| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9967637540453075, | |
| "eval_steps": 500, | |
| "global_step": 231, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004314994606256742, | |
| "grad_norm": 7.241082191467285, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.5885, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008629989212513484, | |
| "grad_norm": 9.26927661895752, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.8217, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.012944983818770227, | |
| "grad_norm": 7.139358997344971, | |
| "learning_rate": 3e-06, | |
| "loss": 2.1334, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.017259978425026967, | |
| "grad_norm": 7.390294551849365, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.5074, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.021574973031283712, | |
| "grad_norm": 6.828765392303467, | |
| "learning_rate": 5e-06, | |
| "loss": 1.5498, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.025889967637540454, | |
| "grad_norm": 9.9520263671875, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.030204962243797196, | |
| "grad_norm": 7.4763312339782715, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 1.7989, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.034519956850053934, | |
| "grad_norm": 6.728983402252197, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.7428, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.038834951456310676, | |
| "grad_norm": 8.699463844299316, | |
| "learning_rate": 9e-06, | |
| "loss": 1.5321, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.043149946062567425, | |
| "grad_norm": 8.82262897491455, | |
| "learning_rate": 1e-05, | |
| "loss": 2.0886, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04746494066882417, | |
| "grad_norm": 11.480703353881836, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 1.6821, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.05177993527508091, | |
| "grad_norm": 9.1630220413208, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.8223, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05609492988133765, | |
| "grad_norm": 19.579421997070312, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 2.0728, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.06040992448759439, | |
| "grad_norm": 10.23599910736084, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.9451, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06472491909385113, | |
| "grad_norm": 35.375518798828125, | |
| "learning_rate": 1.5e-05, | |
| "loss": 2.1366, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06903991370010787, | |
| "grad_norm": 9.47799301147461, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.8541, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.07335490830636461, | |
| "grad_norm": 8.337750434875488, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 1.9606, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.07766990291262135, | |
| "grad_norm": 9.68613338470459, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.8872, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.08198489751887811, | |
| "grad_norm": 10.925890922546387, | |
| "learning_rate": 1.9e-05, | |
| "loss": 1.8552, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08629989212513485, | |
| "grad_norm": 9.537017822265625, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0876, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09061488673139159, | |
| "grad_norm": 10.281198501586914, | |
| "learning_rate": 2.1e-05, | |
| "loss": 1.8258, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.09492988133764833, | |
| "grad_norm": 10.57076644897461, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 2.1992, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.09924487594390508, | |
| "grad_norm": 10.87122631072998, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 2.2141, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.10355987055016182, | |
| "grad_norm": 8.736374855041504, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.6908, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10787486515641856, | |
| "grad_norm": 8.651305198669434, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.866, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1121898597626753, | |
| "grad_norm": 7.790002346038818, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 2.1399, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.11650485436893204, | |
| "grad_norm": 12.541814804077148, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 1.7723, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.12081984897518878, | |
| "grad_norm": 12.116938591003418, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 2.1648, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.12513484358144553, | |
| "grad_norm": 8.52156925201416, | |
| "learning_rate": 2.9e-05, | |
| "loss": 1.8476, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.12944983818770225, | |
| "grad_norm": 14.150349617004395, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7714, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.133764832793959, | |
| "grad_norm": 10.8223876953125, | |
| "learning_rate": 3.1e-05, | |
| "loss": 1.937, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.13807982740021574, | |
| "grad_norm": 16.77346420288086, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.9196, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1423948220064725, | |
| "grad_norm": 9.418197631835938, | |
| "learning_rate": 3.3e-05, | |
| "loss": 1.4786, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.14670981661272922, | |
| "grad_norm": 10.013666152954102, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 2.023, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.15102481121898598, | |
| "grad_norm": 16.585966110229492, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.8047, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1553398058252427, | |
| "grad_norm": 9.47374153137207, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.5705, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.15965480043149946, | |
| "grad_norm": 15.564196586608887, | |
| "learning_rate": 3.7e-05, | |
| "loss": 1.8173, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.16396979503775622, | |
| "grad_norm": 9.214970588684082, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.8441, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.16828478964401294, | |
| "grad_norm": 13.743130683898926, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 1.9396, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.1725997842502697, | |
| "grad_norm": 13.389596939086914, | |
| "learning_rate": 4e-05, | |
| "loss": 2.2324, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.17691477885652643, | |
| "grad_norm": 11.09151554107666, | |
| "learning_rate": 4.1e-05, | |
| "loss": 1.7249, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.18122977346278318, | |
| "grad_norm": 27.17494010925293, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.4567, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1855447680690399, | |
| "grad_norm": 14.562529563903809, | |
| "learning_rate": 4.3e-05, | |
| "loss": 1.8576, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.18985976267529667, | |
| "grad_norm": 19.85394287109375, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.7954, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1941747572815534, | |
| "grad_norm": 15.496027946472168, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.5856, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.19848975188781015, | |
| "grad_norm": 13.179373741149902, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.7754, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.20280474649406688, | |
| "grad_norm": 14.764791488647461, | |
| "learning_rate": 4.7e-05, | |
| "loss": 1.5744, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.20711974110032363, | |
| "grad_norm": 15.0728120803833, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.7136, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.21143473570658036, | |
| "grad_norm": 14.698800086975098, | |
| "learning_rate": 4.9e-05, | |
| "loss": 1.7763, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.21574973031283712, | |
| "grad_norm": 10.139289855957031, | |
| "learning_rate": 5e-05, | |
| "loss": 1.577, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22006472491909385, | |
| "grad_norm": 7.694875717163086, | |
| "learning_rate": 5.1000000000000006e-05, | |
| "loss": 1.1383, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.2243797195253506, | |
| "grad_norm": 22.01158905029297, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 1.1455, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.22869471413160733, | |
| "grad_norm": 14.804813385009766, | |
| "learning_rate": 5.300000000000001e-05, | |
| "loss": 1.2175, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.23300970873786409, | |
| "grad_norm": 9.50756549835205, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.0274, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2373247033441208, | |
| "grad_norm": 14.460721969604492, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.413, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.24163969795037757, | |
| "grad_norm": 11.028923988342285, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.2362, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2459546925566343, | |
| "grad_norm": 10.161205291748047, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 1.1661, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.25026968716289105, | |
| "grad_norm": 33.76395034790039, | |
| "learning_rate": 5.8e-05, | |
| "loss": 1.5243, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2545846817691478, | |
| "grad_norm": 21.20954704284668, | |
| "learning_rate": 5.9e-05, | |
| "loss": 1.2345, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.2588996763754045, | |
| "grad_norm": 13.017189025878906, | |
| "learning_rate": 6e-05, | |
| "loss": 1.1945, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2632146709816613, | |
| "grad_norm": 12.412013053894043, | |
| "learning_rate": 6.1e-05, | |
| "loss": 1.3839, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.267529665587918, | |
| "grad_norm": 11.117060661315918, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.0156, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.27184466019417475, | |
| "grad_norm": 10.622220039367676, | |
| "learning_rate": 6.3e-05, | |
| "loss": 1.0484, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2761596548004315, | |
| "grad_norm": 16.456932067871094, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.2239, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.28047464940668826, | |
| "grad_norm": 15.173189163208008, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.1349, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.284789644012945, | |
| "grad_norm": 13.377814292907715, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.2073, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2891046386192017, | |
| "grad_norm": 12.130473136901855, | |
| "learning_rate": 6.7e-05, | |
| "loss": 0.9574, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.29341963322545844, | |
| "grad_norm": 35.022850036621094, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 1.2098, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2977346278317152, | |
| "grad_norm": 14.555148124694824, | |
| "learning_rate": 6.9e-05, | |
| "loss": 1.3269, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.30204962243797195, | |
| "grad_norm": 26.871200561523438, | |
| "learning_rate": 7e-05, | |
| "loss": 1.5483, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3063646170442287, | |
| "grad_norm": 22.50891876220703, | |
| "learning_rate": 7.1e-05, | |
| "loss": 1.2168, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.3106796116504854, | |
| "grad_norm": 15.199380874633789, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.3749, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3149946062567422, | |
| "grad_norm": 19.34256935119629, | |
| "learning_rate": 7.3e-05, | |
| "loss": 1.12, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.3193096008629989, | |
| "grad_norm": 21.201690673828125, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.1308, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.32362459546925565, | |
| "grad_norm": 16.24036407470703, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.1344, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.32793959007551243, | |
| "grad_norm": 14.903363227844238, | |
| "learning_rate": 7.6e-05, | |
| "loss": 1.4773, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.33225458468176916, | |
| "grad_norm": 13.717198371887207, | |
| "learning_rate": 7.7e-05, | |
| "loss": 0.9703, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3365695792880259, | |
| "grad_norm": 12.899928092956543, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 1.0071, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.3408845738942826, | |
| "grad_norm": 15.445221900939941, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 1.1232, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.3451995685005394, | |
| "grad_norm": 14.587578773498535, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9729, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.34951456310679613, | |
| "grad_norm": 11.019111633300781, | |
| "learning_rate": 8.1e-05, | |
| "loss": 1.157, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.35382955771305286, | |
| "grad_norm": 17.6428165435791, | |
| "learning_rate": 8.2e-05, | |
| "loss": 1.6735, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3581445523193096, | |
| "grad_norm": 17.31894874572754, | |
| "learning_rate": 8.3e-05, | |
| "loss": 1.0333, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.36245954692556637, | |
| "grad_norm": 10.62403678894043, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.0533, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3667745415318231, | |
| "grad_norm": 11.239089012145996, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.064, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3710895361380798, | |
| "grad_norm": 13.075492858886719, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.4484, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.37540453074433655, | |
| "grad_norm": 12.833667755126953, | |
| "learning_rate": 8.7e-05, | |
| "loss": 1.2874, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.37971952535059333, | |
| "grad_norm": 12.494976043701172, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.2693, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.38403451995685006, | |
| "grad_norm": 10.73293685913086, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 1.2738, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3883495145631068, | |
| "grad_norm": 11.104293823242188, | |
| "learning_rate": 9e-05, | |
| "loss": 1.1347, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3926645091693635, | |
| "grad_norm": 16.805377960205078, | |
| "learning_rate": 9.1e-05, | |
| "loss": 1.2275, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.3969795037756203, | |
| "grad_norm": 15.351284980773926, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.2335, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.40129449838187703, | |
| "grad_norm": 11.592009544372559, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 1.2577, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.40560949298813376, | |
| "grad_norm": 20.89798927307129, | |
| "learning_rate": 9.4e-05, | |
| "loss": 1.3747, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.4099244875943905, | |
| "grad_norm": 18.117237091064453, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.5576, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.41423948220064727, | |
| "grad_norm": 22.605777740478516, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.0343, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.418554476806904, | |
| "grad_norm": 19.342384338378906, | |
| "learning_rate": 9.7e-05, | |
| "loss": 1.372, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.4228694714131607, | |
| "grad_norm": 25.61117935180664, | |
| "learning_rate": 9.8e-05, | |
| "loss": 1.246, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.42718446601941745, | |
| "grad_norm": 48.5102424621582, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 1.0849, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.43149946062567424, | |
| "grad_norm": 19.60106086730957, | |
| "learning_rate": 0.0001, | |
| "loss": 1.321, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.43581445523193096, | |
| "grad_norm": 23.769451141357422, | |
| "learning_rate": 9.998562273265785e-05, | |
| "loss": 0.9655, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.4401294498381877, | |
| "grad_norm": 16.307497024536133, | |
| "learning_rate": 9.994249919886402e-05, | |
| "loss": 1.2114, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 22.229373931884766, | |
| "learning_rate": 9.98706541985615e-05, | |
| "loss": 0.9414, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.4487594390507012, | |
| "grad_norm": 20.33111572265625, | |
| "learning_rate": 9.977012904914133e-05, | |
| "loss": 1.1184, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.45307443365695793, | |
| "grad_norm": 20.57488250732422, | |
| "learning_rate": 9.964098156168142e-05, | |
| "loss": 1.2291, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.45738942826321466, | |
| "grad_norm": 25.17799949645996, | |
| "learning_rate": 9.948328600769995e-05, | |
| "loss": 1.547, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4617044228694714, | |
| "grad_norm": 17.313278198242188, | |
| "learning_rate": 9.929713307644244e-05, | |
| "loss": 1.0866, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.46601941747572817, | |
| "grad_norm": 18.543851852416992, | |
| "learning_rate": 9.908262982272724e-05, | |
| "loss": 1.1097, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.4703344120819849, | |
| "grad_norm": 15.473976135253906, | |
| "learning_rate": 9.883989960537933e-05, | |
| "loss": 0.9769, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4746494066882416, | |
| "grad_norm": 25.937881469726562, | |
| "learning_rate": 9.85690820162878e-05, | |
| "loss": 1.1811, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.47896440129449835, | |
| "grad_norm": 41.61457824707031, | |
| "learning_rate": 9.827033280012783e-05, | |
| "loss": 1.5683, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.48327939590075514, | |
| "grad_norm": 21.347333908081055, | |
| "learning_rate": 9.794382376479334e-05, | |
| "loss": 0.9634, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.48759439050701187, | |
| "grad_norm": 31.98029899597168, | |
| "learning_rate": 9.7589742682592e-05, | |
| "loss": 1.1704, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4919093851132686, | |
| "grad_norm": 28.764488220214844, | |
| "learning_rate": 9.720829318225897e-05, | |
| "loss": 1.4165, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.4962243797195254, | |
| "grad_norm": 35.89641571044922, | |
| "learning_rate": 9.6799694631852e-05, | |
| "loss": 1.3282, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5005393743257821, | |
| "grad_norm": 30.28986167907715, | |
| "learning_rate": 9.63641820125949e-05, | |
| "loss": 1.4151, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.5048543689320388, | |
| "grad_norm": 32.468040466308594, | |
| "learning_rate": 9.590200578374198e-05, | |
| "loss": 0.9763, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.5091693635382956, | |
| "grad_norm": 35.83097839355469, | |
| "learning_rate": 9.541343173854127e-05, | |
| "loss": 1.5905, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.5134843581445523, | |
| "grad_norm": 62.54613494873047, | |
| "learning_rate": 9.48987408513794e-05, | |
| "loss": 1.295, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.517799352750809, | |
| "grad_norm": 30.689411163330078, | |
| "learning_rate": 9.435822911619564e-05, | |
| "loss": 1.0429, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5221143473570659, | |
| "grad_norm": 34.693260192871094, | |
| "learning_rate": 9.379220737625877e-05, | |
| "loss": 0.98, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.5264293419633226, | |
| "grad_norm": 20.702442169189453, | |
| "learning_rate": 9.320100114540382e-05, | |
| "loss": 1.0123, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5307443365695793, | |
| "grad_norm": 43.67185974121094, | |
| "learning_rate": 9.258495042083221e-05, | |
| "loss": 1.3323, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.535059331175836, | |
| "grad_norm": 24.4400577545166, | |
| "learning_rate": 9.19444094875825e-05, | |
| "loss": 1.0803, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5393743257820928, | |
| "grad_norm": 31.675437927246094, | |
| "learning_rate": 9.127974671478432e-05, | |
| "loss": 1.1885, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5436893203883495, | |
| "grad_norm": 23.858922958374023, | |
| "learning_rate": 9.059134434381273e-05, | |
| "loss": 1.3453, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5480043149946062, | |
| "grad_norm": 35.00339889526367, | |
| "learning_rate": 8.98795982684648e-05, | |
| "loss": 1.2452, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.552319309600863, | |
| "grad_norm": 20.78013038635254, | |
| "learning_rate": 8.914491780728471e-05, | |
| "loss": 1.0057, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5566343042071198, | |
| "grad_norm": 25.536537170410156, | |
| "learning_rate": 8.838772546816856e-05, | |
| "loss": 0.9765, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5609492988133765, | |
| "grad_norm": 24.167560577392578, | |
| "learning_rate": 8.760845670538387e-05, | |
| "loss": 1.1077, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5652642934196332, | |
| "grad_norm": 22.21959686279297, | |
| "learning_rate": 8.680755966914401e-05, | |
| "loss": 1.0824, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.56957928802589, | |
| "grad_norm": 22.32419776916504, | |
| "learning_rate": 8.598549494788111e-05, | |
| "loss": 1.1707, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5738942826321467, | |
| "grad_norm": 27.373973846435547, | |
| "learning_rate": 8.5142735303366e-05, | |
| "loss": 1.0421, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5782092772384034, | |
| "grad_norm": 20.71979331970215, | |
| "learning_rate": 8.427976539882724e-05, | |
| "loss": 1.5185, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5825242718446602, | |
| "grad_norm": 24.339160919189453, | |
| "learning_rate": 8.339708152022585e-05, | |
| "loss": 1.313, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5868392664509169, | |
| "grad_norm": 15.251348495483398, | |
| "learning_rate": 8.24951912908459e-05, | |
| "loss": 1.0494, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5911542610571737, | |
| "grad_norm": 14.0764799118042, | |
| "learning_rate": 8.157461337936506e-05, | |
| "loss": 1.0703, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.5954692556634305, | |
| "grad_norm": 12.889816284179688, | |
| "learning_rate": 8.063587720157298e-05, | |
| "loss": 1.0582, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.5997842502696872, | |
| "grad_norm": 11.844062805175781, | |
| "learning_rate": 7.967952261590935e-05, | |
| "loss": 1.3497, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.6040992448759439, | |
| "grad_norm": 12.643463134765625, | |
| "learning_rate": 7.870609961299627e-05, | |
| "loss": 0.9812, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6084142394822006, | |
| "grad_norm": 13.69425106048584, | |
| "learning_rate": 7.771616799934371e-05, | |
| "loss": 0.8235, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.6127292340884574, | |
| "grad_norm": 13.072551727294922, | |
| "learning_rate": 7.67102970754101e-05, | |
| "loss": 1.0534, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.6170442286947141, | |
| "grad_norm": 19.078474044799805, | |
| "learning_rate": 7.568906530820282e-05, | |
| "loss": 0.9966, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.6213592233009708, | |
| "grad_norm": 11.433798789978027, | |
| "learning_rate": 7.465305999860728e-05, | |
| "loss": 1.5812, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.6256742179072277, | |
| "grad_norm": 13.047567367553711, | |
| "learning_rate": 7.360287694363566e-05, | |
| "loss": 1.0221, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.6299892125134844, | |
| "grad_norm": 17.899805068969727, | |
| "learning_rate": 7.253912009378953e-05, | |
| "loss": 1.1943, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.6343042071197411, | |
| "grad_norm": 11.809228897094727, | |
| "learning_rate": 7.146240120573358e-05, | |
| "loss": 0.9722, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.6386192017259978, | |
| "grad_norm": 34.78435134887695, | |
| "learning_rate": 7.037333949048005e-05, | |
| "loss": 1.241, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6429341963322546, | |
| "grad_norm": 11.29897403717041, | |
| "learning_rate": 6.927256125728624e-05, | |
| "loss": 0.8864, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6472491909385113, | |
| "grad_norm": 14.957293510437012, | |
| "learning_rate": 6.816069955346985e-05, | |
| "loss": 1.1857, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.651564185544768, | |
| "grad_norm": 10.794960975646973, | |
| "learning_rate": 6.703839380034946e-05, | |
| "loss": 0.835, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6558791801510249, | |
| "grad_norm": 10.843793869018555, | |
| "learning_rate": 6.590628942551909e-05, | |
| "loss": 0.8092, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6601941747572816, | |
| "grad_norm": 10.880565643310547, | |
| "learning_rate": 6.476503749166904e-05, | |
| "loss": 0.8796, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6645091693635383, | |
| "grad_norm": 10.481005668640137, | |
| "learning_rate": 6.361529432216559e-05, | |
| "loss": 1.1045, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.668824163969795, | |
| "grad_norm": 14.980287551879883, | |
| "learning_rate": 6.245772112360568e-05, | |
| "loss": 0.9401, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6731391585760518, | |
| "grad_norm": 8.482874870300293, | |
| "learning_rate": 6.129298360556304e-05, | |
| "loss": 1.0408, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6774541531823085, | |
| "grad_norm": 35.517127990722656, | |
| "learning_rate": 6.012175159774488e-05, | |
| "loss": 0.8204, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6817691477885652, | |
| "grad_norm": 14.304695129394531, | |
| "learning_rate": 5.894469866477905e-05, | |
| "loss": 1.1895, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.686084142394822, | |
| "grad_norm": 10.242888450622559, | |
| "learning_rate": 5.776250171885329e-05, | |
| "loss": 0.8243, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6903991370010788, | |
| "grad_norm": 15.357219696044922, | |
| "learning_rate": 5.6575840630429286e-05, | |
| "loss": 0.9941, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6947141316073355, | |
| "grad_norm": 9.460321426391602, | |
| "learning_rate": 5.5385397837255556e-05, | |
| "loss": 1.1884, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.6990291262135923, | |
| "grad_norm": 7.212802886962891, | |
| "learning_rate": 5.4191857951903826e-05, | |
| "loss": 0.9294, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.703344120819849, | |
| "grad_norm": 10.395973205566406, | |
| "learning_rate": 5.29959073680547e-05, | |
| "loss": 0.9332, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.7076591154261057, | |
| "grad_norm": 8.725715637207031, | |
| "learning_rate": 5.179823386575907e-05, | |
| "loss": 1.1057, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.7119741100323624, | |
| "grad_norm": 11.002646446228027, | |
| "learning_rate": 5.059952621590216e-05, | |
| "loss": 1.0017, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.7162891046386192, | |
| "grad_norm": 10.000289916992188, | |
| "learning_rate": 4.940047378409786e-05, | |
| "loss": 0.799, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.7206040992448759, | |
| "grad_norm": 10.543927192687988, | |
| "learning_rate": 4.820176613424095e-05, | |
| "loss": 1.3095, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.7249190938511327, | |
| "grad_norm": 10.675516128540039, | |
| "learning_rate": 4.700409263194531e-05, | |
| "loss": 1.3398, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.7292340884573895, | |
| "grad_norm": 9.756085395812988, | |
| "learning_rate": 4.580814204809618e-05, | |
| "loss": 1.3962, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.7335490830636462, | |
| "grad_norm": 13.422933578491211, | |
| "learning_rate": 4.461460216274445e-05, | |
| "loss": 1.1023, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7378640776699029, | |
| "grad_norm": 10.959280014038086, | |
| "learning_rate": 4.3424159369570725e-05, | |
| "loss": 1.0259, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.7421790722761596, | |
| "grad_norm": 17.383562088012695, | |
| "learning_rate": 4.223749828114672e-05, | |
| "loss": 1.1365, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7464940668824164, | |
| "grad_norm": 10.744115829467773, | |
| "learning_rate": 4.1055301335220955e-05, | |
| "loss": 1.2487, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7508090614886731, | |
| "grad_norm": 13.598590850830078, | |
| "learning_rate": 3.987824840225512e-05, | |
| "loss": 1.1206, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.7551240560949298, | |
| "grad_norm": 11.32901382446289, | |
| "learning_rate": 3.870701639443698e-05, | |
| "loss": 1.1377, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7594390507011867, | |
| "grad_norm": 13.684596061706543, | |
| "learning_rate": 3.7542278876394336e-05, | |
| "loss": 1.0804, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7637540453074434, | |
| "grad_norm": 12.446979522705078, | |
| "learning_rate": 3.638470567783442e-05, | |
| "loss": 1.3194, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7680690399137001, | |
| "grad_norm": 10.964544296264648, | |
| "learning_rate": 3.5234962508330974e-05, | |
| "loss": 0.9998, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7723840345199569, | |
| "grad_norm": 12.713930130004883, | |
| "learning_rate": 3.409371057448092e-05, | |
| "loss": 0.8016, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.7766990291262136, | |
| "grad_norm": 14.682384490966797, | |
| "learning_rate": 3.296160619965056e-05, | |
| "loss": 1.2947, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7810140237324703, | |
| "grad_norm": 11.022130966186523, | |
| "learning_rate": 3.183930044653014e-05, | |
| "loss": 1.242, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.785329018338727, | |
| "grad_norm": 10.892326354980469, | |
| "learning_rate": 3.072743874271376e-05, | |
| "loss": 1.1655, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7896440129449838, | |
| "grad_norm": 15.45097541809082, | |
| "learning_rate": 2.962666050951997e-05, | |
| "loss": 0.9805, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7939590075512406, | |
| "grad_norm": 16.197540283203125, | |
| "learning_rate": 2.853759879426644e-05, | |
| "loss": 1.1878, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7982740021574973, | |
| "grad_norm": 12.304449081420898, | |
| "learning_rate": 2.7460879906210487e-05, | |
| "loss": 1.218, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.8025889967637541, | |
| "grad_norm": 8.651905059814453, | |
| "learning_rate": 2.6397123056364365e-05, | |
| "loss": 0.8546, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.8069039913700108, | |
| "grad_norm": 11.859578132629395, | |
| "learning_rate": 2.5346940001392728e-05, | |
| "loss": 0.9048, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.8112189859762675, | |
| "grad_norm": 9.403504371643066, | |
| "learning_rate": 2.4310934691797203e-05, | |
| "loss": 0.9499, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.8155339805825242, | |
| "grad_norm": 11.484695434570312, | |
| "learning_rate": 2.3289702924589914e-05, | |
| "loss": 1.0291, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.819848975188781, | |
| "grad_norm": 12.270526885986328, | |
| "learning_rate": 2.22838320006563e-05, | |
| "loss": 1.0698, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8241639697950378, | |
| "grad_norm": 11.357561111450195, | |
| "learning_rate": 2.129390038700374e-05, | |
| "loss": 1.0403, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.8284789644012945, | |
| "grad_norm": 18.774675369262695, | |
| "learning_rate": 2.0320477384090665e-05, | |
| "loss": 1.1037, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.8327939590075513, | |
| "grad_norm": 11.835728645324707, | |
| "learning_rate": 1.936412279842705e-05, | |
| "loss": 1.0071, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.837108953613808, | |
| "grad_norm": 10.06062126159668, | |
| "learning_rate": 1.842538662063496e-05, | |
| "loss": 1.1429, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.8414239482200647, | |
| "grad_norm": 10.284469604492188, | |
| "learning_rate": 1.7504808709154104e-05, | |
| "loss": 0.874, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.8457389428263214, | |
| "grad_norm": 14.327275276184082, | |
| "learning_rate": 1.6602918479774148e-05, | |
| "loss": 1.1121, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.8500539374325782, | |
| "grad_norm": 11.892559051513672, | |
| "learning_rate": 1.5720234601172766e-05, | |
| "loss": 1.2422, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8543689320388349, | |
| "grad_norm": 12.877029418945312, | |
| "learning_rate": 1.485726469663401e-05, | |
| "loss": 1.0604, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.8586839266450917, | |
| "grad_norm": 14.058242797851562, | |
| "learning_rate": 1.4014505052118892e-05, | |
| "loss": 0.9899, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8629989212513485, | |
| "grad_norm": 19.5707950592041, | |
| "learning_rate": 1.3192440330856004e-05, | |
| "loss": 1.0327, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8673139158576052, | |
| "grad_norm": 8.181835174560547, | |
| "learning_rate": 1.2391543294616147e-05, | |
| "loss": 0.9223, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8716289104638619, | |
| "grad_norm": 24.17749786376953, | |
| "learning_rate": 1.1612274531831463e-05, | |
| "loss": 1.0612, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8759439050701187, | |
| "grad_norm": 9.240547180175781, | |
| "learning_rate": 1.0855082192715294e-05, | |
| "loss": 1.172, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8802588996763754, | |
| "grad_norm": 13.899747848510742, | |
| "learning_rate": 1.0120401731535212e-05, | |
| "loss": 1.1667, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8845738942826321, | |
| "grad_norm": 19.95920753479004, | |
| "learning_rate": 9.408655656187282e-06, | |
| "loss": 1.439, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 27.467838287353516, | |
| "learning_rate": 8.720253285215685e-06, | |
| "loss": 1.1132, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.8932038834951457, | |
| "grad_norm": 10.904566764831543, | |
| "learning_rate": 8.0555905124175e-06, | |
| "loss": 0.8483, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.8975188781014024, | |
| "grad_norm": 13.368602752685547, | |
| "learning_rate": 7.415049579167782e-06, | |
| "loss": 0.9571, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.9018338727076591, | |
| "grad_norm": 8.375765800476074, | |
| "learning_rate": 6.798998854596189e-06, | |
| "loss": 0.9591, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.9061488673139159, | |
| "grad_norm": 12.240999221801758, | |
| "learning_rate": 6.207792623741249e-06, | |
| "loss": 1.0154, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9104638619201726, | |
| "grad_norm": 10.665072441101074, | |
| "learning_rate": 5.641770883804365e-06, | |
| "loss": 0.8709, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.9147788565264293, | |
| "grad_norm": 10.916919708251953, | |
| "learning_rate": 5.101259148620619e-06, | |
| "loss": 1.1977, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.919093851132686, | |
| "grad_norm": 8.152597427368164, | |
| "learning_rate": 4.586568261458729e-06, | |
| "loss": 0.963, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.9234088457389428, | |
| "grad_norm": 9.507950782775879, | |
| "learning_rate": 4.097994216258039e-06, | |
| "loss": 1.1745, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.9277238403451996, | |
| "grad_norm": 13.667984962463379, | |
| "learning_rate": 3.63581798740511e-06, | |
| "loss": 1.129, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.9320388349514563, | |
| "grad_norm": 12.18842601776123, | |
| "learning_rate": 3.2003053681480098e-06, | |
| "loss": 1.2118, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.9363538295577131, | |
| "grad_norm": 20.156063079833984, | |
| "learning_rate": 2.791706817741041e-06, | |
| "loss": 1.4603, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.9406688241639698, | |
| "grad_norm": 8.63728141784668, | |
| "learning_rate": 2.41025731740801e-06, | |
| "loss": 1.2983, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.9449838187702265, | |
| "grad_norm": 7.928760528564453, | |
| "learning_rate": 2.056176235206664e-06, | |
| "loss": 0.6431, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.9492988133764833, | |
| "grad_norm": 11.23835277557373, | |
| "learning_rate": 1.729667199872187e-06, | |
| "loss": 0.9134, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.95361380798274, | |
| "grad_norm": 9.602569580078125, | |
| "learning_rate": 1.4309179837122044e-06, | |
| "loss": 1.073, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.9579288025889967, | |
| "grad_norm": 11.985309600830078, | |
| "learning_rate": 1.1601003946206724e-06, | |
| "loss": 0.9185, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.9622437971952535, | |
| "grad_norm": 9.460104942321777, | |
| "learning_rate": 9.17370177272775e-07, | |
| "loss": 1.0697, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9665587918015103, | |
| "grad_norm": 19.731552124023438, | |
| "learning_rate": 7.028669235575714e-07, | |
| "loss": 1.2517, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.970873786407767, | |
| "grad_norm": 13.174097061157227, | |
| "learning_rate": 5.167139923000553e-07, | |
| "loss": 0.8468, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9751887810140237, | |
| "grad_norm": 8.821778297424316, | |
| "learning_rate": 3.5901843831857576e-07, | |
| "loss": 0.8934, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9795037756202805, | |
| "grad_norm": 11.48652458190918, | |
| "learning_rate": 2.2987095085867937e-07, | |
| "loss": 0.9348, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.9838187702265372, | |
| "grad_norm": 11.902351379394531, | |
| "learning_rate": 1.2934580143851295e-07, | |
| "loss": 0.9476, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9881337648327939, | |
| "grad_norm": 9.836396217346191, | |
| "learning_rate": 5.750080113598455e-08, | |
| "loss": 1.109, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.9924487594390508, | |
| "grad_norm": 14.28419017791748, | |
| "learning_rate": 1.4377267342158274e-08, | |
| "loss": 0.7782, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.9967637540453075, | |
| "grad_norm": 18.68885040283203, | |
| "learning_rate": 0.0, | |
| "loss": 1.0927, | |
| "step": 231 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 231, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 239, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.209717719498752e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |