dddraxxx's picture
Upload folder using huggingface_hub
166c81c verified
raw
history blame
39.7 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.00967305088024763,
"eval_steps": 500,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"completion_length": 101.8984375,
"epoch": 9.67305088024763e-05,
"grad_norm": 2.32728163610487,
"kl": 0.0,
"learning_rate": 9.999516347455987e-07,
"loss": 0.0,
"reward": 1.4375,
"reward_std": 0.5256750583648682,
"rewards/accuracy_reward": 0.578125,
"rewards/format_reward": 0.859375,
"step": 1
},
{
"completion_length": 94.296875,
"epoch": 0.0001934610176049526,
"grad_norm": 2.4117220362597704,
"kl": 0.0005168914794921875,
"learning_rate": 9.999032694911974e-07,
"loss": 0.0,
"reward": 1.3359375,
"reward_std": 0.5609308183193207,
"rewards/accuracy_reward": 0.640625,
"rewards/format_reward": 0.6953125,
"step": 2
},
{
"completion_length": 103.6171875,
"epoch": 0.0002901915264074289,
"grad_norm": 2.6870203168696194,
"kl": 0.001102447509765625,
"learning_rate": 9.998549042367962e-07,
"loss": 0.0,
"reward": 1.46875,
"reward_std": 0.5523095726966858,
"rewards/accuracy_reward": 0.5546875,
"rewards/format_reward": 0.9140625,
"step": 3
},
{
"completion_length": 103.2265625,
"epoch": 0.0003869220352099052,
"grad_norm": 3.289268811539874,
"kl": 0.001312255859375,
"learning_rate": 9.998065389823949e-07,
"loss": 0.0001,
"reward": 1.3828125,
"reward_std": 0.5850119292736053,
"rewards/accuracy_reward": 0.5,
"rewards/format_reward": 0.8828125,
"step": 4
},
{
"completion_length": 95.078125,
"epoch": 0.0004836525440123815,
"grad_norm": 3.4676494080868085,
"kl": 0.00283050537109375,
"learning_rate": 9.997581737279938e-07,
"loss": 0.0001,
"reward": 1.640625,
"reward_std": 0.4787983000278473,
"rewards/accuracy_reward": 0.7265625,
"rewards/format_reward": 0.9140625,
"step": 5
},
{
"completion_length": 94.0703125,
"epoch": 0.0005803830528148578,
"grad_norm": 2.136642051715881,
"kl": 0.0055389404296875,
"learning_rate": 9.997098084735926e-07,
"loss": 0.0002,
"reward": 1.625,
"reward_std": 0.5282997637987137,
"rewards/accuracy_reward": 0.7109375,
"rewards/format_reward": 0.9140625,
"step": 6
},
{
"completion_length": 95.5390625,
"epoch": 0.0006771135616173341,
"grad_norm": 2.227802492062461,
"kl": 0.0041351318359375,
"learning_rate": 9.996614432191913e-07,
"loss": 0.0002,
"reward": 1.5703125,
"reward_std": 0.47315794229507446,
"rewards/accuracy_reward": 0.6328125,
"rewards/format_reward": 0.9375,
"step": 7
},
{
"completion_length": 92.734375,
"epoch": 0.0007738440704198104,
"grad_norm": 3.9540910586565667,
"kl": 0.0067596435546875,
"learning_rate": 9.9961307796479e-07,
"loss": 0.0003,
"reward": 1.453125,
"reward_std": 0.5513157844543457,
"rewards/accuracy_reward": 0.578125,
"rewards/format_reward": 0.875,
"step": 8
},
{
"completion_length": 95.4609375,
"epoch": 0.0008705745792222867,
"grad_norm": 2.277169214459152,
"kl": 0.004730224609375,
"learning_rate": 9.995647127103888e-07,
"loss": 0.0002,
"reward": 1.640625,
"reward_std": 0.47120702266693115,
"rewards/accuracy_reward": 0.7109375,
"rewards/format_reward": 0.9296875,
"step": 9
},
{
"completion_length": 95.890625,
"epoch": 0.000967305088024763,
"grad_norm": 2.715836345654194,
"kl": 0.005645751953125,
"learning_rate": 9.995163474559875e-07,
"loss": 0.0002,
"reward": 1.453125,
"reward_std": 0.46409690380096436,
"rewards/accuracy_reward": 0.484375,
"rewards/format_reward": 0.96875,
"step": 10
},
{
"completion_length": 85.4765625,
"epoch": 0.0010640355968272394,
"grad_norm": 6.999726728005914,
"kl": 0.011322021484375,
"learning_rate": 9.994679822015862e-07,
"loss": 0.0005,
"reward": 1.5625,
"reward_std": 0.506556510925293,
"rewards/accuracy_reward": 0.59375,
"rewards/format_reward": 0.96875,
"step": 11
},
{
"completion_length": 75.5703125,
"epoch": 0.0011607661056297156,
"grad_norm": 2.143072298162728,
"kl": 0.00958251953125,
"learning_rate": 9.994196169471852e-07,
"loss": 0.0004,
"reward": 1.671875,
"reward_std": 0.305637001991272,
"rewards/accuracy_reward": 0.6796875,
"rewards/format_reward": 0.9921875,
"step": 12
},
{
"completion_length": 75.15625,
"epoch": 0.0012574966144321918,
"grad_norm": 3.4444032547065477,
"kl": 0.0086669921875,
"learning_rate": 9.99371251692784e-07,
"loss": 0.0003,
"reward": 1.671875,
"reward_std": 0.43638455867767334,
"rewards/accuracy_reward": 0.6953125,
"rewards/format_reward": 0.9765625,
"step": 13
},
{
"completion_length": 73.609375,
"epoch": 0.0013542271232346683,
"grad_norm": 4.318234911676206,
"kl": 0.16168212890625,
"learning_rate": 9.993228864383827e-07,
"loss": 0.0065,
"reward": 1.8203125,
"reward_std": 0.2869499623775482,
"rewards/accuracy_reward": 0.828125,
"rewards/format_reward": 0.9921875,
"step": 14
},
{
"completion_length": 77.53125,
"epoch": 0.0014509576320371445,
"grad_norm": 2.4229540687416615,
"kl": 0.02655029296875,
"learning_rate": 9.992745211839814e-07,
"loss": 0.0011,
"reward": 1.71875,
"reward_std": 0.374386265873909,
"rewards/accuracy_reward": 0.734375,
"rewards/format_reward": 0.984375,
"step": 15
},
{
"completion_length": 76.796875,
"epoch": 0.0015476881408396208,
"grad_norm": 2.1086002986586028,
"kl": 0.019073486328125,
"learning_rate": 9.992261559295801e-07,
"loss": 0.0008,
"reward": 1.8125,
"reward_std": 0.2964432090520859,
"rewards/accuracy_reward": 0.8125,
"rewards/format_reward": 1.0,
"step": 16
},
{
"completion_length": 78.5625,
"epoch": 0.0016444186496420972,
"grad_norm": 7.813042535077011,
"kl": 0.01275634765625,
"learning_rate": 9.991777906751789e-07,
"loss": 0.0005,
"reward": 1.515625,
"reward_std": 0.41770240664482117,
"rewards/accuracy_reward": 0.5234375,
"rewards/format_reward": 0.9921875,
"step": 17
},
{
"completion_length": 72.546875,
"epoch": 0.0017411491584445734,
"grad_norm": 2.058748456561744,
"kl": 0.0169677734375,
"learning_rate": 9.991294254207778e-07,
"loss": 0.0007,
"reward": 1.71875,
"reward_std": 0.40267257392406464,
"rewards/accuracy_reward": 0.734375,
"rewards/format_reward": 0.984375,
"step": 18
},
{
"completion_length": 75.9453125,
"epoch": 0.0018378796672470497,
"grad_norm": 1.9567380549324036,
"kl": 0.01763916015625,
"learning_rate": 9.990810601663763e-07,
"loss": 0.0007,
"reward": 1.71875,
"reward_std": 0.3778578191995621,
"rewards/accuracy_reward": 0.734375,
"rewards/format_reward": 0.984375,
"step": 19
},
{
"completion_length": 70.859375,
"epoch": 0.001934610176049526,
"grad_norm": 2.9734394355786566,
"kl": 0.0279541015625,
"learning_rate": 9.99032694911975e-07,
"loss": 0.0011,
"reward": 1.765625,
"reward_std": 0.36024925112724304,
"rewards/accuracy_reward": 0.7734375,
"rewards/format_reward": 0.9921875,
"step": 20
},
{
"completion_length": 70.9765625,
"epoch": 0.0020313406848520023,
"grad_norm": 5.237584038295644,
"kl": 0.0252685546875,
"learning_rate": 9.98984329657574e-07,
"loss": 0.001,
"reward": 1.7421875,
"reward_std": 0.29719971120357513,
"rewards/accuracy_reward": 0.75,
"rewards/format_reward": 0.9921875,
"step": 21
},
{
"completion_length": 66.875,
"epoch": 0.0021280711936544788,
"grad_norm": 1.8875711559595536,
"kl": 0.02392578125,
"learning_rate": 9.989359644031727e-07,
"loss": 0.001,
"reward": 1.7109375,
"reward_std": 0.2845958471298218,
"rewards/accuracy_reward": 0.71875,
"rewards/format_reward": 0.9921875,
"step": 22
},
{
"completion_length": 67.78125,
"epoch": 0.002224801702456955,
"grad_norm": 2.4127947583544898,
"kl": 0.0750732421875,
"learning_rate": 9.988875991487715e-07,
"loss": 0.003,
"reward": 1.765625,
"reward_std": 0.3253750801086426,
"rewards/accuracy_reward": 0.765625,
"rewards/format_reward": 1.0,
"step": 23
},
{
"completion_length": 69.3125,
"epoch": 0.0023215322112594312,
"grad_norm": 2.0045706926241076,
"kl": 0.0491943359375,
"learning_rate": 9.988392338943702e-07,
"loss": 0.002,
"reward": 1.65625,
"reward_std": 0.3323955237865448,
"rewards/accuracy_reward": 0.6640625,
"rewards/format_reward": 0.9921875,
"step": 24
},
{
"completion_length": 64.4296875,
"epoch": 0.0024182627200619077,
"grad_norm": 2.0518160638069354,
"kl": 0.037109375,
"learning_rate": 9.98790868639969e-07,
"loss": 0.0015,
"reward": 1.78125,
"reward_std": 0.3174794614315033,
"rewards/accuracy_reward": 0.7890625,
"rewards/format_reward": 0.9921875,
"step": 25
},
{
"completion_length": 63.90625,
"epoch": 0.0025149932288643837,
"grad_norm": 2.3307956486799584,
"kl": 0.042724609375,
"learning_rate": 9.987425033855677e-07,
"loss": 0.0017,
"reward": 1.78125,
"reward_std": 0.3350958526134491,
"rewards/accuracy_reward": 0.796875,
"rewards/format_reward": 0.984375,
"step": 26
},
{
"completion_length": 64.4609375,
"epoch": 0.00261172373766686,
"grad_norm": 2.5099214137943027,
"kl": 0.0430908203125,
"learning_rate": 9.986941381311666e-07,
"loss": 0.0017,
"reward": 1.6875,
"reward_std": 0.3600292205810547,
"rewards/accuracy_reward": 0.6953125,
"rewards/format_reward": 0.9921875,
"step": 27
},
{
"completion_length": 59.625,
"epoch": 0.0027084542464693366,
"grad_norm": 2.482566057952426,
"kl": 0.0484619140625,
"learning_rate": 9.986457728767654e-07,
"loss": 0.0019,
"reward": 1.75,
"reward_std": 0.33345144987106323,
"rewards/accuracy_reward": 0.765625,
"rewards/format_reward": 0.984375,
"step": 28
},
{
"completion_length": 64.046875,
"epoch": 0.0028051847552718126,
"grad_norm": 2.3396995268081224,
"kl": 0.0419921875,
"learning_rate": 9.98597407622364e-07,
"loss": 0.0017,
"reward": 1.640625,
"reward_std": 0.42347636818885803,
"rewards/accuracy_reward": 0.6484375,
"rewards/format_reward": 0.9921875,
"step": 29
},
{
"completion_length": 57.796875,
"epoch": 0.002901915264074289,
"grad_norm": 1.9834549525672387,
"kl": 0.03662109375,
"learning_rate": 9.985490423679628e-07,
"loss": 0.0015,
"reward": 1.8515625,
"reward_std": 0.21937325596809387,
"rewards/accuracy_reward": 0.8515625,
"rewards/format_reward": 1.0,
"step": 30
},
{
"completion_length": 59.40625,
"epoch": 0.0029986457728767655,
"grad_norm": 2.1428993036002755,
"kl": 0.0450439453125,
"learning_rate": 9.985006771135616e-07,
"loss": 0.0018,
"reward": 1.8515625,
"reward_std": 0.15991678833961487,
"rewards/accuracy_reward": 0.8515625,
"rewards/format_reward": 1.0,
"step": 31
},
{
"completion_length": 59.0703125,
"epoch": 0.0030953762816792415,
"grad_norm": 1.744931292499253,
"kl": 0.0435791015625,
"learning_rate": 9.984523118591603e-07,
"loss": 0.0017,
"reward": 1.828125,
"reward_std": 0.23934084177017212,
"rewards/accuracy_reward": 0.8359375,
"rewards/format_reward": 0.9921875,
"step": 32
},
{
"completion_length": 60.8984375,
"epoch": 0.003192106790481718,
"grad_norm": 6.227609607513973,
"kl": 0.0377197265625,
"learning_rate": 9.98403946604759e-07,
"loss": 0.0015,
"reward": 1.6796875,
"reward_std": 0.2664504945278168,
"rewards/accuracy_reward": 0.6796875,
"rewards/format_reward": 1.0,
"step": 33
},
{
"completion_length": 61.3359375,
"epoch": 0.0032888372992841944,
"grad_norm": 6.128138900713421,
"kl": 0.038330078125,
"learning_rate": 9.98355581350358e-07,
"loss": 0.0015,
"reward": 1.8515625,
"reward_std": 0.26502610743045807,
"rewards/accuracy_reward": 0.875,
"rewards/format_reward": 0.9765625,
"step": 34
},
{
"completion_length": 55.796875,
"epoch": 0.0033855678080866704,
"grad_norm": 2.268887670294983,
"kl": 0.0474853515625,
"learning_rate": 9.983072160959565e-07,
"loss": 0.0019,
"reward": 1.8359375,
"reward_std": 0.2685721516609192,
"rewards/accuracy_reward": 0.8359375,
"rewards/format_reward": 1.0,
"step": 35
},
{
"completion_length": 58.5625,
"epoch": 0.003482298316889147,
"grad_norm": 1.831179805291792,
"kl": 0.04833984375,
"learning_rate": 9.982588508415555e-07,
"loss": 0.0019,
"reward": 1.875,
"reward_std": 0.25513993203639984,
"rewards/accuracy_reward": 0.8828125,
"rewards/format_reward": 0.9921875,
"step": 36
},
{
"completion_length": 51.8828125,
"epoch": 0.0035790288256916233,
"grad_norm": 1.1643785087500524,
"kl": 0.0455322265625,
"learning_rate": 9.982104855871542e-07,
"loss": 0.0018,
"reward": 1.9140625,
"reward_std": 0.11784427426755428,
"rewards/accuracy_reward": 0.9140625,
"rewards/format_reward": 1.0,
"step": 37
},
{
"completion_length": 57.234375,
"epoch": 0.0036757593344940993,
"grad_norm": 3.851049516152568,
"kl": 0.068359375,
"learning_rate": 9.98162120332753e-07,
"loss": 0.0027,
"reward": 1.796875,
"reward_std": 0.23391559720039368,
"rewards/accuracy_reward": 0.8046875,
"rewards/format_reward": 0.9921875,
"step": 38
},
{
"completion_length": 61.3046875,
"epoch": 0.0037724898432965758,
"grad_norm": 2.407035797036174,
"kl": 0.0482177734375,
"learning_rate": 9.981137550783517e-07,
"loss": 0.0019,
"reward": 1.8203125,
"reward_std": 0.15991678833961487,
"rewards/accuracy_reward": 0.8203125,
"rewards/format_reward": 1.0,
"step": 39
},
{
"completion_length": 57.6328125,
"epoch": 0.003869220352099052,
"grad_norm": 7.117809923901703,
"kl": 0.056640625,
"learning_rate": 9.980653898239504e-07,
"loss": 0.0023,
"reward": 1.8359375,
"reward_std": 0.23805539309978485,
"rewards/accuracy_reward": 0.84375,
"rewards/format_reward": 0.9921875,
"step": 40
},
{
"completion_length": 51.109375,
"epoch": 0.003965950860901528,
"grad_norm": 3.338321735003842,
"kl": 0.066650390625,
"learning_rate": 9.980170245695491e-07,
"loss": 0.0027,
"reward": 1.8828125,
"reward_std": 0.22726888954639435,
"rewards/accuracy_reward": 0.890625,
"rewards/format_reward": 0.9921875,
"step": 41
},
{
"completion_length": 52.2421875,
"epoch": 0.004062681369704005,
"grad_norm": 5.272467713176707,
"kl": 0.05078125,
"learning_rate": 9.979686593151479e-07,
"loss": 0.002,
"reward": 1.875,
"reward_std": 0.18861131370067596,
"rewards/accuracy_reward": 0.8828125,
"rewards/format_reward": 0.9921875,
"step": 42
},
{
"completion_length": 55.8515625,
"epoch": 0.004159411878506481,
"grad_norm": 1.8587770107731327,
"kl": 0.0513916015625,
"learning_rate": 9.979202940607468e-07,
"loss": 0.0021,
"reward": 1.8671875,
"reward_std": 0.21778053790330887,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 1.0,
"step": 43
},
{
"completion_length": 57.75,
"epoch": 0.0042561423873089576,
"grad_norm": 1.3632678648207632,
"kl": 0.0679931640625,
"learning_rate": 9.978719288063455e-07,
"loss": 0.0027,
"reward": 1.8828125,
"reward_std": 0.16439256072044373,
"rewards/accuracy_reward": 0.890625,
"rewards/format_reward": 0.9921875,
"step": 44
},
{
"completion_length": 53.9375,
"epoch": 0.004352872896111433,
"grad_norm": 1.9297448374668549,
"kl": 0.0732421875,
"learning_rate": 9.978235635519443e-07,
"loss": 0.0029,
"reward": 1.7890625,
"reward_std": 0.23609422147274017,
"rewards/accuracy_reward": 0.796875,
"rewards/format_reward": 0.9921875,
"step": 45
},
{
"completion_length": 54.8984375,
"epoch": 0.00444960340491391,
"grad_norm": 1.592387021513692,
"kl": 0.0538330078125,
"learning_rate": 9.97775198297543e-07,
"loss": 0.0022,
"reward": 1.8671875,
"reward_std": 0.154142826795578,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 1.0,
"step": 46
},
{
"completion_length": 50.390625,
"epoch": 0.004546333913716386,
"grad_norm": 2.6280541201699132,
"kl": 0.104736328125,
"learning_rate": 9.977268330431417e-07,
"loss": 0.0042,
"reward": 1.90625,
"reward_std": 0.1643974632024765,
"rewards/accuracy_reward": 0.90625,
"rewards/format_reward": 1.0,
"step": 47
},
{
"completion_length": 54.1875,
"epoch": 0.0046430644225188625,
"grad_norm": 1.9173757301011318,
"kl": 0.0537109375,
"learning_rate": 9.976784677887405e-07,
"loss": 0.0021,
"reward": 1.8828125,
"reward_std": 0.16834919713437557,
"rewards/accuracy_reward": 0.890625,
"rewards/format_reward": 0.9921875,
"step": 48
},
{
"completion_length": 51.5625,
"epoch": 0.004739794931321339,
"grad_norm": 1.8246139342634278,
"kl": 0.074951171875,
"learning_rate": 9.976301025343394e-07,
"loss": 0.003,
"reward": 1.890625,
"reward_std": 0.1841355338692665,
"rewards/accuracy_reward": 0.8984375,
"rewards/format_reward": 0.9921875,
"step": 49
},
{
"completion_length": 54.515625,
"epoch": 0.004836525440123815,
"grad_norm": 3.434717946815584,
"kl": 0.185791015625,
"learning_rate": 9.97581737279938e-07,
"loss": 0.0075,
"reward": 1.875,
"reward_std": 0.2130674198269844,
"rewards/accuracy_reward": 0.8828125,
"rewards/format_reward": 0.9921875,
"step": 50
},
{
"completion_length": 52.9296875,
"epoch": 0.004933255948926291,
"grad_norm": 1.030484859998448,
"kl": 0.0589599609375,
"learning_rate": 9.975333720255367e-07,
"loss": 0.0024,
"reward": 1.9609375,
"reward_std": 0.0765409953892231,
"rewards/accuracy_reward": 0.96875,
"rewards/format_reward": 0.9921875,
"step": 51
},
{
"completion_length": 57.234375,
"epoch": 0.005029986457728767,
"grad_norm": 4.311249728707867,
"kl": 0.06787109375,
"learning_rate": 9.974850067711356e-07,
"loss": 0.0027,
"reward": 1.859375,
"reward_std": 0.19673937559127808,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 0.9921875,
"step": 52
},
{
"completion_length": 55.6328125,
"epoch": 0.005126716966531244,
"grad_norm": 2.1298896861044496,
"kl": 0.066162109375,
"learning_rate": 9.974366415167344e-07,
"loss": 0.0027,
"reward": 1.8515625,
"reward_std": 0.1633366346359253,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 0.984375,
"step": 53
},
{
"completion_length": 54.4765625,
"epoch": 0.00522344747533372,
"grad_norm": 2.8188687448085985,
"kl": 0.0665283203125,
"learning_rate": 9.97388276262333e-07,
"loss": 0.0027,
"reward": 1.8125,
"reward_std": 0.20357907563447952,
"rewards/accuracy_reward": 0.8125,
"rewards/format_reward": 1.0,
"step": 54
},
{
"completion_length": 53.5234375,
"epoch": 0.005320177984136197,
"grad_norm": 3.8394130424569513,
"kl": 0.0771484375,
"learning_rate": 9.973399110079318e-07,
"loss": 0.0031,
"reward": 1.8125,
"reward_std": 0.24758517742156982,
"rewards/accuracy_reward": 0.8203125,
"rewards/format_reward": 0.9921875,
"step": 55
},
{
"completion_length": 55.9609375,
"epoch": 0.005416908492938673,
"grad_norm": 1.6442486623089194,
"kl": 0.0545654296875,
"learning_rate": 9.972915457535306e-07,
"loss": 0.0022,
"reward": 1.890625,
"reward_std": 0.19568344950675964,
"rewards/accuracy_reward": 0.890625,
"rewards/format_reward": 1.0,
"step": 56
},
{
"completion_length": 57.609375,
"epoch": 0.005513639001741149,
"grad_norm": 2.380780380370499,
"kl": 0.083251953125,
"learning_rate": 9.972431804991293e-07,
"loss": 0.0033,
"reward": 1.890625,
"reward_std": 0.2414703369140625,
"rewards/accuracy_reward": 0.8984375,
"rewards/format_reward": 0.9921875,
"step": 57
},
{
"completion_length": 57.15625,
"epoch": 0.005610369510543625,
"grad_norm": 1.8965682333399105,
"kl": 0.064453125,
"learning_rate": 9.971948152447282e-07,
"loss": 0.0026,
"reward": 1.9296875,
"reward_std": 0.14389308914542198,
"rewards/accuracy_reward": 0.9375,
"rewards/format_reward": 0.9921875,
"step": 58
},
{
"completion_length": 57.0625,
"epoch": 0.005707100019346102,
"grad_norm": 1.6399837752342512,
"kl": 0.0811767578125,
"learning_rate": 9.97146449990327e-07,
"loss": 0.0032,
"reward": 1.890625,
"reward_std": 0.13781969994306564,
"rewards/accuracy_reward": 0.90625,
"rewards/format_reward": 0.984375,
"step": 59
},
{
"completion_length": 59.1953125,
"epoch": 0.005803830528148578,
"grad_norm": 2.4532760446589053,
"kl": 0.0611572265625,
"learning_rate": 9.970980847359257e-07,
"loss": 0.0024,
"reward": 1.8203125,
"reward_std": 0.23646268993616104,
"rewards/accuracy_reward": 0.8203125,
"rewards/format_reward": 1.0,
"step": 60
},
{
"completion_length": 56.6953125,
"epoch": 0.0059005610369510545,
"grad_norm": 1.347418830825943,
"kl": 0.0498046875,
"learning_rate": 9.970497194815245e-07,
"loss": 0.002,
"reward": 1.9375,
"reward_std": 0.10888782143592834,
"rewards/accuracy_reward": 0.9375,
"rewards/format_reward": 1.0,
"step": 61
},
{
"completion_length": 61.7109375,
"epoch": 0.005997291545753531,
"grad_norm": 1.7296816337075058,
"kl": 0.0567626953125,
"learning_rate": 9.970013542271232e-07,
"loss": 0.0023,
"reward": 1.8125,
"reward_std": 0.16675157845020294,
"rewards/accuracy_reward": 0.8125,
"rewards/format_reward": 1.0,
"step": 62
},
{
"completion_length": 60.0859375,
"epoch": 0.006094022054556007,
"grad_norm": 1.1174900550936973,
"kl": 0.05126953125,
"learning_rate": 9.96952988972722e-07,
"loss": 0.0021,
"reward": 1.9765625,
"reward_std": 0.06629125960171223,
"rewards/accuracy_reward": 0.9765625,
"rewards/format_reward": 1.0,
"step": 63
},
{
"completion_length": 63.2734375,
"epoch": 0.006190752563358483,
"grad_norm": 1.2212567674010755,
"kl": 0.0545654296875,
"learning_rate": 9.969046237183207e-07,
"loss": 0.0022,
"reward": 1.921875,
"reward_std": 0.12255740165710449,
"rewards/accuracy_reward": 0.921875,
"rewards/format_reward": 1.0,
"step": 64
},
{
"completion_length": 62.90625,
"epoch": 0.0062874830721609595,
"grad_norm": 1.0607062370398306,
"kl": 0.0516357421875,
"learning_rate": 9.968562584639196e-07,
"loss": 0.0021,
"reward": 1.8828125,
"reward_std": 0.07654099725186825,
"rewards/accuracy_reward": 0.8828125,
"rewards/format_reward": 1.0,
"step": 65
},
{
"completion_length": 61.9609375,
"epoch": 0.006384213580963436,
"grad_norm": 3.6380301094080325,
"kl": 0.079345703125,
"learning_rate": 9.968078932095181e-07,
"loss": 0.0032,
"reward": 1.890625,
"reward_std": 0.09863808006048203,
"rewards/accuracy_reward": 0.890625,
"rewards/format_reward": 1.0,
"step": 66
},
{
"completion_length": 65.6015625,
"epoch": 0.006480944089765912,
"grad_norm": 1.2101367300868426,
"kl": 0.0499267578125,
"learning_rate": 9.96759527955117e-07,
"loss": 0.002,
"reward": 1.921875,
"reward_std": 0.0883883461356163,
"rewards/accuracy_reward": 0.921875,
"rewards/format_reward": 1.0,
"step": 67
},
{
"completion_length": 64.9375,
"epoch": 0.006577674598568389,
"grad_norm": 1.4058299605738835,
"kl": 0.05517578125,
"learning_rate": 9.967111627007158e-07,
"loss": 0.0022,
"reward": 1.8359375,
"reward_std": 0.13488983362913132,
"rewards/accuracy_reward": 0.8515625,
"rewards/format_reward": 0.984375,
"step": 68
},
{
"completion_length": 67.6328125,
"epoch": 0.006674405107370864,
"grad_norm": 2.243127302505205,
"kl": 0.040283203125,
"learning_rate": 9.966627974463145e-07,
"loss": 0.0016,
"reward": 1.8984375,
"reward_std": 0.15467960387468338,
"rewards/accuracy_reward": 0.9140625,
"rewards/format_reward": 0.984375,
"step": 69
},
{
"completion_length": 68.9765625,
"epoch": 0.006771135616173341,
"grad_norm": 1.5075765934443708,
"kl": 0.041259765625,
"learning_rate": 9.966144321919133e-07,
"loss": 0.0016,
"reward": 1.953125,
"reward_std": 0.1173202246427536,
"rewards/accuracy_reward": 0.953125,
"rewards/format_reward": 1.0,
"step": 70
},
{
"completion_length": 71.6015625,
"epoch": 0.006867866124975817,
"grad_norm": 1.2737896051744175,
"kl": 0.0438232421875,
"learning_rate": 9.96566066937512e-07,
"loss": 0.0018,
"reward": 1.8359375,
"reward_std": 0.15490422397851944,
"rewards/accuracy_reward": 0.8359375,
"rewards/format_reward": 1.0,
"step": 71
},
{
"completion_length": 72.078125,
"epoch": 0.006964596633778294,
"grad_norm": 3.138596738453845,
"kl": 0.042724609375,
"learning_rate": 9.965177016831107e-07,
"loss": 0.0017,
"reward": 1.8984375,
"reward_std": 0.16834919527173042,
"rewards/accuracy_reward": 0.8984375,
"rewards/format_reward": 1.0,
"step": 72
},
{
"completion_length": 70.0078125,
"epoch": 0.00706132714258077,
"grad_norm": 1.2396648384145925,
"kl": 0.04046630859375,
"learning_rate": 9.964693364287095e-07,
"loss": 0.0016,
"reward": 1.9375,
"reward_std": 0.1462521031498909,
"rewards/accuracy_reward": 0.9453125,
"rewards/format_reward": 0.9921875,
"step": 73
},
{
"completion_length": 72.734375,
"epoch": 0.007158057651383247,
"grad_norm": 1.2894041867000514,
"kl": 0.0477294921875,
"learning_rate": 9.964209711743084e-07,
"loss": 0.0019,
"reward": 1.859375,
"reward_std": 0.12020329385995865,
"rewards/accuracy_reward": 0.859375,
"rewards/format_reward": 1.0,
"step": 74
},
{
"completion_length": 68.5,
"epoch": 0.007254788160185722,
"grad_norm": 2.6400118191598185,
"kl": 0.0372314453125,
"learning_rate": 9.963726059199072e-07,
"loss": 0.0015,
"reward": 1.8828125,
"reward_std": 0.1633366346359253,
"rewards/accuracy_reward": 0.890625,
"rewards/format_reward": 0.9921875,
"step": 75
},
{
"completion_length": 72.4765625,
"epoch": 0.007351518668988199,
"grad_norm": 1.6894479413134922,
"kl": 0.041259765625,
"learning_rate": 9.963242406655059e-07,
"loss": 0.0016,
"reward": 1.8671875,
"reward_std": 0.17282496020197868,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 1.0,
"step": 76
},
{
"completion_length": 78.59375,
"epoch": 0.007448249177790675,
"grad_norm": 2.0103213762329792,
"kl": 0.0479736328125,
"learning_rate": 9.962758754111046e-07,
"loss": 0.0019,
"reward": 1.8046875,
"reward_std": 0.24329256266355515,
"rewards/accuracy_reward": 0.8125,
"rewards/format_reward": 0.9921875,
"step": 77
},
{
"completion_length": 76.9375,
"epoch": 0.0075449796865931515,
"grad_norm": 2.1679810295686592,
"kl": 0.039306640625,
"learning_rate": 9.962275101567034e-07,
"loss": 0.0016,
"reward": 1.8671875,
"reward_std": 0.1965562254190445,
"rewards/accuracy_reward": 0.875,
"rewards/format_reward": 0.9921875,
"step": 78
},
{
"completion_length": 73.890625,
"epoch": 0.007641710195395628,
"grad_norm": 1.7310448885764356,
"kl": 0.0416259765625,
"learning_rate": 9.96179144902302e-07,
"loss": 0.0017,
"reward": 1.921875,
"reward_std": 0.15072787925601006,
"rewards/accuracy_reward": 0.921875,
"rewards/format_reward": 1.0,
"step": 79
},
{
"completion_length": 77.859375,
"epoch": 0.007738440704198104,
"grad_norm": 1.1571932464254209,
"kl": 0.03521728515625,
"learning_rate": 9.961307796479008e-07,
"loss": 0.0014,
"reward": 1.875,
"reward_std": 0.11678344011306763,
"rewards/accuracy_reward": 0.8828125,
"rewards/format_reward": 0.9921875,
"step": 80
},
{
"completion_length": 80.6640625,
"epoch": 0.00783517121300058,
"grad_norm": 1.1846650681630302,
"kl": 0.0355224609375,
"learning_rate": 9.960824143934998e-07,
"loss": 0.0014,
"reward": 1.8984375,
"reward_std": 0.16781240701675415,
"rewards/accuracy_reward": 0.90625,
"rewards/format_reward": 0.9921875,
"step": 81
},
{
"completion_length": 80.9453125,
"epoch": 0.007931901721803056,
"grad_norm": 1.7831378206980077,
"kl": 0.0323486328125,
"learning_rate": 9.960340491390983e-07,
"loss": 0.0013,
"reward": 1.8828125,
"reward_std": 0.1512519307434559,
"rewards/accuracy_reward": 0.8828125,
"rewards/format_reward": 1.0,
"step": 82
},
{
"completion_length": 74.7109375,
"epoch": 0.008028632230605534,
"grad_norm": 1.8651786426826615,
"kl": 0.0574951171875,
"learning_rate": 9.959856838846972e-07,
"loss": 0.0023,
"reward": 1.8203125,
"reward_std": 0.19567855447530746,
"rewards/accuracy_reward": 0.8203125,
"rewards/format_reward": 1.0,
"step": 83
},
{
"completion_length": 82.1484375,
"epoch": 0.00812536273940801,
"grad_norm": 1.2035450615079515,
"kl": 0.0396728515625,
"learning_rate": 9.95937318630296e-07,
"loss": 0.0016,
"reward": 1.8515625,
"reward_std": 0.13888052850961685,
"rewards/accuracy_reward": 0.8515625,
"rewards/format_reward": 1.0,
"step": 84
},
{
"completion_length": 79.6015625,
"epoch": 0.008222093248210485,
"grad_norm": 1.9528011129648917,
"kl": 0.0355224609375,
"learning_rate": 9.958889533758947e-07,
"loss": 0.0014,
"reward": 1.8828125,
"reward_std": 0.12073516473174095,
"rewards/accuracy_reward": 0.8828125,
"rewards/format_reward": 1.0,
"step": 85
},
{
"completion_length": 75.609375,
"epoch": 0.008318823757012962,
"grad_norm": 1.3877474003576578,
"kl": 0.0458984375,
"learning_rate": 9.958405881214934e-07,
"loss": 0.0018,
"reward": 1.8984375,
"reward_std": 0.10547287575900555,
"rewards/accuracy_reward": 0.8984375,
"rewards/format_reward": 1.0,
"step": 86
},
{
"completion_length": 80.34375,
"epoch": 0.008415554265815438,
"grad_norm": 2.220481843420492,
"kl": 0.03564453125,
"learning_rate": 9.957922228670922e-07,
"loss": 0.0014,
"reward": 1.8515625,
"reward_std": 0.11048543266952038,
"rewards/accuracy_reward": 0.859375,
"rewards/format_reward": 0.9921875,
"step": 87
},
{
"completion_length": 76.859375,
"epoch": 0.008512284774617915,
"grad_norm": 1.3296039029130402,
"kl": 0.0399169921875,
"learning_rate": 9.95743857612691e-07,
"loss": 0.0016,
"reward": 1.921875,
"reward_std": 0.12756996229290962,
"rewards/accuracy_reward": 0.921875,
"rewards/format_reward": 1.0,
"step": 88
},
{
"completion_length": 79.5703125,
"epoch": 0.00860901528342039,
"grad_norm": 1.5250275021762014,
"kl": 0.03662109375,
"learning_rate": 9.956954923582899e-07,
"loss": 0.0015,
"reward": 1.828125,
"reward_std": 0.2359209954738617,
"rewards/accuracy_reward": 0.828125,
"rewards/format_reward": 1.0,
"step": 89
},
{
"completion_length": 78.625,
"epoch": 0.008705745792222866,
"grad_norm": 2.1022892308790833,
"kl": 0.0694580078125,
"learning_rate": 9.956471271038886e-07,
"loss": 0.0028,
"reward": 1.890625,
"reward_std": 0.06233953312039375,
"rewards/accuracy_reward": 0.890625,
"rewards/format_reward": 1.0,
"step": 90
},
{
"completion_length": 89.296875,
"epoch": 0.008802476301025344,
"grad_norm": 2.649730494042045,
"kl": 0.119384765625,
"learning_rate": 9.955987618494873e-07,
"loss": 0.0048,
"reward": 1.90625,
"reward_std": 0.09863807819783688,
"rewards/accuracy_reward": 0.9140625,
"rewards/format_reward": 0.9921875,
"step": 91
},
{
"completion_length": 83.671875,
"epoch": 0.00889920680982782,
"grad_norm": 1.328799864584759,
"kl": 0.03448486328125,
"learning_rate": 9.95550396595086e-07,
"loss": 0.0014,
"reward": 1.90625,
"reward_std": 0.1065337061882019,
"rewards/accuracy_reward": 0.921875,
"rewards/format_reward": 0.984375,
"step": 92
},
{
"completion_length": 82.671875,
"epoch": 0.008995937318630296,
"grad_norm": 1.6487837590480994,
"kl": 0.0421142578125,
"learning_rate": 9.955020313406848e-07,
"loss": 0.0017,
"reward": 1.890625,
"reward_std": 0.19044627249240875,
"rewards/accuracy_reward": 0.8984375,
"rewards/format_reward": 0.9921875,
"step": 93
},
{
"completion_length": 85.546875,
"epoch": 0.009092667827432772,
"grad_norm": 0.9521030151806793,
"kl": 0.046142578125,
"learning_rate": 9.954536660862835e-07,
"loss": 0.0018,
"reward": 1.8984375,
"reward_std": 0.10547287575900555,
"rewards/accuracy_reward": 0.8984375,
"rewards/format_reward": 1.0,
"step": 94
},
{
"completion_length": 79.9765625,
"epoch": 0.00918939833623525,
"grad_norm": 0.7454143427025838,
"kl": 0.0355224609375,
"learning_rate": 9.954053008318823e-07,
"loss": 0.0014,
"reward": 1.984375,
"reward_std": 0.04419417306780815,
"rewards/accuracy_reward": 0.984375,
"rewards/format_reward": 1.0,
"step": 95
},
{
"completion_length": 87.4921875,
"epoch": 0.009286128845037725,
"grad_norm": 1.133464625493318,
"kl": 0.0345458984375,
"learning_rate": 9.953569355774812e-07,
"loss": 0.0014,
"reward": 1.859375,
"reward_std": 0.1530819982290268,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 0.9921875,
"step": 96
},
{
"completion_length": 80.859375,
"epoch": 0.0093828593538402,
"grad_norm": 0.7929599617421281,
"kl": 0.03466796875,
"learning_rate": 9.953085703230797e-07,
"loss": 0.0014,
"reward": 1.9765625,
"reward_std": 0.051028965041041374,
"rewards/accuracy_reward": 0.9765625,
"rewards/format_reward": 1.0,
"step": 97
},
{
"completion_length": 85.875,
"epoch": 0.009479589862642678,
"grad_norm": 1.1597056531543017,
"kl": 0.0352783203125,
"learning_rate": 9.952602050686787e-07,
"loss": 0.0014,
"reward": 1.8671875,
"reward_std": 0.1054728701710701,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 1.0,
"step": 98
},
{
"completion_length": 83.0390625,
"epoch": 0.009576320371445153,
"grad_norm": 1.07555295208371,
"kl": 0.03387451171875,
"learning_rate": 9.952118398142774e-07,
"loss": 0.0014,
"reward": 1.9140625,
"reward_std": 0.08443661965429783,
"rewards/accuracy_reward": 0.9140625,
"rewards/format_reward": 1.0,
"step": 99
},
{
"completion_length": 80.453125,
"epoch": 0.00967305088024763,
"grad_norm": 4.643230012328778,
"kl": 0.0306396484375,
"learning_rate": 9.951634745598762e-07,
"loss": 0.0012,
"reward": 1.859375,
"reward_std": 0.16675157845020294,
"rewards/accuracy_reward": 0.8671875,
"rewards/format_reward": 0.9921875,
"step": 100
}
],
"logging_steps": 1.0,
"max_steps": 20676,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}