -0.0719958245754242,
+      -0.009761612862348557,
+      -0.036501675844192505,
+      -0.12502069771289825,
+      -0.0013336818665266037,
+      0.014050843194127083,
+      -0.006407836452126503,
+      0.016416456550359726,
+      0.010326145216822624,
+      -0.04670792445540428,
+      -0.05867944657802582,
+      1.0332092642784119e-05,
+      0.011082771234214306,
+      -0.0830128937959671,
+      -0.0022481046617031097,
+      -0.009278824552893639,
+      -0.1630394458770752,
+      0.014726893045008183,
+      0.009142588824033737,
+      0.01728634163737297,
+      0.006687477231025696,
+      0.0021704640239477158,
+      0.012956775724887848,
+      0.016162145882844925,
+      0.012694327160716057,
+      0.015958743169903755,
+      0.01844569854438305,
+      0.01067356951534748,
+      0.010267311707139015,
+      0.01123642735183239
+    ],
+    "base_loss": [
+      9.25883960723877,
+      6.370208740234375,
+      3.7065303325653076,
+      2.0331058502197266,
+      1.95931077003479,
+      1.8724143505096436,
+      1.8358138799667358,
+      1.7839851379394531,
+      1.8063528537750244,
+      1.4257497787475586,
+      0.8925438523292542,
+      0.7544015049934387,
+      0.6830788850784302,
+      0.6486687064170837,
+      0.49347400665283203,
+      0.3990347981452942,
+      0.41614702343940735,
+      0.41534119844436646,
+      0.3495095670223236,
+      0.2816762924194336,
+      0.3096388876438141,
+      0.2650178074836731,
+      0.23259153962135315,
+      0.26580312848091125,
+      0.21972285211086273,
+      0.25576403737068176,
+      0.2180083841085434,
+      0.19050060212612152,
+      0.1893482506275177,
+      0.18317624926567078,
+      0.15178455412387848,
+      0.1825704127550125,
+      0.20081649720668793,
+      0.16461952030658722,
+      0.14625681936740875,
+      0.11179553717374802,
+      0.15817393362522125,
+      0.12313541024923325,
+      0.15251047909259796,
+      0.11650027334690094,
+      0.10264534503221512,
+      0.1392023265361786,
+      0.09644023329019547,
+      0.10179956257343292,
+      0.11617104709148407,
+      0.08889312297105789,
+      0.1257583051919937,
+      0.10296887904405594,
+      0.11060570180416107,
+      0.11073609441518784,
+      0.1322706639766693,
+      0.0693567544221878,
+      0.07331784814596176,
+      0.06344141066074371,
+      0.06363283097743988,
+      0.06521368771791458,
+      0.07137874513864517,
+      0.03687870875000954,
+      0.0886240154504776,
+      0.0316486619412899,
+      0.04859679192304611,
+      0.04001149162650108,
+      0.0594177171587944,
+      0.03198449686169624,
+      0.05951383709907532,
+      0.0927167758345604,
+      0.020596766844391823,
+      0.04924142360687256,
+      0.04642181843519211,
+      0.04873456805944443,
+      0.041637055575847626,
+      0.02686495892703533,
+      0.026218848302960396,
+      0.04749896004796028,
+      0.031423475593328476,
+      0.026792261749505997,
+      0.02315610647201538,
+      0.027571475133299828,
+      0.01708334870636463,
+      0.019614964723587036,
+      0.02113191783428192,
+      0.02272043190896511,
+      0.04554429650306702,
+      0.02509036846458912,
+      0.05237502604722977,
+      0.011714950203895569,
+      0.024812577292323112,
+      0.03115662932395935,
+      0.01989857666194439,
+      0.024637239053845406,
+      0.020898962393403053,
+      0.023834871128201485,
+      0.014209302142262459,
+      0.011255674995481968,
+      0.009083413518965244,
+      0.011385834775865078,
+      0.0028439622838050127,
+      0.018480319529771805,
+      0.014435835182666779,
+      0.021674534305930138,
+      0.011868111789226532,
+      0.009697256609797478,
+      0.008302876725792885,
+      0.017009999603033066,
+      0.0038561702240258455,
+      0.016001036390662193,
+      0.015324118547141552,
+      0.002748801140114665,
+      0.0010739824501797557,
+      0.006728976033627987,
+      0.00928917434066534,
+      0.0024181229528039694,
+      0.005708904471248388,
+      0.01629350520670414,
+      0.0029613084625452757,
+      0.0003541119222063571,
+      0.002198481233790517,
+      0.0014702372718602419,
+      0.001857394352555275,
+      0.006702867336571217,
+      0.008157839998602867,
+      0.001853235182352364,
+      0.0002782914671115577,
+      0.010882347822189331,
+      0.0027038822881877422,
+      0.0030141728930175304,
+      0.020343586802482605,
+      0.00020745322399307042,
+      9.352602501166984e-05,
+      0.0006757898954674602,
+      0.0005854826304130256,
+      0.0013537565246224403,
+      0.0008722890052013099,
+      0.0004628680762834847,
+      0.0004031405842397362,
+      0.0008272135746665299,
+      0.00024150208628270775,
+      0.0004777032299898565,
+      0.00025579737848602235,
+      0.0005325012025423348
+    ],
+    "info_loss": [
+      -1.3981437683105469,
+      -0.35674333572387695,
+      -0.1073293685913086,
+      -0.02326798439025879,
+      -0.023933053016662598,
+      -0.025423526763916016,
+      -0.09565258026123047,
+      -0.20662808418273926,
+      -0.6795123815536499,
+      -0.7024374008178711,
+      -0.29942911863327026,
+      -0.21352607011795044,
+      -0.23093774914741516,
+      -0.28096023201942444,
+      -0.24447700381278992,
+      -0.2022431492805481,
+      -0.18190807104110718,
+      -0.21125152707099915,
+      -0.16269072890281677,
+      -0.13090260326862335,
+      -0.1593063622713089,
+      -0.16261976957321167,
+      -0.1382463276386261,
+      -0.2155323028564453,
+      -0.17649787664413452,
+      -0.16470617055892944,
+      -0.19673863053321838,
+      -0.16458788514137268,
+      -0.17244882881641388,
+      -0.1751340925693512,
+      -0.14462099969387054,
+      -0.15938325226306915,
+      -0.19329020380973816,
+      -0.1597842425107956,
+      -0.14075155556201935,
+      -0.10603491216897964,
+      -0.15410369634628296,
+      -0.1185174509882927,
+      -0.1451079249382019,
+      -0.1115327849984169,
+      -0.09845148026943207,
+      -0.13466130197048187,
+      -0.09395471960306168,
+      -0.0970274806022644,
+      -0.11261290311813354,
+      -0.08472369611263275,
+      -0.1171511560678482,
+      -0.08429136872291565,
+      -0.10386909544467926,
+      -0.10420701652765274,
+      -0.1310877948999405,
+      -0.06857217848300934,
+      -0.07247594743967056,
+      -0.06123020499944687,
+      -0.062376223504543304,
+      -0.06436779350042343,
+      -0.07016307860612869,
+      -0.03637612238526344,
+      -0.0855904370546341,
+      -0.03112221509218216,
+      -0.040240220725536346,
+      -0.03903472423553467,
+      -0.05904826149344444,
+      -0.03156021237373352,
+      -0.05928679183125496,
+      -0.09059230238199234,
+      -0.020370472222566605,
+      -0.04692664369940758,
+      -0.04496893659234047,
+      -0.04838690906763077,
+      -0.0406218096613884,
+      -0.02652939409017563,
+      -0.025999585166573524,
+      -0.04552473500370979,
+      -0.031129708513617516,
+      -0.025714172050356865,
+      -0.02223960869014263,
+      -0.027326716110110283,
+      -0.01689782179892063,
+      -0.019511861726641655,
+      -0.020856689661741257,
+      -0.02208787202835083,
+      -0.04520589858293533,
+      -0.024929175153374672,
+      -0.05229552835226059,
+      -0.011622090823948383,
+      -0.02474926970899105,
+      -0.030646927654743195,
+      -0.01971360854804516,
+      -0.02455708757042885,
+      -0.02074458636343479,
+      -0.02322465553879738,
+      -0.013449409045279026,
+      -0.011075416579842567,
+      -0.009047647938132286,
+      -0.011344228871166706,
+      -0.002775580622255802,
+      0.006502024829387665,
+      -0.014414124190807343,
+      -0.02149597369134426,
+      -0.011825395748019218,
+      -0.009671507403254509,
+      -0.008267457596957684,
+      -0.016983015462756157,
+      -0.0038351379334926605,
+      -0.01598292402923107,
+      -0.015295109711587429,
+      -0.002731711370870471,
+      -0.0010551607701927423,
+      -0.006713584065437317,
+      -0.009263768792152405,
+      -0.002401233185082674,
+      -0.005693664308637381,
+      -0.01627790927886963,
+      -0.0029315929859876633,
+      -0.0003119634639006108,
+      -0.00217966316267848,
+      -0.0014510895125567913,
+      -0.0018381474073976278,
+      -0.006692532449960709,
+      -0.008143601939082146,
+      -0.0018416679231449962,
+      -0.0002660305181052536,
+      -0.01086888276040554,
+      -0.002692815847694874,
+      -0.002999652875587344,
+      -0.02033308334648609,
+      -0.0001986423449125141,
+      -8.62046581460163e-05,
+      -0.0006662025698460639,
+      -0.0005774407181888819,
+      -0.0013461634516716003,
+      -0.000865207810420543,
+      -0.00045342743396759033,
+      -0.00039656684384681284,
+      -0.0008164274040609598,
+      -0.00023399226483888924,
+      -0.0004711582150775939,
+      -0.000249892589636147,
+      -0.0005257821758277714
+    ],
+    "abs_loss": [
+      3.3791351318359375,
+      3.226520538330078,
+      2.748342275619507,
+      2.7246789932250977,
+      2.734380006790161,
+      2.763914108276367,
+      2.776501417160034,
+      2.782012939453125,
+      2.7223784923553467,
+      2.7750370502471924,
+      2.712878465652466,
+      2.5715386867523193,
+      2.417691469192505,
+      2.2169277667999268,
+      1.8418618440628052,
+      1.4758011102676392,
+      1.445372462272644,
+      1.4330517053604126,
+      1.1268384456634521,
+      1.0863100290298462,
+      0.946248471736908,
+      0.8062345385551453,
+      0.8195869326591492,
+      0.7489545941352844,
+      0.8105387687683105,
+      0.7428224682807922,
+      0.6433082222938538,
+      0.6014466285705566,
+      0.5794520974159241,
+      0.42335280776023865,
+      0.38783881068229675,
+      0.37080153822898865,
+      0.29639407992362976,
+      0.276778906583786,
+      0.2690867781639099,
+      0.23475337028503418,
+      0.24031853675842285,
+      0.18818873167037964,
+      0.1650061011314392,
+      0.146784707903862,
+      0.15803326666355133,
+      0.17478173971176147,
+      0.14593462646007538,
+      0.11577823758125305,
+      0.1437772810459137,
+      0.1562107652425766,
+      0.11654581874608994,
+      0.09483480453491211,
+      0.10927138477563858,
+      0.1195492148399353,
+      0.09249719232320786,
+      0.11310684680938721,
+      0.10394057631492615,
+      0.08904767036437988,
+      0.11636340618133545,
+      0.07780475169420242,
+      0.05981302261352539,
+      0.06853172928094864,
+      0.05035331845283508,
+      0.0653398260474205,
+      0.06224076822400093,
+      0.0763515830039978,
+      0.05397680774331093,
+      0.03979582339525223,
+      0.0471830852329731,
+      0.04971422627568245,
+      0.04336335137486458,
+      0.04716387018561363,
+      0.04091737046837807,
+      0.04822436347603798,
+      0.03870323672890663,
+      0.05136945843696594,
+      0.04786120355129242,
+      0.04661177471280098,
+      0.037042561918497086,
+      0.03333150967955589,
+      0.038183193653821945,
+      0.029593804851174355,
+      0.036583200097084045,
+      0.030410559847950935,
+      0.03435201197862625,
+      0.023689428344368935,
+      0.023031366989016533,
+      0.03888919949531555,
+      0.0251901987940073,
+      0.019065313041210175,
+      0.024439163506031036,
+      0.03290224075317383,
+      0.02057739906013012,
+      0.01778118684887886,
+      0.03099118359386921,
+      0.016362473368644714,
+      0.015957167372107506,
+      0.020016077905893326,
+      0.018773702904582024,
+      0.021185478195548058,
+      0.01584002375602722,
+      0.01608959771692753,
+      0.012648460455238819,
+      0.02667520008981228,
+      0.01431312132626772,
+      0.01208517700433731,
+      0.009955878369510174,
+      0.015077034942805767,
+      0.01430173497647047,
+      0.011124823242425919,
+      0.0111143933609128,
+      0.017690075561404228,
+      0.012765412218868732,
+      0.014279651455581188,
+      0.016503993421792984,
+      0.008516673929989338,
+      0.01265427004545927,
+      0.011229410767555237,
+      0.009121122770011425,
+      0.008324752561748028,
+      0.009793217293918133,
+      0.01735287718474865,
+      0.011216825805604458,
+      0.01033524889498949,
+      0.008808014914393425,
+      0.009041398763656616,
+      0.009020754136145115,
+      0.008437370881438255,
+      0.007728166412562132,
+      0.00990037526935339,
+      0.004430749919265509,
+      0.002050023293122649,
+      0.0034150404389947653,
+      0.005411065649241209,
+      0.002684989245608449,
+      0.003553754882887006,
+      0.002668159082531929,
+      0.004461775068193674,
+      0.0015429150080308318,
+      0.0036014870274811983,
+      0.0015474098036065698,
+      0.0015115175629034638,
+      0.0032429725397378206,
+      0.0029643059242516756
+    ],
+    "zipf_loss": [
+      6.096660614013672,
+      3.0216193199157715,
+      1.1079144477844238,
+      0.8584577441215515,
+      0.8447268009185791,
+      0.8445221185684204,
+      0.8384208679199219,
+      0.836836576461792,
+      0.8415907621383667,
+      0.8431114554405212,
+      0.8413488864898682,
+      0.8400747179985046,
+      0.8227418661117554,
+      0.7981064319610596,
+      0.7291973829269409,
+      0.6602441072463989,
+      0.6061134934425354,
+      0.5841646194458008,
+      0.5259760022163391,
+      0.4913484454154968,
+      0.4189065098762512,
+      0.39345496892929077,
+      0.34449493885040283,
+      0.3470086455345154,
+      0.3458594083786011,
+      0.33897310495376587,
+      0.2877923846244812,
+      0.27711623907089233,
+      0.20436091721057892,
+      0.2219478338956833,
+      0.14254306256771088,
+      0.13022243976593018,
+      0.09485892951488495,
+      0.11160877346992493,
+      0.1104343831539154,
+      0.08637912571430206,
+      0.07919713109731674,
+      0.07661852240562439,
+      0.07685010135173798,
+      0.07177796214818954,
+      0.08191970735788345,
+      0.07486019283533096,
+      0.05786818265914917,
+      0.07439438998699188,
+      0.06588367372751236,
+      0.040220167487859726,
+      0.06548376381397247,
+      0.0668281838297844,
+      0.05891464650630951,
+      0.04942414164543152,
+      0.047688212245702744,
+      0.04744149371981621,
+      0.04269491136074066,
+      0.03331145644187927,
+      0.031559787690639496,
+      0.03323303163051605,
+      0.032001838088035583,
+      0.026231151074171066,
+      0.031076692044734955,
+      0.0467931292951107,
+      0.03033972904086113,
+      0.02189948782324791,
+      0.03687412291765213,
+      0.02876521274447441,
+      0.01940906047821045,
+      0.030548758804798126,
+      0.02594112604856491,
+      0.017772624269127846,
+      0.029773566871881485,
+      0.0215754471719265,
+      0.025409821420907974,
+      0.030210105702280998,
+      0.026356801390647888,
+      0.02681579440832138,
+      0.025273678824305534,
+      0.013666437938809395,
+      0.01964213326573372,
+      0.021178971976041794,
+      0.024462364614009857,
+      0.01623188890516758,
+      0.031845346093177795,
+      0.01793096587061882,
+      0.018547624349594116,
+      0.02520833909511566,
+      0.015166297554969788,
+      0.023529723286628723,
+      0.019764218479394913,
+      0.023304790258407593,
+      0.0221687164157629,
+      0.009310144931077957,
+      0.025749940425157547,
+      0.021359212696552277,
+      0.03811759129166603,
+      0.019252443686127663,
+      0.018562600016593933,
+      0.019488658756017685,
+      0.026913905516266823,
+      0.017620697617530823,
+      0.023938598111271858,
+      0.01437095832079649,
+      0.015183642506599426,
+      0.02061663195490837,
+      0.0314711257815361,
+      0.01461365632712841,
+      0.014332099817693233,
+      0.00978973601013422,
+      0.021664708852767944,
+      0.020017333328723907,
+      0.019807031378149986,
+      0.02138802781701088,
+      0.009702291339635849,
+      0.01098092831671238,
+      0.013460641726851463,
+      0.02034195140004158,
+      0.024108827114105225,
+      0.015983890742063522,
+      0.012210993096232414,
+      0.0277218259871006,
+      0.025728542357683182,
+      0.012481008656322956,
+      0.013717934489250183,
+      0.01566963642835617,
+      0.012562709860503674,
+      0.013949854299426079,
+      0.021203354001045227,
+      0.016713494434952736,
+      0.01950472593307495,
+      0.016300860792398453,
+      0.009569604881107807,
+      0.022731471806764603,
+      0.011607903055846691,
+      0.013922966085374355,
+      0.020469749346375465,
+      0.019787374883890152,
+      0.016102563589811325,
+      0.0229356549680233,
+      0.0203893780708313,
+      0.014756296761333942,
+      0.01218614261597395,
+      0.015665316954255104
+    ],
+    "denoise_loss": [],
+    "ortho_loss": [
+      0.639721691608429,
+      0.23384006321430206,
+      0.14489184319972992,
+      0.10792448371648788,
+      0.08456820994615555,
+      0.08123986423015594,
+      0.0778353363275528,
+      0.11002214252948761,
+      0.12268389016389847,
+      0.15539494156837463,
+      0.16700926423072815,
+      0.17942556738853455,
+      0.18666550517082214,
+      0.19445793330669403,
+      0.2032562643289566,
+      0.20760567486286163,
+      0.21392135322093964,
+      0.21670381724834442,
+      0.2180689126253128,
+      0.22063450515270233,
+      0.21883326768875122,
+      0.21863742172718048,
+      0.217630073428154,
+      0.21256931126117706,
+      0.22088980674743652,
+      0.22135354578495026,
+      0.2220180779695511,
+      0.23056040704250336,
+      0.23146554827690125,
+      0.24043810367584229,
+      0.24833738803863525,
+      0.24956285953521729,
+      0.25005948543548584,
+      0.2468118816614151,
+      0.2451467216014862,
+      0.24257886409759521,
+      0.24182473123073578,
+      0.2443898469209671,
+      0.24258306622505188,
+      0.24195465445518494,
+      0.23636561632156372,
+      0.23654253780841827,
+      0.2325715571641922,
+      0.2336747646331787,
+      0.23286280035972595,
+      0.22832824289798737,
+      0.22483302652835846,
+      0.22698204219341278,
+      0.22653689980506897,
+      0.2247782200574875,
+      0.223842054605484,
+      0.22841455042362213,
+      0.22809086740016937,
+      0.23049242794513702,
+      0.22902648150920868,
+      0.2248598039150238,
+      0.22201740741729736,
+      0.2185370773077011,
+      0.22141091525554657,
+      0.2136271744966507,
+      0.22246746718883514,
+      0.2175038456916809,
+      0.22038504481315613,
+      0.21444186568260193,
+      0.2150769978761673,
+      0.21607351303100586,
+      0.2178366482257843,
+      0.22308430075645447,
+      0.22000181674957275,
+      0.2187521755695343,
+      0.21133314073085785,
+      0.21594126522541046,
+      0.22053566575050354,
+      0.22134697437286377,
+      0.21547871828079224,
+      0.20994195342063904,
+      0.20882248878479004,
+      0.2071162313222885,
+      0.20842047035694122,
+      0.20825868844985962,
+      0.2067190557718277,
+      0.2005504071712494,
+      0.19398145377635956,
+      0.19522309303283691,
+      0.19257421791553497,
+      0.19569028913974762,
+      0.19560210406780243,
+      0.20005181431770325,
+      0.20596961677074432,
+      0.2090260088443756,
+      0.20348960161209106,
+      0.21076665818691254,
+      0.21390055119991302,
+      0.21353191137313843,
+      0.2112823873758316,
+      0.21168112754821777,
+      0.2081877738237381,
+      0.2079249769449234,
+      0.20868492126464844,
+      0.20853272080421448,
+      0.21011212468147278,
+      0.21035552024841309,
+      0.21197006106376648,
+      0.21510401368141174,
+      0.21735182404518127,
+      0.21705609560012817,
+      0.21730081737041473,
+      0.2183588296175003,
+      0.22225020825862885,
+      0.22318315505981445,
+      0.22398485243320465,
+      0.22349779307842255,
+      0.2245921492576599,
+      0.22917932271957397,
+      0.23140773177146912,
+      0.23064255714416504,
+      0.23163443803787231,
+      0.23274460434913635,
+      0.2331482321023941,
+      0.2337607890367508,
+      0.23383651673793793,
+      0.23467832803726196,
+      0.23464395105838776,
+      0.235525444149971,
+      0.235772043466568,
+      0.23610159754753113,
+      0.2369203120470047,
+      0.23657864332199097,
+      0.23652523756027222,
+      0.23683376610279083,
+      0.23720180988311768,
+      0.23807238042354584,
+      0.23984350264072418,
+      0.24047333002090454,
+      0.24087047576904297,
+      0.24072767794132233,
+      0.24080558121204376,
+      0.24085009098052979,
+      0.2407883256673813,
+      0.2406560778617859
+    ],
+    "lr": [
+      1.6752136752136756e-05,
+      3.384615384615385e-05,
+      5.094017094017095e-05,
+      6.803418803418804e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      8e-05,
+      7.987186923279727e-05,
+      7.868547324017929e-05,
+      7.74990772475613e-05,
+      7.631268125494332e-05,
+      7.512628526232533e-05,
+      7.393988926970736e-05,
+      7.178064856314262e-05,
+      7.059425257052466e-05,
+      6.940785657790668e-05,
+      6.822146058528868e-05,
+      6.703506459267071e-05,
+      6.584866860005273e-05,
+      6.466227260743475e-05,
+      6.250303190087003e-05,
+      6.131663590825203e-05,
+      6.0130239915634074e-05,
+      5.894384392301608e-05,
+      5.77574479303981e-05,
+      5.657105193778013e-05,
+      5.538465594516214e-05,
+      5.3225415238597426e-05,
+      5.203901924597944e-05,
+      5.0852623253361464e-05,
+      4.966622726074349e-05,
+      4.8479831268125495e-05,
+      4.729343527550752e-05,
+      4.610703928288953e-05,
+      4.3947798576324816e-05,
+      4.276140258370682e-05,
+      4.157500659108885e-05,
+      4.038861059847088e-05,
+      3.9202214605852886e-05,
+      3.801581861323492e-05,
+      3.6829422620616924e-05,
+      3.46701819140522e-05,
+      3.348378592143423e-05,
+      3.229738992881624e-05,
+      3.111099393619827e-05,
+      2.9924597943580276e-05,
+      2.8738201950962302e-05,
+      2.7551805958344328e-05,
+      2.5392565251779594e-05,
+      2.4206169259161623e-05,
+      2.3019773266543632e-05,
+      2.1833377273925658e-05,
+      2.0646981281307667e-05,
+      1.9460585288689693e-05,
+      1.8274189296071722e-05,
+      1.6114948589506984e-05,
+      1.4928552596889012e-05,
+      1.374215660427102e-05,
+      1.2555760611653047e-05,
+      1.1369364619035074e-05,
+      1.0182968626417082e-05,
+      8.99657263379911e-06
+    ],
+    "emb_lr": [],
+    "eval_step": [
+      350,
+      741,
+      1132,
+      1523,
+      1914,
+      2305,
+      2696,
+      3087,
+      3478,
+      3869,
+      4260,
+      4651,
+      5042,
+      5433,
+      5824,
+      6215,
+      6606,
+      6997,
+      7388,
+      7779
+    ],
+    "eval_accuracy": [
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0,
+      0.0
+    ]
+  },
+  "final_accuracy": 1.0,
+  "sft_eval": {
+    "config": {
+      "ops": "add_sub",
+      "K": null,
+      "mode": "sft",
+      "n_digits": 6,
+      "n_per_split": 100
+    },
+    "splits": {
+      "add_S0": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 605
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 95
+          }
+        }
+      },
+      "add_S1": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 204
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 169
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 31
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 296
+          }
+        }
+      },
+      "add_S2": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 163
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 130
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 87
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 203
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 117
+          }
+        }
+      },
+      "add_S3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 121
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 121
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 49
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 186
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 223
+          }
+        }
+      },
+      "add_S4": {
+        "full_accuracy": 0.93,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 104
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 106
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 23
+          },
+          "UC": {
+            "accuracy": 0.95625,
+            "count": 160
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 307
+          }
+        }
+      },
+      "add_S5": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 400
+          }
+        }
+      },
+      "add_S6": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 500
+          }
+        }
+      },
+      "add_random": {
+        "full_accuracy": 1.0,
+        "n_examples": 200,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 447
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 320
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 56
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 529
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 48
+          }
+        }
+      },
+      "add_C3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 300
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 193
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 107
+          }
+        }
+      },
+      "add_C4": {
+        "full_accuracy": 0.97,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 200
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 0.98828125,
+            "count": 256
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 144
+          }
+        }
+      },
+      "add_C5": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 306
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 194
+          }
+        }
+      },
+      "add_C6": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 366
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 234
+          }
+        }
+      },
+      "sub_M0": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 601
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 99
+          }
+        }
+      },
+      "sub_M1": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 279
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 145
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 24
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 252
+          }
+        }
+      },
+      "sub_M2": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 213
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 113
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 85
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 181
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 108
+          }
+        }
+      },
+      "sub_M3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 179
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 103
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 56
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 149
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 213
+          }
+        }
+      },
+      "sub_M4": {
+        "full_accuracy": 0.85,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 200
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 0.85,
+            "count": 100
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 300
+          }
+        }
+      },
+      "sub_M5": {
+        "full_accuracy": 0.76,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 0.76,
+            "count": 100
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 400
+          }
+        }
+      },
+      "sub_random": {
+        "full_accuracy": 1.0,
+        "n_examples": 200,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 600
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 267
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 53
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 439
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 41
+          }
+        }
+      },
+      "sub_B3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 300
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 197
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 103
+          }
+        }
+      },
+      "sub_B4": {
+        "full_accuracy": 0.95,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 200
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 0.979757085020243,
+            "count": 247
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 153
+          }
+        }
+      },
+      "sub_B5": {
+        "full_accuracy": 0.98,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 0.9932885906040269,
+            "count": 298
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 202
+          }
+        }
+      }
+    },
+    "summary": {
+      "overall_accuracy": 0.9766666666666667,
+      "total_examples": 2400,
+      "n_splits": 22
+    }
+  },
+  "sorl_eval": {
+    "config": {
+      "ops": "add_sub",
+      "K": 1,
+      "mode": "sorl",
+      "n_digits": 6,
+      "n_per_split": 100
+    },
+    "splits": {
+      "add_S0": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 605
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 95
+          }
+        }
+      },
+      "add_S1": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 204
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 169
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 31
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 296
+          }
+        }
+      },
+      "add_S2": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 163
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 130
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 87
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 203
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 117
+          }
+        }
+      },
+      "add_S3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 121
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 121
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 49
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 186
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 223
+          }
+        }
+      },
+      "add_S4": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 104
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 106
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 23
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 160
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 307
+          }
+        }
+      },
+      "add_S5": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 400
+          }
+        }
+      },
+      "add_S6": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 500
+          }
+        }
+      },
+      "add_random": {
+        "full_accuracy": 1.0,
+        "n_examples": 200,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 447
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 320
+          },
+          "SS": {
+            "accuracy": 1.0,
+            "count": 56
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 529
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 48
+          }
+        }
+      },
+      "add_C3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 300
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 193
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 107
+          }
+        }
+      },
+      "add_C4": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 200
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 256
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 144
+          }
+        }
+      },
+      "add_C5": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SA": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 306
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 194
+          }
+        }
+      },
+      "add_C6": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "SC": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UC": {
+            "accuracy": 1.0,
+            "count": 366
+          },
+          "US": {
+            "accuracy": 1.0,
+            "count": 234
+          }
+        }
+      },
+      "sub_M0": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 601
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 99
+          }
+        }
+      },
+      "sub_M1": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 279
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 145
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 24
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 252
+          }
+        }
+      },
+      "sub_M2": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 213
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 113
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 85
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 181
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 108
+          }
+        }
+      },
+      "sub_M3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 179
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 103
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 56
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 149
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 213
+          }
+        }
+      },
+      "sub_M4": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 200
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 300
+          }
+        }
+      },
+      "sub_M5": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 400
+          }
+        }
+      },
+      "sub_random": {
+        "full_accuracy": 1.0,
+        "n_examples": 200,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 600
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 267
+          },
+          "ME": {
+            "accuracy": 1.0,
+            "count": 53
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 439
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 41
+          }
+        }
+      },
+      "sub_B3": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 300
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 197
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 103
+          }
+        }
+      },
+      "sub_B4": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 200
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 247
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 153
+          }
+        }
+      },
+      "sub_B5": {
+        "full_accuracy": 1.0,
+        "n_examples": 100,
+        "per_subtask": {
+          "MD": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "MB": {
+            "accuracy": 1.0,
+            "count": 100
+          },
+          "UB": {
+            "accuracy": 1.0,
+            "count": 298
+          },
+          "UD": {
+            "accuracy": 1.0,
+            "count": 202
+          }
+        }
+      }
+    },
+    "summary": {
+      "overall_accuracy": 1.0,
+      "total_examples": 2400,
+      "n_splits": 22
+    }
+  },
+  "sorl_overall_accuracy": 1.0,
+  "sft_overall_accuracy": 0.9766666666666667
+}

add_sub_sorl_v1_abs30_K1_25K/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10a719b33a7daab70d7e4cf7bc8d44d7cef4ff6fbfd976546301e57fbbf8644e
+size 650385300

add_sub_sorl_v1_abs30_K1_25K/train_config.json

@@ -0,0 +1,84 @@
+{
+  "num_rollouts": 4,
+  "K": 1,
+  "max_iterations": 2,
+  "memory_span_abs": 1792,
+  "memory_span_traj": 1792,
+  "temperature": 1.0,
+  "ar_search": false,
+  "response_only_abs": false,
+  "alpha_info_gain": 10.0,
+  "alpha_abs": 0.1,
+  "alpha_soft_zipf": 1.0,
+  "alpha_ortho": 0.0,
+  "alpha_anchor": 0.0,
+  "alpha_jacobi": 0.0,
+  "decay": 0.8,
+  "target_vocab_util": 0.8,
+  "min_abs_ppl": 0.0,
+  "zipf_alpha": 1.0,
+  "lr": 8e-05,
+  "emb_lr_mult": 1.0,
+  "weight_decay": 0.01,
+  "warmup_steps": 234,
+  "cooldown_frac": 0.4,
+  "max_grad_norm": 1.0,
+  "vq_abs_pretrain_steps": 0,
+  "vq_abs_pretrain_lr": 0.001,
+  "vq_abs_pretrain_layer": -1,
+  "vq_abs_pretrain_batch_size": 256,
+  "vq_abs_pretrain_target_vectors": 20000,
+  "batch_size": 64,
+  "gradient_accumulation_steps": 1,
+  "num_epochs": 20,
+  "emb_warmup_steps": 0,
+  "log_every": 50,
+  "eval_every": 390,
+  "save_every": 999999,
+  "eval_samples": 100,
+  "output_dir": "ckpt/sweep/as_sorl_abs30_K1_25K",
+  "eval_K": 4,
+  "alpha_traj": 0.0,
+  "corrupt_method": "shuffle",
+  "corrupt_ratio": 0.3,
+  "alpha_contrastive": 1.0,
+  "gamma_contrastive": 0.5,
+  "alpha_masked_traj": 0.0,
+  "mask_nl_ratio": 0.3,
+  "mask_nl_mode": "fixed",
+  "mask_nl_fixed_id": 0,
+  "use_ste": true,
+  "n_inner": 1,
+  "random_K": null,
+  "strip_suffix": null,
+  "compress_prefix": null,
+  "random_mem_span": null,
+  "warmup_ratio": 0.03,
+  "beta2": 0.999,
+  "seed": 42,
+  "n_digits": 6,
+  "n_layer": 2,
+  "n_head": 3,
+  "n_embd": 510,
+  "ops": "add_sub",
+  "abs_vocab": 30,
+  "dataset_size": 25000,
+  "mode": "sorl",
+  "device": "cuda",
+  "push_to_hub": true,
+  "no_wandb": false,
+  "n_params": 162519662,
+  "run_name": "add_sub_sorl_v1_abs30_K1_25K",
+  "git_commit": "57deaa28d9c21e39ddac5ef448d6e1be992fba91",
+  "timestamp": "2026-04-13T09:57:41.955984+00:00",
+  "tokenizer": "Qwen/Qwen3-0.6B",
+  "dataset_repo": "thoughtworks/arithmetic-sorl-data",
+  "dataset_config": "add_sub_6digit",
+  "model_repo": "thoughtworks/arithmetic-sorl",
+  "trainer_version": "v1",
+  "wandb_run_id": "cf3zi3ax",
+  "wandb_url": "https://wandb.ai/nlp_and_interpretability/sorl-arithmetic/runs/cf3zi3ax",
+  "final_accuracy": 1.0,
+  "sft_accuracy": 0.9766666666666667,
+  "eval_method": "ArithmeticEvaluator"
+}