Upload add_sub_sorl_v1_abs5_K1_50K

add_sub_sorl_v1_abs5_K1_50K/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": 151649
+}

add_sub_sorl_v1_abs5_K1_50K/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_abs5_K1_50K/metrics.json

@@ -0,0 +1,2257 @@
+{
+  "history": {
+    "step": [
+      50,
+      100,
+      150,
+      200,
+      250,
+      300,
+      350,
+      400,
+      450,
+      500,
+      550,
+      600,
+      650,
+      700,
+      750,
+      832,
+      882,
+      932,
+      982,
+      1032,
+      1082,
+      1132,
+      1182,
+      1232,
+      1282,
+      1332,
+      1382,
+      1432,
+      1482,
+      1532,
+      1614,
+      1664,
+      1714,
+      1764,
+      1814,
+      1864,
+      1914,
+      1964,
+      2014,
+      2064,
+      2114,
+      2164,
+      2214,
+      2264,
+      2314,
+      2396,
+      2446,
+      2496,
+      2546,
+      2596,
+      2646,
+      2696,
+      2746,
+      2796,
+      2846,
+      2896,
+      2946,
+      2996,
+      3046,
+      3096,
+      3178,
+      3228,
+      3278,
+      3328,
+      3378,
+      3428,
+      3478,
+      3528,
+      3578,
+      3628,
+      3678,
+      3728,
+      3778,
+      3828,
+      3878,
+      3960,
+      4010,
+      4060,
+      4110,
+      4160,
+      4210,
+      4260,
+      4310,
+      4360,
+      4410,
+      4460,
+      4510,
+      4560,
+      4610,
+      4660,
+      4742,
+      4792,
+      4842,
+      4892,
+      4942,
+      4992,
+      5042,
+      5092,
+      5142,
+      5192,
+      5242,
+      5292,
+      5342,
+      5392,
+      5442,
+      5524,
+      5574,
+      5624,
+      5674,
+      5724,
+      5774,
+      5824,
+      5874,
+      5924,
+      5974,
+      6024,
+      6074,
+      6124,
+      6174,
+      6224,
+      6306,
+      6356,
+      6406,
+      6456,
+      6506,
+      6556,
+      6606,
+      6656,
+      6706,
+      6756,
+      6806,
+      6856,
+      6906,
+      6956,
+      7006,
+      7088,
+      7138,
+      7188,
+      7238,
+      7288,
+      7338,
+      7388,
+      7438,
+      7488,
+      7538,
+      7588,
+      7638,
+      7688,
+      7738,
+      7788
+    ],
+    "loss": [
+      0.2505636215209961,
+      8.46504020690918,
+      8.983885765075684,
+      5.396197319030762,
+      4.661125183105469,
+      4.198425769805908,
+      3.9200804233551025,
+      4.002684116363525,
+      3.967268705368042,
+      3.5404810905456543,
+      2.8022069931030273,
+      1.8075677156448364,
+      -0.4370992183685303,
+      -3.252620220184326,
+      -6.226922035217285,
+      -6.981635093688965,
+      -8.204410552978516,
+      -8.275205612182617,
+      -9.201150894165039,
+      -8.603435516357422,
+      -9.279670715332031,
+      -8.784062385559082,
+      -8.871679306030273,
+      -6.306247711181641,
+      -5.68474006652832,
+      -3.962198495864868,
+      -3.0238935947418213,
+      -2.9744505882263184,
+      -2.6373329162597656,
+      -2.1695408821105957,
+      -1.080020785331726,
+      -1.360289454460144,
+      -1.0075311660766602,
+      -1.0235530138015747,
+      -0.46744096279144287,
+      -0.3752056360244751,
+      0.04357093572616577,
+      -0.09943628311157227,
+      -0.4955475330352783,
+      -0.2734338641166687,
+      -0.3429461717605591,
+      -0.0041134655475616455,
+      0.04709044098854065,
+      0.1768924593925476,
+      -0.27656251192092896,
+      -0.13877591490745544,
+      -0.12509974837303162,
+      -0.24758702516555786,
+      -0.5813546180725098,
+      -0.5284724831581116,
+      -0.3846120834350586,
+      -0.47701379656791687,
+      -0.459797739982605,
+      -0.1737753003835678,
+      -0.3913196623325348,
+      -0.46543970704078674,
+      -0.4292522370815277,
+      -0.5559454560279846,
+      -0.21943989396095276,
+      -0.34869182109832764,
+      -0.40362223982810974,
+      -0.2524590492248535,
+      -0.2675473988056183,
+      -0.1487952321767807,
+      -0.307020366191864,
+      -0.32443857192993164,
+      -0.6750359535217285,
+      -0.22530588507652283,
+      -0.30566859245300293,
+      -0.33527910709381104,
+      -0.2065487504005432,
+      -0.14366477727890015,
+      -0.13279062509536743,
+      -0.07906606048345566,
+      -0.09167033433914185,
+      -0.0976349413394928,
+      -0.10151524096727371,
+      -0.0636506974697113,
+      -0.11694946885108948,
+      -0.04894915968179703,
+      -0.047724585980176926,
+      -0.011612199246883392,
+      0.018235700204968452,
+      -0.05753451958298683,
+      -0.040516600012779236,
+      -0.009241508319973946,
+      -0.014481740072369576,
+      -0.00770045118406415,
+      -0.14257672429084778,
+      -0.03606821969151497,
+      -0.009480145759880543,
+      -0.015739701688289642,
+      -0.011609197594225407,
+      -0.03947978466749191,
+      0.06050138548016548,
+      -0.01320057176053524,
+      -0.027511538937687874,
+      -0.016208164393901825,
+      -0.0333053320646286,
+      -0.024525409564375877,
+      -0.01065415795892477,
+      -0.02119392529129982,
+      -0.04003360867500305,
+      -0.0028016045689582825,
+      0.000910859671421349,
+      -0.13593541085720062,
+      0.001402085181325674,
+      -0.005520656704902649,
+      -0.046631112694740295,
+      -0.009012637659907341,
+      -0.06137138605117798,
+      -0.00438198447227478,
+      -0.05739019811153412,
+      0.0014041318790987134,
+      0.0018914949614554644,
+      -0.015167666599154472,
+      -0.005372755695134401,
+      -0.03327453136444092,
+      -0.008505746722221375,
+      -0.00026095949579030275,
+      -0.00024359801318496466,
+      0.0003202931256964803,
+      -0.09214256703853607,
+      -0.0035724681802093983,
+      -0.0005808689165860415,
+      0.00034972524736076593,
+      -0.014495953917503357,
+      0.0004223707364872098,
+      -0.02243591845035553,
+      -0.01563563011586666,
+      -0.015591835603117943,
+      0.00010260124690830708,
+      -0.0015123512130230665,
+      0.002046036534011364,
+      -0.0009971989784389734,
+      -0.027398593723773956,
+      0.00042443181155249476,
+      0.0009379229159094393,
+      0.0023569948971271515,
+      0.00271740835160017,
+      -0.004399539437144995,
+      0.00021369947353377938,
+      0.0007292578229680657,
+      0.0008549960912205279,
+      0.000461005256511271,
+      0.0010421416955068707,
+      -0.00028817541897296906,
+      0.0004380897735245526,
+      5.740142660215497e-05,
+      0.0007358812727034092
+    ],
+    "base_loss": [
+      10.585969924926758,
+      7.861696720123291,
+      6.595737457275391,
+      4.664162635803223,
+      2.468219041824341,
+      2.0123753547668457,
+      1.9060745239257812,
+      1.9411739110946655,
+      1.8617933988571167,
+      1.7926157712936401,
+      1.8101921081542969,
+      1.7342417240142822,
+      1.840061902999878,
+      1.817416787147522,
+      1.8616607189178467,
+      1.7341166734695435,
+      1.7626491785049438,
+      1.7099848985671997,
+      1.7449618577957153,
+      1.6113771200180054,
+      1.5788114070892334,
+      1.4948910474777222,
+      1.4506797790527344,
+      1.1060410737991333,
+      1.0262805223464966,
+      0.8133479356765747,
+      0.707499086856842,
+      0.6684346199035645,
+      0.5689376592636108,
+      0.5056790709495544,
+      0.38431841135025024,
+      0.3539868891239166,
+      0.3170618414878845,
+      0.3117201626300812,
+      0.21727757155895233,
+      0.19561822712421417,
+      0.12323100864887238,
+      0.09690731763839722,
+      0.12097124755382538,
+      0.08452977240085602,
+      0.09661463648080826,
+      0.04839548468589783,
+      0.043548472225666046,
+      0.03469983488321304,
+      0.07495895773172379,
+      0.05109835043549538,
+      0.06454946845769882,
+      0.04492994770407677,
+      0.08394144475460052,
+      0.06295685470104218,
+      0.053767867386341095,
+      0.05481096729636192,
+      0.057550448924303055,
+      0.02092977985739708,
+      0.04481356963515282,
+      0.05254203453660011,
+      0.04828425124287605,
+      0.062493324279785156,
+      0.04113643616437912,
+      0.039543312042951584,
+      0.04525873064994812,
+      0.02862536907196045,
+      0.030017653480172157,
+      0.0259102750569582,
+      0.03472105786204338,
+      0.03658299520611763,
+      0.075889952480793,
+      0.025799283757805824,
+      0.03533393517136574,
+      0.03777829185128212,
+      0.023547964170575142,
+      0.01642688177525997,
+      0.015161960385739803,
+      0.008982157334685326,
+      0.010374007746577263,
+      0.011107687838375568,
+      0.011426926590502262,
+      0.00722096860408783,
+      0.01364094763994217,
+      0.005588749889284372,
+      0.005449043121188879,
+      0.0015719280345365405,
+      0.0039480640552937984,
+      0.006539866793900728,
+      0.004745051264762878,
+      0.0011512581259012222,
+      0.0018374177161604166,
+      0.0009692967869341373,
+      0.015946198254823685,
+      0.0041474527679383755,
+      0.0012120535830035806,
+      0.0018659065244719386,
+      0.0014860676601529121,
+      0.004629795905202627,
+      0.005724634509533644,
+      0.0016615678323432803,
+      0.0031864303164184093,
+      0.0019036593148484826,
+      0.0038307576905936003,
+      0.0029207405168563128,
+      0.0015083644539117813,
+      0.0025551484432071447,
+      0.004551176447421312,
+      0.0004310809599701315,
+      9.97485694824718e-05,
+      0.015383963473141193,
+      7.338295836234465e-05,
+      0.0007650703191757202,
+      0.005420635920017958,
+      0.001144574023783207,
+      0.006923999171704054,
+      0.0005839763907715678,
+      0.006508989725261927,
+      4.6668767026858404e-05,
+      0.00011568832269404083,
+      0.0018093654653057456,
+      0.0007959016365930438,
+      0.0038521860260516405,
+      0.001005106489174068,
+      0.00015298707876354456,
+      0.00016213013441301882,
+      8.991890354081988e-05,
+      0.010388972237706184,
+      0.0004740772128570825,
+      0.0001657261891523376,
+      0.0001314091496169567,
+      0.0017237851861864328,
+      0.00014208715583663434,
+      0.0029368281830102205,
+      0.001902392483316362,
+      0.0021551058162003756,
+      8.371058356715366e-05,
+      0.00031782890437170863,
+      7.045956590445712e-05,
+      0.00023993372451514006,
+      0.0033891168422997,
+      3.269490480306558e-05,
+      2.2259604520513676e-05,
+      2.659805068105925e-05,
+      3.853382804663852e-05,
+      0.0007118353387340903,
+      9.246311674360186e-05,
+      4.051219002576545e-05,
+      3.327736703795381e-05,
+      0.00012879837595392019,
+      2.466775185894221e-05,
+      0.00014854212349746376,
+      1.9150504158460535e-05,
+      6.794735963921994e-05,
+      2.4384833523072302e-05
+    ],
+    "info_loss": [
+      -2.183910369873047,
+      -0.8172898292541504,
+      -0.3265523910522461,
+      -0.29645252227783203,
+      -0.09593391418457031,
+      -0.08809053897857666,
+      -0.10290849208831787,
+      -0.09725630283355713,
+      -0.09241962432861328,
+      -0.12810420989990234,
+      -0.20332062244415283,
+      -0.2951948642730713,
+      -0.5304032564163208,
+      -0.8100159168243408,
+      -1.112187385559082,
+      -1.1749544143676758,
+      -1.2994019985198975,
+      -1.2979850769042969,
+      -1.3871701955795288,
+      -1.3036789894104004,
+      -1.349029302597046,
+      -1.2852455377578735,
+      -1.2771252393722534,
+      -0.9778541326522827,
+      -0.8947479724884033,
+      -0.69999760389328,
+      -0.5798141956329346,
+      -0.5599962472915649,
+      -0.5034015774726868,
+      -0.44325190782546997,
+      -0.3099415600299835,
+      -0.3180922269821167,
+      -0.2747928202152252,
+      -0.27571025490760803,
+      -0.19737744331359863,
+      -0.16460935771465302,
+      -0.10986328125,
+      -0.09181390702724457,
+      -0.1143815740942955,
+      -0.07358857244253159,
+      -0.08351294696331024,
+      -0.041542500257492065,
+      -0.03816980496048927,
+      -0.0214642696082592,
+      -0.07020267099142075,
+      -0.049885910004377365,
+      -0.0500519797205925,
+      -0.044273801147937775,
+      -0.06719288229942322,
+      -0.05959625542163849,
+      -0.044250987470149994,
+      -0.05371290445327759,
+      -0.052004072815179825,
+      -0.020543813705444336,
+      -0.04393193870782852,
+      -0.05230949446558952,
+      -0.048010554164648056,
+      -0.0620872937142849,
+      -0.0262104831635952,
+      -0.03927260637283325,
+      -0.045116618275642395,
+      -0.028391283005475998,
+      -0.029934341087937355,
+      -0.01769854500889778,
+      -0.03448145464062691,
+      -0.036520395427942276,
+      -0.07541237026453018,
+      -0.025723034515976906,
+      -0.03477848321199417,
+      -0.037701305001974106,
+      -0.023497894406318665,
+      -0.016341710463166237,
+      -0.01498723216354847,
+      -0.008923816494643688,
+      -0.010336598381400108,
+      -0.011041379533708096,
+      -0.011384105309844017,
+      -0.007182232104241848,
+      -0.013588892295956612,
+      -0.005559454672038555,
+      -0.005420235451310873,
+      -0.0014992294600233436,
+      0.0013668490573763847,
+      -0.00651158019900322,
+      -0.004717795643955469,
+      -0.0011278484016656876,
+      -0.00181629101280123,
+      -0.0009426028118468821,
+      -0.015919113531708717,
+      -0.0041302829049527645,
+      -0.0011922880075871944,
+      -0.0018478219863027334,
+      -0.0014635310508310795,
+      -0.004584491718560457,
+      0.005315172020345926,
+      -0.001626395620405674,
+      -0.003172175493091345,
+      -0.0018896134570240974,
+      -0.0038163100834935904,
+      -0.0029021964874118567,
+      -0.001496299752034247,
+      -0.0025390409864485264,
+      -0.004537238739430904,
+      -0.00041555240750312805,
+      -8.846276614349335e-05,
+      -0.015372051857411861,
+      -6.181209755595773e-05,
+      -0.0007561057573184371,
+      -0.0054064528085291386,
+      -0.001132069039158523,
+      -0.00691541563719511,
+      -0.0005742469220422208,
+      -0.0064970338717103004,
+      -3.8318328734021634e-05,
+      -0.00010713449592003599,
+      -0.0018018593546003103,
+      -0.000784172210842371,
+      -0.0038447838742285967,
+      -0.0009972280822694302,
+      -0.0001449241244699806,
+      -0.00015577537124045193,
+      -8.330684067914262e-05,
+      -0.01038222387433052,
+      -0.00046742818085476756,
+      -0.00015831156633794308,
+      -0.00012537343718577176,
+      -0.0017165575409308076,
+      -0.00013619274250231683,
+      -0.0029318383894860744,
+      -0.001897612470202148,
+      -0.0021497313864529133,
+      -7.87463941378519e-05,
+      -0.00031296690576709807,
+      -6.56055417493917e-05,
+      -0.0002351704315515235,
+      -0.003384436247870326,
+      -2.8172687962069176e-05,
+      -1.711163167783525e-05,
+      -2.2289243133855052e-05,
+      -3.4640113881323487e-05,
+      -0.00070682488149032,
+      -8.828522550174966e-05,
+      -3.625154204200953e-05,
+      -2.956273419840727e-05,
+      -0.0001251215289812535,
+      -2.0451281670830213e-05,
+      -0.0001448101975256577,
+      -1.5480572983506136e-05,
+      -6.413826486095786e-05,
+      -2.0543287973850965e-05
+    ],
+    "abs_loss": [
+      1.6025371551513672,
+      1.5280265808105469,
+      1.3616584539413452,
+      1.1554263830184937,
+      1.090998888015747,
+      1.0798345804214478,
+      1.0821243524551392,
+      1.0772966146469116,
+      1.077260136604309,
+      1.0835458040237427,
+      1.079797387123108,
+      1.0777150392532349,
+      1.0810126066207886,
+      1.0769461393356323,
+      1.0705296993255615,
+      1.0520821809768677,
+      0.9844305515289307,
+      0.8652377128601074,
+      0.7227235436439514,
+      0.5906375050544739,
+      0.4413999617099762,
+      0.3452411890029907,
+      0.29571032524108887,
+      0.23282186686992645,
+      0.14773382246494293,
+      0.13641048967838287,
+      0.08999554067850113,
+      0.11419766396284103,
+      0.08703526109457016,
+      0.07698491960763931,
+      0.07054328173398972,
+      0.08026841282844543,
+      0.07668456435203552,
+      0.10232362896203995,
+      0.07094237953424454,
+      0.07218757271766663,
+      0.07528065890073776,
+      0.08964914083480835,
+      0.09180616587400436,
+      0.0713563784956932,
+      0.08065492659807205,
+      0.05928153917193413,
+      0.09654293209314346,
+      0.18235540390014648,
+      0.08406329900026321,
+      0.04552678391337395,
+      0.046885158866643906,
+      0.042871445417404175,
+      0.039759550243616104,
+      0.029482612386345863,
+      0.02260918915271759,
+      0.029610687866806984,
+      0.002320162719115615,
+      0.001828500535339117,
+      0.0209819208830595,
+      0.028438260778784752,
+      0.0035261947195976973,
+      0.0012173553695902228,
+      0.0015700069488957524,
+      0.013621640391647816,
+      0.0018642926588654518,
+      0.00222670822404325,
+      0.007209902163594961,
+      0.002658851444721222,
+      0.0030335213523358107,
+      0.02170608937740326,
+      0.001575362985022366,
+      0.006342296022921801,
+      0.004211599472910166,
+      0.0006940102321095765,
+      0.0012646445538848639,
+      0.0021566555369645357,
+      0.0002222745679318905,
+      0.0015051417285576463,
+      0.0026516858488321304,
+      0.0001819822209654376,
+      0.00018783057748805732,
+      0.0017949346220120788,
+      0.001176174613647163,
+      0.00023223921016324311,
+      0.0003606785030569881,
+      0.001950734294950962,
+      0.0009348828461952507,
+      0.0011603906750679016,
+      0.0030520979780703783,
+      0.0013620670652016997,
+      0.0009282492683269083,
+      0.00015765779244247824,
+      0.00037819836870767176,
+      9.389529441250488e-05,
+      0.0002264125068904832,
+      0.0009022634476423264,
+      0.0008199801668524742,
+      0.005860028322786093,
+      0.0004964592517353594,
+      0.0003333363856654614,
+      0.0016712062060832977,
+      0.0011127572506666183,
+      0.0008218744187615812,
+      0.00011704187636496499,
+      3.0456712920567952e-05,
+      4.790263483300805e-05,
+      0.0020059377420693636,
+      0.00022615592752117664,
+      0.0014674357371404767,
+      0.0002781325893010944,
+      4.343148248153739e-05,
+      0.0027780078817158937,
+      0.0004207564052194357,
+      0.0005745245143771172,
+      2.9878183340770192e-05,
+      1.2017285371257458e-05,
+      0.001584974699653685,
+      3.3227872336283326e-05,
+      9.690172009868547e-05,
+      0.00015782308764755726,
+      8.321733184857294e-05,
+      1.623890238988679e-05,
+      1.5769308447488584e-05,
+      1.3366395251068752e-05,
+      1.8728727809502743e-05,
+      2.2210122551769018e-05,
+      1.4690090210933704e-05,
+      0.0001451492280466482,
+      3.520072641549632e-05,
+      0.0012909631477668881,
+      0.0012377469101920724,
+      1.613760832697153e-05,
+      1.3538238818000536e-05,
+      1.547469037177507e-05,
+      0.00024891310022212565,
+      1.7161219147965312e-05,
+      0.003930000588297844,
+      1.9172684915247373e-05,
+      3.660873699118383e-05,
+      1.2629378943529446e-05,
+      0.00029429898131638765,
+      3.262208338128403e-05,
+      6.562576891155913e-05,
+      0.0014089597389101982,
+      0.0005585301551036537,
+      3.131064659100957e-05,
+      3.751121766981669e-05,
+      0.0006812455249018967,
+      4.927303234580904e-05,
+      0.0008097690879367292,
+      3.0214263460948132e-05,
+      0.0005298063042573631,
+      1.2007123586954549e-05,
+      1.4416099475056399e-05
+    ],
+    "zipf_loss": [
+      11.343443870544434,
+      8.623438835144043,
+      5.5175065994262695,
+      3.58101749420166,
+      3.043145179748535,
+      2.958972215652466,
+      2.934878349304199,
+      2.9263434410095215,
+      2.921945571899414,
+      2.9205527305603027,
+      2.917241334915161,
+      2.9175031185150146,
+      2.9187698364257812,
+      2.922427177429199,
+      2.9262380599975586,
+      2.928584098815918,
+      2.9285175800323486,
+      2.9081363677978516,
+      2.8533177375793457,
+      2.762913465499878,
+      2.5876712799072266,
+      2.5389773845672607,
+      2.419322967529297,
+      2.342971086502075,
+      2.2216851711273193,
+      2.2107889652252197,
+      2.0577499866485596,
+      1.9456571340560913,
+      1.819041132926941,
+      1.7496004104614258,
+      1.6280220746994019,
+      1.4586189985275269,
+      1.4156665802001953,
+      1.4115968942642212,
+      1.2819616794586182,
+      1.0680509805679321,
+      1.0114446878433228,
+      0.7128306031227112,
+      0.5181163549423218,
+      0.37078648805618286,
+      0.38750314712524414,
+      0.3569878935813904,
+      0.37558573484420776,
+      0.3385998010635376,
+      0.3420988917350769,
+      0.3044321537017822,
+      0.3061820864677429,
+      0.1459338665008545,
+      0.0026568169705569744,
+      0.001584875164553523,
+      0.0018690116703510284,
+      0.0023432141169905663,
+      0.0024605486541986465,
+      0.010550200007855892,
+      0.0010879351757466793,
+      0.0022694002836942673,
+      0.0022164294496178627,
+      0.002312395256012678,
+      0.001371507067233324,
+      0.0031287604942917824,
+      0.0020987577736377716,
+      0.002605732064694166,
+      0.0010573845356702805,
+      0.002014062600210309,
+      0.0027697691693902016,
+      0.00201174383983016,
+      0.0030402797274291515,
+      0.005490951240062714,
+      0.0063611241057515144,
+      0.0038862451910972595,
+      0.004755761474370956,
+      0.0031097866594791412,
+      0.0018975045531988144,
+      0.0010394295677542686,
+      0.001056477427482605,
+      0.0016529643908143044,
+      0.0008801072835922241,
+      0.0007711621001362801,
+      0.005180887877941132,
+      0.001033412292599678,
+      0.0009926604107022285,
+      0.0016130937729030848,
+      0.0005256567383185029,
+      0.0009253760799765587,
+      0.0016110949218273163,
+      0.0007495111785829067,
+      0.0017509274184703827,
+      0.0007405146025121212,
+      0.0006303912959992886,
+      0.0010777697898447514,
+      0.0012080390006303787,
+      0.0007823850028216839,
+      0.0014580467250198126,
+      0.001149333082139492,
+      0.0015753833577036858,
+      0.001368482131510973,
+      0.0008566654287278652,
+      0.0006730356253683567,
+      0.0009448234923183918,
+      0.0015641096979379654,
+      0.0027974294498562813,
+      0.0016365437768399715,
+      0.0005870126187801361,
+      0.0009002229198813438,
+      0.0015489952638745308,
+      0.002373320981860161,
+      0.001942479982972145,
+      0.0009975298307836056,
+      0.0019706999883055687,
+      0.001106026116758585,
+      0.000855783000588417,
+      0.0007753064855933189,
+      0.0009126532822847366,
+      0.0017373235896229744,
+      0.002837461419403553,
+      0.0010257791727781296,
+      0.0016647428274154663,
+      0.0013194996863603592,
+      0.000459850300103426,
+      0.0010339580476284027,
+      0.0011501526460051537,
+      0.0010612215846776962,
+      0.0012892335653305054,
+      0.0006132214330136776,
+      0.0008330005221068859,
+      0.0013429541140794754,
+      0.000822061556391418,
+      0.0016405973583459854,
+      0.003944283351302147,
+      0.0014365538954734802,
+      0.0037254802882671356,
+      0.0008046384900808334,
+      0.0009064888581633568,
+      0.0026297150179743767,
+      0.0011109109036624432,
+      0.003055390901863575,
+      0.0006440338911488652,
+      0.0010835174471139908,
+      0.0025467267259955406,
+      0.0028843795880675316,
+      0.0019010212272405624,
+      0.0010009575635194778,
+      0.0010475099552422762,
+      0.0010492214933037758,
+      0.001578494906425476,
+      0.0011410098522901535,
+      0.0010083629749715328,
+      0.000520764384418726,
+      0.0006296359933912754,
+      0.0009154877043329179
+    ],
+    "denoise_loss": [],
+    "ortho_loss": [
+      0.7514292597770691,
+      0.44231683015823364,
+      0.4044279158115387,
+      0.31896743178367615,
+      0.2748689353466034,
+      0.17610643804073334,
+      0.1378389149904251,
+      0.10322701930999756,
+      0.08539271354675293,
+      0.09809615463018417,
+      0.10976400226354599,
+      0.1415826380252838,
+      0.1451120674610138,
+      0.17278920114040375,
+      0.22727791965007782,
+      0.25248804688453674,
+      0.255528062582016,
+      0.27296891808509827,
+      0.3045079708099365,
+      0.317409485578537,
+      0.33449694514274597,
+      0.3505777418613434,
+      0.3575138747692108,
+      0.35384806990623474,
+      0.3518454134464264,
+      0.3580361008644104,
+      0.3517009913921356,
+      0.33349522948265076,
+      0.33289653062820435,
+      0.31936779618263245,
+      0.3168710172176361,
+      0.3141125738620758,
+      0.30932894349098206,
+      0.3887239098548889,
+      0.3971191942691803,
+      0.41927996277809143,
+      0.4253721833229065,
+      0.44365379214286804,
+      0.44678258895874023,
+      0.4558866620063782,
+      0.4441855549812317,
+      0.4194315969944,
+      0.4099842607975006,
+      0.4103780686855316,
+      0.4245625138282776,
+      0.4197509288787842,
+      0.42175063490867615,
+      0.411930650472641,
+      0.41106978058815,
+      0.4170016944408417,
+      0.41532015800476074,
+      0.41429853439331055,
+      0.41899004578590393,
+      0.42344552278518677,
+      0.4204721450805664,
+      0.42217817902565,
+      0.4267289340496063,
+      0.4268219470977783,
+      0.42429524660110474,
+      0.4305848181247711,
+      0.4328194260597229,
+      0.431699275970459,
+      0.4296744465827942,
+      0.4308277666568756,
+      0.4295274317264557,
+      0.4337575137615204,
+      0.4304613173007965,
+      0.43323755264282227,
+      0.42645522952079773,
+      0.42617306113243103,
+      0.4247611463069916,
+      0.42557570338249207,
+      0.427293986082077,
+      0.4365611672401428,
+      0.4403775632381439,
+      0.4325670897960663,
+      0.4502088725566864,
+      0.45125389099121094,
+      0.4531281590461731,
+      0.45476648211479187,
+      0.4559301435947418,
+      0.45802441239356995,
+      0.44424352049827576,
+      0.44277673959732056,
+      0.44663745164871216,
+      0.45184335112571716,
+      0.44507738947868347,
+      0.4511640965938568,
+      0.4508034288883209,
+      0.4560854136943817,
+      0.4548608362674713,
+      0.4576341211795807,
+      0.4547962248325348,
+      0.4557437598705292,
+      0.42322593927383423,
+      0.3989056348800659,
+      0.40023231506347656,
+      0.4040592610836029,
+      0.40395256876945496,
+      0.4029274880886078,
+      0.40079614520072937,
+      0.3979244828224182,
+      0.3972989022731781,
+      0.4015220105648041,
+      0.3990614116191864,
+      0.39628490805625916,
+      0.395735502243042,
+      0.39462000131607056,
+      0.3914993703365326,
+      0.3908904492855072,
+      0.388724148273468,
+      0.38905712962150574,
+      0.3822788596153259,
+      0.38399258255958557,
+      0.38412877917289734,
+      0.3843962252140045,
+      0.38461238145828247,
+      0.3833766579627991,
+      0.38389861583709717,
+      0.3849593698978424,
+      0.3849909007549286,
+      0.386178582906723,
+      0.3854115903377533,
+      0.3844901919364929,
+      0.3846789300441742,
+      0.38446399569511414,
+      0.38404810428619385,
+      0.3841705024242401,
+      0.38369980454444885,
+      0.38355860114097595,
+      0.38261696696281433,
+      0.3843592703342438,
+      0.3838559091091156,
+      0.38580289483070374,
+      0.3865111172199249,
+      0.3902479410171509,
+      0.39057838916778564,
+      0.390543133020401,
+      0.39057406783103943,
+      0.3909136652946472,
+      0.39078858494758606,
+      0.39051759243011475,
+      0.3902330994606018,
+      0.3898555338382721,
+      0.389696329832077,
+      0.3901805877685547,
+      0.390118807554245,
+      0.38976243138313293,
+      0.38940849900245667,
+      0.38956424593925476
+    ],
+    "lr": [
+      8.376068376068378e-06,
+      1.6923076923076924e-05,
+      2.5470085470085475e-05,
+      3.401709401709402e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      4e-05,
+      3.993593461639863e-05,
+      3.934273662008964e-05,
+      3.874953862378065e-05,
+      3.815634062747166e-05,
+      3.7563142631162665e-05,
+      3.696994463485368e-05,
+      3.637674663854469e-05,
+      3.5783548642235693e-05,
+      3.519035064592671e-05,
+      3.459715264961771e-05,
+      3.400395465330873e-05,
+      3.341075665699973e-05,
+      3.281755866069075e-05,
+      3.2224360664381764e-05,
+      3.1251515950435014e-05,
+      3.065831795412602e-05,
+      3.0065119957817037e-05,
+      2.947192196150804e-05,
+      2.887872396519905e-05,
+      2.8285525968890065e-05,
+      2.769232797258107e-05,
+      2.709912997627208e-05,
+      2.6505931979963087e-05,
+      2.5912733983654104e-05,
+      2.5319535987345117e-05,
+      2.472633799103612e-05,
+      2.4133139994727132e-05,
+      2.3539941998418135e-05,
+      2.2946744002109148e-05,
+      2.1973899288162408e-05,
+      2.138070129185341e-05,
+      2.0787503295544424e-05,
+      2.019430529923544e-05,
+      1.9601107302926443e-05,
+      1.900790930661746e-05,
+      1.8414711310308462e-05,
+      1.7821513313999475e-05,
+      1.7228315317690488e-05,
+      1.663511732138149e-05,
+      1.6041919325072507e-05,
+      1.544872132876351e-05,
+      1.4855523332454524e-05,
+      1.4262325336145537e-05,
+      1.366912733983654e-05,
+      1.2696282625889797e-05,
+      1.2103084629580812e-05,
+      1.1509886633271816e-05,
+      1.0916688636962829e-05,
+      1.0323490640653833e-05,
+      9.730292644344846e-06,
+      9.137094648035861e-06,
+      8.543896651726864e-06,
+      7.950698655417878e-06,
+      7.357500659108881e-06,
+      6.764302662799895e-06,
+      6.171104666490909e-06,
+      5.577906670181913e-06,
+      4.984708673872927e-06,
+      4.391510677563931e-06
+    ],
+    "emb_lr": [],
+    "eval_step": [
+      750,
+      1532,
+      2314,
+      3096,
+      3878,
+      4660,
+      5442,
+      6224,
+      7006,
+      7788
+    ],
+    "eval_accuracy": [
+      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": 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_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": 1.0
+}

add_sub_sorl_v1_abs5_K1_50K/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:320da49221baacc16c296960d0855142b5f676072ebecf1bd64b79536ba9ee50
+size 650283250

add_sub_sorl_v1_abs5_K1_50K/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": 4e-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": 10,
+  "emb_warmup_steps": 0,
+  "log_every": 50,
+  "eval_every": 781,
+  "save_every": 999999,
+  "eval_samples": 100,
+  "output_dir": "ckpt/sweep/as_sorl_abs5_K1_50K",
+  "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": 5,
+  "dataset_size": 50000,
+  "mode": "sorl",
+  "device": "cuda",
+  "push_to_hub": true,
+  "no_wandb": false,
+  "n_params": 162494162,
+  "run_name": "add_sub_sorl_v1_abs5_K1_50K",
+  "git_commit": "f447da529caceac8c7d256cbb2cd185cbc50feac",
+  "timestamp": "2026-04-12T10:05:47.875905+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": "iulz6nra",
+  "wandb_url": "https://wandb.ai/nlp_and_interpretability/sorl-arithmetic/runs/iulz6nra",
+  "final_accuracy": 1.0,
+  "sft_accuracy": 1.0,
+  "eval_method": "ArithmeticEvaluator"
+}