| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9152, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 1.7099160686484538e-05, | |
| "learning_rate": 0.0, | |
| "loss": 2.8702, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 1.8180451661464758e-05, | |
| "learning_rate": 4e-05, | |
| "loss": 2.7103, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 1.6317009794875048e-05, | |
| "learning_rate": 8e-05, | |
| "loss": 2.7065, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 1.9236362277297303e-05, | |
| "learning_rate": 0.00012, | |
| "loss": 2.8041, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 1.9754068489419296e-05, | |
| "learning_rate": 0.00016, | |
| "loss": 2.7484, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 1.956670530489646e-05, | |
| "learning_rate": 0.0002, | |
| "loss": 2.6335, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 1.8515776901040226e-05, | |
| "learning_rate": 0.0001993220338983051, | |
| "loss": 2.5606, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 1.424604943167651e-05, | |
| "learning_rate": 0.00019864406779661017, | |
| "loss": 2.49, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 1.6697169485269114e-05, | |
| "learning_rate": 0.00019796610169491526, | |
| "loss": 2.3816, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 1.5206620446406305e-05, | |
| "learning_rate": 0.00019728813559322035, | |
| "loss": 2.3784, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 1.8514156181481667e-05, | |
| "learning_rate": 0.00019661016949152545, | |
| "loss": 2.2394, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 1.9562941815820523e-05, | |
| "learning_rate": 0.0001959322033898305, | |
| "loss": 2.1889, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 1.5121709111554082e-05, | |
| "learning_rate": 0.0001952542372881356, | |
| "loss": 2.152, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 1.7433398170396686e-05, | |
| "learning_rate": 0.0001945762711864407, | |
| "loss": 2.2189, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 1.6279864212265238e-05, | |
| "learning_rate": 0.0001938983050847458, | |
| "loss": 2.2305, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 1.6447980669909157e-05, | |
| "learning_rate": 0.00019322033898305085, | |
| "loss": 2.2208, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 1.414388225384755e-05, | |
| "learning_rate": 0.00019254237288135595, | |
| "loss": 2.3029, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 1.2313284059928264e-05, | |
| "learning_rate": 0.000191864406779661, | |
| "loss": 2.0473, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 1.828079803090077e-05, | |
| "learning_rate": 0.0001911864406779661, | |
| "loss": 2.0803, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 1.7634034520597197e-05, | |
| "learning_rate": 0.0001905084745762712, | |
| "loss": 2.104, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 2.085095184156671e-05, | |
| "learning_rate": 0.0001898305084745763, | |
| "loss": 2.022, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 1.3288636182551272e-05, | |
| "learning_rate": 0.00018915254237288136, | |
| "loss": 1.9308, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 1.620672810531687e-05, | |
| "learning_rate": 0.00018847457627118645, | |
| "loss": 2.0044, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 1.86624783964362e-05, | |
| "learning_rate": 0.00018779661016949151, | |
| "loss": 1.9754, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.754058393999003e-05, | |
| "learning_rate": 0.00018711864406779663, | |
| "loss": 1.9435, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.7156691683339886e-05, | |
| "learning_rate": 0.0001864406779661017, | |
| "loss": 1.8329, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.7366985048283823e-05, | |
| "learning_rate": 0.0001857627118644068, | |
| "loss": 1.8393, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 1.7832513549365103e-05, | |
| "learning_rate": 0.00018508474576271186, | |
| "loss": 1.8218, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1856, | |
| "grad_norm": 2.0902105461573228e-05, | |
| "learning_rate": 0.00018440677966101695, | |
| "loss": 1.8353, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.7098485841415823e-05, | |
| "learning_rate": 0.00018372881355932204, | |
| "loss": 1.7535, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1984, | |
| "grad_norm": 1.7024976841639727e-05, | |
| "learning_rate": 0.00018305084745762714, | |
| "loss": 1.857, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.6189233065233566e-05, | |
| "learning_rate": 0.0001823728813559322, | |
| "loss": 1.7131, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.6069398043327965e-05, | |
| "learning_rate": 0.0001816949152542373, | |
| "loss": 1.7581, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.238884033227805e-05, | |
| "learning_rate": 0.00018101694915254239, | |
| "loss": 1.5205, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 1.185688961413689e-05, | |
| "learning_rate": 0.00018033898305084748, | |
| "loss": 1.6754, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 8.686945875524543e-06, | |
| "learning_rate": 0.00017966101694915257, | |
| "loss": 1.7046, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2368, | |
| "grad_norm": 9.034667527885176e-06, | |
| "learning_rate": 0.00017898305084745764, | |
| "loss": 1.7919, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 8.488589628541376e-06, | |
| "learning_rate": 0.00017830508474576273, | |
| "loss": 1.7843, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 9.044196303875651e-06, | |
| "learning_rate": 0.0001776271186440678, | |
| "loss": 1.6485, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 8.335708116646856e-06, | |
| "learning_rate": 0.0001769491525423729, | |
| "loss": 1.6605, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2624, | |
| "grad_norm": 1.0821303476404864e-05, | |
| "learning_rate": 0.00017627118644067798, | |
| "loss": 1.5653, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 9.278046491090208e-06, | |
| "learning_rate": 0.00017559322033898307, | |
| "loss": 1.5807, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2752, | |
| "grad_norm": 9.212875738739967e-06, | |
| "learning_rate": 0.00017491525423728814, | |
| "loss": 1.5894, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 9.06234527064953e-06, | |
| "learning_rate": 0.00017423728813559323, | |
| "loss": 1.4922, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 8.783045814197976e-06, | |
| "learning_rate": 0.0001735593220338983, | |
| "loss": 1.6767, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 8.542882824258413e-06, | |
| "learning_rate": 0.00017288135593220342, | |
| "loss": 1.6121, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3008, | |
| "grad_norm": 1.0660360203473829e-05, | |
| "learning_rate": 0.00017220338983050848, | |
| "loss": 1.497, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 2.9956399885122664e-05, | |
| "learning_rate": 0.00017152542372881357, | |
| "loss": 1.6358, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3136, | |
| "grad_norm": 9.688827049103566e-06, | |
| "learning_rate": 0.00017084745762711864, | |
| "loss": 1.7185, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 8.132267794280779e-06, | |
| "learning_rate": 0.00017016949152542373, | |
| "loss": 1.5845, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 1.0473886504769325e-05, | |
| "learning_rate": 0.00016949152542372882, | |
| "loss": 1.4661, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 8.518994945916347e-06, | |
| "learning_rate": 0.00016881355932203392, | |
| "loss": 1.7661, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3392, | |
| "grad_norm": 1.0024953553511295e-05, | |
| "learning_rate": 0.00016813559322033898, | |
| "loss": 1.5461, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 8.88018803379964e-06, | |
| "learning_rate": 0.00016745762711864408, | |
| "loss": 1.5655, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 8.150585017574485e-06, | |
| "learning_rate": 0.00016677966101694914, | |
| "loss": 1.5083, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 7.710960744589102e-06, | |
| "learning_rate": 0.00016610169491525423, | |
| "loss": 1.4909, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 7.61793444326031e-06, | |
| "learning_rate": 0.00016542372881355933, | |
| "loss": 1.6288, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 8.035198334255256e-06, | |
| "learning_rate": 0.00016474576271186442, | |
| "loss": 1.5901, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3776, | |
| "grad_norm": 9.163947652268689e-06, | |
| "learning_rate": 0.00016406779661016948, | |
| "loss": 1.5237, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.4977602404542267e-05, | |
| "learning_rate": 0.00016338983050847458, | |
| "loss": 1.6332, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3904, | |
| "grad_norm": 1.1450239981058985e-05, | |
| "learning_rate": 0.00016271186440677967, | |
| "loss": 1.5703, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 9.222550033882726e-06, | |
| "learning_rate": 0.00016203389830508476, | |
| "loss": 1.6719, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 9.58750842983136e-06, | |
| "learning_rate": 0.00016135593220338985, | |
| "loss": 1.5141, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 8.320729648403358e-06, | |
| "learning_rate": 0.00016067796610169492, | |
| "loss": 1.5551, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 1.0597662367217708e-05, | |
| "learning_rate": 0.00016, | |
| "loss": 1.6389, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 7.677650501136668e-06, | |
| "learning_rate": 0.00015932203389830508, | |
| "loss": 1.4681, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4288, | |
| "grad_norm": 7.769591320538893e-06, | |
| "learning_rate": 0.0001586440677966102, | |
| "loss": 1.6009, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 8.620034350315109e-06, | |
| "learning_rate": 0.00015796610169491526, | |
| "loss": 1.6013, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 8.902542504074518e-06, | |
| "learning_rate": 0.00015728813559322036, | |
| "loss": 1.3544, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.3868037967768032e-05, | |
| "learning_rate": 0.00015661016949152542, | |
| "loss": 1.6156, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4544, | |
| "grad_norm": 9.34260697249556e-06, | |
| "learning_rate": 0.00015593220338983051, | |
| "loss": 1.578, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 9.92396871879464e-06, | |
| "learning_rate": 0.0001552542372881356, | |
| "loss": 1.5632, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.4672, | |
| "grad_norm": 9.71571535046678e-06, | |
| "learning_rate": 0.0001545762711864407, | |
| "loss": 1.6627, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.1435681699367706e-05, | |
| "learning_rate": 0.00015389830508474577, | |
| "loss": 1.4347, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.1326865205774084e-05, | |
| "learning_rate": 0.00015322033898305086, | |
| "loss": 1.5778, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 9.508394214208238e-06, | |
| "learning_rate": 0.00015254237288135592, | |
| "loss": 1.5227, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4928, | |
| "grad_norm": 9.118397429119796e-06, | |
| "learning_rate": 0.00015186440677966102, | |
| "loss": 1.4213, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.0663600733096246e-05, | |
| "learning_rate": 0.0001511864406779661, | |
| "loss": 1.4303, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5056, | |
| "grad_norm": 1.0753619790193625e-05, | |
| "learning_rate": 0.0001505084745762712, | |
| "loss": 1.558, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 8.718889148440212e-06, | |
| "learning_rate": 0.00014983050847457627, | |
| "loss": 1.4912, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 1.0215439942840021e-05, | |
| "learning_rate": 0.00014915254237288136, | |
| "loss": 1.3035, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 9.638847586757038e-06, | |
| "learning_rate": 0.00014847457627118645, | |
| "loss": 1.5648, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5312, | |
| "grad_norm": 9.258928912458941e-06, | |
| "learning_rate": 0.00014779661016949154, | |
| "loss": 1.671, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.3850747563992627e-05, | |
| "learning_rate": 0.0001471186440677966, | |
| "loss": 1.5068, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 3.328406819491647e-05, | |
| "learning_rate": 0.0001464406779661017, | |
| "loss": 1.5843, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 9.368337487103418e-06, | |
| "learning_rate": 0.00014576271186440677, | |
| "loss": 1.6765, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 1.0530213330639526e-05, | |
| "learning_rate": 0.00014508474576271186, | |
| "loss": 1.4233, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 8.799790521152318e-06, | |
| "learning_rate": 0.00014440677966101695, | |
| "loss": 1.3518, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5696, | |
| "grad_norm": 9.422311450180132e-06, | |
| "learning_rate": 0.00014372881355932205, | |
| "loss": 1.4354, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.0388404916739091e-05, | |
| "learning_rate": 0.00014305084745762714, | |
| "loss": 1.4614, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5824, | |
| "grad_norm": 1.0160752026422415e-05, | |
| "learning_rate": 0.0001423728813559322, | |
| "loss": 1.52, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 9.5980967671494e-06, | |
| "learning_rate": 0.0001416949152542373, | |
| "loss": 1.4679, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 9.245924957212992e-06, | |
| "learning_rate": 0.0001410169491525424, | |
| "loss": 1.5605, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 1.3253756151243579e-05, | |
| "learning_rate": 0.00014033898305084748, | |
| "loss": 1.4912, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 9.214073543262202e-06, | |
| "learning_rate": 0.00013966101694915255, | |
| "loss": 1.4879, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 9.253428288502619e-06, | |
| "learning_rate": 0.00013898305084745764, | |
| "loss": 1.3903, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6208, | |
| "grad_norm": 1.098973098123679e-05, | |
| "learning_rate": 0.0001383050847457627, | |
| "loss": 1.5223, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 9.57713564275764e-06, | |
| "learning_rate": 0.0001376271186440678, | |
| "loss": 1.4031, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 9.398447218700312e-06, | |
| "learning_rate": 0.0001369491525423729, | |
| "loss": 1.4427, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.282966240978567e-05, | |
| "learning_rate": 0.00013627118644067798, | |
| "loss": 1.4614, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6464, | |
| "grad_norm": 1.0865313015528955e-05, | |
| "learning_rate": 0.00013559322033898305, | |
| "loss": 1.5453, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 1.0939942512777634e-05, | |
| "learning_rate": 0.00013491525423728814, | |
| "loss": 1.5382, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6592, | |
| "grad_norm": 1.0936541002593003e-05, | |
| "learning_rate": 0.0001342372881355932, | |
| "loss": 1.5441, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 8.611181328888051e-06, | |
| "learning_rate": 0.00013355932203389833, | |
| "loss": 1.4941, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 9.062526260095183e-06, | |
| "learning_rate": 0.0001328813559322034, | |
| "loss": 1.3454, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 1.031234296533512e-05, | |
| "learning_rate": 0.00013220338983050849, | |
| "loss": 1.5647, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6848, | |
| "grad_norm": 1.1043604899896309e-05, | |
| "learning_rate": 0.00013152542372881355, | |
| "loss": 1.498, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.024298671836732e-05, | |
| "learning_rate": 0.00013084745762711864, | |
| "loss": 1.5241, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6976, | |
| "grad_norm": 8.635867743578274e-06, | |
| "learning_rate": 0.00013016949152542374, | |
| "loss": 1.4648, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.0859222129511181e-05, | |
| "learning_rate": 0.00012949152542372883, | |
| "loss": 1.5595, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 8.587241609347984e-06, | |
| "learning_rate": 0.0001288135593220339, | |
| "loss": 1.4229, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 9.152197890216485e-06, | |
| "learning_rate": 0.000128135593220339, | |
| "loss": 1.3533, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7232, | |
| "grad_norm": 9.574167961545754e-06, | |
| "learning_rate": 0.00012745762711864405, | |
| "loss": 1.3756, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 9.985215911001433e-06, | |
| "learning_rate": 0.00012677966101694917, | |
| "loss": 1.4022, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 9.489425792708062e-06, | |
| "learning_rate": 0.00012610169491525426, | |
| "loss": 1.4269, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 1.0372446922701783e-05, | |
| "learning_rate": 0.00012542372881355933, | |
| "loss": 1.4425, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 1.1192561032657977e-05, | |
| "learning_rate": 0.00012474576271186442, | |
| "loss": 1.5135, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.091165449906839e-05, | |
| "learning_rate": 0.0001240677966101695, | |
| "loss": 1.6096, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7616, | |
| "grad_norm": 8.728093234822154e-06, | |
| "learning_rate": 0.00012338983050847458, | |
| "loss": 1.3809, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.1227505638089497e-05, | |
| "learning_rate": 0.00012271186440677967, | |
| "loss": 1.3501, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7744, | |
| "grad_norm": 8.897823136067018e-06, | |
| "learning_rate": 0.00012203389830508477, | |
| "loss": 1.5278, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 9.632080036681145e-06, | |
| "learning_rate": 0.00012135593220338983, | |
| "loss": 1.4077, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 7.88291981734801e-06, | |
| "learning_rate": 0.00012067796610169492, | |
| "loss": 1.3762, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 8.036658073251601e-06, | |
| "learning_rate": 0.00012, | |
| "loss": 1.356, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 8.462204277748242e-06, | |
| "learning_rate": 0.0001193220338983051, | |
| "loss": 1.4397, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 9.903522368404083e-06, | |
| "learning_rate": 0.00011864406779661017, | |
| "loss": 1.5393, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8128, | |
| "grad_norm": 7.63375010137679e-06, | |
| "learning_rate": 0.00011796610169491527, | |
| "loss": 1.3557, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 8.234695087594446e-06, | |
| "learning_rate": 0.00011728813559322033, | |
| "loss": 1.394, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 8.248807716881856e-06, | |
| "learning_rate": 0.00011661016949152544, | |
| "loss": 1.4227, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 7.1838967414805666e-06, | |
| "learning_rate": 0.0001159322033898305, | |
| "loss": 1.4423, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8384, | |
| "grad_norm": 8.325100679940078e-06, | |
| "learning_rate": 0.0001152542372881356, | |
| "loss": 1.4575, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 8.66544451127993e-06, | |
| "learning_rate": 0.00011457627118644068, | |
| "loss": 1.5005, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8512, | |
| "grad_norm": 6.907540864631301e-06, | |
| "learning_rate": 0.00011389830508474577, | |
| "loss": 1.3423, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 8.388161404582206e-06, | |
| "learning_rate": 0.00011322033898305085, | |
| "loss": 1.3535, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 9.66962306847563e-06, | |
| "learning_rate": 0.00011254237288135594, | |
| "loss": 1.4762, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 8.090796654869337e-06, | |
| "learning_rate": 0.00011186440677966102, | |
| "loss": 1.3992, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8768, | |
| "grad_norm": 8.612944839114789e-06, | |
| "learning_rate": 0.00011118644067796611, | |
| "loss": 1.3203, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 8.227701982832514e-06, | |
| "learning_rate": 0.00011050847457627118, | |
| "loss": 1.3659, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8896, | |
| "grad_norm": 7.986993296071887e-06, | |
| "learning_rate": 0.00010983050847457627, | |
| "loss": 1.385, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.0559672773524653e-05, | |
| "learning_rate": 0.00010915254237288135, | |
| "loss": 1.4261, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 8.250416613009293e-06, | |
| "learning_rate": 0.00010847457627118644, | |
| "loss": 1.4343, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 1.0122534149559215e-05, | |
| "learning_rate": 0.00010779661016949153, | |
| "loss": 1.3774, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9152, | |
| "grad_norm": 1.064319621946197e-05, | |
| "learning_rate": 0.00010711864406779661, | |
| "loss": 1.4693, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 9.444781426282134e-06, | |
| "learning_rate": 0.0001064406779661017, | |
| "loss": 1.4116, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 9.153848623100203e-06, | |
| "learning_rate": 0.00010576271186440679, | |
| "loss": 1.4077, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 9.204763955494855e-06, | |
| "learning_rate": 0.00010508474576271188, | |
| "loss": 1.3664, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 8.885639545042068e-06, | |
| "learning_rate": 0.00010440677966101696, | |
| "loss": 1.3931, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 1.0133306204807013e-05, | |
| "learning_rate": 0.00010372881355932205, | |
| "loss": 1.4038, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.9536, | |
| "grad_norm": 9.189009688270744e-06, | |
| "learning_rate": 0.00010305084745762712, | |
| "loss": 1.3064, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 9.257129022444133e-06, | |
| "learning_rate": 0.00010237288135593222, | |
| "loss": 1.4308, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9664, | |
| "grad_norm": 8.574080311518628e-06, | |
| "learning_rate": 0.00010169491525423729, | |
| "loss": 1.4926, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 8.716850061318837e-06, | |
| "learning_rate": 0.00010101694915254238, | |
| "loss": 1.432, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 8.194006113626529e-06, | |
| "learning_rate": 0.00010033898305084746, | |
| "loss": 1.4217, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 8.842445822665468e-06, | |
| "learning_rate": 9.966101694915255e-05, | |
| "loss": 1.5852, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 8.663917469675653e-06, | |
| "learning_rate": 9.898305084745763e-05, | |
| "loss": 1.4492, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 9.816942110774107e-06, | |
| "learning_rate": 9.830508474576272e-05, | |
| "loss": 1.2898, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.8757211364572868e-05, | |
| "learning_rate": 9.76271186440678e-05, | |
| "loss": 1.7219, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0064, | |
| "grad_norm": 8.92592106538359e-06, | |
| "learning_rate": 9.69491525423729e-05, | |
| "loss": 1.2517, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0128, | |
| "grad_norm": 8.74044872034574e-06, | |
| "learning_rate": 9.627118644067797e-05, | |
| "loss": 1.3098, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0192, | |
| "grad_norm": 8.749080734560266e-06, | |
| "learning_rate": 9.559322033898305e-05, | |
| "loss": 1.3814, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0256, | |
| "grad_norm": 9.595259143679868e-06, | |
| "learning_rate": 9.491525423728815e-05, | |
| "loss": 1.4628, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 9.312339898315258e-06, | |
| "learning_rate": 9.423728813559322e-05, | |
| "loss": 1.3214, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0384, | |
| "grad_norm": 8.024624548852444e-06, | |
| "learning_rate": 9.355932203389832e-05, | |
| "loss": 1.4862, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.0448, | |
| "grad_norm": 8.171000445145182e-06, | |
| "learning_rate": 9.28813559322034e-05, | |
| "loss": 1.3171, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0512, | |
| "grad_norm": 7.481769443984376e-06, | |
| "learning_rate": 9.220338983050847e-05, | |
| "loss": 1.4089, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.0576, | |
| "grad_norm": 8.156550393323414e-06, | |
| "learning_rate": 9.152542372881357e-05, | |
| "loss": 1.5559, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 8.366131623915862e-06, | |
| "learning_rate": 9.084745762711865e-05, | |
| "loss": 1.5331, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.0704, | |
| "grad_norm": 7.666075362067204e-06, | |
| "learning_rate": 9.016949152542374e-05, | |
| "loss": 1.4877, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.0768, | |
| "grad_norm": 8.464576239930466e-06, | |
| "learning_rate": 8.949152542372882e-05, | |
| "loss": 1.4605, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0832, | |
| "grad_norm": 8.2285159805906e-06, | |
| "learning_rate": 8.88135593220339e-05, | |
| "loss": 1.3632, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0896, | |
| "grad_norm": 8.318780601257458e-06, | |
| "learning_rate": 8.813559322033899e-05, | |
| "loss": 1.4802, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.096, | |
| "grad_norm": 9.67723099165596e-06, | |
| "learning_rate": 8.745762711864407e-05, | |
| "loss": 1.3884, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1024, | |
| "grad_norm": 9.269687325286213e-06, | |
| "learning_rate": 8.677966101694915e-05, | |
| "loss": 1.4619, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1088, | |
| "grad_norm": 9.004014827951323e-06, | |
| "learning_rate": 8.610169491525424e-05, | |
| "loss": 1.4982, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1152, | |
| "grad_norm": 7.967708370415494e-06, | |
| "learning_rate": 8.542372881355932e-05, | |
| "loss": 1.5949, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1216, | |
| "grad_norm": 8.402012099395506e-06, | |
| "learning_rate": 8.474576271186441e-05, | |
| "loss": 1.5464, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 8.84342171048047e-06, | |
| "learning_rate": 8.406779661016949e-05, | |
| "loss": 1.3175, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1344, | |
| "grad_norm": 9.028297426993959e-06, | |
| "learning_rate": 8.338983050847457e-05, | |
| "loss": 1.3999, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.1408, | |
| "grad_norm": 1.0762661986518651e-05, | |
| "learning_rate": 8.271186440677966e-05, | |
| "loss": 1.3264, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.1472, | |
| "grad_norm": 9.566549124429002e-06, | |
| "learning_rate": 8.203389830508474e-05, | |
| "loss": 1.4701, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1536, | |
| "grad_norm": 8.58042221807409e-06, | |
| "learning_rate": 8.135593220338983e-05, | |
| "loss": 1.2315, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 8.269061254395638e-06, | |
| "learning_rate": 8.067796610169493e-05, | |
| "loss": 1.5126, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.1663999999999999, | |
| "grad_norm": 8.415841875830665e-06, | |
| "learning_rate": 8e-05, | |
| "loss": 1.3945, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.1728, | |
| "grad_norm": 9.450182005821262e-06, | |
| "learning_rate": 7.93220338983051e-05, | |
| "loss": 1.4273, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.1792, | |
| "grad_norm": 8.444080776826013e-06, | |
| "learning_rate": 7.864406779661018e-05, | |
| "loss": 1.2723, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.1856, | |
| "grad_norm": 8.979165613709483e-06, | |
| "learning_rate": 7.796610169491526e-05, | |
| "loss": 1.3041, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.192, | |
| "grad_norm": 1.0844120879482944e-05, | |
| "learning_rate": 7.728813559322035e-05, | |
| "loss": 1.4845, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.1984, | |
| "grad_norm": 8.827923011267558e-06, | |
| "learning_rate": 7.661016949152543e-05, | |
| "loss": 1.3927, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2048, | |
| "grad_norm": 9.292446520703379e-06, | |
| "learning_rate": 7.593220338983051e-05, | |
| "loss": 1.4373, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2112, | |
| "grad_norm": 9.616896022635046e-06, | |
| "learning_rate": 7.52542372881356e-05, | |
| "loss": 1.5701, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2176, | |
| "grad_norm": 9.281196071242448e-06, | |
| "learning_rate": 7.457627118644068e-05, | |
| "loss": 1.4267, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 8.995242751552723e-06, | |
| "learning_rate": 7.389830508474577e-05, | |
| "loss": 1.5239, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2304, | |
| "grad_norm": 8.705556865606923e-06, | |
| "learning_rate": 7.322033898305085e-05, | |
| "loss": 1.4249, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.2368000000000001, | |
| "grad_norm": 1.0153792572964448e-05, | |
| "learning_rate": 7.254237288135593e-05, | |
| "loss": 1.3514, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.2432, | |
| "grad_norm": 9.569729627401102e-06, | |
| "learning_rate": 7.186440677966102e-05, | |
| "loss": 1.4394, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.2496, | |
| "grad_norm": 9.4219876700663e-06, | |
| "learning_rate": 7.11864406779661e-05, | |
| "loss": 1.4073, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 1.0102286069013644e-05, | |
| "learning_rate": 7.05084745762712e-05, | |
| "loss": 1.4447, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.2624, | |
| "grad_norm": 8.233397238655016e-06, | |
| "learning_rate": 6.983050847457627e-05, | |
| "loss": 1.4248, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.2688, | |
| "grad_norm": 1.0032190402853303e-05, | |
| "learning_rate": 6.915254237288135e-05, | |
| "loss": 1.2962, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.2752, | |
| "grad_norm": 9.319129276263993e-06, | |
| "learning_rate": 6.847457627118645e-05, | |
| "loss": 1.4528, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2816, | |
| "grad_norm": 9.403600415680557e-06, | |
| "learning_rate": 6.779661016949152e-05, | |
| "loss": 1.3798, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.288, | |
| "grad_norm": 9.744459930516314e-06, | |
| "learning_rate": 6.71186440677966e-05, | |
| "loss": 1.43, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.2944, | |
| "grad_norm": 9.615372619009577e-06, | |
| "learning_rate": 6.64406779661017e-05, | |
| "loss": 1.455, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.3008, | |
| "grad_norm": 9.919615877151955e-06, | |
| "learning_rate": 6.576271186440678e-05, | |
| "loss": 1.5919, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.3072, | |
| "grad_norm": 9.284139196097385e-06, | |
| "learning_rate": 6.508474576271187e-05, | |
| "loss": 1.398, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.3136, | |
| "grad_norm": 9.643885277910158e-06, | |
| "learning_rate": 6.440677966101695e-05, | |
| "loss": 1.3843, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 8.82606218510773e-06, | |
| "learning_rate": 6.372881355932203e-05, | |
| "loss": 1.4858, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.3264, | |
| "grad_norm": 9.312198017141782e-06, | |
| "learning_rate": 6.305084745762713e-05, | |
| "loss": 1.3652, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.3328, | |
| "grad_norm": 9.111182407650631e-06, | |
| "learning_rate": 6.237288135593221e-05, | |
| "loss": 1.3979, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.3392, | |
| "grad_norm": 8.83075608726358e-06, | |
| "learning_rate": 6.169491525423729e-05, | |
| "loss": 1.5379, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3456000000000001, | |
| "grad_norm": 1.0460000339662656e-05, | |
| "learning_rate": 6.101694915254238e-05, | |
| "loss": 1.4637, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 1.0193573871219996e-05, | |
| "learning_rate": 6.033898305084746e-05, | |
| "loss": 1.4097, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.3584, | |
| "grad_norm": 1.0683503205655143e-05, | |
| "learning_rate": 5.966101694915255e-05, | |
| "loss": 1.3366, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.3648, | |
| "grad_norm": 1.099618475564057e-05, | |
| "learning_rate": 5.8983050847457634e-05, | |
| "loss": 1.3535, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.3712, | |
| "grad_norm": 8.492717824992724e-06, | |
| "learning_rate": 5.830508474576272e-05, | |
| "loss": 1.4466, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.3776, | |
| "grad_norm": 9.0052853920497e-06, | |
| "learning_rate": 5.76271186440678e-05, | |
| "loss": 1.3828, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.384, | |
| "grad_norm": 8.887329386197962e-06, | |
| "learning_rate": 5.6949152542372884e-05, | |
| "loss": 1.4676, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.3904, | |
| "grad_norm": 9.442865120945498e-06, | |
| "learning_rate": 5.627118644067797e-05, | |
| "loss": 1.3521, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.3968, | |
| "grad_norm": 9.431981197849382e-06, | |
| "learning_rate": 5.5593220338983056e-05, | |
| "loss": 1.4575, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.4032, | |
| "grad_norm": 9.556749319017399e-06, | |
| "learning_rate": 5.4915254237288135e-05, | |
| "loss": 1.382, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.4096, | |
| "grad_norm": 9.188460353470873e-06, | |
| "learning_rate": 5.423728813559322e-05, | |
| "loss": 1.4435, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 1.2848876394855324e-05, | |
| "learning_rate": 5.355932203389831e-05, | |
| "loss": 1.4886, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.4224, | |
| "grad_norm": 8.7191392594832e-06, | |
| "learning_rate": 5.288135593220339e-05, | |
| "loss": 1.575, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.4288, | |
| "grad_norm": 1.0412767551315483e-05, | |
| "learning_rate": 5.220338983050848e-05, | |
| "loss": 1.3876, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.4352, | |
| "grad_norm": 1.0227216989733279e-05, | |
| "learning_rate": 5.152542372881356e-05, | |
| "loss": 1.5142, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.4416, | |
| "grad_norm": 9.834668162511662e-06, | |
| "learning_rate": 5.0847457627118643e-05, | |
| "loss": 1.4878, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 9.492201570537873e-06, | |
| "learning_rate": 5.016949152542373e-05, | |
| "loss": 1.5036, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.4544000000000001, | |
| "grad_norm": 9.64340961218113e-06, | |
| "learning_rate": 4.9491525423728815e-05, | |
| "loss": 1.291, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.4607999999999999, | |
| "grad_norm": 1.0210535947408061e-05, | |
| "learning_rate": 4.88135593220339e-05, | |
| "loss": 1.3764, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.4672, | |
| "grad_norm": 1.0165509593207389e-05, | |
| "learning_rate": 4.813559322033899e-05, | |
| "loss": 1.3909, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.4736, | |
| "grad_norm": 1.1141897630295716e-05, | |
| "learning_rate": 4.745762711864407e-05, | |
| "loss": 1.5371, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.0782315257529262e-05, | |
| "learning_rate": 4.677966101694916e-05, | |
| "loss": 1.3645, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.4864, | |
| "grad_norm": 9.489853255217895e-06, | |
| "learning_rate": 4.610169491525424e-05, | |
| "loss": 1.3953, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.4928, | |
| "grad_norm": 1.0764860235212836e-05, | |
| "learning_rate": 4.542372881355932e-05, | |
| "loss": 1.4158, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.4992, | |
| "grad_norm": 9.499301086179912e-06, | |
| "learning_rate": 4.474576271186441e-05, | |
| "loss": 1.4482, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.5056, | |
| "grad_norm": 9.37463300942909e-06, | |
| "learning_rate": 4.4067796610169495e-05, | |
| "loss": 1.3295, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 8.773986337473616e-06, | |
| "learning_rate": 4.3389830508474574e-05, | |
| "loss": 1.4853, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.5184, | |
| "grad_norm": 9.019207027449738e-06, | |
| "learning_rate": 4.271186440677966e-05, | |
| "loss": 1.4459, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.5248, | |
| "grad_norm": 9.02684678294463e-06, | |
| "learning_rate": 4.2033898305084746e-05, | |
| "loss": 1.409, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.5312000000000001, | |
| "grad_norm": 9.173134458251297e-06, | |
| "learning_rate": 4.135593220338983e-05, | |
| "loss": 1.44, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.5375999999999999, | |
| "grad_norm": 8.426234671787824e-06, | |
| "learning_rate": 4.067796610169492e-05, | |
| "loss": 1.3803, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 9.337900337413885e-06, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3766, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.5504, | |
| "grad_norm": 9.299056728195865e-06, | |
| "learning_rate": 3.932203389830509e-05, | |
| "loss": 1.4039, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.5568, | |
| "grad_norm": 9.916229828377254e-06, | |
| "learning_rate": 3.8644067796610175e-05, | |
| "loss": 1.345, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.5632000000000001, | |
| "grad_norm": 9.195588063448668e-06, | |
| "learning_rate": 3.7966101694915254e-05, | |
| "loss": 1.2379, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.5695999999999999, | |
| "grad_norm": 1.062252249539597e-05, | |
| "learning_rate": 3.728813559322034e-05, | |
| "loss": 1.4613, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.576, | |
| "grad_norm": 1.0199491953244433e-05, | |
| "learning_rate": 3.6610169491525426e-05, | |
| "loss": 1.3225, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.5824, | |
| "grad_norm": 1.1027201253455132e-05, | |
| "learning_rate": 3.593220338983051e-05, | |
| "loss": 1.2296, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.5888, | |
| "grad_norm": 9.499065527052153e-06, | |
| "learning_rate": 3.52542372881356e-05, | |
| "loss": 1.2963, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.5952, | |
| "grad_norm": 1.012124903354561e-05, | |
| "learning_rate": 3.4576271186440676e-05, | |
| "loss": 1.57, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6016, | |
| "grad_norm": 7.979737347341143e-06, | |
| "learning_rate": 3.389830508474576e-05, | |
| "loss": 1.2537, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 1.0675358680600766e-05, | |
| "learning_rate": 3.322033898305085e-05, | |
| "loss": 1.3929, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.6143999999999998, | |
| "grad_norm": 8.171174158633221e-06, | |
| "learning_rate": 3.2542372881355934e-05, | |
| "loss": 1.4492, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.6208, | |
| "grad_norm": 9.480341759626754e-06, | |
| "learning_rate": 3.186440677966101e-05, | |
| "loss": 1.5049, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.6272, | |
| "grad_norm": 9.047450475918595e-06, | |
| "learning_rate": 3.1186440677966106e-05, | |
| "loss": 1.331, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.6336, | |
| "grad_norm": 8.962598258221988e-06, | |
| "learning_rate": 3.050847457627119e-05, | |
| "loss": 1.515, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 1.0434688192617614e-05, | |
| "learning_rate": 2.9830508474576274e-05, | |
| "loss": 1.5539, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.6463999999999999, | |
| "grad_norm": 9.877274351310916e-06, | |
| "learning_rate": 2.915254237288136e-05, | |
| "loss": 1.4858, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.6528, | |
| "grad_norm": 1.055645952874329e-05, | |
| "learning_rate": 2.8474576271186442e-05, | |
| "loss": 1.4561, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.6592, | |
| "grad_norm": 1.05274293673574e-05, | |
| "learning_rate": 2.7796610169491528e-05, | |
| "loss": 1.3357, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.6656, | |
| "grad_norm": 9.783374480321072e-06, | |
| "learning_rate": 2.711864406779661e-05, | |
| "loss": 1.3287, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.6720000000000002, | |
| "grad_norm": 1.0285318239766639e-05, | |
| "learning_rate": 2.6440677966101696e-05, | |
| "loss": 1.5907, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.6784, | |
| "grad_norm": 1.0502969416847918e-05, | |
| "learning_rate": 2.576271186440678e-05, | |
| "loss": 1.5564, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.6848, | |
| "grad_norm": 9.352576853416394e-06, | |
| "learning_rate": 2.5084745762711865e-05, | |
| "loss": 1.3834, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.6912, | |
| "grad_norm": 1.0319121429347433e-05, | |
| "learning_rate": 2.440677966101695e-05, | |
| "loss": 1.4144, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.6976, | |
| "grad_norm": 1.0816199392138515e-05, | |
| "learning_rate": 2.3728813559322036e-05, | |
| "loss": 1.4279, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 9.127339581027627e-06, | |
| "learning_rate": 2.305084745762712e-05, | |
| "loss": 1.2921, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.7104, | |
| "grad_norm": 1.0231586202280596e-05, | |
| "learning_rate": 2.2372881355932205e-05, | |
| "loss": 1.4732, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.7168, | |
| "grad_norm": 1.0996167475241236e-05, | |
| "learning_rate": 2.1694915254237287e-05, | |
| "loss": 1.4838, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.7231999999999998, | |
| "grad_norm": 9.594040420779493e-06, | |
| "learning_rate": 2.1016949152542373e-05, | |
| "loss": 1.4376, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.7296, | |
| "grad_norm": 1.0523729542910587e-05, | |
| "learning_rate": 2.033898305084746e-05, | |
| "loss": 1.3195, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 9.395639608555939e-06, | |
| "learning_rate": 1.9661016949152545e-05, | |
| "loss": 1.3746, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.7424, | |
| "grad_norm": 1.078158038581023e-05, | |
| "learning_rate": 1.8983050847457627e-05, | |
| "loss": 1.3984, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.7488000000000001, | |
| "grad_norm": 9.743846021592617e-06, | |
| "learning_rate": 1.8305084745762713e-05, | |
| "loss": 1.3992, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.7551999999999999, | |
| "grad_norm": 1.0031043530034367e-05, | |
| "learning_rate": 1.76271186440678e-05, | |
| "loss": 1.2568, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.7616, | |
| "grad_norm": 9.389639672008343e-06, | |
| "learning_rate": 1.694915254237288e-05, | |
| "loss": 1.4703, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.768, | |
| "grad_norm": 8.901612090994604e-06, | |
| "learning_rate": 1.6271186440677967e-05, | |
| "loss": 1.2991, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.7744, | |
| "grad_norm": 9.43897248362191e-06, | |
| "learning_rate": 1.5593220338983053e-05, | |
| "loss": 1.3101, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.7808000000000002, | |
| "grad_norm": 9.650594620325137e-06, | |
| "learning_rate": 1.4915254237288137e-05, | |
| "loss": 1.4399, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.7872, | |
| "grad_norm": 1.0018439752457198e-05, | |
| "learning_rate": 1.4237288135593221e-05, | |
| "loss": 1.5083, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.7936, | |
| "grad_norm": 1.0699907761591021e-05, | |
| "learning_rate": 1.3559322033898305e-05, | |
| "loss": 1.328, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.0454657967784442e-05, | |
| "learning_rate": 1.288135593220339e-05, | |
| "loss": 1.4945, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.8064, | |
| "grad_norm": 1.0842384654097259e-05, | |
| "learning_rate": 1.2203389830508475e-05, | |
| "loss": 1.4409, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.8128, | |
| "grad_norm": 1.3086712897347752e-05, | |
| "learning_rate": 1.152542372881356e-05, | |
| "loss": 1.3681, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.8192, | |
| "grad_norm": 1.085216263163602e-05, | |
| "learning_rate": 1.0847457627118644e-05, | |
| "loss": 1.3303, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.8256000000000001, | |
| "grad_norm": 9.203211448038928e-06, | |
| "learning_rate": 1.016949152542373e-05, | |
| "loss": 1.4076, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 1.1280193575657904e-05, | |
| "learning_rate": 9.491525423728814e-06, | |
| "loss": 1.3461, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.8384, | |
| "grad_norm": 1.0468848813616205e-05, | |
| "learning_rate": 8.8135593220339e-06, | |
| "loss": 1.4124, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.8448, | |
| "grad_norm": 9.741090252646245e-06, | |
| "learning_rate": 8.135593220338983e-06, | |
| "loss": 1.3195, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.8512, | |
| "grad_norm": 1.0476109309820458e-05, | |
| "learning_rate": 7.4576271186440685e-06, | |
| "loss": 1.3275, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.8576000000000001, | |
| "grad_norm": 1.0405379725852981e-05, | |
| "learning_rate": 6.779661016949153e-06, | |
| "loss": 1.3069, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.8639999999999999, | |
| "grad_norm": 9.575111107551493e-06, | |
| "learning_rate": 6.101694915254238e-06, | |
| "loss": 1.4803, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.8704, | |
| "grad_norm": 1.0216043847321998e-05, | |
| "learning_rate": 5.423728813559322e-06, | |
| "loss": 1.4113, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.8768, | |
| "grad_norm": 1.198629706777865e-05, | |
| "learning_rate": 4.745762711864407e-06, | |
| "loss": 1.3817, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.8832, | |
| "grad_norm": 9.737221262184903e-06, | |
| "learning_rate": 4.067796610169492e-06, | |
| "loss": 1.3605, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.8896, | |
| "grad_norm": 1.0807473699969705e-05, | |
| "learning_rate": 3.3898305084745763e-06, | |
| "loss": 1.3617, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 1.0247385034745093e-05, | |
| "learning_rate": 2.711864406779661e-06, | |
| "loss": 1.4461, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.9024, | |
| "grad_norm": 9.495196536590811e-06, | |
| "learning_rate": 2.033898305084746e-06, | |
| "loss": 1.5681, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.9088, | |
| "grad_norm": 1.012134180200519e-05, | |
| "learning_rate": 1.3559322033898304e-06, | |
| "loss": 1.4065, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.9152, | |
| "grad_norm": 9.227464943251107e-06, | |
| "learning_rate": 6.779661016949152e-07, | |
| "loss": 1.4213, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.919460493180928e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |
Xet Storage Details
- Size:
- 50.1 kB
- Xet hash:
- b798acee4a1ef0ee6278eb4db8f6b30056e4f51388dd61efac14b1d6baf27000
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.