prxy5605's picture
Training in progress, step 200, checkpoint
82fcdc9 verified
{
"best_metric": 0.9981580972671509,
"best_model_checkpoint": "miner_id_24/checkpoint-150",
"epoch": 0.8629989212513485,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004314994606256742,
"grad_norm": 0.2654043436050415,
"learning_rate": 1e-05,
"loss": 0.8217,
"step": 1
},
{
"epoch": 0.004314994606256742,
"eval_loss": 1.1413410902023315,
"eval_runtime": 27.1745,
"eval_samples_per_second": 14.352,
"eval_steps_per_second": 3.606,
"step": 1
},
{
"epoch": 0.008629989212513484,
"grad_norm": 0.2317211627960205,
"learning_rate": 2e-05,
"loss": 1.4718,
"step": 2
},
{
"epoch": 0.012944983818770227,
"grad_norm": 0.2540183365345001,
"learning_rate": 3e-05,
"loss": 1.5512,
"step": 3
},
{
"epoch": 0.017259978425026967,
"grad_norm": 0.24836920201778412,
"learning_rate": 4e-05,
"loss": 1.5605,
"step": 4
},
{
"epoch": 0.021574973031283712,
"grad_norm": 0.21192340552806854,
"learning_rate": 5e-05,
"loss": 1.644,
"step": 5
},
{
"epoch": 0.025889967637540454,
"grad_norm": 0.20893314480781555,
"learning_rate": 6e-05,
"loss": 1.4608,
"step": 6
},
{
"epoch": 0.030204962243797196,
"grad_norm": 0.2030942738056183,
"learning_rate": 7e-05,
"loss": 1.5844,
"step": 7
},
{
"epoch": 0.034519956850053934,
"grad_norm": 0.21079899370670319,
"learning_rate": 8e-05,
"loss": 1.5795,
"step": 8
},
{
"epoch": 0.038834951456310676,
"grad_norm": 0.23301894962787628,
"learning_rate": 9e-05,
"loss": 1.5572,
"step": 9
},
{
"epoch": 0.043149946062567425,
"grad_norm": 0.21977968513965607,
"learning_rate": 0.0001,
"loss": 1.4419,
"step": 10
},
{
"epoch": 0.04746494066882417,
"grad_norm": 0.21727196872234344,
"learning_rate": 9.999316524962345e-05,
"loss": 1.5206,
"step": 11
},
{
"epoch": 0.05177993527508091,
"grad_norm": 0.2866550385951996,
"learning_rate": 9.997266286704631e-05,
"loss": 1.3769,
"step": 12
},
{
"epoch": 0.05609492988133765,
"grad_norm": 0.23287613689899445,
"learning_rate": 9.993849845741524e-05,
"loss": 1.4165,
"step": 13
},
{
"epoch": 0.06040992448759439,
"grad_norm": 0.19964908063411713,
"learning_rate": 9.989068136093873e-05,
"loss": 1.2461,
"step": 14
},
{
"epoch": 0.06472491909385113,
"grad_norm": 0.20694532990455627,
"learning_rate": 9.98292246503335e-05,
"loss": 1.276,
"step": 15
},
{
"epoch": 0.06903991370010787,
"grad_norm": 0.1831284761428833,
"learning_rate": 9.975414512725057e-05,
"loss": 1.3878,
"step": 16
},
{
"epoch": 0.07335490830636461,
"grad_norm": 0.1903073489665985,
"learning_rate": 9.966546331768191e-05,
"loss": 1.2079,
"step": 17
},
{
"epoch": 0.07766990291262135,
"grad_norm": 0.19069454073905945,
"learning_rate": 9.956320346634876e-05,
"loss": 1.2313,
"step": 18
},
{
"epoch": 0.08198489751887811,
"grad_norm": 0.18895849585533142,
"learning_rate": 9.944739353007344e-05,
"loss": 1.2313,
"step": 19
},
{
"epoch": 0.08629989212513485,
"grad_norm": 0.20607298612594604,
"learning_rate": 9.931806517013612e-05,
"loss": 1.1666,
"step": 20
},
{
"epoch": 0.09061488673139159,
"grad_norm": 0.21161971986293793,
"learning_rate": 9.917525374361912e-05,
"loss": 1.1593,
"step": 21
},
{
"epoch": 0.09492988133764833,
"grad_norm": 0.21831297874450684,
"learning_rate": 9.901899829374047e-05,
"loss": 1.1235,
"step": 22
},
{
"epoch": 0.09924487594390508,
"grad_norm": 0.23938357830047607,
"learning_rate": 9.884934153917997e-05,
"loss": 0.9985,
"step": 23
},
{
"epoch": 0.10355987055016182,
"grad_norm": 0.22694087028503418,
"learning_rate": 9.86663298624003e-05,
"loss": 1.0481,
"step": 24
},
{
"epoch": 0.10787486515641856,
"grad_norm": 0.2468775361776352,
"learning_rate": 9.847001329696653e-05,
"loss": 0.9692,
"step": 25
},
{
"epoch": 0.1121898597626753,
"grad_norm": 0.26769381761550903,
"learning_rate": 9.826044551386744e-05,
"loss": 1.0162,
"step": 26
},
{
"epoch": 0.11650485436893204,
"grad_norm": 0.285277783870697,
"learning_rate": 9.803768380684242e-05,
"loss": 0.9497,
"step": 27
},
{
"epoch": 0.12081984897518878,
"grad_norm": 0.2573104202747345,
"learning_rate": 9.780178907671789e-05,
"loss": 0.847,
"step": 28
},
{
"epoch": 0.12513484358144553,
"grad_norm": 0.2511522173881531,
"learning_rate": 9.755282581475769e-05,
"loss": 0.9037,
"step": 29
},
{
"epoch": 0.12944983818770225,
"grad_norm": 0.25227245688438416,
"learning_rate": 9.729086208503174e-05,
"loss": 0.9912,
"step": 30
},
{
"epoch": 0.133764832793959,
"grad_norm": 0.2540568709373474,
"learning_rate": 9.701596950580806e-05,
"loss": 0.8676,
"step": 31
},
{
"epoch": 0.13807982740021574,
"grad_norm": 0.2567998468875885,
"learning_rate": 9.672822322997305e-05,
"loss": 0.8828,
"step": 32
},
{
"epoch": 0.1423948220064725,
"grad_norm": 0.24519716203212738,
"learning_rate": 9.642770192448536e-05,
"loss": 0.9211,
"step": 33
},
{
"epoch": 0.14670981661272922,
"grad_norm": 0.23400765657424927,
"learning_rate": 9.611448774886924e-05,
"loss": 0.8223,
"step": 34
},
{
"epoch": 0.15102481121898598,
"grad_norm": 0.22917449474334717,
"learning_rate": 9.578866633275288e-05,
"loss": 0.9494,
"step": 35
},
{
"epoch": 0.1553398058252427,
"grad_norm": 0.23426009714603424,
"learning_rate": 9.545032675245813e-05,
"loss": 0.8989,
"step": 36
},
{
"epoch": 0.15965480043149946,
"grad_norm": 0.23744282126426697,
"learning_rate": 9.509956150664796e-05,
"loss": 1.0092,
"step": 37
},
{
"epoch": 0.16396979503775622,
"grad_norm": 0.24764712154865265,
"learning_rate": 9.473646649103818e-05,
"loss": 0.7986,
"step": 38
},
{
"epoch": 0.16828478964401294,
"grad_norm": 0.24543999135494232,
"learning_rate": 9.43611409721806e-05,
"loss": 0.813,
"step": 39
},
{
"epoch": 0.1725997842502697,
"grad_norm": 0.24142329394817352,
"learning_rate": 9.397368756032445e-05,
"loss": 0.8867,
"step": 40
},
{
"epoch": 0.17691477885652643,
"grad_norm": 0.27213943004608154,
"learning_rate": 9.357421218136386e-05,
"loss": 0.7335,
"step": 41
},
{
"epoch": 0.18122977346278318,
"grad_norm": 0.2508591413497925,
"learning_rate": 9.316282404787871e-05,
"loss": 0.6742,
"step": 42
},
{
"epoch": 0.1855447680690399,
"grad_norm": 0.24565966427326202,
"learning_rate": 9.273963562927695e-05,
"loss": 0.6454,
"step": 43
},
{
"epoch": 0.18985976267529667,
"grad_norm": 0.2706315517425537,
"learning_rate": 9.230476262104677e-05,
"loss": 0.6801,
"step": 44
},
{
"epoch": 0.1941747572815534,
"grad_norm": 0.25447723269462585,
"learning_rate": 9.185832391312644e-05,
"loss": 0.7541,
"step": 45
},
{
"epoch": 0.19848975188781015,
"grad_norm": 0.25922590494155884,
"learning_rate": 9.140044155740101e-05,
"loss": 0.7839,
"step": 46
},
{
"epoch": 0.20280474649406688,
"grad_norm": 0.36019840836524963,
"learning_rate": 9.093124073433463e-05,
"loss": 1.0946,
"step": 47
},
{
"epoch": 0.20711974110032363,
"grad_norm": 0.37877389788627625,
"learning_rate": 9.045084971874738e-05,
"loss": 1.2572,
"step": 48
},
{
"epoch": 0.21143473570658036,
"grad_norm": 0.3896636962890625,
"learning_rate": 8.995939984474624e-05,
"loss": 1.2254,
"step": 49
},
{
"epoch": 0.21574973031283712,
"grad_norm": 0.48014289140701294,
"learning_rate": 8.945702546981969e-05,
"loss": 1.496,
"step": 50
},
{
"epoch": 0.21574973031283712,
"eval_loss": 1.0348436832427979,
"eval_runtime": 27.7389,
"eval_samples_per_second": 14.06,
"eval_steps_per_second": 3.533,
"step": 50
},
{
"epoch": 0.22006472491909385,
"grad_norm": 0.2100718915462494,
"learning_rate": 8.894386393810563e-05,
"loss": 1.1813,
"step": 51
},
{
"epoch": 0.2243797195253506,
"grad_norm": 0.2261289805173874,
"learning_rate": 8.842005554284296e-05,
"loss": 1.3932,
"step": 52
},
{
"epoch": 0.22869471413160733,
"grad_norm": 0.28611212968826294,
"learning_rate": 8.788574348801675e-05,
"loss": 1.4921,
"step": 53
},
{
"epoch": 0.23300970873786409,
"grad_norm": 0.21410827338695526,
"learning_rate": 8.73410738492077e-05,
"loss": 1.5079,
"step": 54
},
{
"epoch": 0.2373247033441208,
"grad_norm": 0.1962806135416031,
"learning_rate": 8.678619553365659e-05,
"loss": 1.4757,
"step": 55
},
{
"epoch": 0.24163969795037757,
"grad_norm": 0.16831356287002563,
"learning_rate": 8.622126023955446e-05,
"loss": 1.4364,
"step": 56
},
{
"epoch": 0.2459546925566343,
"grad_norm": 0.18131586909294128,
"learning_rate": 8.564642241456986e-05,
"loss": 1.3536,
"step": 57
},
{
"epoch": 0.25026968716289105,
"grad_norm": 0.19793841242790222,
"learning_rate": 8.506183921362443e-05,
"loss": 1.4726,
"step": 58
},
{
"epoch": 0.2545846817691478,
"grad_norm": 0.22674345970153809,
"learning_rate": 8.44676704559283e-05,
"loss": 1.597,
"step": 59
},
{
"epoch": 0.2588996763754045,
"grad_norm": 0.19330620765686035,
"learning_rate": 8.386407858128706e-05,
"loss": 1.519,
"step": 60
},
{
"epoch": 0.2632146709816613,
"grad_norm": 0.1925821602344513,
"learning_rate": 8.32512286056924e-05,
"loss": 1.4618,
"step": 61
},
{
"epoch": 0.267529665587918,
"grad_norm": 0.17810407280921936,
"learning_rate": 8.262928807620843e-05,
"loss": 1.3706,
"step": 62
},
{
"epoch": 0.27184466019417475,
"grad_norm": 0.17821729183197021,
"learning_rate": 8.199842702516583e-05,
"loss": 1.3594,
"step": 63
},
{
"epoch": 0.2761596548004315,
"grad_norm": 0.18764498829841614,
"learning_rate": 8.135881792367686e-05,
"loss": 1.2776,
"step": 64
},
{
"epoch": 0.28047464940668826,
"grad_norm": 0.18041546642780304,
"learning_rate": 8.07106356344834e-05,
"loss": 1.3856,
"step": 65
},
{
"epoch": 0.284789644012945,
"grad_norm": 0.19698365032672882,
"learning_rate": 8.005405736415126e-05,
"loss": 1.192,
"step": 66
},
{
"epoch": 0.2891046386192017,
"grad_norm": 0.19803135097026825,
"learning_rate": 7.938926261462366e-05,
"loss": 1.2134,
"step": 67
},
{
"epoch": 0.29341963322545844,
"grad_norm": 0.20288443565368652,
"learning_rate": 7.871643313414718e-05,
"loss": 1.1692,
"step": 68
},
{
"epoch": 0.2977346278317152,
"grad_norm": 0.20673032104969025,
"learning_rate": 7.803575286758364e-05,
"loss": 1.1733,
"step": 69
},
{
"epoch": 0.30204962243797195,
"grad_norm": 0.21595843136310577,
"learning_rate": 7.734740790612136e-05,
"loss": 1.1435,
"step": 70
},
{
"epoch": 0.3063646170442287,
"grad_norm": 0.21774247288703918,
"learning_rate": 7.66515864363997e-05,
"loss": 1.0255,
"step": 71
},
{
"epoch": 0.3106796116504854,
"grad_norm": 0.2304445505142212,
"learning_rate": 7.594847868906076e-05,
"loss": 0.8884,
"step": 72
},
{
"epoch": 0.3149946062567422,
"grad_norm": 0.2240193486213684,
"learning_rate": 7.52382768867422e-05,
"loss": 1.0888,
"step": 73
},
{
"epoch": 0.3193096008629989,
"grad_norm": 0.22037872672080994,
"learning_rate": 7.452117519152542e-05,
"loss": 1.0014,
"step": 74
},
{
"epoch": 0.32362459546925565,
"grad_norm": 0.21170316636562347,
"learning_rate": 7.379736965185368e-05,
"loss": 0.9183,
"step": 75
},
{
"epoch": 0.32793959007551243,
"grad_norm": 0.2662046253681183,
"learning_rate": 7.30670581489344e-05,
"loss": 0.8797,
"step": 76
},
{
"epoch": 0.33225458468176916,
"grad_norm": 0.23489448428153992,
"learning_rate": 7.233044034264034e-05,
"loss": 0.8412,
"step": 77
},
{
"epoch": 0.3365695792880259,
"grad_norm": 0.2394036501646042,
"learning_rate": 7.158771761692464e-05,
"loss": 0.791,
"step": 78
},
{
"epoch": 0.3408845738942826,
"grad_norm": 0.23320372402668,
"learning_rate": 7.083909302476453e-05,
"loss": 0.8137,
"step": 79
},
{
"epoch": 0.3451995685005394,
"grad_norm": 0.2332010567188263,
"learning_rate": 7.008477123264848e-05,
"loss": 0.7717,
"step": 80
},
{
"epoch": 0.34951456310679613,
"grad_norm": 0.22427760064601898,
"learning_rate": 6.932495846462261e-05,
"loss": 0.852,
"step": 81
},
{
"epoch": 0.35382955771305286,
"grad_norm": 0.2259577065706253,
"learning_rate": 6.855986244591104e-05,
"loss": 0.6904,
"step": 82
},
{
"epoch": 0.3581445523193096,
"grad_norm": 0.21389418840408325,
"learning_rate": 6.778969234612584e-05,
"loss": 0.8132,
"step": 83
},
{
"epoch": 0.36245954692556637,
"grad_norm": 0.23344309628009796,
"learning_rate": 6.701465872208216e-05,
"loss": 0.8384,
"step": 84
},
{
"epoch": 0.3667745415318231,
"grad_norm": 0.22995668649673462,
"learning_rate": 6.623497346023418e-05,
"loss": 0.8363,
"step": 85
},
{
"epoch": 0.3710895361380798,
"grad_norm": 0.23668435215950012,
"learning_rate": 6.545084971874738e-05,
"loss": 0.7859,
"step": 86
},
{
"epoch": 0.37540453074433655,
"grad_norm": 0.23146218061447144,
"learning_rate": 6.466250186922325e-05,
"loss": 0.9469,
"step": 87
},
{
"epoch": 0.37971952535059333,
"grad_norm": 0.22001032531261444,
"learning_rate": 6.387014543809223e-05,
"loss": 0.9658,
"step": 88
},
{
"epoch": 0.38403451995685006,
"grad_norm": 0.23594342172145844,
"learning_rate": 6.307399704769099e-05,
"loss": 0.7287,
"step": 89
},
{
"epoch": 0.3883495145631068,
"grad_norm": 0.24331915378570557,
"learning_rate": 6.227427435703997e-05,
"loss": 0.703,
"step": 90
},
{
"epoch": 0.3926645091693635,
"grad_norm": 0.2396632581949234,
"learning_rate": 6.147119600233758e-05,
"loss": 0.7592,
"step": 91
},
{
"epoch": 0.3969795037756203,
"grad_norm": 0.23275357484817505,
"learning_rate": 6.066498153718735e-05,
"loss": 0.7608,
"step": 92
},
{
"epoch": 0.40129449838187703,
"grad_norm": 0.24012094736099243,
"learning_rate": 5.985585137257401e-05,
"loss": 0.6974,
"step": 93
},
{
"epoch": 0.40560949298813376,
"grad_norm": 0.23205330967903137,
"learning_rate": 5.90440267166055e-05,
"loss": 0.7119,
"step": 94
},
{
"epoch": 0.4099244875943905,
"grad_norm": 0.2341987043619156,
"learning_rate": 5.8229729514036705e-05,
"loss": 0.7015,
"step": 95
},
{
"epoch": 0.41423948220064727,
"grad_norm": 0.2407538741827011,
"learning_rate": 5.74131823855921e-05,
"loss": 0.704,
"step": 96
},
{
"epoch": 0.418554476806904,
"grad_norm": 0.2716779112815857,
"learning_rate": 5.6594608567103456e-05,
"loss": 0.8122,
"step": 97
},
{
"epoch": 0.4228694714131607,
"grad_norm": 0.3880912959575653,
"learning_rate": 5.577423184847932e-05,
"loss": 1.1928,
"step": 98
},
{
"epoch": 0.42718446601941745,
"grad_norm": 0.4338785707950592,
"learning_rate": 5.495227651252315e-05,
"loss": 1.1404,
"step": 99
},
{
"epoch": 0.43149946062567424,
"grad_norm": 0.4976014494895935,
"learning_rate": 5.4128967273616625e-05,
"loss": 1.4886,
"step": 100
},
{
"epoch": 0.43149946062567424,
"eval_loss": 1.0101395845413208,
"eval_runtime": 27.73,
"eval_samples_per_second": 14.064,
"eval_steps_per_second": 3.534,
"step": 100
},
{
"epoch": 0.43581445523193096,
"grad_norm": 0.3170960545539856,
"learning_rate": 5.330452921628497e-05,
"loss": 0.9696,
"step": 101
},
{
"epoch": 0.4401294498381877,
"grad_norm": 0.22053739428520203,
"learning_rate": 5.247918773366112e-05,
"loss": 1.4546,
"step": 102
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.1882530003786087,
"learning_rate": 5.165316846586541e-05,
"loss": 1.4966,
"step": 103
},
{
"epoch": 0.4487594390507012,
"grad_norm": 0.1807342767715454,
"learning_rate": 5.0826697238317935e-05,
"loss": 1.4385,
"step": 104
},
{
"epoch": 0.45307443365695793,
"grad_norm": 0.17745080590248108,
"learning_rate": 5e-05,
"loss": 1.5556,
"step": 105
},
{
"epoch": 0.45738942826321466,
"grad_norm": 0.17979130148887634,
"learning_rate": 4.917330276168208e-05,
"loss": 1.3778,
"step": 106
},
{
"epoch": 0.4617044228694714,
"grad_norm": 0.17751434445381165,
"learning_rate": 4.834683153413459e-05,
"loss": 1.5271,
"step": 107
},
{
"epoch": 0.46601941747572817,
"grad_norm": 0.1694229394197464,
"learning_rate": 4.7520812266338885e-05,
"loss": 1.4857,
"step": 108
},
{
"epoch": 0.4703344120819849,
"grad_norm": 0.17555078864097595,
"learning_rate": 4.669547078371504e-05,
"loss": 1.5168,
"step": 109
},
{
"epoch": 0.4746494066882416,
"grad_norm": 0.17622652649879456,
"learning_rate": 4.5871032726383386e-05,
"loss": 1.4157,
"step": 110
},
{
"epoch": 0.47896440129449835,
"grad_norm": 0.17620591819286346,
"learning_rate": 4.504772348747687e-05,
"loss": 1.3489,
"step": 111
},
{
"epoch": 0.48327939590075514,
"grad_norm": 0.17374135553836823,
"learning_rate": 4.4225768151520694e-05,
"loss": 1.2986,
"step": 112
},
{
"epoch": 0.48759439050701187,
"grad_norm": 0.18013916909694672,
"learning_rate": 4.3405391432896555e-05,
"loss": 1.3957,
"step": 113
},
{
"epoch": 0.4919093851132686,
"grad_norm": 0.18480269610881805,
"learning_rate": 4.2586817614407895e-05,
"loss": 1.3487,
"step": 114
},
{
"epoch": 0.4962243797195254,
"grad_norm": 0.18366093933582306,
"learning_rate": 4.17702704859633e-05,
"loss": 1.2979,
"step": 115
},
{
"epoch": 0.5005393743257821,
"grad_norm": 0.1900927573442459,
"learning_rate": 4.095597328339452e-05,
"loss": 1.2878,
"step": 116
},
{
"epoch": 0.5048543689320388,
"grad_norm": 0.18152514100074768,
"learning_rate": 4.0144148627425993e-05,
"loss": 1.151,
"step": 117
},
{
"epoch": 0.5091693635382956,
"grad_norm": 0.19675031304359436,
"learning_rate": 3.933501846281267e-05,
"loss": 1.2083,
"step": 118
},
{
"epoch": 0.5134843581445523,
"grad_norm": 0.1963890641927719,
"learning_rate": 3.852880399766243e-05,
"loss": 1.0812,
"step": 119
},
{
"epoch": 0.517799352750809,
"grad_norm": 0.20205821096897125,
"learning_rate": 3.772572564296005e-05,
"loss": 1.2502,
"step": 120
},
{
"epoch": 0.5221143473570659,
"grad_norm": 0.21323993802070618,
"learning_rate": 3.6926002952309016e-05,
"loss": 1.0091,
"step": 121
},
{
"epoch": 0.5264293419633226,
"grad_norm": 0.22073538601398468,
"learning_rate": 3.612985456190778e-05,
"loss": 1.025,
"step": 122
},
{
"epoch": 0.5307443365695793,
"grad_norm": 0.2334536910057068,
"learning_rate": 3.533749813077677e-05,
"loss": 0.9833,
"step": 123
},
{
"epoch": 0.535059331175836,
"grad_norm": 0.2239895462989807,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.9613,
"step": 124
},
{
"epoch": 0.5393743257820928,
"grad_norm": 0.22818151116371155,
"learning_rate": 3.3765026539765834e-05,
"loss": 0.7748,
"step": 125
},
{
"epoch": 0.5436893203883495,
"grad_norm": 0.23672465980052948,
"learning_rate": 3.298534127791785e-05,
"loss": 0.9676,
"step": 126
},
{
"epoch": 0.5480043149946062,
"grad_norm": 0.22591139376163483,
"learning_rate": 3.221030765387417e-05,
"loss": 0.8771,
"step": 127
},
{
"epoch": 0.552319309600863,
"grad_norm": 0.2156669944524765,
"learning_rate": 3.144013755408895e-05,
"loss": 0.8455,
"step": 128
},
{
"epoch": 0.5566343042071198,
"grad_norm": 0.25983619689941406,
"learning_rate": 3.0675041535377405e-05,
"loss": 0.8735,
"step": 129
},
{
"epoch": 0.5609492988133765,
"grad_norm": 0.23810257017612457,
"learning_rate": 2.991522876735154e-05,
"loss": 0.9294,
"step": 130
},
{
"epoch": 0.5652642934196332,
"grad_norm": 0.23978839814662933,
"learning_rate": 2.916090697523549e-05,
"loss": 0.8999,
"step": 131
},
{
"epoch": 0.56957928802589,
"grad_norm": 0.23231779038906097,
"learning_rate": 2.8412282383075363e-05,
"loss": 0.8487,
"step": 132
},
{
"epoch": 0.5738942826321467,
"grad_norm": 0.2106061726808548,
"learning_rate": 2.766955965735968e-05,
"loss": 0.7526,
"step": 133
},
{
"epoch": 0.5782092772384034,
"grad_norm": 0.22033323347568512,
"learning_rate": 2.693294185106562e-05,
"loss": 0.6419,
"step": 134
},
{
"epoch": 0.5825242718446602,
"grad_norm": 0.22057168185710907,
"learning_rate": 2.6202630348146324e-05,
"loss": 0.7992,
"step": 135
},
{
"epoch": 0.5868392664509169,
"grad_norm": 0.2245577722787857,
"learning_rate": 2.547882480847461e-05,
"loss": 0.9545,
"step": 136
},
{
"epoch": 0.5911542610571737,
"grad_norm": 0.21244750916957855,
"learning_rate": 2.476172311325783e-05,
"loss": 0.7485,
"step": 137
},
{
"epoch": 0.5954692556634305,
"grad_norm": 0.21956144273281097,
"learning_rate": 2.405152131093926e-05,
"loss": 0.8313,
"step": 138
},
{
"epoch": 0.5997842502696872,
"grad_norm": 0.23018768429756165,
"learning_rate": 2.3348413563600325e-05,
"loss": 0.7689,
"step": 139
},
{
"epoch": 0.6040992448759439,
"grad_norm": 0.22504369914531708,
"learning_rate": 2.2652592093878666e-05,
"loss": 0.733,
"step": 140
},
{
"epoch": 0.6084142394822006,
"grad_norm": 0.22533173859119415,
"learning_rate": 2.196424713241637e-05,
"loss": 0.6765,
"step": 141
},
{
"epoch": 0.6127292340884574,
"grad_norm": 0.22474339604377747,
"learning_rate": 2.128356686585282e-05,
"loss": 0.6602,
"step": 142
},
{
"epoch": 0.6170442286947141,
"grad_norm": 0.2265556901693344,
"learning_rate": 2.061073738537635e-05,
"loss": 0.7123,
"step": 143
},
{
"epoch": 0.6213592233009708,
"grad_norm": 0.2438202053308487,
"learning_rate": 1.9945942635848748e-05,
"loss": 0.7158,
"step": 144
},
{
"epoch": 0.6256742179072277,
"grad_norm": 0.22435429692268372,
"learning_rate": 1.928936436551661e-05,
"loss": 0.6925,
"step": 145
},
{
"epoch": 0.6299892125134844,
"grad_norm": 0.2435307502746582,
"learning_rate": 1.8641182076323148e-05,
"loss": 0.7221,
"step": 146
},
{
"epoch": 0.6343042071197411,
"grad_norm": 0.26046523451805115,
"learning_rate": 1.800157297483417e-05,
"loss": 1.1229,
"step": 147
},
{
"epoch": 0.6386192017259978,
"grad_norm": 0.3411429524421692,
"learning_rate": 1.7370711923791567e-05,
"loss": 1.0276,
"step": 148
},
{
"epoch": 0.6429341963322546,
"grad_norm": 0.4017745852470398,
"learning_rate": 1.6748771394307585e-05,
"loss": 1.2985,
"step": 149
},
{
"epoch": 0.6472491909385113,
"grad_norm": 0.47186926007270813,
"learning_rate": 1.6135921418712956e-05,
"loss": 1.4719,
"step": 150
},
{
"epoch": 0.6472491909385113,
"eval_loss": 0.9981580972671509,
"eval_runtime": 27.7363,
"eval_samples_per_second": 14.061,
"eval_steps_per_second": 3.533,
"step": 150
},
{
"epoch": 0.651564185544768,
"grad_norm": 0.25626233220100403,
"learning_rate": 1.553232954407171e-05,
"loss": 0.7416,
"step": 151
},
{
"epoch": 0.6558791801510249,
"grad_norm": 0.27391961216926575,
"learning_rate": 1.4938160786375572e-05,
"loss": 1.3329,
"step": 152
},
{
"epoch": 0.6601941747572816,
"grad_norm": 0.2119150608778,
"learning_rate": 1.435357758543015e-05,
"loss": 1.4031,
"step": 153
},
{
"epoch": 0.6645091693635383,
"grad_norm": 0.21125996112823486,
"learning_rate": 1.3778739760445552e-05,
"loss": 1.5411,
"step": 154
},
{
"epoch": 0.668824163969795,
"grad_norm": 0.20459342002868652,
"learning_rate": 1.3213804466343421e-05,
"loss": 1.5654,
"step": 155
},
{
"epoch": 0.6731391585760518,
"grad_norm": 0.20502012968063354,
"learning_rate": 1.2658926150792322e-05,
"loss": 1.5269,
"step": 156
},
{
"epoch": 0.6774541531823085,
"grad_norm": 0.18337692320346832,
"learning_rate": 1.2114256511983274e-05,
"loss": 1.4455,
"step": 157
},
{
"epoch": 0.6817691477885652,
"grad_norm": 0.17909735441207886,
"learning_rate": 1.157994445715706e-05,
"loss": 1.4208,
"step": 158
},
{
"epoch": 0.686084142394822,
"grad_norm": 0.17741838097572327,
"learning_rate": 1.1056136061894384e-05,
"loss": 1.5558,
"step": 159
},
{
"epoch": 0.6903991370010788,
"grad_norm": 0.16822201013565063,
"learning_rate": 1.0542974530180327e-05,
"loss": 1.3754,
"step": 160
},
{
"epoch": 0.6947141316073355,
"grad_norm": 0.17058689892292023,
"learning_rate": 1.0040600155253765e-05,
"loss": 1.364,
"step": 161
},
{
"epoch": 0.6990291262135923,
"grad_norm": 0.1698887050151825,
"learning_rate": 9.549150281252633e-06,
"loss": 1.4269,
"step": 162
},
{
"epoch": 0.703344120819849,
"grad_norm": 0.17073321342468262,
"learning_rate": 9.068759265665384e-06,
"loss": 1.3197,
"step": 163
},
{
"epoch": 0.7076591154261057,
"grad_norm": 0.1761074960231781,
"learning_rate": 8.599558442598998e-06,
"loss": 1.3182,
"step": 164
},
{
"epoch": 0.7119741100323624,
"grad_norm": 0.17921654880046844,
"learning_rate": 8.141676086873572e-06,
"loss": 1.3117,
"step": 165
},
{
"epoch": 0.7162891046386192,
"grad_norm": 0.18380288779735565,
"learning_rate": 7.695237378953223e-06,
"loss": 1.1363,
"step": 166
},
{
"epoch": 0.7206040992448759,
"grad_norm": 0.1892642080783844,
"learning_rate": 7.260364370723044e-06,
"loss": 1.3208,
"step": 167
},
{
"epoch": 0.7249190938511327,
"grad_norm": 0.19996848702430725,
"learning_rate": 6.837175952121306e-06,
"loss": 1.1256,
"step": 168
},
{
"epoch": 0.7292340884573895,
"grad_norm": 0.19909560680389404,
"learning_rate": 6.425787818636131e-06,
"loss": 1.2357,
"step": 169
},
{
"epoch": 0.7335490830636462,
"grad_norm": 0.21045197546482086,
"learning_rate": 6.026312439675552e-06,
"loss": 1.0831,
"step": 170
},
{
"epoch": 0.7378640776699029,
"grad_norm": 0.2156793177127838,
"learning_rate": 5.6388590278194096e-06,
"loss": 1.0637,
"step": 171
},
{
"epoch": 0.7421790722761596,
"grad_norm": 0.2282598465681076,
"learning_rate": 5.263533508961827e-06,
"loss": 1.0763,
"step": 172
},
{
"epoch": 0.7464940668824164,
"grad_norm": 0.2251502424478531,
"learning_rate": 4.900438493352055e-06,
"loss": 1.0099,
"step": 173
},
{
"epoch": 0.7508090614886731,
"grad_norm": 0.22714872658252716,
"learning_rate": 4.549673247541875e-06,
"loss": 0.9487,
"step": 174
},
{
"epoch": 0.7551240560949298,
"grad_norm": 0.20686142146587372,
"learning_rate": 4.2113336672471245e-06,
"loss": 1.0784,
"step": 175
},
{
"epoch": 0.7594390507011867,
"grad_norm": 0.2434941977262497,
"learning_rate": 3.885512251130763e-06,
"loss": 0.9793,
"step": 176
},
{
"epoch": 0.7637540453074434,
"grad_norm": 0.2319464087486267,
"learning_rate": 3.5722980755146517e-06,
"loss": 0.8571,
"step": 177
},
{
"epoch": 0.7680690399137001,
"grad_norm": 0.23625849187374115,
"learning_rate": 3.271776770026963e-06,
"loss": 0.7877,
"step": 178
},
{
"epoch": 0.7723840345199569,
"grad_norm": 0.22388741374015808,
"learning_rate": 2.9840304941919415e-06,
"loss": 0.8952,
"step": 179
},
{
"epoch": 0.7766990291262136,
"grad_norm": 0.23320500552654266,
"learning_rate": 2.7091379149682685e-06,
"loss": 0.9672,
"step": 180
},
{
"epoch": 0.7810140237324703,
"grad_norm": 0.23083074390888214,
"learning_rate": 2.4471741852423237e-06,
"loss": 0.9844,
"step": 181
},
{
"epoch": 0.785329018338727,
"grad_norm": 0.22144003212451935,
"learning_rate": 2.1982109232821178e-06,
"loss": 0.7437,
"step": 182
},
{
"epoch": 0.7896440129449838,
"grad_norm": 0.2303430140018463,
"learning_rate": 1.962316193157593e-06,
"loss": 0.777,
"step": 183
},
{
"epoch": 0.7939590075512406,
"grad_norm": 0.22940360009670258,
"learning_rate": 1.7395544861325718e-06,
"loss": 0.7314,
"step": 184
},
{
"epoch": 0.7982740021574973,
"grad_norm": 0.2274598479270935,
"learning_rate": 1.5299867030334814e-06,
"loss": 0.7879,
"step": 185
},
{
"epoch": 0.8025889967637541,
"grad_norm": 0.22800025343894958,
"learning_rate": 1.333670137599713e-06,
"loss": 0.8764,
"step": 186
},
{
"epoch": 0.8069039913700108,
"grad_norm": 0.23226092755794525,
"learning_rate": 1.1506584608200367e-06,
"loss": 1.0145,
"step": 187
},
{
"epoch": 0.8112189859762675,
"grad_norm": 0.2228040099143982,
"learning_rate": 9.810017062595322e-07,
"loss": 0.751,
"step": 188
},
{
"epoch": 0.8155339805825242,
"grad_norm": 0.21984310448169708,
"learning_rate": 8.247462563808817e-07,
"loss": 0.73,
"step": 189
},
{
"epoch": 0.819848975188781,
"grad_norm": 0.22310470044612885,
"learning_rate": 6.819348298638839e-07,
"loss": 0.7142,
"step": 190
},
{
"epoch": 0.8241639697950378,
"grad_norm": 0.23268581926822662,
"learning_rate": 5.526064699265753e-07,
"loss": 0.7824,
"step": 191
},
{
"epoch": 0.8284789644012945,
"grad_norm": 0.23572231829166412,
"learning_rate": 4.367965336512403e-07,
"loss": 0.7614,
"step": 192
},
{
"epoch": 0.8327939590075513,
"grad_norm": 0.23753508925437927,
"learning_rate": 3.3453668231809286e-07,
"loss": 0.7199,
"step": 193
},
{
"epoch": 0.837108953613808,
"grad_norm": 0.2294815182685852,
"learning_rate": 2.458548727494292e-07,
"loss": 0.6749,
"step": 194
},
{
"epoch": 0.8414239482200647,
"grad_norm": 0.238484725356102,
"learning_rate": 1.7077534966650766e-07,
"loss": 0.7044,
"step": 195
},
{
"epoch": 0.8457389428263214,
"grad_norm": 0.2378358095884323,
"learning_rate": 1.0931863906127327e-07,
"loss": 0.7034,
"step": 196
},
{
"epoch": 0.8500539374325782,
"grad_norm": 0.2880394160747528,
"learning_rate": 6.150154258476315e-08,
"loss": 1.12,
"step": 197
},
{
"epoch": 0.8543689320388349,
"grad_norm": 0.3312850892543793,
"learning_rate": 2.7337132953697554e-08,
"loss": 0.9537,
"step": 198
},
{
"epoch": 0.8586839266450917,
"grad_norm": 0.3399243950843811,
"learning_rate": 6.834750376549792e-09,
"loss": 1.0413,
"step": 199
},
{
"epoch": 0.8629989212513485,
"grad_norm": 0.43729159235954285,
"learning_rate": 0.0,
"loss": 1.2303,
"step": 200
},
{
"epoch": 0.8629989212513485,
"eval_loss": 0.9987650513648987,
"eval_runtime": 27.7231,
"eval_samples_per_second": 14.068,
"eval_steps_per_second": 3.535,
"step": 200
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 1
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.84377737068544e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}