| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.9967637540453075, |
| "eval_steps": 500, |
| "global_step": 154, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.006472491909385114, |
| "grad_norm": 5.306360721588135, |
| "learning_rate": 0.0, |
| "loss": 1.1501, |
| "num_tokens": 963559.0, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.012944983818770227, |
| "grad_norm": 5.30866003036499, |
| "learning_rate": 6.25e-08, |
| "loss": 1.1455, |
| "num_tokens": 1927760.0, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.019417475728155338, |
| "grad_norm": 5.259037971496582, |
| "learning_rate": 1.25e-07, |
| "loss": 1.1393, |
| "num_tokens": 2896369.0, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.025889967637540454, |
| "grad_norm": 5.178048610687256, |
| "learning_rate": 1.875e-07, |
| "loss": 1.133, |
| "num_tokens": 3886583.0, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.032362459546925564, |
| "grad_norm": 5.211574554443359, |
| "learning_rate": 2.5e-07, |
| "loss": 1.1368, |
| "num_tokens": 4864476.0, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.038834951456310676, |
| "grad_norm": 5.201828479766846, |
| "learning_rate": 3.1249999999999997e-07, |
| "loss": 1.1455, |
| "num_tokens": 5873295.0, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.045307443365695796, |
| "grad_norm": 5.131554126739502, |
| "learning_rate": 3.75e-07, |
| "loss": 1.1259, |
| "num_tokens": 6876484.0, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.05177993527508091, |
| "grad_norm": 5.083103179931641, |
| "learning_rate": 4.375e-07, |
| "loss": 1.1167, |
| "num_tokens": 7871664.0, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.05825242718446602, |
| "grad_norm": 5.106949806213379, |
| "learning_rate": 5e-07, |
| "loss": 1.1324, |
| "num_tokens": 8869931.0, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.06472491909385113, |
| "grad_norm": 4.861895561218262, |
| "learning_rate": 5.625e-07, |
| "loss": 1.1082, |
| "num_tokens": 9874687.0, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.07119741100323625, |
| "grad_norm": 4.900995254516602, |
| "learning_rate": 6.249999999999999e-07, |
| "loss": 1.1035, |
| "num_tokens": 10842628.0, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.07766990291262135, |
| "grad_norm": 4.93906307220459, |
| "learning_rate": 6.875e-07, |
| "loss": 1.1207, |
| "num_tokens": 11815438.0, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.08414239482200647, |
| "grad_norm": 4.3877153396606445, |
| "learning_rate": 7.5e-07, |
| "loss": 1.0708, |
| "num_tokens": 12788222.0, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.09061488673139159, |
| "grad_norm": 4.357248306274414, |
| "learning_rate": 8.125e-07, |
| "loss": 1.0679, |
| "num_tokens": 13756270.0, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0970873786407767, |
| "grad_norm": 4.3989644050598145, |
| "learning_rate": 8.75e-07, |
| "loss": 1.068, |
| "num_tokens": 14689813.0, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.10355987055016182, |
| "grad_norm": 4.206644535064697, |
| "learning_rate": 9.374999999999999e-07, |
| "loss": 1.0545, |
| "num_tokens": 15670657.0, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.11003236245954692, |
| "grad_norm": 4.120600700378418, |
| "learning_rate": 1e-06, |
| "loss": 1.0357, |
| "num_tokens": 16640491.0, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.11650485436893204, |
| "grad_norm": 3.066406011581421, |
| "learning_rate": 1e-06, |
| "loss": 0.9571, |
| "num_tokens": 17611388.0, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.12297734627831715, |
| "grad_norm": 3.0390701293945312, |
| "learning_rate": 1e-06, |
| "loss": 0.9557, |
| "num_tokens": 18582653.0, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.12944983818770225, |
| "grad_norm": 2.996507406234741, |
| "learning_rate": 1e-06, |
| "loss": 0.9467, |
| "num_tokens": 19557459.0, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.13592233009708737, |
| "grad_norm": 2.852745771408081, |
| "learning_rate": 1e-06, |
| "loss": 0.9414, |
| "num_tokens": 20524158.0, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.1423948220064725, |
| "grad_norm": 2.7312402725219727, |
| "learning_rate": 1e-06, |
| "loss": 0.9161, |
| "num_tokens": 21502674.0, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.1488673139158576, |
| "grad_norm": 2.712592124938965, |
| "learning_rate": 1e-06, |
| "loss": 0.9201, |
| "num_tokens": 22446294.0, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.1553398058252427, |
| "grad_norm": 2.51646089553833, |
| "learning_rate": 1e-06, |
| "loss": 0.8971, |
| "num_tokens": 23430225.0, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.16181229773462782, |
| "grad_norm": 1.9695550203323364, |
| "learning_rate": 1e-06, |
| "loss": 0.8188, |
| "num_tokens": 24380419.0, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.16828478964401294, |
| "grad_norm": 1.7356574535369873, |
| "learning_rate": 1e-06, |
| "loss": 0.7798, |
| "num_tokens": 25343499.0, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.17475728155339806, |
| "grad_norm": 1.5638344287872314, |
| "learning_rate": 1e-06, |
| "loss": 0.7747, |
| "num_tokens": 26315921.0, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.18122977346278318, |
| "grad_norm": 1.524434208869934, |
| "learning_rate": 1e-06, |
| "loss": 0.7787, |
| "num_tokens": 27255686.0, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.18770226537216828, |
| "grad_norm": 1.3938937187194824, |
| "learning_rate": 1e-06, |
| "loss": 0.7647, |
| "num_tokens": 28232449.0, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.1941747572815534, |
| "grad_norm": 1.3155360221862793, |
| "learning_rate": 1e-06, |
| "loss": 0.7544, |
| "num_tokens": 29213099.0, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.20064724919093851, |
| "grad_norm": 1.1851674318313599, |
| "learning_rate": 1e-06, |
| "loss": 0.751, |
| "num_tokens": 30210546.0, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.20711974110032363, |
| "grad_norm": 1.0744044780731201, |
| "learning_rate": 1e-06, |
| "loss": 0.7319, |
| "num_tokens": 31190215.0, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.21359223300970873, |
| "grad_norm": 1.027134895324707, |
| "learning_rate": 1e-06, |
| "loss": 0.7286, |
| "num_tokens": 32152368.0, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.22006472491909385, |
| "grad_norm": 0.9877621531486511, |
| "learning_rate": 1e-06, |
| "loss": 0.7382, |
| "num_tokens": 33138410.0, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.22653721682847897, |
| "grad_norm": 0.9485065937042236, |
| "learning_rate": 1e-06, |
| "loss": 0.7276, |
| "num_tokens": 34103559.0, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.23300970873786409, |
| "grad_norm": 0.8876718282699585, |
| "learning_rate": 1e-06, |
| "loss": 0.7243, |
| "num_tokens": 35099702.0, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.23948220064724918, |
| "grad_norm": 0.781726598739624, |
| "learning_rate": 1e-06, |
| "loss": 0.715, |
| "num_tokens": 36067246.0, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.2459546925566343, |
| "grad_norm": 0.6435908675193787, |
| "learning_rate": 1e-06, |
| "loss": 0.6923, |
| "num_tokens": 37074157.0, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.2524271844660194, |
| "grad_norm": 0.5644338130950928, |
| "learning_rate": 1e-06, |
| "loss": 0.6791, |
| "num_tokens": 38026864.0, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.2588996763754045, |
| "grad_norm": 0.5346189141273499, |
| "learning_rate": 1e-06, |
| "loss": 0.6658, |
| "num_tokens": 39007950.0, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.26537216828478966, |
| "grad_norm": 0.536400556564331, |
| "learning_rate": 1e-06, |
| "loss": 0.673, |
| "num_tokens": 40004404.0, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.27184466019417475, |
| "grad_norm": 0.5293827652931213, |
| "learning_rate": 1e-06, |
| "loss": 0.6625, |
| "num_tokens": 40979367.0, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.2783171521035599, |
| "grad_norm": 0.5181026458740234, |
| "learning_rate": 1e-06, |
| "loss": 0.6512, |
| "num_tokens": 41954716.0, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.284789644012945, |
| "grad_norm": 0.4992174506187439, |
| "learning_rate": 1e-06, |
| "loss": 0.6621, |
| "num_tokens": 42927994.0, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.2912621359223301, |
| "grad_norm": 0.47395315766334534, |
| "learning_rate": 1e-06, |
| "loss": 0.6478, |
| "num_tokens": 43870458.0, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.2977346278317152, |
| "grad_norm": 0.41670018434524536, |
| "learning_rate": 1e-06, |
| "loss": 0.6411, |
| "num_tokens": 44845205.0, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.3042071197411003, |
| "grad_norm": 0.3809607923030853, |
| "learning_rate": 1e-06, |
| "loss": 0.6372, |
| "num_tokens": 45817785.0, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.3106796116504854, |
| "grad_norm": 0.36747702956199646, |
| "learning_rate": 1e-06, |
| "loss": 0.6403, |
| "num_tokens": 46768552.0, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.31715210355987056, |
| "grad_norm": 0.34375864267349243, |
| "learning_rate": 1e-06, |
| "loss": 0.6193, |
| "num_tokens": 47720390.0, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.32362459546925565, |
| "grad_norm": 0.3278081715106964, |
| "learning_rate": 1e-06, |
| "loss": 0.6277, |
| "num_tokens": 48659119.0, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.3300970873786408, |
| "grad_norm": 0.3173641860485077, |
| "learning_rate": 1e-06, |
| "loss": 0.6262, |
| "num_tokens": 49635896.0, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.3365695792880259, |
| "grad_norm": 0.3114546239376068, |
| "learning_rate": 1e-06, |
| "loss": 0.6229, |
| "num_tokens": 50617481.0, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.343042071197411, |
| "grad_norm": 0.31268301606178284, |
| "learning_rate": 1e-06, |
| "loss": 0.6187, |
| "num_tokens": 51580322.0, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.34951456310679613, |
| "grad_norm": 0.29753509163856506, |
| "learning_rate": 1e-06, |
| "loss": 0.6176, |
| "num_tokens": 52544983.0, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.3559870550161812, |
| "grad_norm": 0.2916501760482788, |
| "learning_rate": 1e-06, |
| "loss": 0.6086, |
| "num_tokens": 53538431.0, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.36245954692556637, |
| "grad_norm": 0.2825452983379364, |
| "learning_rate": 1e-06, |
| "loss": 0.6077, |
| "num_tokens": 54497097.0, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.36893203883495146, |
| "grad_norm": 0.264753133058548, |
| "learning_rate": 1e-06, |
| "loss": 0.5966, |
| "num_tokens": 55515695.0, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.37540453074433655, |
| "grad_norm": 0.25653982162475586, |
| "learning_rate": 1e-06, |
| "loss": 0.5997, |
| "num_tokens": 56471100.0, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.3818770226537217, |
| "grad_norm": 0.2426663488149643, |
| "learning_rate": 1e-06, |
| "loss": 0.603, |
| "num_tokens": 57460948.0, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.3883495145631068, |
| "grad_norm": 0.22974367439746857, |
| "learning_rate": 1e-06, |
| "loss": 0.5993, |
| "num_tokens": 58450663.0, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.3948220064724919, |
| "grad_norm": 0.21884475648403168, |
| "learning_rate": 1e-06, |
| "loss": 0.5884, |
| "num_tokens": 59407638.0, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.40129449838187703, |
| "grad_norm": 0.21299949288368225, |
| "learning_rate": 1e-06, |
| "loss": 0.5862, |
| "num_tokens": 60358388.0, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.4077669902912621, |
| "grad_norm": 0.21099086105823517, |
| "learning_rate": 1e-06, |
| "loss": 0.5787, |
| "num_tokens": 61306137.0, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.41423948220064727, |
| "grad_norm": 0.20446249842643738, |
| "learning_rate": 1e-06, |
| "loss": 0.5888, |
| "num_tokens": 62298616.0, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.42071197411003236, |
| "grad_norm": 0.20021338760852814, |
| "learning_rate": 1e-06, |
| "loss": 0.5759, |
| "num_tokens": 63242620.0, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.42718446601941745, |
| "grad_norm": 0.20259256660938263, |
| "learning_rate": 1e-06, |
| "loss": 0.5939, |
| "num_tokens": 64231555.0, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.4336569579288026, |
| "grad_norm": 0.19663812220096588, |
| "learning_rate": 1e-06, |
| "loss": 0.5788, |
| "num_tokens": 65190302.0, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.4401294498381877, |
| "grad_norm": 0.1927073448896408, |
| "learning_rate": 1e-06, |
| "loss": 0.5755, |
| "num_tokens": 66175194.0, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.44660194174757284, |
| "grad_norm": 0.18935494124889374, |
| "learning_rate": 1e-06, |
| "loss": 0.573, |
| "num_tokens": 67146659.0, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.45307443365695793, |
| "grad_norm": 0.18019847571849823, |
| "learning_rate": 1e-06, |
| "loss": 0.5539, |
| "num_tokens": 68115421.0, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.459546925566343, |
| "grad_norm": 0.17649590969085693, |
| "learning_rate": 1e-06, |
| "loss": 0.5506, |
| "num_tokens": 69078962.0, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.46601941747572817, |
| "grad_norm": 0.16998673975467682, |
| "learning_rate": 1e-06, |
| "loss": 0.573, |
| "num_tokens": 70056552.0, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.47249190938511326, |
| "grad_norm": 0.16949011385440826, |
| "learning_rate": 1e-06, |
| "loss": 0.5674, |
| "num_tokens": 71056405.0, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.47896440129449835, |
| "grad_norm": 0.1706332266330719, |
| "learning_rate": 1e-06, |
| "loss": 0.5755, |
| "num_tokens": 72011963.0, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.4854368932038835, |
| "grad_norm": 0.16622762382030487, |
| "learning_rate": 1e-06, |
| "loss": 0.563, |
| "num_tokens": 72974088.0, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.4919093851132686, |
| "grad_norm": 0.16113753616809845, |
| "learning_rate": 1e-06, |
| "loss": 0.5545, |
| "num_tokens": 73963066.0, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.49838187702265374, |
| "grad_norm": 0.15716572105884552, |
| "learning_rate": 1e-06, |
| "loss": 0.5465, |
| "num_tokens": 74927507.0, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.5048543689320388, |
| "grad_norm": 0.15952879190444946, |
| "learning_rate": 1e-06, |
| "loss": 0.5513, |
| "num_tokens": 75913302.0, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.511326860841424, |
| "grad_norm": 0.1577850878238678, |
| "learning_rate": 1e-06, |
| "loss": 0.559, |
| "num_tokens": 76871637.0, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.517799352750809, |
| "grad_norm": 0.1505461186170578, |
| "learning_rate": 1e-06, |
| "loss": 0.5509, |
| "num_tokens": 77844738.0, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.5242718446601942, |
| "grad_norm": 0.1514582335948944, |
| "learning_rate": 1e-06, |
| "loss": 0.5453, |
| "num_tokens": 78789735.0, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.5307443365695793, |
| "grad_norm": 0.15150229632854462, |
| "learning_rate": 1e-06, |
| "loss": 0.5417, |
| "num_tokens": 79740057.0, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.5372168284789643, |
| "grad_norm": 0.14974889159202576, |
| "learning_rate": 1e-06, |
| "loss": 0.5498, |
| "num_tokens": 80691344.0, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.5436893203883495, |
| "grad_norm": 0.14659319818019867, |
| "learning_rate": 1e-06, |
| "loss": 0.5456, |
| "num_tokens": 81695065.0, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.5501618122977346, |
| "grad_norm": 0.14839723706245422, |
| "learning_rate": 1e-06, |
| "loss": 0.5446, |
| "num_tokens": 82670591.0, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.5566343042071198, |
| "grad_norm": 0.1449030339717865, |
| "learning_rate": 1e-06, |
| "loss": 0.5434, |
| "num_tokens": 83619525.0, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.5631067961165048, |
| "grad_norm": 0.14636728167533875, |
| "learning_rate": 1e-06, |
| "loss": 0.5427, |
| "num_tokens": 84583829.0, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.56957928802589, |
| "grad_norm": 0.14410880208015442, |
| "learning_rate": 1e-06, |
| "loss": 0.54, |
| "num_tokens": 85554557.0, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.5760517799352751, |
| "grad_norm": 0.14123332500457764, |
| "learning_rate": 1e-06, |
| "loss": 0.5412, |
| "num_tokens": 86534800.0, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.5825242718446602, |
| "grad_norm": 0.14631222188472748, |
| "learning_rate": 1e-06, |
| "loss": 0.5451, |
| "num_tokens": 87479322.0, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.5889967637540453, |
| "grad_norm": 0.141328364610672, |
| "learning_rate": 1e-06, |
| "loss": 0.543, |
| "num_tokens": 88440604.0, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.5954692556634305, |
| "grad_norm": 0.14410269260406494, |
| "learning_rate": 1e-06, |
| "loss": 0.5389, |
| "num_tokens": 89440902.0, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.6019417475728155, |
| "grad_norm": 0.14109647274017334, |
| "learning_rate": 1e-06, |
| "loss": 0.5312, |
| "num_tokens": 90420292.0, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.6084142394822006, |
| "grad_norm": 0.14093320071697235, |
| "learning_rate": 1e-06, |
| "loss": 0.5432, |
| "num_tokens": 91365927.0, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.6148867313915858, |
| "grad_norm": 0.1392253190279007, |
| "learning_rate": 1e-06, |
| "loss": 0.5342, |
| "num_tokens": 92339481.0, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.6213592233009708, |
| "grad_norm": 0.14118197560310364, |
| "learning_rate": 1e-06, |
| "loss": 0.5291, |
| "num_tokens": 93277920.0, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.627831715210356, |
| "grad_norm": 0.14340411126613617, |
| "learning_rate": 1e-06, |
| "loss": 0.527, |
| "num_tokens": 94233058.0, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.6343042071197411, |
| "grad_norm": 0.13777285814285278, |
| "learning_rate": 1e-06, |
| "loss": 0.5407, |
| "num_tokens": 95195865.0, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.6407766990291263, |
| "grad_norm": 0.14087024331092834, |
| "learning_rate": 1e-06, |
| "loss": 0.5219, |
| "num_tokens": 96165533.0, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.6472491909385113, |
| "grad_norm": 0.13480006158351898, |
| "learning_rate": 1e-06, |
| "loss": 0.5229, |
| "num_tokens": 97128908.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.6537216828478964, |
| "grad_norm": 0.13443224132061005, |
| "learning_rate": 1e-06, |
| "loss": 0.5235, |
| "num_tokens": 98137812.0, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.6601941747572816, |
| "grad_norm": 0.13486695289611816, |
| "learning_rate": 1e-06, |
| "loss": 0.521, |
| "num_tokens": 99133445.0, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.6666666666666666, |
| "grad_norm": 0.13491222262382507, |
| "learning_rate": 1e-06, |
| "loss": 0.5118, |
| "num_tokens": 100106034.0, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.6731391585760518, |
| "grad_norm": 0.13977712392807007, |
| "learning_rate": 1e-06, |
| "loss": 0.5394, |
| "num_tokens": 101060830.0, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.6796116504854369, |
| "grad_norm": 0.1345338374376297, |
| "learning_rate": 1e-06, |
| "loss": 0.5226, |
| "num_tokens": 102032708.0, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.686084142394822, |
| "grad_norm": 0.13142578303813934, |
| "learning_rate": 1e-06, |
| "loss": 0.5061, |
| "num_tokens": 103020888.0, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.6925566343042071, |
| "grad_norm": 0.14058642089366913, |
| "learning_rate": 1e-06, |
| "loss": 0.5315, |
| "num_tokens": 103986883.0, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.6990291262135923, |
| "grad_norm": 0.13689804077148438, |
| "learning_rate": 1e-06, |
| "loss": 0.5275, |
| "num_tokens": 104978514.0, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.7055016181229773, |
| "grad_norm": 0.1382923722267151, |
| "learning_rate": 1e-06, |
| "loss": 0.5159, |
| "num_tokens": 105938542.0, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.7119741100323624, |
| "grad_norm": 0.13341419398784637, |
| "learning_rate": 1e-06, |
| "loss": 0.5175, |
| "num_tokens": 106909993.0, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.7184466019417476, |
| "grad_norm": 0.13735052943229675, |
| "learning_rate": 1e-06, |
| "loss": 0.5204, |
| "num_tokens": 107872313.0, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.7249190938511327, |
| "grad_norm": 0.14685890078544617, |
| "learning_rate": 1e-06, |
| "loss": 0.5154, |
| "num_tokens": 108876000.0, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.7313915857605178, |
| "grad_norm": 0.13520538806915283, |
| "learning_rate": 1e-06, |
| "loss": 0.5163, |
| "num_tokens": 109857790.0, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.7378640776699029, |
| "grad_norm": 0.1330995261669159, |
| "learning_rate": 1e-06, |
| "loss": 0.5155, |
| "num_tokens": 110826497.0, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.7443365695792881, |
| "grad_norm": 0.134913370013237, |
| "learning_rate": 1e-06, |
| "loss": 0.5142, |
| "num_tokens": 111775780.0, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.7508090614886731, |
| "grad_norm": 0.13504451513290405, |
| "learning_rate": 1e-06, |
| "loss": 0.5161, |
| "num_tokens": 112755203.0, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.7572815533980582, |
| "grad_norm": 0.13805583119392395, |
| "learning_rate": 1e-06, |
| "loss": 0.5166, |
| "num_tokens": 113720983.0, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.7637540453074434, |
| "grad_norm": 0.137771874666214, |
| "learning_rate": 1e-06, |
| "loss": 0.5057, |
| "num_tokens": 114690500.0, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.7702265372168284, |
| "grad_norm": 0.13603690266609192, |
| "learning_rate": 1e-06, |
| "loss": 0.5059, |
| "num_tokens": 115663752.0, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.7766990291262136, |
| "grad_norm": 0.1295483261346817, |
| "learning_rate": 1e-06, |
| "loss": 0.5031, |
| "num_tokens": 116656118.0, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.7831715210355987, |
| "grad_norm": 0.13289949297904968, |
| "learning_rate": 1e-06, |
| "loss": 0.5127, |
| "num_tokens": 117647006.0, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.7896440129449838, |
| "grad_norm": 0.13150979578495026, |
| "learning_rate": 1e-06, |
| "loss": 0.5023, |
| "num_tokens": 118637898.0, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.7961165048543689, |
| "grad_norm": 0.13312214612960815, |
| "learning_rate": 1e-06, |
| "loss": 0.5143, |
| "num_tokens": 119611539.0, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.8025889967637541, |
| "grad_norm": 0.13679178059101105, |
| "learning_rate": 1e-06, |
| "loss": 0.5149, |
| "num_tokens": 120576485.0, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.8090614886731392, |
| "grad_norm": 0.13002805411815643, |
| "learning_rate": 1e-06, |
| "loss": 0.5143, |
| "num_tokens": 121567862.0, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.8155339805825242, |
| "grad_norm": 0.13436037302017212, |
| "learning_rate": 1e-06, |
| "loss": 0.5028, |
| "num_tokens": 122530893.0, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.8220064724919094, |
| "grad_norm": 0.13467973470687866, |
| "learning_rate": 1e-06, |
| "loss": 0.5069, |
| "num_tokens": 123480227.0, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.8284789644012945, |
| "grad_norm": 0.13601741194725037, |
| "learning_rate": 1e-06, |
| "loss": 0.5194, |
| "num_tokens": 124462070.0, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.8349514563106796, |
| "grad_norm": 0.1381833255290985, |
| "learning_rate": 1e-06, |
| "loss": 0.5142, |
| "num_tokens": 125440428.0, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.8414239482200647, |
| "grad_norm": 0.1346794068813324, |
| "learning_rate": 1e-06, |
| "loss": 0.5115, |
| "num_tokens": 126416310.0, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.8478964401294499, |
| "grad_norm": 0.13682374358177185, |
| "learning_rate": 1e-06, |
| "loss": 0.512, |
| "num_tokens": 127406436.0, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.8543689320388349, |
| "grad_norm": 0.13239574432373047, |
| "learning_rate": 1e-06, |
| "loss": 0.5068, |
| "num_tokens": 128372197.0, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.86084142394822, |
| "grad_norm": 0.13347913324832916, |
| "learning_rate": 1e-06, |
| "loss": 0.5062, |
| "num_tokens": 129371264.0, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.8673139158576052, |
| "grad_norm": 0.1342724710702896, |
| "learning_rate": 1e-06, |
| "loss": 0.4917, |
| "num_tokens": 130306080.0, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.8737864077669902, |
| "grad_norm": 0.13179434835910797, |
| "learning_rate": 1e-06, |
| "loss": 0.5107, |
| "num_tokens": 131261937.0, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.8802588996763754, |
| "grad_norm": 0.13348270952701569, |
| "learning_rate": 1e-06, |
| "loss": 0.5044, |
| "num_tokens": 132223182.0, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.8867313915857605, |
| "grad_norm": 0.22263851761817932, |
| "learning_rate": 1e-06, |
| "loss": 0.5082, |
| "num_tokens": 133211562.0, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.8932038834951457, |
| "grad_norm": 0.13348044455051422, |
| "learning_rate": 1e-06, |
| "loss": 0.5043, |
| "num_tokens": 134208133.0, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.8996763754045307, |
| "grad_norm": 0.13663546741008759, |
| "learning_rate": 1e-06, |
| "loss": 0.5083, |
| "num_tokens": 135194807.0, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.9061488673139159, |
| "grad_norm": 0.1293790638446808, |
| "learning_rate": 1e-06, |
| "loss": 0.5046, |
| "num_tokens": 136175941.0, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.912621359223301, |
| "grad_norm": 0.13469921052455902, |
| "learning_rate": 1e-06, |
| "loss": 0.5183, |
| "num_tokens": 137117638.0, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.919093851132686, |
| "grad_norm": 0.1336260437965393, |
| "learning_rate": 1e-06, |
| "loss": 0.4941, |
| "num_tokens": 138102412.0, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.9255663430420712, |
| "grad_norm": 0.1314634084701538, |
| "learning_rate": 1e-06, |
| "loss": 0.497, |
| "num_tokens": 139056304.0, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.9320388349514563, |
| "grad_norm": 0.13598428666591644, |
| "learning_rate": 1e-06, |
| "loss": 0.5026, |
| "num_tokens": 140002988.0, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.9385113268608414, |
| "grad_norm": 0.13465505838394165, |
| "learning_rate": 1e-06, |
| "loss": 0.5051, |
| "num_tokens": 140939323.0, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.9449838187702265, |
| "grad_norm": 0.13320808112621307, |
| "learning_rate": 1e-06, |
| "loss": 0.5007, |
| "num_tokens": 141898246.0, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.9514563106796117, |
| "grad_norm": 0.13584767282009125, |
| "learning_rate": 1e-06, |
| "loss": 0.5074, |
| "num_tokens": 142861118.0, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.9579288025889967, |
| "grad_norm": 0.13394641876220703, |
| "learning_rate": 1e-06, |
| "loss": 0.5002, |
| "num_tokens": 143845628.0, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.9644012944983819, |
| "grad_norm": 0.13663163781166077, |
| "learning_rate": 1e-06, |
| "loss": 0.5129, |
| "num_tokens": 144816454.0, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.970873786407767, |
| "grad_norm": 0.12985925376415253, |
| "learning_rate": 1e-06, |
| "loss": 0.5006, |
| "num_tokens": 145791528.0, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.9773462783171522, |
| "grad_norm": 0.1344696283340454, |
| "learning_rate": 1e-06, |
| "loss": 0.5136, |
| "num_tokens": 146776225.0, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.9838187702265372, |
| "grad_norm": 0.1393062323331833, |
| "learning_rate": 1e-06, |
| "loss": 0.4963, |
| "num_tokens": 147716038.0, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.9902912621359223, |
| "grad_norm": 0.13568711280822754, |
| "learning_rate": 1e-06, |
| "loss": 0.4935, |
| "num_tokens": 148641036.0, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.9967637540453075, |
| "grad_norm": 0.12916181981563568, |
| "learning_rate": 1e-06, |
| "loss": 0.4964, |
| "num_tokens": 149611850.0, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.9967637540453075, |
| "step": 154, |
| "total_flos": 6.182777521931551e+18, |
| "train_loss": 0.6460230693027571, |
| "train_runtime": 6408.7951, |
| "train_samples_per_second": 10.787, |
| "train_steps_per_second": 0.024 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 154, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 8, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 6.182777521931551e+18, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
