| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.984375, | |
| "eval_steps": 500, | |
| "global_step": 170, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01171875, | |
| "grad_norm": 36.23282241821289, | |
| "learning_rate": 5.0000000000000004e-08, | |
| "loss": 2.3839, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0234375, | |
| "grad_norm": 35.918636322021484, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 2.3798, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03515625, | |
| "grad_norm": 35.62618637084961, | |
| "learning_rate": 1.5000000000000002e-07, | |
| "loss": 2.386, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.046875, | |
| "grad_norm": 35.966087341308594, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 2.3803, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05859375, | |
| "grad_norm": 35.38177490234375, | |
| "learning_rate": 2.5000000000000004e-07, | |
| "loss": 2.3937, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0703125, | |
| "grad_norm": 35.99677658081055, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 2.3906, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.08203125, | |
| "grad_norm": 35.44341278076172, | |
| "learning_rate": 3.5000000000000004e-07, | |
| "loss": 2.3539, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.09375, | |
| "grad_norm": 35.300697326660156, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 2.3459, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.10546875, | |
| "grad_norm": 34.092952728271484, | |
| "learning_rate": 4.5000000000000003e-07, | |
| "loss": 2.2959, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1171875, | |
| "grad_norm": 34.46371841430664, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 2.2661, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12890625, | |
| "grad_norm": 34.62260818481445, | |
| "learning_rate": 5.5e-07, | |
| "loss": 2.2918, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.140625, | |
| "grad_norm": 33.790374755859375, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 2.223, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.15234375, | |
| "grad_norm": 33.766536712646484, | |
| "learning_rate": 6.5e-07, | |
| "loss": 2.2267, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1640625, | |
| "grad_norm": 33.894081115722656, | |
| "learning_rate": 7.000000000000001e-07, | |
| "loss": 2.1465, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.17578125, | |
| "grad_norm": 33.162452697753906, | |
| "learning_rate": 7.5e-07, | |
| "loss": 2.0495, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1875, | |
| "grad_norm": 32.954341888427734, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 1.9627, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.19921875, | |
| "grad_norm": 33.96324157714844, | |
| "learning_rate": 8.500000000000001e-07, | |
| "loss": 1.8867, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2109375, | |
| "grad_norm": 33.81139373779297, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 1.7752, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.22265625, | |
| "grad_norm": 34.87086868286133, | |
| "learning_rate": 9.500000000000001e-07, | |
| "loss": 1.6944, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.234375, | |
| "grad_norm": 34.84965133666992, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.5707, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.24609375, | |
| "grad_norm": 35.227317810058594, | |
| "learning_rate": 1.0500000000000001e-06, | |
| "loss": 1.4369, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2578125, | |
| "grad_norm": 34.91344451904297, | |
| "learning_rate": 1.1e-06, | |
| "loss": 1.3202, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.26953125, | |
| "grad_norm": 31.7376766204834, | |
| "learning_rate": 1.1500000000000002e-06, | |
| "loss": 1.1398, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.28125, | |
| "grad_norm": 30.24741554260254, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 1.0421, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.29296875, | |
| "grad_norm": 28.292400360107422, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.8817, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3046875, | |
| "grad_norm": 30.44672393798828, | |
| "learning_rate": 1.3e-06, | |
| "loss": 0.7073, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.31640625, | |
| "grad_norm": 29.416427612304688, | |
| "learning_rate": 1.3500000000000002e-06, | |
| "loss": 0.5444, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.328125, | |
| "grad_norm": 24.820096969604492, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.4025, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.33984375, | |
| "grad_norm": 21.023277282714844, | |
| "learning_rate": 1.45e-06, | |
| "loss": 0.307, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3515625, | |
| "grad_norm": 19.656967163085938, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.2151, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.36328125, | |
| "grad_norm": 14.91929817199707, | |
| "learning_rate": 1.5500000000000002e-06, | |
| "loss": 0.1448, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 5.083199977874756, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.09, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.38671875, | |
| "grad_norm": 2.320681571960449, | |
| "learning_rate": 1.6500000000000003e-06, | |
| "loss": 0.0641, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3984375, | |
| "grad_norm": 1.6233159303665161, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 0.0584, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.41015625, | |
| "grad_norm": 1.6057201623916626, | |
| "learning_rate": 1.75e-06, | |
| "loss": 0.0626, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.421875, | |
| "grad_norm": 1.8360320329666138, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.0563, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.43359375, | |
| "grad_norm": 1.736350178718567, | |
| "learning_rate": 1.85e-06, | |
| "loss": 0.0609, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4453125, | |
| "grad_norm": 1.1473922729492188, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 0.0541, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.45703125, | |
| "grad_norm": 1.1722168922424316, | |
| "learning_rate": 1.9500000000000004e-06, | |
| "loss": 0.0534, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.46875, | |
| "grad_norm": 1.356987714767456, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.0496, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.48046875, | |
| "grad_norm": 0.8023216724395752, | |
| "learning_rate": 2.05e-06, | |
| "loss": 0.0527, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4921875, | |
| "grad_norm": 0.9803515672683716, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 0.0478, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.50390625, | |
| "grad_norm": 0.8733468651771545, | |
| "learning_rate": 2.15e-06, | |
| "loss": 0.052, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.515625, | |
| "grad_norm": 0.8213743567466736, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.0448, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.52734375, | |
| "grad_norm": 0.843189537525177, | |
| "learning_rate": 2.25e-06, | |
| "loss": 0.0498, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5390625, | |
| "grad_norm": 0.8801079392433167, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 0.0408, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.55078125, | |
| "grad_norm": 0.7131401300430298, | |
| "learning_rate": 2.35e-06, | |
| "loss": 0.0405, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5625, | |
| "grad_norm": 0.8996126651763916, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.0525, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.57421875, | |
| "grad_norm": 0.8606986403465271, | |
| "learning_rate": 2.4500000000000003e-06, | |
| "loss": 0.0438, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5859375, | |
| "grad_norm": 0.6918051838874817, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0394, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.59765625, | |
| "grad_norm": 0.6177802085876465, | |
| "learning_rate": 2.55e-06, | |
| "loss": 0.0387, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.609375, | |
| "grad_norm": 0.7042555809020996, | |
| "learning_rate": 2.6e-06, | |
| "loss": 0.0434, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.62109375, | |
| "grad_norm": 0.6537717580795288, | |
| "learning_rate": 2.6500000000000005e-06, | |
| "loss": 0.0396, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6328125, | |
| "grad_norm": 0.7834082841873169, | |
| "learning_rate": 2.7000000000000004e-06, | |
| "loss": 0.0411, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.64453125, | |
| "grad_norm": 0.7287272810935974, | |
| "learning_rate": 2.7500000000000004e-06, | |
| "loss": 0.0408, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.65625, | |
| "grad_norm": 0.7186263203620911, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.0394, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.66796875, | |
| "grad_norm": 0.7264899611473083, | |
| "learning_rate": 2.85e-06, | |
| "loss": 0.0427, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6796875, | |
| "grad_norm": 0.7665618062019348, | |
| "learning_rate": 2.9e-06, | |
| "loss": 0.0368, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.69140625, | |
| "grad_norm": 0.7222962379455566, | |
| "learning_rate": 2.95e-06, | |
| "loss": 0.0412, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.703125, | |
| "grad_norm": 0.7061101794242859, | |
| "learning_rate": 3e-06, | |
| "loss": 0.0377, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.71484375, | |
| "grad_norm": 0.5724324584007263, | |
| "learning_rate": 3.05e-06, | |
| "loss": 0.0387, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7265625, | |
| "grad_norm": 0.5535506010055542, | |
| "learning_rate": 3.1000000000000004e-06, | |
| "loss": 0.0403, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.73828125, | |
| "grad_norm": 0.6553678512573242, | |
| "learning_rate": 3.1500000000000003e-06, | |
| "loss": 0.0415, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.6137285828590393, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.0383, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.76171875, | |
| "grad_norm": 0.5985754132270813, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 0.0355, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7734375, | |
| "grad_norm": 0.5903909802436829, | |
| "learning_rate": 3.3000000000000006e-06, | |
| "loss": 0.0374, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.78515625, | |
| "grad_norm": 0.5718765258789062, | |
| "learning_rate": 3.3500000000000005e-06, | |
| "loss": 0.0339, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.796875, | |
| "grad_norm": 0.6844965815544128, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.0405, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.80859375, | |
| "grad_norm": 0.5959618091583252, | |
| "learning_rate": 3.45e-06, | |
| "loss": 0.0338, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.8203125, | |
| "grad_norm": 0.6095123291015625, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.0362, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.83203125, | |
| "grad_norm": 0.543708086013794, | |
| "learning_rate": 3.5500000000000003e-06, | |
| "loss": 0.0355, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.84375, | |
| "grad_norm": 0.6969983577728271, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.0325, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.85546875, | |
| "grad_norm": 0.6022969484329224, | |
| "learning_rate": 3.65e-06, | |
| "loss": 0.0342, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8671875, | |
| "grad_norm": 0.6262147426605225, | |
| "learning_rate": 3.7e-06, | |
| "loss": 0.0348, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.87890625, | |
| "grad_norm": 0.5729933381080627, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.0318, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.890625, | |
| "grad_norm": 0.5846775770187378, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.0309, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.90234375, | |
| "grad_norm": 0.6469219923019409, | |
| "learning_rate": 3.85e-06, | |
| "loss": 0.0324, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9140625, | |
| "grad_norm": 0.6574859023094177, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 0.0325, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.92578125, | |
| "grad_norm": 0.5833832025527954, | |
| "learning_rate": 3.95e-06, | |
| "loss": 0.0232, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.9375, | |
| "grad_norm": 0.7503570318222046, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.0267, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.94921875, | |
| "grad_norm": 0.7181633114814758, | |
| "learning_rate": 4.05e-06, | |
| "loss": 0.0304, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.9609375, | |
| "grad_norm": 0.6477274298667908, | |
| "learning_rate": 4.1e-06, | |
| "loss": 0.0297, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.97265625, | |
| "grad_norm": 0.6768563389778137, | |
| "learning_rate": 4.15e-06, | |
| "loss": 0.0279, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.984375, | |
| "grad_norm": 0.7905837297439575, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.0301, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.99609375, | |
| "grad_norm": 0.5576608777046204, | |
| "learning_rate": 4.25e-06, | |
| "loss": 0.0322, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5576608777046204, | |
| "learning_rate": 4.3e-06, | |
| "loss": 0.0226, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.01171875, | |
| "grad_norm": 1.0774812698364258, | |
| "learning_rate": 4.350000000000001e-06, | |
| "loss": 0.0215, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.0234375, | |
| "grad_norm": 0.47373324632644653, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.0235, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.03515625, | |
| "grad_norm": 0.7665970325469971, | |
| "learning_rate": 4.450000000000001e-06, | |
| "loss": 0.0242, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.046875, | |
| "grad_norm": 0.6290147304534912, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.0209, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.05859375, | |
| "grad_norm": 0.5703024864196777, | |
| "learning_rate": 4.5500000000000005e-06, | |
| "loss": 0.0192, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0703125, | |
| "grad_norm": 0.6099259853363037, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.0181, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.08203125, | |
| "grad_norm": 0.6570988297462463, | |
| "learning_rate": 4.65e-06, | |
| "loss": 0.0201, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.09375, | |
| "grad_norm": 0.7848325371742249, | |
| "learning_rate": 4.7e-06, | |
| "loss": 0.0253, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.10546875, | |
| "grad_norm": 0.6759209036827087, | |
| "learning_rate": 4.75e-06, | |
| "loss": 0.0195, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.1171875, | |
| "grad_norm": 0.4861151874065399, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.0191, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.12890625, | |
| "grad_norm": 0.6268576383590698, | |
| "learning_rate": 4.85e-06, | |
| "loss": 0.0211, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.140625, | |
| "grad_norm": 0.5862017869949341, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 0.0177, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.15234375, | |
| "grad_norm": 0.4569724202156067, | |
| "learning_rate": 4.95e-06, | |
| "loss": 0.0164, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.1640625, | |
| "grad_norm": 0.4539048969745636, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0152, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.17578125, | |
| "grad_norm": 0.4553528428077698, | |
| "learning_rate": 4.999926609487568e-06, | |
| "loss": 0.0208, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.1875, | |
| "grad_norm": 0.5182592272758484, | |
| "learning_rate": 4.999706442259205e-06, | |
| "loss": 0.0154, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.19921875, | |
| "grad_norm": 0.5602673888206482, | |
| "learning_rate": 4.999339511241458e-06, | |
| "loss": 0.0196, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.2109375, | |
| "grad_norm": 0.7579494118690491, | |
| "learning_rate": 4.9988258379777334e-06, | |
| "loss": 0.0198, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.22265625, | |
| "grad_norm": 0.603757381439209, | |
| "learning_rate": 4.998165452627025e-06, | |
| "loss": 0.0185, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.234375, | |
| "grad_norm": 0.5520291924476624, | |
| "learning_rate": 4.99735839396215e-06, | |
| "loss": 0.018, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.24609375, | |
| "grad_norm": 0.55808424949646, | |
| "learning_rate": 4.996404709367466e-06, | |
| "loss": 0.0159, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.2578125, | |
| "grad_norm": 0.47174298763275146, | |
| "learning_rate": 4.995304454836095e-06, | |
| "loss": 0.0122, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.26953125, | |
| "grad_norm": 0.5289337038993835, | |
| "learning_rate": 4.994057694966632e-06, | |
| "loss": 0.0168, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.28125, | |
| "grad_norm": 0.5390430092811584, | |
| "learning_rate": 4.992664502959351e-06, | |
| "loss": 0.017, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.29296875, | |
| "grad_norm": 0.4966451823711395, | |
| "learning_rate": 4.991124960611916e-06, | |
| "loss": 0.0145, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.3046875, | |
| "grad_norm": 0.6148604154586792, | |
| "learning_rate": 4.989439158314566e-06, | |
| "loss": 0.0139, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.31640625, | |
| "grad_norm": 0.6303534507751465, | |
| "learning_rate": 4.9876071950448185e-06, | |
| "loss": 0.0118, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.328125, | |
| "grad_norm": 0.5410207509994507, | |
| "learning_rate": 4.98562917836165e-06, | |
| "loss": 0.0094, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.33984375, | |
| "grad_norm": 0.5350080132484436, | |
| "learning_rate": 4.983505224399188e-06, | |
| "loss": 0.0158, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.3515625, | |
| "grad_norm": 1.017317295074463, | |
| "learning_rate": 4.9812354578598876e-06, | |
| "loss": 0.0201, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.36328125, | |
| "grad_norm": 0.6891007423400879, | |
| "learning_rate": 4.978820012007213e-06, | |
| "loss": 0.0127, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.4756389260292053, | |
| "learning_rate": 4.976259028657812e-06, | |
| "loss": 0.0188, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.38671875, | |
| "grad_norm": 0.5957350730895996, | |
| "learning_rate": 4.973552658173186e-06, | |
| "loss": 0.011, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3984375, | |
| "grad_norm": 0.5012223720550537, | |
| "learning_rate": 4.970701059450872e-06, | |
| "loss": 0.0138, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.41015625, | |
| "grad_norm": 0.4408419132232666, | |
| "learning_rate": 4.9677043999151e-06, | |
| "loss": 0.0144, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.421875, | |
| "grad_norm": 0.5721736550331116, | |
| "learning_rate": 4.964562855506976e-06, | |
| "loss": 0.0135, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.43359375, | |
| "grad_norm": 0.5479208827018738, | |
| "learning_rate": 4.961276610674141e-06, | |
| "loss": 0.0128, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.4453125, | |
| "grad_norm": 1.0117675065994263, | |
| "learning_rate": 4.9578458583599495e-06, | |
| "loss": 0.0111, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.45703125, | |
| "grad_norm": 0.5504026412963867, | |
| "learning_rate": 4.954270799992138e-06, | |
| "loss": 0.0083, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.46875, | |
| "grad_norm": 0.48403099179267883, | |
| "learning_rate": 4.950551645470998e-06, | |
| "loss": 0.0083, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.48046875, | |
| "grad_norm": 0.6866800785064697, | |
| "learning_rate": 4.9466886131570565e-06, | |
| "loss": 0.0085, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.4921875, | |
| "grad_norm": 0.872557520866394, | |
| "learning_rate": 4.942681929858249e-06, | |
| "loss": 0.0102, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.50390625, | |
| "grad_norm": 0.6924716234207153, | |
| "learning_rate": 4.9385318308166065e-06, | |
| "loss": 0.012, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.515625, | |
| "grad_norm": 0.5060118436813354, | |
| "learning_rate": 4.934238559694448e-06, | |
| "loss": 0.0084, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.52734375, | |
| "grad_norm": 0.6256171464920044, | |
| "learning_rate": 4.929802368560066e-06, | |
| "loss": 0.0081, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.5390625, | |
| "grad_norm": 0.5422537922859192, | |
| "learning_rate": 4.925223517872934e-06, | |
| "loss": 0.0077, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.55078125, | |
| "grad_norm": 0.953416109085083, | |
| "learning_rate": 4.920502276468408e-06, | |
| "loss": 0.0078, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.5625, | |
| "grad_norm": 0.4540804624557495, | |
| "learning_rate": 4.915638921541952e-06, | |
| "loss": 0.0097, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.57421875, | |
| "grad_norm": 0.3773641884326935, | |
| "learning_rate": 4.9106337386328524e-06, | |
| "loss": 0.0098, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.5859375, | |
| "grad_norm": 0.7970175743103027, | |
| "learning_rate": 4.905487021607462e-06, | |
| "loss": 0.0056, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.59765625, | |
| "grad_norm": 0.45197635889053345, | |
| "learning_rate": 4.900199072641937e-06, | |
| "loss": 0.0078, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.609375, | |
| "grad_norm": 0.38231438398361206, | |
| "learning_rate": 4.894770202204509e-06, | |
| "loss": 0.0072, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.62109375, | |
| "grad_norm": 0.2945426404476166, | |
| "learning_rate": 4.889200729037241e-06, | |
| "loss": 0.0086, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.6328125, | |
| "grad_norm": 0.49699363112449646, | |
| "learning_rate": 4.883490980137327e-06, | |
| "loss": 0.0073, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.64453125, | |
| "grad_norm": 0.38112956285476685, | |
| "learning_rate": 4.8776412907378845e-06, | |
| "loss": 0.0056, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.65625, | |
| "grad_norm": 0.46780407428741455, | |
| "learning_rate": 4.871652004288275e-06, | |
| "loss": 0.0078, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.66796875, | |
| "grad_norm": 0.43764325976371765, | |
| "learning_rate": 4.865523472433942e-06, | |
| "loss": 0.005, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.6796875, | |
| "grad_norm": 0.3445664644241333, | |
| "learning_rate": 4.859256054995758e-06, | |
| "loss": 0.0069, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.69140625, | |
| "grad_norm": 0.40410447120666504, | |
| "learning_rate": 4.8528501199489045e-06, | |
| "loss": 0.0088, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.703125, | |
| "grad_norm": 0.5876736640930176, | |
| "learning_rate": 4.846306043401268e-06, | |
| "loss": 0.0057, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.71484375, | |
| "grad_norm": 0.5149250626564026, | |
| "learning_rate": 4.839624209571352e-06, | |
| "loss": 0.0056, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.7265625, | |
| "grad_norm": 0.7009180784225464, | |
| "learning_rate": 4.832805010765724e-06, | |
| "loss": 0.0088, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.73828125, | |
| "grad_norm": 0.42258334159851074, | |
| "learning_rate": 4.8258488473559794e-06, | |
| "loss": 0.004, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.39231887459754944, | |
| "learning_rate": 4.8187561277552376e-06, | |
| "loss": 0.005, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.76171875, | |
| "grad_norm": 0.3317432701587677, | |
| "learning_rate": 4.811527268394157e-06, | |
| "loss": 0.0038, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.7734375, | |
| "grad_norm": 0.5022267699241638, | |
| "learning_rate": 4.804162693696494e-06, | |
| "loss": 0.0056, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.78515625, | |
| "grad_norm": 0.39019322395324707, | |
| "learning_rate": 4.796662836054176e-06, | |
| "loss": 0.0053, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.796875, | |
| "grad_norm": 0.5674042701721191, | |
| "learning_rate": 4.789028135801919e-06, | |
| "loss": 0.007, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.80859375, | |
| "grad_norm": 0.5690024495124817, | |
| "learning_rate": 4.7812590411913755e-06, | |
| "loss": 0.0053, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.8203125, | |
| "grad_norm": 0.23775412142276764, | |
| "learning_rate": 4.773356008364812e-06, | |
| "loss": 0.0031, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.83203125, | |
| "grad_norm": 0.4698558747768402, | |
| "learning_rate": 4.765319501328332e-06, | |
| "loss": 0.0021, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.84375, | |
| "grad_norm": 0.21603639423847198, | |
| "learning_rate": 4.757149991924633e-06, | |
| "loss": 0.0046, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.85546875, | |
| "grad_norm": 0.33830726146698, | |
| "learning_rate": 4.748847959805297e-06, | |
| "loss": 0.0022, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.8671875, | |
| "grad_norm": 0.44919782876968384, | |
| "learning_rate": 4.740413892402639e-06, | |
| "loss": 0.0032, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.87890625, | |
| "grad_norm": 0.5119614601135254, | |
| "learning_rate": 4.731848284901082e-06, | |
| "loss": 0.006, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.890625, | |
| "grad_norm": 0.3875437080860138, | |
| "learning_rate": 4.723151640208084e-06, | |
| "loss": 0.0024, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.90234375, | |
| "grad_norm": 0.3179910182952881, | |
| "learning_rate": 4.714324468924614e-06, | |
| "loss": 0.0037, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.9140625, | |
| "grad_norm": 0.43395644426345825, | |
| "learning_rate": 4.705367289315172e-06, | |
| "loss": 0.0027, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.92578125, | |
| "grad_norm": 0.3703945577144623, | |
| "learning_rate": 4.696280627277356e-06, | |
| "loss": 0.0047, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.9375, | |
| "grad_norm": 0.2503529191017151, | |
| "learning_rate": 4.687065016310996e-06, | |
| "loss": 0.0052, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.94921875, | |
| "grad_norm": 0.3613075315952301, | |
| "learning_rate": 4.6777209974868194e-06, | |
| "loss": 0.0034, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.9609375, | |
| "grad_norm": 0.3578515350818634, | |
| "learning_rate": 4.668249119414692e-06, | |
| "loss": 0.0021, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.97265625, | |
| "grad_norm": 0.1784515529870987, | |
| "learning_rate": 4.6586499382113985e-06, | |
| "loss": 0.0018, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.984375, | |
| "grad_norm": 0.259198397397995, | |
| "learning_rate": 4.648924017468003e-06, | |
| "loss": 0.0009, | |
| "step": 170 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 510, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 85, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.277422139847475e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |