| { | |
| "best_metric": 0.765149712562561, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 198, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015151515151515152, | |
| "grad_norm": 0.8728165626525879, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8243, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.015151515151515152, | |
| "eval_loss": 0.8913009166717529, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 56.06, | |
| "eval_steps_per_second": 14.141, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 0.9878870248794556, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8533, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 0.9583198428153992, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8803, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 0.929862380027771, | |
| "learning_rate": 4e-05, | |
| "loss": 0.953, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07575757575757576, | |
| "grad_norm": 0.8759853839874268, | |
| "learning_rate": 5e-05, | |
| "loss": 0.908, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 0.9699501395225525, | |
| "learning_rate": 6e-05, | |
| "loss": 0.9424, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.10606060606060606, | |
| "grad_norm": 1.0141632556915283, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9098, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 0.944428026676178, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8006, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 0.9591452479362488, | |
| "learning_rate": 9e-05, | |
| "loss": 0.7831, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 1.165297269821167, | |
| "learning_rate": 0.0001, | |
| "loss": 0.851, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 0.9556719064712524, | |
| "learning_rate": 9.999301905929286e-05, | |
| "loss": 0.8079, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.9887181520462036, | |
| "learning_rate": 9.997207818651274e-05, | |
| "loss": 0.8619, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.19696969696969696, | |
| "grad_norm": 0.8687989115715027, | |
| "learning_rate": 9.99371832291393e-05, | |
| "loss": 0.699, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 1.0262181758880615, | |
| "learning_rate": 9.988834393115767e-05, | |
| "loss": 0.7811, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 0.9539054036140442, | |
| "learning_rate": 9.982557393033758e-05, | |
| "loss": 0.6579, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 1.3137527704238892, | |
| "learning_rate": 9.974889075442521e-05, | |
| "loss": 0.8023, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.25757575757575757, | |
| "grad_norm": 0.7480142712593079, | |
| "learning_rate": 9.965831581624871e-05, | |
| "loss": 0.6959, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.869584858417511, | |
| "learning_rate": 9.9553874407739e-05, | |
| "loss": 0.8288, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2878787878787879, | |
| "grad_norm": 0.6878750920295715, | |
| "learning_rate": 9.94355956928673e-05, | |
| "loss": 0.8074, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 0.6765142679214478, | |
| "learning_rate": 9.930351269950143e-05, | |
| "loss": 0.7684, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 0.6932944655418396, | |
| "learning_rate": 9.915766231018318e-05, | |
| "loss": 0.7873, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.719284176826477, | |
| "learning_rate": 9.899808525182935e-05, | |
| "loss": 0.8531, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3484848484848485, | |
| "grad_norm": 0.7493268251419067, | |
| "learning_rate": 9.882482608435923e-05, | |
| "loss": 0.8127, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.7806094884872437, | |
| "learning_rate": 9.863793318825186e-05, | |
| "loss": 0.779, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3787878787878788, | |
| "grad_norm": 0.724729061126709, | |
| "learning_rate": 9.843745875103627e-05, | |
| "loss": 0.7675, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 0.7515143156051636, | |
| "learning_rate": 9.822345875271883e-05, | |
| "loss": 0.7964, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 0.7275444865226746, | |
| "learning_rate": 9.799599295015154e-05, | |
| "loss": 0.7636, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 0.7667958736419678, | |
| "learning_rate": 9.775512486034563e-05, | |
| "loss": 0.8213, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.4393939393939394, | |
| "grad_norm": 0.799657940864563, | |
| "learning_rate": 9.750092174273521e-05, | |
| "loss": 0.8064, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.7070261240005493, | |
| "learning_rate": 9.723345458039594e-05, | |
| "loss": 0.6593, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4696969696969697, | |
| "grad_norm": 0.8530207276344299, | |
| "learning_rate": 9.69527980602239e-05, | |
| "loss": 0.6984, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 1.3286330699920654, | |
| "learning_rate": 9.665903055208014e-05, | |
| "loss": 0.7172, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.5906330347061157, | |
| "learning_rate": 9.635223408690688e-05, | |
| "loss": 0.6883, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 0.6573566198348999, | |
| "learning_rate": 9.603249433382144e-05, | |
| "loss": 0.8076, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5303030303030303, | |
| "grad_norm": 0.6441850066184998, | |
| "learning_rate": 9.569990057619414e-05, | |
| "loss": 0.8064, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.6237654089927673, | |
| "learning_rate": 9.535454568671704e-05, | |
| "loss": 0.7729, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5606060606060606, | |
| "grad_norm": 0.6479617953300476, | |
| "learning_rate": 9.49965261014704e-05, | |
| "loss": 0.9037, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 0.6237462162971497, | |
| "learning_rate": 9.462594179299406e-05, | |
| "loss": 0.8116, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5909090909090909, | |
| "grad_norm": 0.6595411896705627, | |
| "learning_rate": 9.424289624237144e-05, | |
| "loss": 0.804, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 0.7035729289054871, | |
| "learning_rate": 9.384749641033359e-05, | |
| "loss": 0.8202, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6212121212121212, | |
| "grad_norm": 0.7246851921081543, | |
| "learning_rate": 9.343985270739182e-05, | |
| "loss": 0.8076, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 0.7167584300041199, | |
| "learning_rate": 9.302007896300698e-05, | |
| "loss": 0.8257, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6515151515151515, | |
| "grad_norm": 0.6961088180541992, | |
| "learning_rate": 9.25882923938038e-05, | |
| "loss": 0.7563, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.7266330122947693, | |
| "learning_rate": 9.214461357083985e-05, | |
| "loss": 0.8316, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.6818181818181818, | |
| "grad_norm": 0.7156293392181396, | |
| "learning_rate": 9.168916638593736e-05, | |
| "loss": 0.7674, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.696969696969697, | |
| "grad_norm": 0.7188352346420288, | |
| "learning_rate": 9.122207801708802e-05, | |
| "loss": 0.7153, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7121212121212122, | |
| "grad_norm": 0.7589750289916992, | |
| "learning_rate": 9.074347889294016e-05, | |
| "loss": 0.6744, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.054231882095337, | |
| "learning_rate": 9.025350265637815e-05, | |
| "loss": 0.6719, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7424242424242424, | |
| "grad_norm": 0.5536164045333862, | |
| "learning_rate": 8.975228612720416e-05, | |
| "loss": 0.7251, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 0.6009869575500488, | |
| "learning_rate": 8.923996926393305e-05, | |
| "loss": 0.7412, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "eval_loss": 0.7730022072792053, | |
| "eval_runtime": 1.9828, | |
| "eval_samples_per_second": 55.98, | |
| "eval_steps_per_second": 14.121, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7727272727272727, | |
| "grad_norm": 0.5975300073623657, | |
| "learning_rate": 8.871669512471068e-05, | |
| "loss": 0.7294, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 0.626697301864624, | |
| "learning_rate": 8.818260982736661e-05, | |
| "loss": 0.8291, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.803030303030303, | |
| "grad_norm": 0.6658130288124084, | |
| "learning_rate": 8.763786250861256e-05, | |
| "loss": 0.8711, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.604573667049408, | |
| "learning_rate": 8.708260528239788e-05, | |
| "loss": 0.7987, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 0.6139510273933411, | |
| "learning_rate": 8.651699319743347e-05, | |
| "loss": 0.7789, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 0.6184104681015015, | |
| "learning_rate": 8.594118419389647e-05, | |
| "loss": 0.7976, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8636363636363636, | |
| "grad_norm": 0.6175539493560791, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.7657, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.8787878787878788, | |
| "grad_norm": 0.6376726627349854, | |
| "learning_rate": 8.475962138373213e-05, | |
| "loss": 0.7227, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.8939393939393939, | |
| "grad_norm": 0.7012733221054077, | |
| "learning_rate": 8.415419751390155e-05, | |
| "loss": 0.7794, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.6630960702896118, | |
| "learning_rate": 8.353923650696118e-05, | |
| "loss": 0.7118, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9242424242424242, | |
| "grad_norm": 0.7102512121200562, | |
| "learning_rate": 8.291491008316409e-05, | |
| "loss": 0.7812, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9393939393939394, | |
| "grad_norm": 0.7596468329429626, | |
| "learning_rate": 8.228139257794012e-05, | |
| "loss": 0.7273, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9545454545454546, | |
| "grad_norm": 0.7956432700157166, | |
| "learning_rate": 8.163886089321493e-05, | |
| "loss": 0.6815, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 1.2094839811325073, | |
| "learning_rate": 8.098749444801224e-05, | |
| "loss": 0.7167, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.9848484848484849, | |
| "grad_norm": 0.5577558875083923, | |
| "learning_rate": 8.032747512835337e-05, | |
| "loss": 0.7765, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7755289077758789, | |
| "learning_rate": 7.965898723646776e-05, | |
| "loss": 0.7962, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0151515151515151, | |
| "grad_norm": 0.46732282638549805, | |
| "learning_rate": 7.898221743932888e-05, | |
| "loss": 0.6422, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 0.5171197652816772, | |
| "learning_rate": 7.829735471652978e-05, | |
| "loss": 0.698, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0454545454545454, | |
| "grad_norm": 0.5156084299087524, | |
| "learning_rate": 7.760459030751284e-05, | |
| "loss": 0.6649, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0606060606060606, | |
| "grad_norm": 0.5573360323905945, | |
| "learning_rate": 7.690411765816864e-05, | |
| "loss": 0.7386, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0757575757575757, | |
| "grad_norm": 0.5586574077606201, | |
| "learning_rate": 7.619613236681843e-05, | |
| "loss": 0.7219, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.5802133679389954, | |
| "learning_rate": 7.548083212959588e-05, | |
| "loss": 0.7142, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.106060606060606, | |
| "grad_norm": 0.5919927358627319, | |
| "learning_rate": 7.475841668524268e-05, | |
| "loss": 0.7303, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.121212121212121, | |
| "grad_norm": 0.5782243013381958, | |
| "learning_rate": 7.402908775933419e-05, | |
| "loss": 0.6328, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 0.6222028732299805, | |
| "learning_rate": 7.329304900794991e-05, | |
| "loss": 0.6729, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 0.5943819880485535, | |
| "learning_rate": 7.255050596080509e-05, | |
| "loss": 0.6003, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 0.6687825918197632, | |
| "learning_rate": 7.180166596385914e-05, | |
| "loss": 0.691, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 0.6929337978363037, | |
| "learning_rate": 7.104673812141675e-05, | |
| "loss": 0.6827, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.196969696969697, | |
| "grad_norm": 0.6936319470405579, | |
| "learning_rate": 7.02859332377382e-05, | |
| "loss": 0.5991, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 0.6937688589096069, | |
| "learning_rate": 6.951946375817474e-05, | |
| "loss": 0.533, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2272727272727273, | |
| "grad_norm": 0.7830970883369446, | |
| "learning_rate": 6.874754370984606e-05, | |
| "loss": 0.55, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2424242424242424, | |
| "grad_norm": 0.9644233584403992, | |
| "learning_rate": 6.797038864187564e-05, | |
| "loss": 0.5428, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2575757575757576, | |
| "grad_norm": 0.5709019303321838, | |
| "learning_rate": 6.718821556520151e-05, | |
| "loss": 0.5639, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.6445361375808716, | |
| "learning_rate": 6.640124289197845e-05, | |
| "loss": 0.7243, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.2878787878787878, | |
| "grad_norm": 0.6321728825569153, | |
| "learning_rate": 6.560969037458933e-05, | |
| "loss": 0.7057, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.303030303030303, | |
| "grad_norm": 0.6092734932899475, | |
| "learning_rate": 6.481377904428171e-05, | |
| "loss": 0.6839, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.3181818181818181, | |
| "grad_norm": 0.6245138049125671, | |
| "learning_rate": 6.401373114944781e-05, | |
| "loss": 0.7332, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.6300857067108154, | |
| "learning_rate": 6.320977009356431e-05, | |
| "loss": 0.752, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.3484848484848486, | |
| "grad_norm": 0.6165737509727478, | |
| "learning_rate": 6.240212037280966e-05, | |
| "loss": 0.7212, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.6268854141235352, | |
| "learning_rate": 6.159100751337642e-05, | |
| "loss": 0.7093, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.378787878787879, | |
| "grad_norm": 0.627152681350708, | |
| "learning_rate": 6.077665800849568e-05, | |
| "loss": 0.6636, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.393939393939394, | |
| "grad_norm": 0.6311261653900146, | |
| "learning_rate": 5.99592992551918e-05, | |
| "loss": 0.6638, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4090909090909092, | |
| "grad_norm": 0.6642055511474609, | |
| "learning_rate": 5.913915949078452e-05, | |
| "loss": 0.6817, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.4242424242424243, | |
| "grad_norm": 0.669463038444519, | |
| "learning_rate": 5.831646772915651e-05, | |
| "loss": 0.6835, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.4393939393939394, | |
| "grad_norm": 0.7148963809013367, | |
| "learning_rate": 5.749145369680407e-05, | |
| "loss": 0.6229, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.7326797246932983, | |
| "learning_rate": 5.666434776868895e-05, | |
| "loss": 0.5968, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.4696969696969697, | |
| "grad_norm": 0.6959696412086487, | |
| "learning_rate": 5.583538090390882e-05, | |
| "loss": 0.4222, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.4848484848484849, | |
| "grad_norm": 0.9565514922142029, | |
| "learning_rate": 5.5004784581204927e-05, | |
| "loss": 0.4486, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.544596791267395, | |
| "learning_rate": 5.41727907343245e-05, | |
| "loss": 0.662, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 0.6116472482681274, | |
| "learning_rate": 5.3339631687256084e-05, | |
| "loss": 0.6868, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "eval_loss": 0.765149712562561, | |
| "eval_runtime": 1.9996, | |
| "eval_samples_per_second": 55.51, | |
| "eval_steps_per_second": 14.003, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5303030303030303, | |
| "grad_norm": 0.6392404437065125, | |
| "learning_rate": 5.250554008935596e-05, | |
| "loss": 0.7089, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 0.6536526083946228, | |
| "learning_rate": 5.167074885038373e-05, | |
| "loss": 0.6887, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.5606060606060606, | |
| "grad_norm": 0.6564599275588989, | |
| "learning_rate": 5.0835491075465045e-05, | |
| "loss": 0.7512, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 0.6332715153694153, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6863, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.5909090909090908, | |
| "grad_norm": 0.651404857635498, | |
| "learning_rate": 4.916450892453495e-05, | |
| "loss": 0.6958, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.606060606060606, | |
| "grad_norm": 0.6749990582466125, | |
| "learning_rate": 4.832925114961629e-05, | |
| "loss": 0.7117, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.621212121212121, | |
| "grad_norm": 0.6680457592010498, | |
| "learning_rate": 4.749445991064404e-05, | |
| "loss": 0.6995, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.6417652368545532, | |
| "learning_rate": 4.666036831274392e-05, | |
| "loss": 0.6246, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.6515151515151514, | |
| "grad_norm": 0.6964362859725952, | |
| "learning_rate": 4.582720926567552e-05, | |
| "loss": 0.6556, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.6826117038726807, | |
| "learning_rate": 4.4995215418795085e-05, | |
| "loss": 0.5866, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.6818181818181817, | |
| "grad_norm": 0.6842771172523499, | |
| "learning_rate": 4.416461909609119e-05, | |
| "loss": 0.6205, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 0.7190883755683899, | |
| "learning_rate": 4.333565223131107e-05, | |
| "loss": 0.5752, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.7121212121212122, | |
| "grad_norm": 0.7393652200698853, | |
| "learning_rate": 4.250854630319593e-05, | |
| "loss": 0.4992, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 0.9857649803161621, | |
| "learning_rate": 4.1683532270843504e-05, | |
| "loss": 0.4295, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.7424242424242424, | |
| "grad_norm": 0.5148379802703857, | |
| "learning_rate": 4.0860840509215496e-05, | |
| "loss": 0.664, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.7575757575757576, | |
| "grad_norm": 0.5785648226737976, | |
| "learning_rate": 4.0040700744808204e-05, | |
| "loss": 0.7172, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.7727272727272727, | |
| "grad_norm": 0.5515451431274414, | |
| "learning_rate": 3.922334199150432e-05, | |
| "loss": 0.6258, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.7878787878787878, | |
| "grad_norm": 0.6018675565719604, | |
| "learning_rate": 3.840899248662358e-05, | |
| "loss": 0.6799, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.803030303030303, | |
| "grad_norm": 0.6062435507774353, | |
| "learning_rate": 3.7597879627190334e-05, | |
| "loss": 0.7231, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.6036238074302673, | |
| "learning_rate": 3.6790229906435705e-05, | |
| "loss": 0.6664, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 0.620060384273529, | |
| "learning_rate": 3.598626885055219e-05, | |
| "loss": 0.6652, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.8484848484848486, | |
| "grad_norm": 0.6296847462654114, | |
| "learning_rate": 3.5186220955718306e-05, | |
| "loss": 0.6872, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.8636363636363638, | |
| "grad_norm": 0.6178891658782959, | |
| "learning_rate": 3.4390309625410686e-05, | |
| "loss": 0.5877, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.878787878787879, | |
| "grad_norm": 0.6331610679626465, | |
| "learning_rate": 3.3598757108021546e-05, | |
| "loss": 0.6341, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.893939393939394, | |
| "grad_norm": 0.6539804339408875, | |
| "learning_rate": 3.281178443479852e-05, | |
| "loss": 0.5815, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 0.700760006904602, | |
| "learning_rate": 3.202961135812437e-05, | |
| "loss": 0.6294, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.9242424242424243, | |
| "grad_norm": 0.733697235584259, | |
| "learning_rate": 3.1252456290153954e-05, | |
| "loss": 0.6539, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 0.7333973050117493, | |
| "learning_rate": 3.0480536241825263e-05, | |
| "loss": 0.5779, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.9545454545454546, | |
| "grad_norm": 0.7873007655143738, | |
| "learning_rate": 2.9714066762261823e-05, | |
| "loss": 0.5233, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.9696969696969697, | |
| "grad_norm": 1.0427298545837402, | |
| "learning_rate": 2.895326187858326e-05, | |
| "loss": 0.5342, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.9848484848484849, | |
| "grad_norm": 0.5865942239761353, | |
| "learning_rate": 2.8198334036140874e-05, | |
| "loss": 0.6578, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8073463439941406, | |
| "learning_rate": 2.74494940391949e-05, | |
| "loss": 0.5917, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.015151515151515, | |
| "grad_norm": 0.48573416471481323, | |
| "learning_rate": 2.6706950992050094e-05, | |
| "loss": 0.6578, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.0303030303030303, | |
| "grad_norm": 0.5394845008850098, | |
| "learning_rate": 2.5970912240665813e-05, | |
| "loss": 0.6499, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0454545454545454, | |
| "grad_norm": 0.5604919195175171, | |
| "learning_rate": 2.5241583314757327e-05, | |
| "loss": 0.6638, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 0.5522144436836243, | |
| "learning_rate": 2.4519167870404125e-05, | |
| "loss": 0.5831, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.0757575757575757, | |
| "grad_norm": 0.591879665851593, | |
| "learning_rate": 2.3803867633181574e-05, | |
| "loss": 0.6475, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.090909090909091, | |
| "grad_norm": 0.5956389307975769, | |
| "learning_rate": 2.3095882341831372e-05, | |
| "loss": 0.6239, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.106060606060606, | |
| "grad_norm": 0.6151282787322998, | |
| "learning_rate": 2.2395409692487175e-05, | |
| "loss": 0.6135, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 0.6187341213226318, | |
| "learning_rate": 2.1702645283470236e-05, | |
| "loss": 0.6041, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.1363636363636362, | |
| "grad_norm": 0.6386223435401917, | |
| "learning_rate": 2.1017782560671123e-05, | |
| "loss": 0.5876, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.1515151515151514, | |
| "grad_norm": 0.6396329402923584, | |
| "learning_rate": 2.0341012763532243e-05, | |
| "loss": 0.5575, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.1666666666666665, | |
| "grad_norm": 0.6682023406028748, | |
| "learning_rate": 1.967252487164663e-05, | |
| "loss": 0.5834, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.6912877559661865, | |
| "learning_rate": 1.9012505551987765e-05, | |
| "loss": 0.5732, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.196969696969697, | |
| "grad_norm": 0.71503084897995, | |
| "learning_rate": 1.836113910678507e-05, | |
| "loss": 0.5498, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.212121212121212, | |
| "grad_norm": 0.7472315430641174, | |
| "learning_rate": 1.771860742205988e-05, | |
| "loss": 0.5746, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.227272727272727, | |
| "grad_norm": 0.7133623361587524, | |
| "learning_rate": 1.7085089916835923e-05, | |
| "loss": 0.3859, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.242424242424242, | |
| "grad_norm": 0.9030516743659973, | |
| "learning_rate": 1.646076349303884e-05, | |
| "loss": 0.4046, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.257575757575758, | |
| "grad_norm": 0.5055866837501526, | |
| "learning_rate": 1.584580248609846e-05, | |
| "loss": 0.5224, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 0.6145337224006653, | |
| "learning_rate": 1.5240378616267886e-05, | |
| "loss": 0.6049, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "eval_loss": 0.7746257185935974, | |
| "eval_runtime": 1.9852, | |
| "eval_samples_per_second": 55.913, | |
| "eval_steps_per_second": 14.104, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.287878787878788, | |
| "grad_norm": 0.6109030246734619, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.5784, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 0.6387854814529419, | |
| "learning_rate": 1.4058815806103542e-05, | |
| "loss": 0.625, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.3181818181818183, | |
| "grad_norm": 0.659665048122406, | |
| "learning_rate": 1.3483006802566544e-05, | |
| "loss": 0.6757, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.6564140319824219, | |
| "learning_rate": 1.2917394717602121e-05, | |
| "loss": 0.6371, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.3484848484848486, | |
| "grad_norm": 0.6709268689155579, | |
| "learning_rate": 1.2362137491387432e-05, | |
| "loss": 0.6282, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.6831066012382507, | |
| "learning_rate": 1.1817390172633403e-05, | |
| "loss": 0.6034, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.378787878787879, | |
| "grad_norm": 0.665997326374054, | |
| "learning_rate": 1.1283304875289336e-05, | |
| "loss": 0.568, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.393939393939394, | |
| "grad_norm": 0.6755225658416748, | |
| "learning_rate": 1.0760030736066951e-05, | |
| "loss": 0.5589, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.409090909090909, | |
| "grad_norm": 0.7002132534980774, | |
| "learning_rate": 1.024771387279585e-05, | |
| "loss": 0.5588, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 0.716273307800293, | |
| "learning_rate": 9.746497343621857e-06, | |
| "loss": 0.5729, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.4393939393939394, | |
| "grad_norm": 0.717272162437439, | |
| "learning_rate": 9.256521107059834e-06, | |
| "loss": 0.5198, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.4545454545454546, | |
| "grad_norm": 0.7685539722442627, | |
| "learning_rate": 8.777921982911996e-06, | |
| "loss": 0.5231, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.4696969696969697, | |
| "grad_norm": 0.7818545699119568, | |
| "learning_rate": 8.310833614062651e-06, | |
| "loss": 0.4314, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.484848484848485, | |
| "grad_norm": 0.9245851039886475, | |
| "learning_rate": 7.85538642916015e-06, | |
| "loss": 0.3269, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.5567438006401062, | |
| "learning_rate": 7.4117076061961885e-06, | |
| "loss": 0.5873, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.515151515151515, | |
| "grad_norm": 0.5879104733467102, | |
| "learning_rate": 6.979921036993042e-06, | |
| "loss": 0.6398, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.5303030303030303, | |
| "grad_norm": 0.6203252077102661, | |
| "learning_rate": 6.5601472926081766e-06, | |
| "loss": 0.6471, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.6278418302536011, | |
| "learning_rate": 6.152503589666425e-06, | |
| "loss": 0.6167, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.5606060606060606, | |
| "grad_norm": 0.6504820585250854, | |
| "learning_rate": 5.757103757628573e-06, | |
| "loss": 0.6541, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.5757575757575757, | |
| "grad_norm": 0.6489107012748718, | |
| "learning_rate": 5.374058207005944e-06, | |
| "loss": 0.6618, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.590909090909091, | |
| "grad_norm": 0.6700452566146851, | |
| "learning_rate": 5.0034738985296095e-06, | |
| "loss": 0.6294, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.606060606060606, | |
| "grad_norm": 0.651507556438446, | |
| "learning_rate": 4.645454313282965e-06, | |
| "loss": 0.5573, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.621212121212121, | |
| "grad_norm": 0.6590191721916199, | |
| "learning_rate": 4.3000994238058644e-06, | |
| "loss": 0.5596, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.6363636363636362, | |
| "grad_norm": 0.6777567863464355, | |
| "learning_rate": 3.967505666178556e-06, | |
| "loss": 0.5544, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.6515151515151514, | |
| "grad_norm": 0.691135585308075, | |
| "learning_rate": 3.647765913093132e-06, | |
| "loss": 0.5732, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.7173659205436707, | |
| "learning_rate": 3.340969447919873e-06, | |
| "loss": 0.5273, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.6818181818181817, | |
| "grad_norm": 0.7347037196159363, | |
| "learning_rate": 3.0472019397761064e-06, | |
| "loss": 0.4899, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.6969696969696972, | |
| "grad_norm": 0.724153995513916, | |
| "learning_rate": 2.7665454196040664e-06, | |
| "loss": 0.4094, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.712121212121212, | |
| "grad_norm": 0.7561957836151123, | |
| "learning_rate": 2.4990782572647975e-06, | |
| "loss": 0.3645, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.9447779059410095, | |
| "learning_rate": 2.2448751396543787e-06, | |
| "loss": 0.3088, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.742424242424242, | |
| "grad_norm": 0.519648551940918, | |
| "learning_rate": 2.004007049848461e-06, | |
| "loss": 0.5594, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.757575757575758, | |
| "grad_norm": 0.5930406451225281, | |
| "learning_rate": 1.7765412472811771e-06, | |
| "loss": 0.5954, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.7727272727272725, | |
| "grad_norm": 0.6299811601638794, | |
| "learning_rate": 1.5625412489637337e-06, | |
| "loss": 0.658, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 0.6249811053276062, | |
| "learning_rate": 1.3620668117481472e-06, | |
| "loss": 0.6387, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.8030303030303028, | |
| "grad_norm": 0.6202895641326904, | |
| "learning_rate": 1.1751739156407649e-06, | |
| "loss": 0.5981, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.8181818181818183, | |
| "grad_norm": 0.6588528156280518, | |
| "learning_rate": 1.0019147481706625e-06, | |
| "loss": 0.624, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.8333333333333335, | |
| "grad_norm": 0.6690031290054321, | |
| "learning_rate": 8.423376898168245e-07, | |
| "loss": 0.6368, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.8484848484848486, | |
| "grad_norm": 0.6783432364463806, | |
| "learning_rate": 6.964873004985717e-07, | |
| "loss": 0.6003, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.8636363636363638, | |
| "grad_norm": 0.6635212302207947, | |
| "learning_rate": 5.644043071326932e-07, | |
| "loss": 0.5507, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.878787878787879, | |
| "grad_norm": 0.6875348687171936, | |
| "learning_rate": 4.461255922609986e-07, | |
| "loss": 0.5495, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.893939393939394, | |
| "grad_norm": 0.6876786351203918, | |
| "learning_rate": 3.416841837512952e-07, | |
| "loss": 0.5434, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.7225180268287659, | |
| "learning_rate": 2.511092455747932e-07, | |
| "loss": 0.5445, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.9242424242424243, | |
| "grad_norm": 0.7289068698883057, | |
| "learning_rate": 1.7442606966242004e-07, | |
| "loss": 0.5363, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.9393939393939394, | |
| "grad_norm": 0.7535192370414734, | |
| "learning_rate": 1.1165606884234181e-07, | |
| "loss": 0.4994, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.9545454545454546, | |
| "grad_norm": 0.7919872403144836, | |
| "learning_rate": 6.281677086071303e-08, | |
| "loss": 0.4289, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.9696969696969697, | |
| "grad_norm": 1.0089852809906006, | |
| "learning_rate": 2.792181348726941e-08, | |
| "loss": 0.3698, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.984848484848485, | |
| "grad_norm": 0.6285845637321472, | |
| "learning_rate": 6.980940707146389e-09, | |
| "loss": 0.6425, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8173950910568237, | |
| "learning_rate": 0.0, | |
| "loss": 0.5206, | |
| "step": 198 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 198, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5266167953555456e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |