| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 154, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012987012987012988, | |
| "grad_norm": 0.7649294288089501, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0642, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.025974025974025976, | |
| "grad_norm": 0.16744224046876113, | |
| "learning_rate": 4.999472998758978e-05, | |
| "loss": 1.0811, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03896103896103896, | |
| "grad_norm": 0.1506244473047185, | |
| "learning_rate": 4.99789221722016e-05, | |
| "loss": 1.0819, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05194805194805195, | |
| "grad_norm": 0.15366192883175545, | |
| "learning_rate": 4.995258321842611e-05, | |
| "loss": 1.1197, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06493506493506493, | |
| "grad_norm": 0.11707148712561326, | |
| "learning_rate": 4.991572423079236e-05, | |
| "loss": 1.0702, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07792207792207792, | |
| "grad_norm": 0.13073876345749866, | |
| "learning_rate": 4.986836074908616e-05, | |
| "loss": 1.0454, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 0.14345020845358708, | |
| "learning_rate": 4.98105127417984e-05, | |
| "loss": 1.0145, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1038961038961039, | |
| "grad_norm": 0.12284158867400687, | |
| "learning_rate": 4.974220459770639e-05, | |
| "loss": 0.9749, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11688311688311688, | |
| "grad_norm": 0.1316122136803943, | |
| "learning_rate": 4.966346511559149e-05, | |
| "loss": 0.9844, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.12987012987012986, | |
| "grad_norm": 0.12129084788561034, | |
| "learning_rate": 4.957432749209755e-05, | |
| "loss": 0.937, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 0.11402446255258117, | |
| "learning_rate": 4.9474829307735115e-05, | |
| "loss": 0.9119, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15584415584415584, | |
| "grad_norm": 0.13670373476126882, | |
| "learning_rate": 4.9365012511037514e-05, | |
| "loss": 0.9048, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.16883116883116883, | |
| "grad_norm": 0.17167936568998776, | |
| "learning_rate": 4.9244923400875245e-05, | |
| "loss": 0.8975, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.13651826129682004, | |
| "learning_rate": 4.911461260693638e-05, | |
| "loss": 0.815, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19480519480519481, | |
| "grad_norm": 0.1973535264967678, | |
| "learning_rate": 4.8974135068381036e-05, | |
| "loss": 0.7728, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2077922077922078, | |
| "grad_norm": 0.17362622068648217, | |
| "learning_rate": 4.882355001067892e-05, | |
| "loss": 0.7792, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.22077922077922077, | |
| "grad_norm": 0.1972301833866661, | |
| "learning_rate": 4.8662920920639866e-05, | |
| "loss": 0.7213, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23376623376623376, | |
| "grad_norm": 0.20220023509712592, | |
| "learning_rate": 4.849231551964771e-05, | |
| "loss": 0.6796, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.24675324675324675, | |
| "grad_norm": 0.21311631984189247, | |
| "learning_rate": 4.8311805735108894e-05, | |
| "loss": 0.6554, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 0.2369381772825553, | |
| "learning_rate": 4.81214676701278e-05, | |
| "loss": 0.563, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.26507656791198936, | |
| "learning_rate": 4.792138157142158e-05, | |
| "loss": 0.5178, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 0.26498752467037984, | |
| "learning_rate": 4.7711631795488096e-05, | |
| "loss": 0.4661, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2987012987012987, | |
| "grad_norm": 0.24893170518457297, | |
| "learning_rate": 4.749230677304114e-05, | |
| "loss": 0.4262, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3116883116883117, | |
| "grad_norm": 0.22992415120842585, | |
| "learning_rate": 4.726349897172791e-05, | |
| "loss": 0.3719, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3246753246753247, | |
| "grad_norm": 0.17800562793366614, | |
| "learning_rate": 4.702530485714461e-05, | |
| "loss": 0.3395, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.33766233766233766, | |
| "grad_norm": 0.1678374692636547, | |
| "learning_rate": 4.677782485216644e-05, | |
| "loss": 0.3205, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.35064935064935066, | |
| "grad_norm": 0.14760981845430549, | |
| "learning_rate": 4.6521163294609196e-05, | |
| "loss": 0.297, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.12851418569866138, | |
| "learning_rate": 4.625542839324036e-05, | |
| "loss": 0.3212, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.37662337662337664, | |
| "grad_norm": 0.12667316817519966, | |
| "learning_rate": 4.598073218215817e-05, | |
| "loss": 0.3153, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.38961038961038963, | |
| "grad_norm": 0.11051907979123522, | |
| "learning_rate": 4.5697190473557946e-05, | |
| "loss": 0.2752, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4025974025974026, | |
| "grad_norm": 0.1268943489369776, | |
| "learning_rate": 4.540492280890555e-05, | |
| "loss": 0.2866, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4155844155844156, | |
| "grad_norm": 0.12871191005606444, | |
| "learning_rate": 4.510405240853854e-05, | |
| "loss": 0.2705, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 0.13732367872755494, | |
| "learning_rate": 4.4794706119716455e-05, | |
| "loss": 0.2961, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.44155844155844154, | |
| "grad_norm": 0.12777955862078405, | |
| "learning_rate": 4.447701436314176e-05, | |
| "loss": 0.2868, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.12098600481328763, | |
| "learning_rate": 4.415111107797445e-05, | |
| "loss": 0.256, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4675324675324675, | |
| "grad_norm": 0.1284047348249079, | |
| "learning_rate": 4.381713366536311e-05, | |
| "loss": 0.2679, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4805194805194805, | |
| "grad_norm": 0.11163107591942552, | |
| "learning_rate": 4.347522293051648e-05, | |
| "loss": 0.2457, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4935064935064935, | |
| "grad_norm": 0.14312777584685432, | |
| "learning_rate": 4.312552302333982e-05, | |
| "loss": 0.2615, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5064935064935064, | |
| "grad_norm": 0.14239613661086223, | |
| "learning_rate": 4.276818137766118e-05, | |
| "loss": 0.2722, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 0.12803795478640206, | |
| "learning_rate": 4.2403348649073174e-05, | |
| "loss": 0.2662, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5324675324675324, | |
| "grad_norm": 0.10402277357673333, | |
| "learning_rate": 4.203117865141635e-05, | |
| "loss": 0.2454, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.10586149604773042, | |
| "learning_rate": 4.1651828291931264e-05, | |
| "loss": 0.2684, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5584415584415584, | |
| "grad_norm": 0.1212655510313763, | |
| "learning_rate": 4.126545750510605e-05, | |
| "loss": 0.258, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 0.11476596613325958, | |
| "learning_rate": 4.0872229185248075e-05, | |
| "loss": 0.2716, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5844155844155844, | |
| "grad_norm": 0.11875414394829936, | |
| "learning_rate": 4.047230911780737e-05, | |
| "loss": 0.2255, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5974025974025974, | |
| "grad_norm": 0.11819788647354397, | |
| "learning_rate": 4.0065865909481417e-05, | |
| "loss": 0.2492, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6103896103896104, | |
| "grad_norm": 0.13638858604883952, | |
| "learning_rate": 3.965307091713037e-05, | |
| "loss": 0.2494, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6233766233766234, | |
| "grad_norm": 0.115798492930415, | |
| "learning_rate": 3.923409817553284e-05, | |
| "loss": 0.2326, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 0.10245941653264991, | |
| "learning_rate": 3.880912432401265e-05, | |
| "loss": 0.235, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6493506493506493, | |
| "grad_norm": 0.09819274391733306, | |
| "learning_rate": 3.837832853196751e-05, | |
| "loss": 0.246, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6623376623376623, | |
| "grad_norm": 0.0979091297539171, | |
| "learning_rate": 3.794189242333106e-05, | |
| "loss": 0.2588, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6753246753246753, | |
| "grad_norm": 0.11797293731222826, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.2333, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6883116883116883, | |
| "grad_norm": 0.10853973446787894, | |
| "learning_rate": 3.705283756425872e-05, | |
| "loss": 0.2564, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.7012987012987013, | |
| "grad_norm": 0.09739725780688968, | |
| "learning_rate": 3.6600593640234086e-05, | |
| "loss": 0.2467, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 0.13136414642559058, | |
| "learning_rate": 3.6143458894413465e-05, | |
| "loss": 0.2489, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.10623115343843, | |
| "learning_rate": 3.568162605525953e-05, | |
| "loss": 0.2445, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7402597402597403, | |
| "grad_norm": 0.1023205505796002, | |
| "learning_rate": 3.5215289831955786e-05, | |
| "loss": 0.2499, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7532467532467533, | |
| "grad_norm": 0.1002792505451181, | |
| "learning_rate": 3.474464683231698e-05, | |
| "loss": 0.2261, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7662337662337663, | |
| "grad_norm": 0.11002829565848689, | |
| "learning_rate": 3.426989547989902e-05, | |
| "loss": 0.2513, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 0.10861164319015883, | |
| "learning_rate": 3.379123593034342e-05, | |
| "loss": 0.2252, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7922077922077922, | |
| "grad_norm": 0.09758507903096522, | |
| "learning_rate": 3.330886998699149e-05, | |
| "loss": 0.2272, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.8051948051948052, | |
| "grad_norm": 0.10548706725662228, | |
| "learning_rate": 3.282300101580386e-05, | |
| "loss": 0.2393, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.09385065956952136, | |
| "learning_rate": 3.2333833859621153e-05, | |
| "loss": 0.2179, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8311688311688312, | |
| "grad_norm": 0.1608126010863833, | |
| "learning_rate": 3.1841574751802076e-05, | |
| "loss": 0.2341, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.8441558441558441, | |
| "grad_norm": 0.10253273432410331, | |
| "learning_rate": 3.13464312292752e-05, | |
| "loss": 0.2258, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.11329982372255963, | |
| "learning_rate": 3.084861204504122e-05, | |
| "loss": 0.2223, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8701298701298701, | |
| "grad_norm": 0.13350806545341848, | |
| "learning_rate": 3.0348327080162435e-05, | |
| "loss": 0.2456, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8831168831168831, | |
| "grad_norm": 0.10995445961760825, | |
| "learning_rate": 2.9845787255276753e-05, | |
| "loss": 0.2321, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8961038961038961, | |
| "grad_norm": 0.1037165714945774, | |
| "learning_rate": 2.9341204441673266e-05, | |
| "loss": 0.2171, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.10919814710978967, | |
| "learning_rate": 2.8834791371967142e-05, | |
| "loss": 0.2472, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.922077922077922, | |
| "grad_norm": 0.09681778664572314, | |
| "learning_rate": 2.8326761550411345e-05, | |
| "loss": 0.2364, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.935064935064935, | |
| "grad_norm": 0.1074259579148838, | |
| "learning_rate": 2.781732916288303e-05, | |
| "loss": 0.2391, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.948051948051948, | |
| "grad_norm": 0.1054775490154967, | |
| "learning_rate": 2.7306708986582553e-05, | |
| "loss": 0.2595, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.961038961038961, | |
| "grad_norm": 0.09815954447405159, | |
| "learning_rate": 2.679511629948319e-05, | |
| "loss": 0.2426, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.974025974025974, | |
| "grad_norm": 0.09306732537511961, | |
| "learning_rate": 2.628276678956974e-05, | |
| "loss": 0.2423, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.987012987012987, | |
| "grad_norm": 0.11169931049945912, | |
| "learning_rate": 2.5769876463904265e-05, | |
| "loss": 0.2348, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.10138472477596566, | |
| "learning_rate": 2.5256661557557247e-05, | |
| "loss": 0.2387, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.0129870129870129, | |
| "grad_norm": 0.09555843194904001, | |
| "learning_rate": 2.4743338442442755e-05, | |
| "loss": 0.2323, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.025974025974026, | |
| "grad_norm": 0.1181816997966792, | |
| "learning_rate": 2.4230123536095748e-05, | |
| "loss": 0.2436, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.0389610389610389, | |
| "grad_norm": 0.08998379085998691, | |
| "learning_rate": 2.3717233210430256e-05, | |
| "loss": 0.2208, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.051948051948052, | |
| "grad_norm": 0.09999438916245779, | |
| "learning_rate": 2.3204883700516812e-05, | |
| "loss": 0.236, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0649350649350648, | |
| "grad_norm": 0.10658908314699578, | |
| "learning_rate": 2.2693291013417453e-05, | |
| "loss": 0.2267, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.077922077922078, | |
| "grad_norm": 0.1017206414672713, | |
| "learning_rate": 2.2182670837116975e-05, | |
| "loss": 0.2476, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.10154042905245364, | |
| "learning_rate": 2.1673238449588668e-05, | |
| "loss": 0.2455, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.103896103896104, | |
| "grad_norm": 0.09488446331830326, | |
| "learning_rate": 2.116520862803286e-05, | |
| "loss": 0.2066, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1168831168831168, | |
| "grad_norm": 0.10202897851829487, | |
| "learning_rate": 2.0658795558326743e-05, | |
| "loss": 0.2007, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.12987012987013, | |
| "grad_norm": 0.1183943667827613, | |
| "learning_rate": 2.015421274472325e-05, | |
| "loss": 0.2026, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.09921431477749063, | |
| "learning_rate": 1.965167291983757e-05, | |
| "loss": 0.2104, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.155844155844156, | |
| "grad_norm": 0.10093734214597283, | |
| "learning_rate": 1.9151387954958794e-05, | |
| "loss": 0.2046, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.1688311688311688, | |
| "grad_norm": 0.09360429992532161, | |
| "learning_rate": 1.8653568770724806e-05, | |
| "loss": 0.2131, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 0.12020649129202392, | |
| "learning_rate": 1.815842524819793e-05, | |
| "loss": 0.1989, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1948051948051948, | |
| "grad_norm": 0.10680381676149939, | |
| "learning_rate": 1.7666166140378852e-05, | |
| "loss": 0.2078, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.2077922077922079, | |
| "grad_norm": 0.09568299370975282, | |
| "learning_rate": 1.7176998984196146e-05, | |
| "loss": 0.2101, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2207792207792207, | |
| "grad_norm": 0.12879980775489425, | |
| "learning_rate": 1.6691130013008514e-05, | |
| "loss": 0.2166, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.2337662337662338, | |
| "grad_norm": 0.09751174219709864, | |
| "learning_rate": 1.620876406965658e-05, | |
| "loss": 0.216, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2467532467532467, | |
| "grad_norm": 0.12307588066370435, | |
| "learning_rate": 1.5730104520100982e-05, | |
| "loss": 0.2355, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2597402597402598, | |
| "grad_norm": 0.0984790807665978, | |
| "learning_rate": 1.5255353167683017e-05, | |
| "loss": 0.2283, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.11200751310649058, | |
| "learning_rate": 1.4784710168044213e-05, | |
| "loss": 0.2401, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.10465475693157676, | |
| "learning_rate": 1.4318373944740484e-05, | |
| "loss": 0.1995, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.2987012987012987, | |
| "grad_norm": 0.10184278002913967, | |
| "learning_rate": 1.3856541105586545e-05, | |
| "loss": 0.2068, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.3116883116883118, | |
| "grad_norm": 0.11277983230555819, | |
| "learning_rate": 1.339940635976592e-05, | |
| "loss": 0.1994, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3246753246753247, | |
| "grad_norm": 0.10951913478829725, | |
| "learning_rate": 1.2947162435741278e-05, | |
| "loss": 0.2446, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3376623376623376, | |
| "grad_norm": 0.10294119354543721, | |
| "learning_rate": 1.2500000000000006e-05, | |
| "loss": 0.2223, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3506493506493507, | |
| "grad_norm": 0.10683807554925215, | |
| "learning_rate": 1.205810757666894e-05, | |
| "loss": 0.2225, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.10650031782405078, | |
| "learning_rate": 1.1621671468032493e-05, | |
| "loss": 0.2026, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3766233766233766, | |
| "grad_norm": 0.10799384532554564, | |
| "learning_rate": 1.1190875675987356e-05, | |
| "loss": 0.2395, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3896103896103895, | |
| "grad_norm": 0.1099680907946163, | |
| "learning_rate": 1.0765901824467167e-05, | |
| "loss": 0.2266, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.4025974025974026, | |
| "grad_norm": 0.10063986485693889, | |
| "learning_rate": 1.0346929082869641e-05, | |
| "loss": 0.2244, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.4155844155844157, | |
| "grad_norm": 0.09898088073892639, | |
| "learning_rate": 9.934134090518593e-06, | |
| "loss": 0.1933, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.10525578859634234, | |
| "learning_rate": 9.527690882192636e-06, | |
| "loss": 0.2202, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4415584415584415, | |
| "grad_norm": 0.1087751424503037, | |
| "learning_rate": 9.127770814751933e-06, | |
| "loss": 0.2164, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.10675193255605053, | |
| "learning_rate": 8.734542494893955e-06, | |
| "loss": 0.2557, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4675324675324675, | |
| "grad_norm": 0.09433236678416125, | |
| "learning_rate": 8.348171708068747e-06, | |
| "loss": 0.2047, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4805194805194806, | |
| "grad_norm": 0.10552231126115215, | |
| "learning_rate": 7.968821348583644e-06, | |
| "loss": 0.2354, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.4935064935064934, | |
| "grad_norm": 0.1090179336847755, | |
| "learning_rate": 7.5966513509268365e-06, | |
| "loss": 0.1903, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.5064935064935066, | |
| "grad_norm": 0.10671951695163927, | |
| "learning_rate": 7.231818622338823e-06, | |
| "loss": 0.2296, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.5194805194805194, | |
| "grad_norm": 0.10006542534798159, | |
| "learning_rate": 6.8744769766601854e-06, | |
| "loss": 0.2204, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5324675324675323, | |
| "grad_norm": 0.10963040390997872, | |
| "learning_rate": 6.524777069483526e-06, | |
| "loss": 0.2086, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 0.09907281406279463, | |
| "learning_rate": 6.182866334636889e-06, | |
| "loss": 0.1897, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.5584415584415585, | |
| "grad_norm": 0.10508976076834439, | |
| "learning_rate": 5.848888922025553e-06, | |
| "loss": 0.2282, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.134246328671402, | |
| "learning_rate": 5.522985636858239e-06, | |
| "loss": 0.2264, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5844155844155843, | |
| "grad_norm": 0.1059879782400766, | |
| "learning_rate": 5.205293880283552e-06, | |
| "loss": 0.1896, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5974025974025974, | |
| "grad_norm": 0.10587695869132589, | |
| "learning_rate": 4.8959475914614554e-06, | |
| "loss": 0.2229, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.6103896103896105, | |
| "grad_norm": 0.12969941386231867, | |
| "learning_rate": 4.5950771910944605e-06, | |
| "loss": 0.2168, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6233766233766234, | |
| "grad_norm": 0.23291844260878844, | |
| "learning_rate": 4.3028095264420535e-06, | |
| "loss": 0.2216, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.12515572322774973, | |
| "learning_rate": 4.019267817841835e-06, | |
| "loss": 0.2287, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6493506493506493, | |
| "grad_norm": 0.10857583181073842, | |
| "learning_rate": 3.7445716067596503e-06, | |
| "loss": 0.1996, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6623376623376624, | |
| "grad_norm": 0.11253155054511829, | |
| "learning_rate": 3.478836705390809e-06, | |
| "loss": 0.2334, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6753246753246753, | |
| "grad_norm": 0.10767090625172462, | |
| "learning_rate": 3.222175147833556e-06, | |
| "loss": 0.2188, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6883116883116882, | |
| "grad_norm": 0.11985791222342175, | |
| "learning_rate": 2.974695142855388e-06, | |
| "loss": 0.2386, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.7012987012987013, | |
| "grad_norm": 0.10876307711132034, | |
| "learning_rate": 2.7365010282720952e-06, | |
| "loss": 0.2153, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.12451466520756925, | |
| "learning_rate": 2.507693226958871e-06, | |
| "loss": 0.2031, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 0.100099707537048, | |
| "learning_rate": 2.2883682045119063e-06, | |
| "loss": 0.1956, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7402597402597402, | |
| "grad_norm": 0.11691453452769905, | |
| "learning_rate": 2.0786184285784297e-06, | |
| "loss": 0.2099, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.7532467532467533, | |
| "grad_norm": 0.13343643028694882, | |
| "learning_rate": 1.8785323298722097e-06, | |
| "loss": 0.2442, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7662337662337664, | |
| "grad_norm": 0.10550388815035454, | |
| "learning_rate": 1.6881942648911076e-06, | |
| "loss": 0.2089, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7792207792207793, | |
| "grad_norm": 0.09864757317858987, | |
| "learning_rate": 1.5076844803522922e-06, | |
| "loss": 0.2116, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7922077922077921, | |
| "grad_norm": 0.11851794361322249, | |
| "learning_rate": 1.3370790793601373e-06, | |
| "loss": 0.2316, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.8051948051948052, | |
| "grad_norm": 0.13304376665401985, | |
| "learning_rate": 1.1764499893210878e-06, | |
| "loss": 0.1851, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.11193724349100125, | |
| "learning_rate": 1.0258649316189722e-06, | |
| "loss": 0.1954, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8311688311688312, | |
| "grad_norm": 0.09758738574949906, | |
| "learning_rate": 8.85387393063622e-07, | |
| "loss": 0.2079, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.844155844155844, | |
| "grad_norm": 0.14859065932159254, | |
| "learning_rate": 7.550765991247654e-07, | |
| "loss": 0.2141, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.11555838452156621, | |
| "learning_rate": 6.349874889624962e-07, | |
| "loss": 0.2076, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.87012987012987, | |
| "grad_norm": 0.09704689285498966, | |
| "learning_rate": 5.25170692264887e-07, | |
| "loss": 0.2014, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.883116883116883, | |
| "grad_norm": 0.13020262018312262, | |
| "learning_rate": 4.256725079024554e-07, | |
| "loss": 0.2265, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.896103896103896, | |
| "grad_norm": 0.12798337534645113, | |
| "learning_rate": 3.3653488440851255e-07, | |
| "loss": 0.224, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 0.10348239456828633, | |
| "learning_rate": 2.5779540229361745e-07, | |
| "loss": 0.21, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.922077922077922, | |
| "grad_norm": 0.0993975437154903, | |
| "learning_rate": 1.8948725820160662e-07, | |
| "loss": 0.1995, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.935064935064935, | |
| "grad_norm": 0.11597426567128678, | |
| "learning_rate": 1.3163925091384533e-07, | |
| "loss": 0.2236, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.948051948051948, | |
| "grad_norm": 0.10428192963009006, | |
| "learning_rate": 8.427576920763958e-08, | |
| "loss": 0.1893, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9610389610389611, | |
| "grad_norm": 0.1077579100092834, | |
| "learning_rate": 4.741678157389739e-08, | |
| "loss": 0.1985, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.974025974025974, | |
| "grad_norm": 0.1058358938118017, | |
| "learning_rate": 2.1077827798404726e-08, | |
| "loss": 0.2055, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.987012987012987, | |
| "grad_norm": 0.1523430022468915, | |
| "learning_rate": 5.270012410216185e-09, | |
| "loss": 0.1858, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.11217948034066874, | |
| "learning_rate": 0.0, | |
| "loss": 0.1922, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 154, | |
| "total_flos": 98079354388480.0, | |
| "train_loss": 0.32445608553561295, | |
| "train_runtime": 1604.0651, | |
| "train_samples_per_second": 0.382, | |
| "train_steps_per_second": 0.096 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 154, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 98079354388480.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |